ICs for Communications
DMA Supported Serial Communication Controller with 4 Channels
DSCC4
PEB 20534 Version 2.0
Data Sheet 09.98 DS 2
PEB 20534
Revision History: Current Version: 09.98
Previous Version: Data Sheet 07.97 (V 1.1)
Page
(in previou s
Version)
Page
(in current
Version)
Subjects (major changes since last revision)
- - This Data Sheet is completely reorganized.
For questions on technology, delivery and prices please contact the Semiconductor
Group Offices in Germany or the Siemens Companies and Representatives worldwide:
see our webpage at http://www.siemens.de/semiconductor/communication.
Edition 09.98
Published by Siemens AG,
HL SC,
Balans tr aß e 73,
81541 M ünchen
© Siemens AG 1998.
All Rights Re se rv ed.
Attention please!
As far as patents or other rights of third parties are concerned, liability is only assumed for components, not for
applica tio ns , proc es s es and c irc uit s i mp lem ented within com ponents or assem blies.
The inform at ion describes the t yp e of co m ponent and shall not be c ons idered as assure d c haracteristics.
Terms of delivery and right s to ch ange design reserv ed.
Due to technical requirements components may cont ain dangerous substances. For information on the t y pes in
question please conta ct yo ur nearest Sieme ns Offic e, Sem ic onductor Group.
Siemens AG is an approved CECC manufacturer.
Packing
Please use the rec ycling operators k now n t o y ou. W e ca n als o help you – get in touch w it h yo ur nearest sales
office. By agreement we will take packing material back, if it is sorted. You must bear the costs of transport.
For packing material that is returned to us unsorted or which we are not obliged to accept, we shall have to invoice
you for any costs incurred.
Components used in life-support devices or systems must be expressly authorized for such purpose!
Critical com ponents1 of the Sem ic onductor Group of Siemens AG, may only be us ed in life-supp ort dev ic es or
systems2 with the express written approval of the Sem ic onductor Group of Siemens AG.
1 A critical component is a component used in a life-support device or system whose failure can reasonably be
expected to cause the failure of that life-support device or system, or to affect its safety or effectiveness of that
device or system .
2 Life support devices or systems are intended (a) to be implanted in the human body, or (b) to support and/or
maintain and s us t ain human life. If they fail, it is rea so nable to assum e th at the healt h of th e us er m ay be en-
dangered.
IOM®, IOM®-1, IOM®-2, SICOFI®, SICOFI®-2, SICOF I®-4, SICOFI®-4µC, SLICOFI®, ARCO FI® , ARC OFI®-BA,
ARCOFI®-SP, EPIC®-1, EPIC®-S, EL IC®, IPAT®-2, ITA C ®, ISAC®-S, ISAC ®-S T E, IS AC ®-P, ISAC®-P TE, IDEC®,
SICAT®, OCTAT®-P, QUAT®-S are registered tra dem arks of Sieme ns AG.
MUSAC™-A , FALC ™54, IWE™, SA RE™, UTPT™, ASM™, ASP™, DigiTape™ are trade m ark s of Si em ens AG .
All other brand or product names, Hardware or Software names are trademarks or registered trademarks of their
respective c om panies or organ iz at ions .
PEB 20534
Semiconductor Group 3 Data Sheet 09.98
Preface
The
DMA Supported Serial Communication Controller with 4 Channels
(DSCC4) is a
Multi Protocol Controller for a wide range of data communication and telecommunication
applicati ons. Th is docume nt provi des co mplete refere nce inf ormation on hard ware an d
software related issues as well as on general operation.
Organization of this Document
This Data Sheet is divided into 16 chapters. It is organized as follows:
• Chapter 1, Overview
Gives a g eneral descripti on of the product, li sts the key feature s, a nd p r ese nts som e
typical applications.
• Chapter 2, Pin Description
Lists pin location s wi th associ ated sign als, categoriz es signal s acc ording to func tion,
and describes signals.
• Chapters 3, 4,5 ,6,7 Func tion al Descriptio n
These chapters provide detailed descriptions of all DSCC4 internal function blocks.
• Chapter 8, Detailed Protocol Descriptions
Gives a detailed description of all protocols supported by the serial communication
controllers SCCs.
• Chapter 9, Reset and Initializat ion Procedure
Gives examples for DSCC4 initialization procedure and operation.
• Chapter 10, Detailed Register Description
Gives a detailed description of all DSCC4 on chip registers.
• Chapter 11, Host Memory Organization
Provides an overview of all DSCC4 data structures located in the shared memory
• Chapter 12, JTAG Boundary Scan
Gives a detailed description of the boundary scan unit.
• Chapter 13 Application Examples
These chapters will provide additional information on special applications and bus
utilization analysis.
• Chapter 14, Electrical Characteristics
Gives a detailed de scription of all electrical DC and AC c haracteristi cs and prov ides
PEB 20534
Semiconductor Group 4 Data Sheet 09.98
timing diagrams and values for all interfaces.
• Chapter 15, Package Outline
PEB 20534
Table of Contents Page
Semiconductor Group 5 Data Sheet 09.98
1Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18
1.1 Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19
1.2 Differences Between the DSCC4 and the ESCC Family . . . . . . . . . . . . . . .22
1.2.1 Enhancements to the ESCC Serial Core . . . . . . . . . . . . . . . . . . . . . . . . .22
1.2.2 Simplifications to the ESCC Serial Core . . . . . . . . . . . . . . . . . . . . . . . . .22
1.3 Logic Symbol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23
1.4 Typical Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24
1.4.1 Application Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26
1.4.1.1 HSSI Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26
1.4.1.2 HSSI Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27
1.4.1.3 General Data Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28
2Pin Descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29
2.1 Pin Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29
2.2 Pin Definitions and Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30
3Functional Description. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .48
4Microprocessor Bus Interface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .51
4.1 PCI Bus Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .51
4.1.1 Supported PCI Transactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .51
4.1.2 PCI Configuration Space Register Overview . . . . . . . . . . . . . . . . . . . . . .52
4.2 De-multiplexed Bus Interface Extension . . . . . . . . . . . . . . . . . . . . . . . . . . .53
5DMA Controller and Central FIFOs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57
5.1 DMAC Operational Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59
5.1.1 DMAC Register Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59
5.1.2 DMAC Control and Data Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64
5.1.2.1 DMAC Transmit Descriptor Lists . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66
5.1.2.2 DMAC Receive Descriptor Lists . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71
5.1.2.3 DMAC Operation Using Hold-Bit Control Mechanism . . . . . . . . . . . . .76
5.1.2.4 DMAC Operation Using Last Descriptor Address Control Mode . . . . .78
5.1.3 DMAC Interrupt Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81
5.2 Central FIFOs Operational Description . . . . . . . . . . . . . . . . . . . . . . . . . . . .85
5.2.1 Central FIFO Register Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85
5.2.2 Central Transmit FIFO (TFIFO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89
5.2.3 Central Receive FIFO (RFIFO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .92
5.2.4 DMAC Internal Arbitration Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . .93
5.2.5 DMAC Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .94
5.2.6 Little / Big Endian Byte Swap Convention . . . . . . . . . . . . . . . . . . . . . . . .95
6Multi Function Port (MFP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .96
6.1 Local Bus Interface (LBI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98
PEB 20534
Semiconductor Group 6 Data Sheet 09.98
6.1.1 LBI Bus Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .99
6.1.1.1 LBI External Bus Controller (EBC) . . . . . . . . . . . . . . . . . . . . . . . . . . .99
6.1.1.2 Multiplexed Local Bus Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .99
6.1.1.3 De-multiplexed Local Bus Modes . . . . . . . . . . . . . . . . . . . . . . . . . . .100
6.1.1.4 Programmable Bus Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . .102
6.1.1.5 Ready Signal Controlled Bus Cycles . . . . . . . . . . . . . . . . . . . . . . . . .104
6.1.1.6 LBI (EBC) Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .106
6.1.2 LBI Bus Arbitration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .107
6.1.3 PCI to Local Bus Bridge Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . .113
6.1.4 LBI Interrupt Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .114
6.2 Synchronous Serial Control (SSC) Interface . . . . . . . . . . . . . . . . . . . . . . .115
6.2.1 SSC Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .115
6.2.1.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .115
6.2.1.2 Operational Mode: Full-Duplex Operation: . . . . . . . . . . . . . . . . . . . .119
6.2.1.3 Operational Mode: Half Duplex Operation: . . . . . . . . . . . . . . . . . . . .122
6.2.1.4 Baud Rate Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .124
6.2.1.5 Error Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .124
6.2.2 SSC Interrupt (Vector) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .126
6.3 General Purpose Port (GPP) Interface . . . . . . . . . . . . . . . . . . . . . . . . . . .127
6.3.1 GPP Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .127
6.3.2 GPP Interrupt Vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .127
7Serial Communication Controller (SCC) Cores . . . . . . . . . . . . . . . . . . .128
7.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .128
7.2 Protocol Modes Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .132
7.3 SCC FIFOs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .133
7.3.1 SCC Transmit FIFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .133
7.3.2 SCC Receive FIFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .134
7.4 Clocking System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .137
7.4.1 Clock Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .141
7.4.1.1 Clock Mode 0 (0a/0b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .141
7.4.1.2 Clock Mode 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .142
7.4.1.3 Clock Mode 2 (2a/2b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .143
7.4.1.4 Clock Mode 3 (3a/3b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .144
7.4.1.5 Clock Mode 4 (High Speed Interface Clock Mode) . . . . . . . . . . . . . .145
7.4.1.6 Clock Mode 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .147
7.4.1.7 Clock Mode 6 (6a/6b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .152
7.4.1.8 Clock Mode 7 (7a/7b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .153
7.4.2 Baud Rate Generator (BRG) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .154
7.4.3 Clock Recovery (DPLL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .155
7.5 SCC Interrupt Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .158
7.6 High Speed Channel Operation (PEB 20534H-52 only) . . . . . . . . . . . . . .159
7.7 Serial Bus Configuration Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .159
PEB 20534
Semiconductor Group 7 Data Sheet 09.98
7.7.1 Serial Bus Access Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .160
7.7.2 Serial Bus Collisions and Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . .160
7.7.3 Serial Bus Access Priority Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . .161
7.7.4 Serial Bus Configuration Timing Modes . . . . . . . . . . . . . . . . . . . . . . . .161
7.7.5 Functions Of Signal RTS in HDLC Mode . . . . . . . . . . . . . . . . . . . . . . . .162
7.8 Data Encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .162
7.8.1 NRZ and NRZI Encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .162
7.8.2 FM0 and FM1 Encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .163
7.8.3 Manchester Encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .164
7.9 Modem Control Signals (RTS, CTS, CD) . . . . . . . . . . . . . . . . . . . . . . . . .164
7.9.1 RTS/CTS Handshaking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .164
7.9.2 Carrier Detect (CD) Receiver Control . . . . . . . . . . . . . . . . . . . . . . . . . .166
7.10 Local Loop Test Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .166
8Detailed Protocol Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .167
8.1 HDLC/SDLC Protocol Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .169
8.1.1 HDLC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .169
8.1.1.1 Auto Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .170
8.1.1.2 Non Auto Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .170
8.1.1.3 Address Mode 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .171
8.1.1.4 Address Mode 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .171
8.1.2 HDLC/PPP Protocol Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .171
8.1.2.1 Bit Synchronous PPP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .171
8.1.2.2 Octet Synchronous PPP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .172
8.1.2.3 Aynchronous PPP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .172
8.1.3 Extended Transparent Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .172
8.1.4 HDLC Receive Data Processing Overview . . . . . . . . . . . . . . . . . . . . . .172
8.1.5 HDLC Transmit Data Processing Overview . . . . . . . . . . . . . . . . . . . . . .174
8.1.6 Procedural Support (Layer-2 Functions) . . . . . . . . . . . . . . . . . . . . . . . .176
8.1.7 Full-Duplex LAPB/LAPD Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . .177
8.1.8 Half-Duplex SDLC-NRM Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . .182
8.1.9 Special Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .185
8.1.9.1 Extended Transparent Transmission and Reception . . . . . . . . . . . . .185
8.1.9.2 Receive Address Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .185
8.1.9.3 Shared Flags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .185
8.1.9.4 One Bit Insertion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .186
8.1.9.5 Preamble Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .186
8.1.9.6 CRC Generation and Checking . . . . . . . . . . . . . . . . . . . . . . . . . . . . .186
8.1.9.7 Data Transparency in PPP Mode . . . . . . . . . . . . . . . . . . . . . . . . . . .188
8.1.9.8 Receive Length Check Feature . . . . . . . . . . . . . . . . . . . . . . . . . . . . .190
8.2 Asynchronous (ASYNC) Protocol Mode . . . . . . . . . . . . . . . . . . . . . . . . . .191
8.2.1 Character Framing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .191
8.2.2 Data Reception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .191
PEB 20534
Semiconductor Group 8 Data Sheet 09.98
8.2.2.1 Asynchronous Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .191
8.2.2.2 Isochronous Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .192
8.2.2.3 Storage of Receive Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .192
8.2.3 Data Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .193
8.2.4 Special Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .193
8.2.4.1 Break Detection/Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .193
8.2.4.2 In-band Flow Control by XON/XOFF Characters . . . . . . . . . . . . . . .193
8.2.4.3 Out-of-band Flow Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .195
8.3 Character Oriented Synchronous (BISYNC) Protocol Mode . . . . . . . . . . .198
8.3.1 Character Framing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .198
8.3.2 Data Reception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .198
8.3.3 Data Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .199
8.3.4 Special Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .200
8.3.4.1 Preamble Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .200
8.3.4.2 CRC Parity Inhibit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .200
9Reset and Initialization Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . .201
9.1 Reset and Power-On . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .201
9.2 Initialization Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .203
9.3 Start of Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .207
9.4 Initialization Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .214
9.4.1 Test Loop For Data Transfer in HDLC Address Mode 0 . . . . . . . . . . . .214
10 Detailed Register Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .220
10.1 Register Range Overview and Address Mapping . . . . . . . . . . . . . . . . . . .221
10.2 PCI Configuration Space - Detailed Register Description . . . . . . . . . . . . .222
10.3 On-Chip Registers Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .228
10.3.1 Global Registers - Detailed Register Description . . . . . . . . . . . . . . . . .228
10.3.1.1 Global Registers Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .228
10.3.1.2 Global Registers Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .231
10.3.2 SCC Registers - Detailed Register Description . . . . . . . . . . . . . . . . . . .272
10.3.2.1 SCC Registers Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .272
10.3.2.2 SCC Registers Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .274
10.3.3 Peripheral Registers - Detailed Register Description . . . . . . . . . . . . . .349
10.3.3.1 Peripheral Registers Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . .349
10.3.3.2 LBI Registers Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .351
10.3.3.3 SSC Registers Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .355
10.3.3.4 GPP Registers Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .368
11 Host Memory Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .374
11.1 Linked List Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .374
11.1.1 Transmit Descriptor Lists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .374
11.1.1.1 Transmit Descriptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .374
PEB 20534
Semiconductor Group 9 Data Sheet 09.98
11.1.2 Receive Descriptor Lists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .378
11.1.2.1 Receive Descriptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .379
11.1.2.2 Receive Data Section Status Byte (HDLC Mode) . . . . . . . . . . . . . . .383
11.1.2.3 Receive Data Section Status Byte (ASYNC/BISYNC Modes) . . . . . .386
11.2 Interrupt Queue Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .387
11.2.1 Interrupt Queue Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .387
11.2.2 Interrupt Vector Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .389
11.2.2.1 Configuration Interrupt Vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .390
11.2.2.2 DMA Controller Interrupt Vector . . . . . . . . . . . . . . . . . . . . . . . . . . . .391
11.2.2.3 SCC Interrupt Vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .393
11.2.2.4 SSC Interrupt Vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .395
11.2.2.5 LBI Interrupt Vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .397
11.2.2.6 GPP Interrupt Vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .398
12 Test Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .399
12.1 JTAG Boundary Scan Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .399
13 Application Hints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .406
13.1 DSCC4 PCI Bus Utilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .406
13.2 Application Example: De-multiplexed Bus Interface . . . . . . . . . . . . . . . . .407
13.3 SSC Example: Connecting A Serial EEPROM . . . . . . . . . . . . . . . . . . . . .408
14 Electrical Characteristics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .409
14.1 Important Electrical Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .409
14.2 Absolute Maximum Ratings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .411
14.3 Thermal Package Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .411
14.4 DC Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .412
14.5 Capacitances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .413
14.6 AC Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .414
14.6.1 PCI Bus Interface Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .414
14.6.1.1 PCI Read Transaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .415
14.6.1.2 PCI Write Transaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .417
14.6.2 De-multiplexed Bus Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .421
14.6.3 Local Bus Interface Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .423
14.6.4 Local Bus Interface Arbitration Timing . . . . . . . . . . . . . . . . . . . . . . . . . .428
14.6.5 PCM Serial Interface Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .429
14.6.5.1 Clock Input Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .429
14.6.5.2 Receive Cycle Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .431
14.6.5.3 Transmit Cycle Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .433
14.6.5.4 Strobe Timing (clock mode 1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .435
14.6.5.5 Frame Synchronisation Timing (clock mode 5) . . . . . . . . . . . . . . . . .436
14.6.5.6 High Speed Receive Cycle Timing . . . . . . . . . . . . . . . . . . . . . . . . . .437
14.6.5.7 High Speed Transmit Cycle Timing . . . . . . . . . . . . . . . . . . . . . . . . . .438
PEB 20534
Semiconductor Group 10 Data Sheet 09.98
14.6.6 Reset Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .439
14.6.7 JTAG-Boundary Scan Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .440
14.6.8 SSC Serial Interface Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .441
15 Package Outlines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .442
PEB 20534
List of Figures Page
Semiconductor Group 11 Data Sheet 09.98
Figure 1 Logic Symbol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Figure 2 General System Integration (PCI Bus Interface) . . . . . . . . . . . . . . . . . 24
Figure 3 General System Integration (De-multiplexed Interface). . . . . . . . . . . . 25
Figure 4 HSSI Application - DCE Adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Figure 5 HSSI Application - DTE Adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Figure 6 General Data Application. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Figure 7 Pin Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Figure 8 DSCC4 Functional Block Diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Figure 9 Master Single READ Transaction followed by a Master Single WRITE
Transaction in De-multiplexed Configuration. . . . . . . . . . . . . . . . . . . . 54
Figure 10 Master Burst WRITE/READ Transaction in De-multiplexed
Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Figure 11 DMA Controller Block Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
Figure 12 DMA Data Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Figure 13 Transmit Descriptor List Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Figure 14 Transmit Descriptor Memory Example. . . . . . . . . . . . . . . . . . . . . . . . . 70
Figure 15 Receive Descriptor List Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Figure 16 Receive Descriptor Memory Example . . . . . . . . . . . . . . . . . . . . . . . . . 75
Figure 17 Data Transfer controlled via first and last descriptor addresses . . . . . 78
Figure 18 Example: ’Chain Jump’ Handling per ’Dummy Descriptor’. . . . . . . . . . 80
Figure 19 DSCC4 Logical Interrupt Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Figure 20 Central Transmit FIFO Section Thresholds . . . . . . . . . . . . . . . . . . . . . 90
Figure 21 Central Receive FIFO Threshold . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Figure 22 Little/Big Endian Byte Swapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Figure 23 MFP Configurations Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Figure 24 Multiplexed Bus Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Figure 25 De-multiplexed Bus Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Figure 26 Memory Cycle Time. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Figure 27 LRDY Controlled Bus Cycles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Figure 28 LRDY Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Figure 29 External Bus Arbitration (Releasing the Bus). . . . . . . . . . . . . . . . . . . 108
Figure 30 External Bus Arbitration (Regaining the Bus) . . . . . . . . . . . . . . . . . . 109
Figure 31 Connection of the Master and Slave Bus Arbitration Signals . . . . . . 110
Figure 32 Bus Arbitration Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Figure 33 Registers and Port Pins associated with the SSC . . . . . . . . . . . . . . . 116
Figure 34 Synchronous Serial Channel SSC Block Diagram. . . . . . . . . . . . . . . 117
Figure 35 Serial Clock Phase and Polarity Options . . . . . . . . . . . . . . . . . . . . . . 119
Figure 36 SSC Full Duplex Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Figure 37 SSC Half Duplex Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Figure 38 SSC Error Interrupt Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
Figure 39 SCC Transmit FIFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Figure 40 SCC Receive FIFO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
PEB 20534
Semiconductor Group 12 Data Sheet 09.98
Figure 41 Clock Supply Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
Figure 42 Clock Mode 0a/0b Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Figure 43 Clock Mode 1 Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
Figure 44 Clock Mode 2a/2b Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Figure 45 Clock Mode 3a/3b Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
Figure 46 Clock Mode 4 (High Speed) Configuration . . . . . . . . . . . . . . . . . . . . 145
Figure 47 Selecting one time-slot of programmable delay and width . . . . . . . . 148
Figure 48 Selecting one or more time-slots of 8-bit width . . . . . . . . . . . . . . . . . 150
Figure 49 Clock Mode 5 Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Figure 50 Clock Mode 6a/6b Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
Figure 51 Clock Mode 7a/7b Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
Figure 52 DPLL Algorithm for NRZ and NRZI
Coding with Phase Shift Enabled (CCR0:PSD = ‘0’) . . . . . . . . . . . . . 156
Figure 53 DPLL Algorithm for NRZ and NRZI
Encoding with Phase Shift Disabled (CCR0:PSD = ‘1’). . . . . . . . . . . 156
Figure 54 DPLL Algorithm for FM0, FM1 and Manchester Coding . . . . . . . . . . 157
Figure 55 Request-to-Send in Bus Operation . . . . . . . . . . . . . . . . . . . . . . . . . . 162
Figure 56 NRZ and NRZI Data Encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Figure 57 FM0 and FM1 Data Encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Figure 58 Manchester Data Encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
Figure 59 RTS/CTS Handshaking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
Figure 60 SCC Test Loop . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
Figure 61 SCC Receive Data Flow (HDLC Modes) part a) . . . . . . . . . . . . . . . . 173
Figure 62 SCC Receive Data Flow (HDLC Modes) part b) . . . . . . . . . . . . . . . . 174
Figure 63 SCC Transmit Data Flow (HDLC Modes) . . . . . . . . . . . . . . . . . . . . . 175
Figure 64 Processing of Received Frames in Auto Mode . . . . . . . . . . . . . . . . . 178
Figure 65 Timer Procedure/Poll Cycle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
Figure 66 Transmission/Reception of I-Frames and Flow Control. . . . . . . . . . . 181
Figure 67 Flow Control: Reception of S-Commands and Protocol Errors . . . . . 181
Figure 68 No Data to Send: Data Reception/Transmission. . . . . . . . . . . . . . . . 184
Figure 69 Data Transmission (without error), Data Transmission (with error). . 184
Figure 70 PPP Mapping/Unmapping Example . . . . . . . . . . . . . . . . . . . . . . . . . 189
Figure 71 Asynchronous Character Frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Figure 72 Out-of-Band DTE-DTE Bi-directional Flow Control . . . . . . . . . . . . . . 196
Figure 73 Out-of-Band DTE-DCE Bi-directional Flow Control . . . . . . . . . . . . . . 197
Figure 74 BISYNC Message Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
Figure 75 Overview of Data Stuctures in shared Memory before Transmission 214
Figure 76 Overview of Data Stuctures in shared Memory after Transmission. . 219
Figure 77 Transmit Descriptor List Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . 374
Figure 78 Receive Descriptor List Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . 378
Figure 79 ASYNC/BISYNC Receive Status Character Format . . . . . . . . . . . . . 386
Figure 80 DSCC4 Logical Interrupt Structure . . . . . . . . . . . . . . . . . . . . . . . . . . 388
Figure 81 Interrupt Vector Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
PEB 20534
Semiconductor Group 13 Data Sheet 09.98
Figure 82 Block Diagram of Test Access Port and Boundary Scan Unit . . . . . . 399
Figure 83 Power-up and Power-down scenarios . . . . . . . . . . . . . . . . . . . . . . . . 409
Figure 84 Power-Failure scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410
Figure 85 Input/Output Waveform for AC Tests. . . . . . . . . . . . . . . . . . . . . . . . . 414
Figure 86 PCI Output Timing Measurement Waveforms . . . . . . . . . . . . . . . . . . 414
Figure 87 PCI Input Timing Measurement Waveforms . . . . . . . . . . . . . . . . . . . 415
Figure 88 PCI Read Transaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
Figure 89 PCI Write Transaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417
Figure 90 PCI Clock Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
Figure 91 Master Single READ Transaction followed by a Master Single WRITE
Transaction in De-multiplexed Bus Configuration . . . . . . . . . . . . . . . 421
Figure 92 Master Burst WRITE/READ Access in De-multiplexed
Bus Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421
Figure 93 Synchronous LBI Read Cycle Timing Multiplexed Bus . . . . . . . . . . . 423
Figure 94 Synchronous LBI Write Cycle Timing Multiplexed Bus . . . . . . . . . . . 423
Figure 95 Synchronous LBI Read Cycle Timing De-multiplexed Bus . . . . . . . . 425
Figure 96 Synchronous LBI Write Cycle Timing De-multiplexed Bus . . . . . . . . 425
Figure 97 LRDY Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
Figure 98 LBI Arbitration Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
Figure 99 Clock Input Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429
Figure 100 Receive Cycle Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
Figure 101 Transmit Cycle Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
Figure 102 Strobe Timing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
Figure 103 Frame Synchronisation Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
Figure 104 High Speed Receive Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
Figure 105 High Speed Transmit Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
Figure 106 Reset Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
Figure 107 JTAG-Boundary Scan Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440
Figure 108 SSC Interface Timing (Master) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441
Figure 109 Package Outline. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442
PEB 20534
List of Tables Page
Semiconductor Group 14 Data Sheet 09.98
Table 1 PCI Bus Interface(DEMUX Interface) . . . . . . . . . . . . . . . . . . . . . . . . . .31
Table 2 Dedicated Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37
Table 3 JTAG Test Port for Boundary Scan according to IEEE 1149.1 . . . . . . .38
Table 4 Local Bus Interface (LBI) / General Purpose Port (GPP) / Synchronous
Serial Control (SSC) Interface Pins . . . . . . . . . . . . . . . . . . . . . . . . . . . .39
Table 5 Serial Communication Controller (SCC) Signals . . . . . . . . . . . . . . . . . .43
Table 6 PCI Configuration Space Registers. . . . . . . . . . . . . . . . . . . . . . . . . . . .52
Table 7 Non-PCI Signal Extension in the De-multiplexed Bus Interface Mode. .53
Table 8 DEMUX Mode Related Register and Bit-Fields . . . . . . . . . . . . . . . . . . .54
Table 9 Supported Commands in De-multiplexed Bus Mode. . . . . . . . . . . . . . .56
Table 10 DMA Controller Register Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59
Table 11 DMAC Commands. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .62
Table 12 Transmit Descriptor Bit Field Description. . . . . . . . . . . . . . . . . . . . . . . .67
Table 13 Meaning of ADD in Little/Big Endian Mode . . . . . . . . . . . . . . . . . . . . . .69
Table 14 Receive Descriptor Bit Field Description . . . . . . . . . . . . . . . . . . . . . . . .72
Table 15 Receive Data Buffer Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .74
Table 16 Central FIFO Control Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85
Table 17 MFP Configuration via GMODE Register, Bit Field ’PERCFG’:. . . . . . .97
Table 18 LBI Peripheral Transaction Options. . . . . . . . . . . . . . . . . . . . . . . . . . . .98
Table 19 Protocol Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .132
Table 20 Overview of Clock Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .137
Table 21 Clock Modes of the SCCs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .138
Table 22 BRR Register and Bit-Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .154
Table 23 Protocol Mode Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .168
Table 24 Address Comparison Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .169
Table 25 Error Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .182
Table 26 Status after Hardware Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .201
Table 27 Global Configuration of DSCC4 and Initialization of DMAC
(Interrupt Channel). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .203
Table 28 Initialization of DMAC (Data Channels) . . . . . . . . . . . . . . . . . . . . . . . .204
Table 29 Initialization of the SCC(s) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .204
Table 30 Initialization of the MFP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .206
Table 31 Activation of DMAC and SCC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .207
Table 32 Continuous Operation of Data Transfer. . . . . . . . . . . . . . . . . . . . . . . .209
Table 33 Stop Data Transmission. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .210
Table 34 Stop Data Reception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .210
Table 35 Exceptional handling in Case of Receive Data Overflow. . . . . . . . . . .212
Table 36 Exceptional handling in Case of Transmit Data Underrun. . . . . . . . . .212
Table 37 Register Initialization for HDLC Transparent Mode 0, Test Loop. . . . .215
Table 38 Register Range and Address Mapping . . . . . . . . . . . . . . . . . . . . . . . .221
Table 39 DSCC4: PCI Configuration Space Register Set . . . . . . . . . . . . . . . . .222
Table 40 PCI Base Address Ranges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .223
PEB 20534
Semiconductor Group 15 Data Sheet 09.98
Table 41 PCI Configuration Space: Status/Command Register . . . . . . . . . . . . .224
Table 42 Status and Command register bits. . . . . . . . . . . . . . . . . . . . . . . . . . . .225
Table 43 DSCC4 Global Register Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . .228
Table 44 GCMDR: Global Command Register. . . . . . . . . . . . . . . . . . . . . . . . . .232
Table 45 GSTAR: Global Status Register. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .237
Table 46 GMODE: Global Mode Register. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .241
Table 47 IQLENR1: Interrupt Queue Length Register 1. . . . . . . . . . . . . . . . . . .247
Table 48 IQLENR2: Interrupt Queue Length Register 2. . . . . . . . . . . . . . . . . . .249
Table 49 IQSCCiRXBAR:
Interrupt Queue SCCi Receiver Base Address Register (i=0...3) . . . .251
Table 50 IQSCCiTXBAR:
Interrupt Queue SCCi Receiver Base Address Register (i=0...3) . . . .252
Table 51 IQCFGBAR:
Interrupt Queue Configuration Base Address Register . . . . . . . . . . . .253
Table 52 IQPBAR:
Interrupt Queue Peripheral Base Address Register. . . . . . . . . . . . . . .254
Table 53 FIFOCR1: FIFO Control Register 1 . . . . . . . . . . . . . . . . . . . . . . . . . . .255
Table 54 FIFOCR2: FIFO Control Register 2 . . . . . . . . . . . . . . . . . . . . . . . . . . .256
Table 55 FIFOCR3: FIFO Control Register 3 . . . . . . . . . . . . . . . . . . . . . . . . . . .258
Table 56 FIFOCR4: FIFO Control Register 4 . . . . . . . . . . . . . . . . . . . . . . . . . . .260
Table 57 CHiCFG: Channel i Configuration Register (i=3...0) . . . . . . . . . . . . . .262
Table 58 CHiBRDA:
Channel i Base Receive Descriptor Address Register (i=3...0) . . . . . .264
Table 59 CHiBTDA:
Channel i Base Transmit Descriptor Address Register (i=3...0) . . . . .265
Table 60 CHiFRDA: Channe l i First (Curre nt)
Receive Descriptor Address Register (i=3...0). . . . . . . . . . . . . . . . . . .266
Table 61 CHiFTDA: Channel i First (Current)
Transmit Descriptor Address Register (i=3...0) . . . . . . . . . . . . . . . . . .267
Table 62 CHiLRDA:
Channel i Last Receive Descriptor Address Register (i=3...0). . . . . . .268
Table 63 CHiLTDA:
Channel i Last Transmit Descriptor Address Register (i=3...0) . . . . . .270
Table 64 SCC Register Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .273
Table 65 CMDR: Command Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .275
Table 66 STAR: Status Register. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .279
Table 67 CCR0: Channel Configuration Register 0 . . . . . . . . . . . . . . . . . . . . . .283
Table 68 CCR1: Channel Configuration Register 1 . . . . . . . . . . . . . . . . . . . . . .288
Table 69 CCR2: Channel Configuration Register 2 . . . . . . . . . . . . . . . . . . . . . .296
Table 70 ACCM: PPP ASYNC Control Character Map . . . . . . . . . . . . . . . . . . .306
Table 71 UDAC: User Defined PPP ASYNC Control Character Map. . . . . . . . .308
Table 72 TTSA: Transmit Time Slot Assignment Register . . . . . . . . . . . . . . . . .310
Table 73 RTSA: Receive Time Slot Assignment Register . . . . . . . . . . . . . . . . .312
PEB 20534
Semiconductor Group 16 Data Sheet 09.98
Table 74 PCMMTX: PCM Mask for Transmit Direction . . . . . . . . . . . . . . . . . . .314
Table 75 PCMMRX: PCM Mask for Receive Direction. . . . . . . . . . . . . . . . . . . .316
Table 76 BRR: Baud Rate Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .318
Table 77 TIMR: Timer Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .320
Table 78 XADR: Transmit Address Register . . . . . . . . . . . . . . . . . . . . . . . . . . .323
Table 79 RADR: Receive Address Register. . . . . . . . . . . . . . . . . . . . . . . . . . . .325
Table 80 RAMR: Receive Address Mask Register. . . . . . . . . . . . . . . . . . . . . . .327
Table 81 RLCR: Receive Length Check Register. . . . . . . . . . . . . . . . . . . . . . . .329
Table 82 XNXF: XON/XOFF In-Band Flow Control Character Register. . . . . . .331
Table 83 TCR: Termination Character Register . . . . . . . . . . . . . . . . . . . . . . . . .334
Table 84 TICR: Transmit Immediate Character Register . . . . . . . . . . . . . . . . . .336
Table 85 SYNCR: Synchronization Character Register . . . . . . . . . . . . . . . . . . .338
Table 86 IMR: Interrupt Mask Register. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .340
Table 87 ISR: Interrupt Status Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .342
Table 88 DSCC4 Peripheral Register Overview. . . . . . . . . . . . . . . . . . . . . . . . .350
Table 89 LCONF: LBI Configuration Register . . . . . . . . . . . . . . . . . . . . . . . . . .351
Table 90 SSCCON: SSC Control Register. . . . . . . . . . . . . . . . . . . . . . . . . . . . .355
Table 91 SSCBR: SSC Baud Rate Register. . . . . . . . . . . . . . . . . . . . . . . . . . . .360
Table 92 SSC Baud Rate Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .361
Table 93 SSCTB: SSC Transmit Buffer Register . . . . . . . . . . . . . . . . . . . . . . . .362
Table 94 SSCRB: SSC Receive Buffer Register . . . . . . . . . . . . . . . . . . . . . . . .363
Table 95 SSCCSE: SSC Chip Select Enable Register. . . . . . . . . . . . . . . . . . . .364
Table 96 SSCIM: SSC Interrupt Mask Register . . . . . . . . . . . . . . . . . . . . . . . . .366
Table 97 GPDIR: GPP Direction Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . .368
Table 98 GPDATA: GPP Data Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .370
Table 99 GPIM: GPP Interrupt Mask Register . . . . . . . . . . . . . . . . . . . . . . . . . .372
Table 100 Transmit Descriptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .375
Table 101 Receive Descriptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .379
Table 102 CFGIV: Configuration Interrupt Vectori . . . . . . . . . . . . . . . . . . . . . . . .390
Table 103 DMA Controller Interrupt Vectori . . . . . . . . . . . . . . . . . . . . . . . . . . . . .391
Table 104 SCC Interrupt Vectori. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .393
Table 105 SSC Interrupt Vectori. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .395
Table 106 SSC Interrupt Vectori. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .397
Table 107 GPP Interrupt Vectori. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .398
Table 108 Boundary Scan Sequence of the DSCC4 . . . . . . . . . . . . . . . . . . . . . .400
Table 109 Boundary Scan Test Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .404
Table 110 Absolute Maximum Ratings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .411
Table 111 Thermal Package Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . .411
Table 112 DC Characteristics
(Non-PCI Interface Pins and Power Supply Pins) . . . . . . . . . . . . . . . .412
Table 113 DC Characteristics PCI Interface Pins. . . . . . . . . . . . . . . . . . . . . . . . .413
Table 114 Capacitances (Non-PCI Interface Pins). . . . . . . . . . . . . . . . . . . . . . . .413
Table 115 PCI Input and Output Measurement Conditions . . . . . . . . . . . . . . . . .415
PEB 20534
Semiconductor Group 17 Data Sheet 09.98
Table 116 Number of Wait States Inserted by the DSCC4 as Initiator . . . . . . . . .418
Table 117 Number of Wait States Inserted by the DSCC4 as Slave . . . . . . . . . .418
Table 118 PCI Clock Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .419
Table 119 PCI Interface Signal Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . .420
Table 120 Additional De-multiplexed Interface Signal Characteristics . . . . . . . . .422
Table 121 LBI Timing (synchronous, multiplexed bus). . . . . . . . . . . . . . . . . . . . .424
Table 122 LBI Timing (synchronous, de-multiplexed bus) . . . . . . . . . . . . . . . . . .426
Table 123 LBI LRDY Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .427
Table 124 LBI Arbitration Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .428
Table 125 Clock Input Timing (non high speed modes) . . . . . . . . . . . . . . . . . . . .429
Table 126 Clock Input Timing (high speed mode) . . . . . . . . . . . . . . . . . . . . . . . .430
Table 127 Receive Cycle Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .432
Table 128 Transmit Cycle Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .434
Table 129 Strobe Timing (clock mode 1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .435
Table 130 Frame Synchronisation Timing (clock mode 5) . . . . . . . . . . . . . . . . . .436
Table 131 High Speed Receive Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .437
Table 132 High Speed Transmit Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .438
Table 133 Reset Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .439
Table 134 JTAG-Boundary Scan Timing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .440
Table 135 SSC Interface Timing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .441
PEB 20534
Overview
Semiconductor Group 18 Data Sheet 09.98
1 Overview
The DSCC4 is a DMA Supported Serial Communication Controller with four independent
serial channels1). The serial channels are derived from updated protocol logic of the
ESCC device family providing a large set of protocol support and variety in serial
interface configuration. This allows easy integration to different environments and
applications.
A 33-MHz/32-bit PCI bus Master/Slave interface with integrated high performance DMA
controllers provides data transfer from or to host memory with low bus utilization and
easy software handshaking.
An additional de-multiplexed bus interface mode is provided for integration in non-PCI
bus environments with little glue-logic depending on the bus type.
The DMA Controller operates on linked lists which are optimized for data communication
applications. Different control mechanisms allow easy software development well
adapted to the needs of special applications.
Large onch ip FIFO s i n co mbination with e nha nce d thre sho ld c ontro l mec han ism s al low
decoupli ng of traf fic requireme nts on host bu s and seria l interfaces with little e xception
probabilities such as data underuns or overflows.
In a PCI bus application an integrated Local Bus Interface (LBI) provides bridging
functionality to non PCI peripherals such as framers or line interface units (LIUs).
A Synchronous Serial Control (SSC) interface as well as a General Purpose Port (GPP)
allows covering application specific requirements without additional controllers.
Each of the four Serial Communication Controllers (SCC) contains an independent Baud
Rate Generator, DPLL, programmable protocol processing (HDLC, BISYNC, ASYNC
and PPP). Data rates of up to 2 Mbit/s (DPLL assisted modes, ASYNC, BISYNC),
10 Mb it/s (HDLC, PPP) and 52 Mbit/s (H-52 version) are suppo rted. The channels ca n
also handle a large set of layer-2 protocol functions reducing bus and host CPU load.
Four channel specific timers are provided to support protocol functions.
The DSCC4 devices can be used in LAN-WAN inter-networking applications such as
Routers, Switches and Trunk cards and support the common V.35, ISDN BRI (S/T) or
Asynchronous Dial-up interfaces. Its new features provide powerful hardware and
software interfaces to develop high performance systems.
1) The serial ch annels are also calle d ’ports’ or ’cores’ depending on the cont ext.
P-FQFP-208-7
DMA Supported Serial Communication Controller with
4 Cha n nels
DSCC4
PEB 20534
Semiconductor Group 19 Data Sheet 09.98
Version 2.0 CMOS
Type Package
PEB 20534 H-10 P-FQFP-208-7
PEB 20534 H-52 P-FQFP-208-7
1.1 Features
• Serial communication controllers (SCCs)
- Four independent channels
- Full duplex data rates on each channel of up to
10 Mbit/s sync - 2 Mbit/s with DPLL, 2 Mbit/s
async
- Full duplex data rate of up to 52 Mbit/s on any two
channels in high speed mode (HDLC: Address
Mode 0 and extended transparent protocol
mode); up to 45 Mbit/s on any two channels in
high speed mode (HDLC: PPP modes). The
aggregate bandwith for all channels is limited to
108 Mbit/s per direction.
- 17 DWORDs deep receive FIFO per SCC
(+ 128 DWORDs central receive FIFO).
- 8 DWORDs deep transmit FIFO per SCC
(+ 128 DWORDs central transmit FIFO).
• Serial Interface
- On chip clock generation or external clock sources
- On chip DPLLs for clock recovery
- Baud rate generator
- Clock gating signals
- Clock gapping capability
- Programmable time-slot capability for connection to TDM interfaces (e.g. T1, E1)
- NRZ, NRZI, FM and Manchester data encoding
- Optional data flow control using modem control lines (RTS, CTS, CD)
- Support of bus configuration by collision detection and resolution
PEB 20534
Overview
Semiconductor Group 20 Data Sheet 09.98
• Bit Processor Functions
- HDLC/SDLC Mode
- Automatic flag detection and transmission
- Shared opening and closing flag
- Generation of interframe-time fill ’1’s or flags
- Detection of receive line status
- Zero bit insertion and deletion
- CRC generation and checking (CRC-CCITT or CRC-32)
- Transparent CRC option per channel and/or per frame
- Programmable Preamble (8 bit) with selectable repetition rate
- Error detection (abort, long frame, CRC error, short frames)
- ASYNC Mode
- Selectable character length (5 to 8 bits)
- Even, odd, forced or no parity generation/checking
- 1 or 2 stop bits
- Break detecti on/g enera tion
- Flow control by XON/XOFF
- Immediate character insertion
- Termination character detection for end of block identification
- Time out detection
- Error detection (parity error, framing error)
- BISYNC Mode
- Programmable 6/8 bit SYN pattern (MONOSYNC)
- Programmable 12/16 bit SYN pattern (BISYNC)
- Selectable character length (5 to 8 bits)
- Even, odd, forced or no parity generation/checking
- Generation of interframe-time fill ’1’s or SYN characters
- CRC generation (CRC-16 or CRC-CCITT)
- Transparent CRC option per channel and/or per frame
- Programmable Preamble (8 bit) with selectable repetition rate
- Termination character detection for end of block identification
- Error detection (parity error, framing error)
• Protocol Support (provided in HDLC/SDLC Mode)
- Address mode 0
- No address recognition
- Address mode 1
- 8-bit (high byte) address recognition
- Non-auto mode
- 8-bit (low byte) or 16-bit (high and low byte) address recognition
- Auto mode
- 8-bit or 16-bit address generatio n/rec ogn itio n
- Support of LAPB/LAPD
- Automatic processing of control byte(s)
PEB 20534
Overview
Semiconductor Group 21 Data Sheet 09.98
- Automatic handling of S- and I-frames
- Automatic processing of control byte(s)
- Modulo-8 or modulo-128 operation
- Programmable time-out and retry conditions
- Normal Response Mode operation for slave
- Asynchronous PPP mode
- Character oriented transmission of HDLC frame (flag, data, CRC, flag)
- Start/stop bit framing of single character
- No zero bit insertion/deletion
- Programmable character map of 32 hard-wired characters (00H-1FH)
- Four programmable characters for additional mapping
- Insertion/deletion of control characters corresponding to character maps
- Octet Synchronous PPP mode
- Octet oriented transmission of HDLC frame (flag, data, CRC, flag)
- Programmable character map of 32 hard-wired characters (00H-1FH)
- Four programmable characters for additional mapping
- Insertion/deletion of control characters corresponding to character maps
- Zero bit insertion/deletion
- Bit Synchronous PPP mode
- Bit oriented transmission of HDLC frame (flag, data, CRC, flag)
- Zero bit insertion/deletion
- 15 consecutive ’1’ bits abort sequence
- Extended transparent mode
- Fully bit transparent (no framing, no bit manipulation)
• Protocol and Mode Independent
- Data bit inversion
- Data over- and underflow detection
-Timer
• Microprocessor Interface
- 33 MHz/32-bit PCI bus interface option.
- 33 MHz/32-bit De-multiplexed bus interface option.
- 8-channel DMA controller with buffer chaining capability.
Master 15-word burst read and write capability (PCI Mode).
Master 4-word burst read and write capability (DEMUX Mode).
Slave single-w ord re ad and write capabili ty.
- Circular interrupt queues with variable size.
- Maskable interrupts for each channel
• 8-/16-bi t optional Loc al Bus Interface (LBI) for driving non-PC I peripherals in an PCI
environment.
• Synchronous Serial Control interface (SSC) for controlling peripherals.
• 16-bit General Purpose Port (GPP).
PEB 20534
Overview
Semiconductor Group 22 Data Sheet 09.98
• General
- On chip Rx and Tx data buffer; the buffer size is 128 32-bit words each.
- Programmable buffer size in transmit direction per channel; buffer allocation in
receive direction on request.
- Programmable watermark for each receive channel to control transfer of receive
data to host memory.
- Two programmable watermarks for each transmit channel. One controlling data
loading from host memory and one controlling transfer of transmit data to the
corresponding Serial Communication Controller (SCC).
- Internal test loop capability.
- JTAG boundary scan test according to IEEE 1149.1
- Advanced low-power CMOS technology
- TTL-compatible inputs/outputs
- 3.3-V & 5-V power supply
- 3.3-V interfaces (TTL levels; 5 V tolerant in 5 V environment)
- P-FQFP-208-7 package
1.2 Differences Between the DSCC4 and the ESCC Family
This chapter is useful for all being familiar with the SIEMENS ESCC family.
1.2.1 Enhancements to the ESCC Serial Core
The DSC C4 SCC cores contain the core l ogic of the ESCC2 V3.2A as t he heart o f the
device. Some enhancements are incorporated in the SCCs. These are:
• Asynchronous PPP protocol support as in Internet RFC-1662
• Octet and Bit Synchronous PPP protocol support as in Internet RFC-1662
• 16-Kbyte packet length byte counter
• Enhanced address filtering (16-bit maskable)
• Enhanced time slot assigner
• Support of high data rates (45 Mbit/s for DS3 or 52 Mbit/s for OC1). Protocol support
limited to HDLC Sub-modes without address recognition.
1.2.2 Simplifications to the ESCC Serial Core
The following features of the ESCC core have been removed:
• SDLC Loop mode
• Extended transparent mode 0
(this mode provided octet buffered data reception without usage of FIFOs; the DSCC4
supports octet buffered reception via appropriate threshold configurations for the SCC
receive FIFOs)
PEB 20534
Overview
Semiconductor Group 23 Data Sheet 09.98
1.3 Logic Symbol
Figure 1 Logic Symbol
AD(31:0)
C/BE(3:0)
PAR
FRAME
IRDY
TRDY
STOP
IDSEL
DEVSEL
PERR
SERR
REQ
GNT
CLK
RST
INTA
DEMUX
W/R
V
SS
V
DD3
V
DD5
TEST
TCK
TMS
TDI
TDO
TRST
A(31:0)
(de-multiplexed address bus) LD(15:0), LA(15:0)
or
LAD(15:0), GP(15:0)
or
LD(15:0), GP(15:8), LA(7:0)
or
LAD(15:0), GP(15:8), SSC interf.
or
PCI
BUS
LBI
Control Signals
JTAG Test
Interface
Control and Address Bus Extension
for De-multiplexed Bus Interface
Depending on Configuration:
Demuxed Local Bus Interface (LBI)
or
Muxed Local Bus Interface (LBI) + 16 Bit GPP
or
Demuxed LBI (8 Bit Address) + 8 Bit GPP
or
Muxed LBI + 8 Bit GPP + Synchronous Serial
Controller (SSC) interface
TxD0
RxD0
RTS0
CTS0
CD0/RCG0
TxCLK0
RxCLK0
Serial
Channel 0
(SCC0)
Serial
Channel 1
(SCC1)
Serial
Channel 2
(SCC2)
Serial
Channel 3
(SCC3)
DSCC4
PEB 20534
XTAL1
XTAL2
P-FQFP-208-7
PEB 20534
Overview
Semiconductor Group 24 Data Sheet 09.98
1.4 Typical Applications
The DSCC4 is designed to handle up to 4 serial data ports in various configurations,
depending on the application. It transfers the data between the serial ports and a shared
memory via its 32 bit/33 MHz PCI Bus Interface which can optionally be configured as a
generic 32 bit de-multiplexed bus interface in the case that no PCI bus is applicable.
Figure 2 provi des a general overview upon system integration of the DSCC4 in a PCI
bus environment:
Figure 2 General System Integration (PCI Bus Interface)
PCI
Bridge
CPU RAM
Bank
PEB 20534
DSCC4
.
.
.
...
. . .
. . .
. . .
. . .
PCI
Arbiter
REQ
GNT
PCI Bus
Local Peripheral Bus
Transceiver,
Framer
Host Bus
Arbiter
Host Bus
PEB 20534
Overview
Semiconductor Group 25 Data Sheet 09.98
Connection of DSCC4 to PCI Bus according to PCI Specification Rev. 2.1 is free of any
glue-logic.
Figure 3 provide s an overview upon system integration in a non PCI bus environment
by the example of a Motorola 68360 CPU bus:
Figure 3 General System Integration (De-multiplexed Interface)
The glue-logic depends on the host bus which the DSCC4 should be connected to. The
exampl e in Figure 3 shows the glue-lo gic fo r conn ection to an Mo torola 68360 like de-
multiplexed 32 bit bus.
See APPENDIX_xx for more detailed example.
TRDY
IRDY
FRAME
STOP
DEVSEL
Glue Logic
AS
DS
DSACK
A[31..2]
AD[31..0]
A[31..2]
D[31..0]
W/R
R/W
V
DD3
A1
A0
DEMUX
PAR
V
SS
RAM
Bank
IDESL
Chip
Select
Decoder MOTOROLA
68360
Bus
Arbiter
REQ
GNT
BREQ
BGNT
BACK
Transceiver,
Framer
PEB 20534
DSCC4
PEB 20534
Overview
Semiconductor Group 26 Data Sheet 09.98
1.4.1 Application Examples
1.4.1.1 HSSI Application
Figure 4 HSSI Application - DCE Adapter
Clock
Unit
TxCLK0
TxD1
TxCLKO1
RxD1
RxCLK1
SCC1
Loop Back
MUX
Loop
Back
Control
GPP
GPi (input)
TTL
ECL
Line1
Loop Ctrl.
PEB 20534
DSCC4
Loop Back
MUX
TTL
ECL
Line0
SCC0
TxD0
TxCLKO0
RxD0
RxCLK0
TxCLK1
PEB 20534
Overview
Semiconductor Group 27 Data Sheet 09.98
1.4.1. 2 HSSI Application
Figure 5 HSSI Application - DTE Adapter
TxD
TxCLK
RxD
RxCLK
TTL
ECL
Line
Loop Ctrl.
SCCi
GPP
DSCC4
PEB 20534
Overview
Semiconductor Group 28 Data Sheet 09.98
1.4.1.3 General Data Application
Figure 6 General Data Application
Clock
Unit
TxCLK2,
TxCLK3
GPP
Line1
PEB 20534
DSCC4
Line Transceiver
(RS232, RS485, ...)
ASYNC
Line0
SCC0
TxD0
RTS0
RxD0
CTS0
SCC1
TxD1
RTS1
RxD1
CTS1
XTAL1
XTAL2
Control (Loop, ...)
SCC2 SCC3
Line3
Line Transceiver
SYNC/HDLC
Line2
TxD2
TxCLKO2
RxD2
RxCLK2
TxD3
TxCLKO3
RxD3
RxCLK3
Control (Loop, ...)
PEB 20534
Pin Descriptions
Semiconductor Group 29 Data Sheet 09.98
2 P in Descriptions
2.1 Pin Diag ram
(top view)
Figure 7 Pin Configuration
ITP10573
208 7 1013161922252831343740
199
196
193
190
187
184
181
178
175
172
169
166
157 111114117120123126129132135138141144147150
53
62
65
68
71
74
77
80
83
86
89
92
95
104
DSCC4
NC20
59
56
4643
V
SS
NC1
41
202
205
V
SS
160
163
156 153 108 105
101
98
49 52
V
DD3
SS
V
V
DD3
NC12
LHOLD
SS
V
NC19
DD3
V
NC18
NC17
LBHE
LWR
LRD
RESERVED1
LCSO
LBREQ
LHLDA
DD3
V
LD0/A16
LD1/A17
LD2/A18
LD3/A19
LD4/A20
LD5/A21
LD6/A22
LD7/A23
LD8/A24
SS
V
V
DD3
SS
V
DD5
V
SS
V
DD3
V
LD9/A25
LD10/A26
LD11/A27
LD12/A28
LD13/A29
LD14/A30
LD15/A31
SS
V
DD3
V
LA0/A0/GP2/MCS0
LA1/A1/GP1/MCS1
LA2/A2/GP2/MCS2
LA3/A3/GP3/MCS3
LA4/A4/GP4/MRST
LA5/A5/GP5/MTSR
LA6/A6/GP6/MSCLK
LA7/A7/GP7/Mx
V
SS
DD3
V
LA8/A8/GP8
LA9/A9/GP9
NC15
NC14
NC13
NC2
NC3/TRST
AD25
AD24
C/BE3
AD26
DD3
V
IDSEL
AD23
AD22
AD21
AD20
V
SS
AD19
AD18
AD17
AD16
DD3
V
SS
V
C/BE2
FRAME
IRDY
DD5
V
SS
V
DD3
V
TRDY
SS
V
DEVSEL
STOP
PERR
SERR
PAR
C/BE1
AD15
AD14
AD13
DD3
V
SS
V
AD12
DD3
V
AD11
SS
V
AD10
AD9
AD8
AD7
NC4
NC5
NC6
LA10/A10/GP10
LA11/A11/GP11
LA12/A12/GP12
LA13/A13/GP13
LA14/A14/GP14
LA15/A15/GP15
NC11
SS
V
DD3
V
RTS1
CD1/FSC1/RCG1
CTS1/CxD1/TCG1
TxD1
RxD1
TxCLK1
RxCLK1
NC10
TEST1
V
SS
DD3
V
RTS0
CD0/FSC0/RCG0
CTS0/CxD0/TCG0
TxD0
RxD0
TxCLK0
RxCLK0
TDO
TMS
TDI
TCK
SS
V
DD3
V
INTA
SS
V
AD0
DD3
V
AD1
AD2
AD3
AD4
AD5
AD6
C/BE0
SS
V
DD3
V
NC9
NC8
NC7
LRDY
LINTI1
RESERVED2
LCLK
LALE
DD3
VV
SS
RxCLK2
TxCLK2
RxD2
TxD2
CTS2/CxD2/TCG2
CD2/FSC2/RCG2
RTS2
DD3
V
SS
V
XTAL1
XTAL2
RxCLK3
TxCLK3
RxD3
TxD3
CTS3/CxD3/TCG3
CD3/FSC3/RCG3
RTS3
DD3
V
SS
V
NC21
DEMUX
W/R
NC22
RST
DD3
V
CLK
SS
V
GNT
REQ
AD31
V
DD3
AD30
SS
V
AD29
AD28
DD3
V
SS
V
AD27
NC23
NC24
NC25
PEB 20534
PEB 20534
Pin Descriptions
Semiconductor Group 30 Data Sheet 09.98
2.2 Pin Definitions and Functions
Signal Type Definitions:
The following signal type definitions are mainly taken from the PCI Specification
Revision 2.1:
IInput is a standard input-only signal.
OTotem Pole Output is a standard active driver.
t/s, I/O Tri-State or I/O is a bi-directional, tri-state input/output pin.
s/t/s Sustained Tri-State is an active low tri-state signal owned and driven
by one and only one agent at a time. (For further information refer to
the PCI Specification Revision 2.1)
o/d Open Drain allows multiple devices to share as a wire-OR. A pull-up
is required to sustain the inactive state until another agent drives it,
and must be provided by the central resource.
Signal Name Conventions:
NC
Not Connected Pin
These pins are not bonded with the silicon. Although any potential at
these pins will not impact the device it is recommended to leave them
unco nnected. NC pins mi ght be used for a dditiona l function ality i n later
versions of the device. Leaving them unconnected will guarentee
hardware compatibility to later device versions.
Reserved Reserved pins are for vendor specific use only and should be connected
as recommended to guarantee normal operation.
PEB 20534
Pin Descriptions
Semiconductor Group 31 Data Sheet 09.98
Table 1 PCI Bus Interface(DEMUX Interface)
Pin No. Symbol Input (I)
Output (O) Function
197, 199,
201, 202,
205, 4...6,
11...13, 15,
17...20,
37...39, 42,
44, 46...49,
59...64, 66
AD(31:0) t/s Address/Data Bus
A bus transaction consists of an address
phase followed by one or more data phases.
When DSCC4 is Master, AD(31:0) are outputs
in the address phase of a transaction. During
the data phases, AD(31:0) remain outputs for
write transactions, and become inputs for read
transactions.
When DSCC4 is Slave, AD(31:0) are inputs in
the address phase of a transaction. During the
data phases, AD(31:0) remain inputs for write
transactions, and become outputs for read
transactions.
AD(31:0) are updated and sampled on the
rising edge of CLK.
7, 23, 36, 58 C/BE(3:0) t/s Command/Byte Enable
During the address phase of a transaction,
C/BE(3:0) define the bus command. During
the data phase, C/BE(3:0) are used as Byte
Enables. The Byte Enables are valid for the
entire data phase and determine which byte
lanes carry meaningful data. C/BE0 applies to
byte 0 (lsb) and C/BE3 applies to byte 3 (msb).
When DSCC4 is Master, C/BE(3:0) are
outputs. When DSCC4 is Slave, C/BE(3:0) are
inputs.
C/BE(3:0) are updated and sampled on the
rising edge of CLK.
Note: The bus command cycle is not
generated (initiator) or evaluated
(target) in DEMUX mode.
PEB 20534
Pin Descriptions
Semiconductor Group 32 Data Sheet 09.98
35 PAR t/s Parity
PAR is even parity across AD(31:0) and
C/BE(3:0). PAR is stable and valid one clock
after the address phase. PAR has the same
timing as AD(31:0) but delayed by one clock.
When DSCC4 is Master, PAR is output during
address phase and write data phases. When
DSCC4 is Slave, PAR is output during read
data phases.
Parity errors detected by the DSCC4 are
indicated on PERR output.
PAR is updated and sampled on the rising
edge of CLK.
Note: PAR is not generated in DEMUX mode
and remains ’0’.A Pull-Down resistor to
V
SS
is recommended.
24 FRAME s/t/s Frame
FRAME indicates the beginning and end of an
access. FRAME is asserted to indicate a bus
transaction is beginning. While FRAME is
asserted, data transfers continue. When
FRAME is deasserted, the transaction is in the
final phase.
When DSCC4 is Ma ster, FRAME is an output.
When DSCC4 is Slave, FRAME is an input.
FRAME is updated and sampled on the rising
edge of CLK.
Tab le 1 P CI Bu s Int er f ac e (DEMUX Interface) (cont’d)
Pin No. Symbol Input (I)
Output (O) Function
PEB 20534
Pin Descriptions
Semiconductor Group 33 Data Sheet 09.98
25 IRDY s/t/s Initiator Ready
IRDY indicates the bus master’s ability to
complete the current data phase of the
transaction. It is used in conjunction with
TRDY. A data phase is completed on any
clock where both IRDY and TRDY are
sampled asserted. During a write, IRDY
indicates that valid data is present on
AD(31:0). During a read, it indicates the
master is prepared to accept data. Wait cycles
are inserted until both IRDY and TRDY are
asse rted together.
When DSCC4 is Master, IRDY is an output.
When DSCC4 is Slave, IRDY is an input.
IRDY is updated and sampled on the rising
edge of CLK.
29 TRDY s/t/s Target Ready
TRDY indicates a sla ve’s abilit y to complete
the current data phase of the transaction.
During a read, TRDY indicates that valid data
is present on AD(31:0). During a wri te, it
indicates the target is prepared to accept data.
When DSCC4 is Master, TRDY is an input.
When DSCC4 is Slave, TRDY is an output.
TRDY is updated and sampled on the rising
edge of CLK.
32 STOP s/t/s Stop Signal
STOP is used by a slave to request the current
master to stop the current bus transaction.
When DSCC4 is Master, STOP is an input.
When DSCC4 is Slave, STOP is an output.
STOP is updated and sampled on the rising
edge of CLK.
Table 1 PCI Bus Interface(DEMUX Interface) (cont’d)
Pin No. Symbol Input (I)
Output (O) Function
PEB 20534
Pin Descriptions
Semiconductor Group 34 Data Sheet 09.98
9 IDSEL I Initialization Device Select
When DSCC4 is slave in a transaction and if
IDSEL is active in the address phase and
C/BE(3:0) indicates an config read or write
command, the DSCC4 assumes a read or
write to a configuration space register. In
response, the DSCC4 asserts DEVSEL during
the subsequent CLK cycle.
IDSEL is sampled on the rising edge of CLK.
Note: In DEMUX mode IDSEL is a chipselect
for the configuration space registers.
31 DEVSEL s/t/s Device Select
When activated by a slave, it indicates to the
current bus master that the slave has decoded
its address as the target of the current
transaction. If no bus slave activates DEVSEL
within six bus CLK cycles, the master should
abort the transaction.
When DSCC4 is master, DEVSEL is input. If
DEVSEL is not activated within six clock
cycles after an address is output on AD(31:0),
the DSCC4 aborts the transaction and
generates an INTA.
When DSCC4 is slave, DEVSEL is output.
Note: DEVSEL is also valid in DEMUX mode.
33 PERR s/t/s Parity Error
When activated, indicates a parity error over
the AD(31:0) and C/BE(3:0) signals
(compared to the PAR input). It has a delay of
two CLK cycles with respect to AD and
C/BE(3:0) (i.e., it is valid for the cycle
immediately following the corresponding PAR
cycle).
PERR is asserted relative to the rising edge of
CLK.
Tab le 1 P CI Bu s Int er f ac e (DEMUX Interface) (cont’d)
Pin No. Symbol Input (I)
Output (O) Function
PEB 20534
Pin Descriptions
Semiconductor Group 35 Data Sheet 09.98
34 SERR o/d System Error
The DSCC4 asserts this signal to indicate a
fatal system error.
SERR is activated on the rising edge of CLK.
196 REQ t/s Request
Used by the DSCC4 to request control of the
PCI.
REQ is activated on the rising edge of CLK.
195 GNT t/s Grant
This signal is asserted by the arbiter to grant
control of the PCI to the DSCC4 in response to
a bus request via REQ. After GNT is asserted,
the DSCC4 will begin a bus transaction only
after the current bus Master has deasserted
the FRAME sig nal.
GNT is sampled on the rising edge of CLK.
193 CLK I Clock
Provides timing for all PCI transactions. Most
PCI signals are sampled or output driven
relative to the rising edge of CLK. The
maximum CLK frequency is 33 MHz.
191 RST IReset
An active RST signal brings all PCI registers,
sequencers and signals into a consistent
state.
RST also resets all other blocks beside PCI to
thei r init ial sta te.
During RESET
- all PCI output signals are driven to their
benign state
- the TxDn (n=0,..,3) output signals are in high
impe dan ce stat e
- the RTSn (n=0,..,3) output signals are
inactive
- all bi-directi ona l signa ls are input s.
Table 1 PCI Bus Interface(DEMUX Interface) (cont’d)
Pin No. Symbol Input (I)
Output (O) Function
PEB 20534
Pin Descriptions
Semiconductor Group 36 Data Sheet 09.98
68 INTA o/d Interrupt Request
When an interrupt status is active and
unmasked, the DSCC4 activates this open-
drain output. Examples of interrupt sources
are transmission/reception error, completion
of transmit or receive packets etc. The DSCC4
deactivates INTA when the interrupt status is
acknowledged via an appropriate action (e.g.,
specific register write) and no other unmasked
interrupt statuses are active.
INTA is activated/ deactivated asynchronous
to the CLK.
Note: PCI control signals (typ e s/t/s) always require pull-up resistors. For the system
dependent pull-up recommendation please refer to PCI Specification Revision
2.1 chapter 4.3.3
Note: The function of PCI Bus Interface signals is the same in DEMUX mode where
the DSCC4 is operated in a non-PCI bus environment. All recommendations and
signal characteristics also apply in DEMUX mode.
Signal PAR is not generated and remains ’0’ in DEMUX mode.
In DEMUX mode signal IDSEL is a chipselect for the PCI Configuration Space
(no bus commands exists in DEMUX mode to address Configuration Space
access in addition to IDSEL)
Additional DEMUX Interface Control Signals
188 DEMUX I PCI/De-multiplexed Mode Select
DEMUX = ’0’ selects normal PCI operation.
DEMUX = ’1’ selects operation in de-
multiplexed (DEMUX) mode.
(Pull-Up/Down resistors or direct connection
to VSS/VDD3 possible)
189 W/R I/O Write/Read Control
This signal distinguishes between write and
read operations in the De-multiplexed mode.
It is tristate when the DSCC4 is in PCI mode.
A Pull-Up resistor to VDD3 is recommended
for PCI operation mode (DEMUX = VSS).
Tab le 1 P CI Bu s Int er f ac e (DEMUX Interface ) (cont’d)
Pin No. Symbol Input (I)
Output (O) Function
PEB 20534
Pin Descriptions
Semiconductor Group 37 Data Sheet 09.98
Table 2 Dedicated Signals
Pin No. Symbol Input (I)
Output (O) Function
1, 2, 50, 51,
52, 53, 54,
55, 85, 95,
104, 105,
106, 107,
154, 155,
156, 157,
187, 190,
206, 207,
208
NC1,
NC2,
NC4...15
NC17...25
-No-connect Pin 1, 2, 4…15
No-connect Pin 17...25
These pins must be left unconnected.
194, 200,
204, 10, 16,
22 ,27, 30,
41, 45, 57,
67, 70, 83,
94, 103,
111, 121,
130, 132,
134, 145,
159, 166,
175, 186
VSS Ground (0 V)
All pins must be connected to the same
voltage potential.
192, 198,
203, 8, 14,
21, 28, 40,
43, 56, 65,
69, 82, 93,
102, 110,
120, 129,
133, 144,
158, 165,
174, 185
VDD3 Supply Voltage 3.3 V ±0.3 V
All pins must be connected to the same
voltage potential.
26, 131 VDD5 Supply Voltage 5V±0.25 V
Both pins must be connected to the same
voltage potential.
PEB 20534
Pin Descriptions
Semiconductor Group 38 Data Sheet 09.98
84 TEST I Test Input
When connected to VDD3 the DSCC4 works in
a vendor specific test mode.
It is recommended to connect this pin to
VSS.
150 Reserved
1- Reserved Pin 1
A Pull-Up resistor to VDD3 is required.
162 Reserved
2- Reserved Pin 2
A Pull-Up resistor to VDD3 is required.
Table 3 JTAG Test Port for Boundary Scan according to IEEE 1149.1
Pin No. Symbol Input (I)
Output (O) Function
71 TCK I JTAG Test Clock
A Pull-Up resistor to VDD3 is recommended
if boundary scan unit is not used.
73 TMS I JTAG Test Mode Select
A Pull-Up resistor to VDD3 is recommended
if boundary scan unit is not used.
72 TDI I JTAG Test Data Input
A Pull-Up resistor to VDD3 is recommended
if boundary scan unit is not used.
74 TDO O JTAG Test Data Output
3TRST
IJTAG Reset Pin
For normal device operation this pin should be
connected to VSS or logical ’0’ to force
deactivation of the boundary scan unit.
For boundary scan mode TRST should be
connected to high level.
Note: An internal pull-up transistor forces
boundary scan operation mode if this
pin remains unconnected.
Table 2 Dedicated Signals (cont’d)
Pin No. Symbol Input (I)
Output (O) Function
PEB 20534
Pin Descriptions
Semiconductor Group 39 Data Sheet 09.98
Table 4 Local Bus Interface (LBI) / General Purpose Port (GPP) /
Synchronous Serial Control (SSC) Interface Pins
Pin No. Symbol Input (I)
Output (O) Function
96...101,
108, 109,
112...119 LA(15:0)
or
GP(15:0)
I/O
I/O
PCI Mode (DEMUX connected to VSS):
LBI Address Bus
(GMODE.PERCFG=0002)
These pins provide the 16 bit Address bus for
the Local Bus Interface.
General Purpose Port
(GMODE.PERCFG=1002)
A general purpose 16-bit bi-directional parallel
port is provided on pins GP(15...0). Every pin
is individually programmable via register
GPDIR to operate as an output or an input.
If defined as output, the state of the pin is
directly controlled via the register GPDATA.
If defined as input, its current status can be
read via GPDATA. All changes may be
indicated via an interrupt vector. The
interrupts for each single pin can be masked
via mask register GPIM.
PEB 20534
Pin Descriptions
Semiconductor Group 40 Data Sheet 09.98
96...101,
108, 109,
112
113
114
115
116
117
118
119
96...101,
108, 109,
112...119
96...101,
108, 109,
112...119
or
GP (15:8) I/O 8 bit General Purpose Port
(GMODE.PERCFG=0112)
MX
MSCLK
MTSR
MRST
MCS3
MCS2
MCS1
MCS0
or
GP (15:8)
LA(7:0)
or
A(15:0)
I/O
I/O
I/O
SSC Interface (GMODE.PERCFG=0112):
Dont’ Care Pin
This pin is reserved in SSC mode.
A Pull-Up resistor to VDD3 is
recommended.
SSC Shift Clock Input/Output
SSC Master Transmit / Slave Receive
SSC Master Receive / Slave Transmit
SSC Chip select 3
SSC Chip select 2
SSC Chip select 1
SSC Chip select 0
8 bit General Purpose Port
(GMODE.PERCFG=0102)
LBI Address Bus
These pins provide the 8 bit Address bus for
the Local Bus Interface.
De-multiplexed Mode (DEMUX connected
to VDD3) :
DEMUX Address Bus (15:0)
These pins provide the 16 least significant
address lines for the de-multiplexed interface.
Signals A(1:0) are unused due to 32 bit
address alignment. A Pull-Up resistor to
VDD3 is recommended.
Table 4 Local Bus Interface (LBI) / General Purpose Port (GPP) /
Synchronous Serial Control (SSC) Interface Pins (cont’d)
Pin No. Symbol Input (I)
Output (O) Function
PEB 20534
Pin Descriptions
Semiconductor Group 41 Data Sheet 09.98
122...128,
135...143 LD (15:0)
or
A(31:16)
I/O
I/O
PCI Mode (DEMUX connected to VSS):
LBI Data
(GMODE.PERCFG=0XX2)
These pins provide the 16 bit Data bus for the
Local Bus Interface.
LBI Address/Data
(GMODE.PERCFG=1002)
These pins provide the 16 bit multiplexed
Address/Data bus for the Local Bus Interface.
Demultiplexed Mode (DEMUX connected
to VDD3) :
DEMUX Address Bus (31:16)
These pins provide the 16 most significant
address lines for the de-multiplexed interface.
146 LHOLD ILBI Hold Request
LHOLD = ’1’ is used for normal bus drive
mode.
LHOLD = ’0’ requests LBI to enter hold mode.
A Pull-Up resistor to VDD3 is recommended
if LBI is not used.
148 LBREQ OLBI Bus Request
The DSCC4 asserts LBREQ = ’0’ to request
the local bus and deasserts the signal
LBREQ = ’1’ after regaining bus.
Table 4 Local Bus Interface (LBI) / General Purpose Port (GPP) /
Synchronous Serial Control (SSC) Interface Pins (cont’d)
Pin No. Symbol Input (I)
Output (O) Function
PEB 20534
Pin Descriptions
Semiconductor Group 42 Data Sheet 09.98
147 LHLDA I/O LBI Hold Status
(The function depends on whether slave or
master mode is selected in register LCONF)
As an output, LHLDA = ’0’ confirms that the
LBI bus is in HOLD mode (master mode).
As an input, LHLDA = ’1’ means that DSCC4
must remain in hold mode (slave mode).
A Pull-Up resistor to VDD3 is recommended
if LBI is not used.
149 LCSO OLBI Chip Select Output
Used to select LBI external peripheral
164 LALE O LBI Address Latch Enable
A Pull-Down resistor to VSS is
recommended if LBI is not used or
operated in demultiplexed LBI
configuration.
151 LRD OLBI Read strobe
(see page 423 f. timing figures)
152 LWR OLBI Write strobe
(see page 423 f. timing figures)
153 LBHE OLBI Byte high enable
(see page 423 f. timing figures)
160 LRDY ILBI Ready strobe
Control signal for extended bus cycles. This
signal is asserted by the bus target to insert
wait states.
Table 4 Local Bus Interface (LBI) / General Purpose Port (GPP) /
Synchronous Serial Control (SSC) Interface Pins (cont’d)
Pin No. Symbol Input (I)
Output (O) Function
PEB 20534
Pin Descriptions
Semiconductor Group 43 Data Sheet 09.98
161 LINTI1 I LBI Interrupt Input from Peripheral 1
163 LCLK O LBI Clock Output
This is signal provides the internal LBI clock
which is frequency of signal CLK divided by n
(n can be configured in register LCONF).
LCLK is provided for connection of
synchronous peripherals.
Note: The duty cycle (high to low in %) of
LCLK depends on the clock devision
factor:
n=2: 25%
n=4: 12,5%
n=8: 6,25%
n=x: 100/( 2*x) %
The high phase is always T
CLK
/2, i.e.
typically 15ns at 33 MHz system clock
frequency.
Table 5 Serial Communication Controller (SCC) Signals
Pin No. Symbol Input (I)
Output (O) Function
75
86
167
178
RxCLK0
RxCLK1
RxCLK2
RxCLK3
IReceive Clock
The function of these pins depends on the
selected clock mode.
In each channel, RxCLKn may supply either
– the receive clock (clock mode 0, 4), or
– the receive and transmit clock (clock mode
1, 5), or
– the clock input for the baud rate generator
(clock mode 2, 3).
77
88
169
180
RxD0
RxD1
RxD2
RxD3
IReceive Data
Serial data is received on these pins.
Table 4 Local Bus Interface (LBI) / General Purpose Port (GPP) /
Synchronous Serial Control (SSC) Interface Pins (cont’d)
Pin No. Symbol Input (I)
Output (O) Function
PEB 20534
Pin Descriptions
Semiconductor Group 44 Data Sheet 09.98
76
87
168
179
TxCLK0
TxCLK1
TxCLK2
TxCLK3
I/O Transmit Clock
The function of this pin depends on the
selected clock mode and the value of the
SSEL bit (CCR0 register). If programmed as
an input (bit CCR0.TOE=’0’), this pin supplies
either
– the transmit clock for the channel (clock
mode 0, 2, 4, 6; SSEL bit in CCR0 is reset),
or
– a transmit strobe signal for the channel
(clock mode 1).
If programmed as an output
(bit CCR0.TOE=’1’), this pin supplies the
transmit clock for the channel which is
generated either
– from the baud rate generator (clock mode 2,
3, 6, 7; SSEL bit in CCR0 is set), or
– from the DPLL circuit (clock mode 3, 7;
SSEL bit in CCR0 is reset).
In clock mode 5 an active-low tri-state control
signal marks the programmed transmit time-
slot if bit CCR2.TOE is set.
78
89
170
181
TxD0
TxD1
TxD2
TxD3
O,
o/d Transmit Data
Transmit data is shifted out via these pins.
They can be programmed to be either push-
pull or open drain output to support bus
configurations (register CCR1).
Table 5 Serial Communication Controller (SCC) Signals (cont ’d)
Pin No. Symbol Input (I)
Output (O) Function
PEB 20534
Pin Descriptions
Semiconductor Group 45 Data Sheet 09.98
81
92
173
184
RTS0
RTS1
RTS2
RTS3
or
TxCLKO0
TxCLKO1
TxCLKO2
TxCLKO3
ORequest to Send
When the RTS bit in the CCR1 register is set,
the RTS signal goes low. When the RTS bit is
reset, the signal goes high if the transmitter
has finished and there is no further request for
a transmission.
In bus configuration, RTS can be programmed
via CCR1 to:
– go low during the actual transmission of a
frame shifted by one clock period, excluding
collision bits.
– go low during reception of a data frame.
– stay always high (RTS disabl ed).
In ASYNC mode, RTS can be programmed
via CCR1 to:
– to be controlled autonomously by the SCCn
and be activated when a frame
transmission starts and deactivated when
transmission is completed (default state).
– to be controlled autonomously by the SCCn
for bi-directional flow control and to be
forced active when shadow part of RFIFO is
empty and forced in-active when RFIFO has
reached a threshold.
– to be controlled by the host.
Transmit Clock Out
In clock mode 4 the internal transmit clock is
switched to this pin if bit CCR1.TCLKO set.
TxCLKOi will be in phase with the
corresponding transmit data signal TxDi with
regard to the High Speed timing
Characteristics.
Note: It is recommended to ignore signal RTS
in Hi gh Sp eed m ode if C C R1 .TCL K O i s
not set. Nevertheless the signal remains
an active push/pull output.
Table 5 Serial Communication Controller (SCC) Signals (cont’d)
Pin No. Symbol Input (I)
Output (O) Function
PEB 20534
Pin Descriptions
Semiconductor Group 46 Data Sheet 09.98
79
90
171
182
CTS0
CTS1
CTS2
CTS3
or
CxD0
CxD1
CxD2
CxD3
or
TCG0
TCG1
TCG2
TCG3
IClear to Send
A low on the CTSn input enables the
respective transmitter.
Additionally, an inter rupt may be issu ed if a
state transition occurs at the CTSn pin
(programmable feature).
If no ’Clear To Send’ function is required, the
CTSn inputs can be directly connected to VSS.
Collision Data
In a bus configuration, the external serial bus
must be connected to the corresponding CxD
pin for collision detection.
A collision is detected whenever a logical ’1’ is
driven on the open drain TxD output but a
logical ’0’ is detected via CxD input.
Transmit Clock Gating
In clock mode 4 these pins are used as
Transmit Clock Gating signals.
Table 5 Serial Communication Controller (SCC) Signals (cont ’d)
Pin No. Symbol Input (I)
Output (O) Function
PEB 20534
Pin Descriptions
Semiconductor Group 47 Data Sheet 09.98
Note: A value of 4.7k is recommended for all Pull-Down/Pull-Up resistors to V
DD3
/V
SS
.
This re co mme ndation applies to Table 2 to Table 5.
80
91
172
183
CD0
CD1
CD2
CD3
or
FSC0
FSC1
FSC2
FSC3
or
RCG0
RCG1
RCG2
RCG3
ICarrier Detect
The function of this pin depends on the
selected clock mode.
It can supply
– either a modem control or a general
purpose i nput (c lock m odes 0 , 2, 3, 6, 7). If
auto-start is programmed, it functions as a
receiver enable signal.
– or a receive strobe signal ( clock mode 1).
Additionally, an interrupt may be issued if a
state transition occurs at the CDn pin
(programmable feature).
A Pull-Down resistor to VSS is
recommended if CD is not used.
Frame Synchronization
When SCCn is in time-slot mode (e.g. PCM
mode) these pins supply the Frame
Synchronization inputs (clock mode 5).
Receive Clock Gating
In clock mode 4 these pins are used as
Receive Clock Gating signals.
176
177 XTAL1
XTAL2 I
(O) Crystal Connection
If the internal oscillator is used for clock
generation (clock mode 0b, 6, 7) the external
crystal has to be connected to these pins.
Moreover, XTAL1 may be used as common
clock input for all SCCs provided by an
external clo ck generato r (oscill ator).
A Pull-Down resistor to VSS is
recommended if XTAL1 is not used.
Table 5 Serial Communication Controller (SCC) Signals (cont’d)
Pin No. Symbol Input (I)
Output (O) Function
PEB 20534
Functional Description
Semiconductor Group 48 Data Sheet 09.98
3 Functional Description
Figure 8 DSCC4 Functional Block Diagram
Protocol
Machine
SCC0
RFIFO
17
DWORDs
TFIFO
8
DWORDs
SCC1
SCC2
SCC3
Central
TFIFO
128
DWORDs
Central
RFIFO
128
DWORDs
Central
Int. Queue
16
DWORDs
4 Tx
DMACs 4 Rc
DMACs 1 Int.
DMAC
Central
Int.
Controller
32 bit / 33 MHz
PCI Bus Interface DEMUX Bus
Extension
internal slave bus
internal int.
bus
internal master bus
internal de-multiplexed address bus
JTAG
7
PCI to Local Bus
Bridge (LBI) GPP SSC
5 50 2
SCCi
Interfaces
32
13
LBI Control
Interface Configuration Dependent
Interface
(LBI, GPP, SSC, DEMUX
Address Bus)
Multi Function PortCtrl. Port
PCI Bus
Interface
Test
Interface DEMUX Ctrl.
Interface
76
Misc.
2
Crystal/
Oscillator
PEB 20534
Functional Description
Semiconductor Group 49 Data Sheet 09.98
Abbreviations
Acronyms used in the block diagram are explained in the following:
DMAC - DMA Controller
GPP - General Purpose Port
Int. Queue - Interrupt Queue
JTAG - Joint Test Action Group
LBI - Local Bus Interface
MFP - Multi Function Port
PCI - Peripheral Components Interface
Rc- Receive
RFIFO - (Central) Receive FIFO
SCC3...0 - Serial Communication Controller 3...0
SSC - Synchronous Serial Controller
TFIFO - (Central) Transmit FIFO
Tx - Transmit
The functional blocks of the DSCC4 can be partitioned into three different groups:
– The first group provides the data transfer between the shared memory and on chip
registers/FIFOs. It consists of the PCI System Bus Interface, the 9 DMACs, the
Central RFIFO, the Central TFIFO and the Central Interrupt Queue.
– The second group supports the multiport and multiprotocol serial data communication.
It consists of four SCCs.
– The third group provides control functions for external peripherals using the LBI and/or
low speed communication/control functions using SSC and/or GP. The third group
includes the peripheral blocks: GPP, SSC, LBI. These peripheral blocks can be used
in PCI mode configuration only.
Note: In demul tiplexed bus operation mode, t he MFP only s upplies th e De-multiplex ed
address bus extension of the host interface.
PEB 20534
Functional Description
Semiconductor Group 50 Data Sheet 09.98
General Data Flow Description
For register read/write transaction, each single block can be accessed by the host CPU
via the PCI (or de-multiplexed) bus interface.
The DSCC4 central TFIFO and RFIFO provide space for 128 DWORDs each to be
shared by four serial channels (4 SCCs).
Four DMA Controller Channels (DMACs) transfer data associated with the serial
channels from sh ared mem ory v ia the PCI inte rfac e into the centra l TFIFO . Th ese da ta
are forwarded from the TFIFO to the four SCCs.
In receive direction the serial channels (4 SCCs) deliver received data into the central
RFIFO. Another four DMACs transfer these data from the central RFIFO via the PCI
interface into shared memory locations, associated with the serial ports.
The SCCs as well as the peripheral blocks (LBI, GPP, SSC) are able to generate
interrup ts. Co rresp onding interrupt vec tors are delive r ed by any singl e bl ock the cent ra l
interrupt controller and its 16 DWORDs interrupt queue. A DMAC transfers the interrupt
vectors from the interrupt queue into appropriate locations in the shared memory.
Data and interrupt buffering exists in each block as its interface is contending for the
internal bu sses. The access to the internal busses is controlled by arbiters t hat are not
shown in the block diagram.
PEB 20534
Microprocessor Bus Interface
Semiconductor Group 51 Data Sheet 09.98
4 Microprocessor Bus Interface
The DSCC4 may be configured either for 33 MHz/32-bit PCI operation or for a
33 MHz/32-bit De-multiplexed bus interface.
The DSCC4’s DEMUX input pin is used to select the desired configuration:
The De-multiplexed bus interface mode provides an additional address bus and W/R
control signal extension to the PCI bus interface.
4.1 PCI Bus Interface
In this confi guration, the DSCC4 int erfaces directly to a 33 MH z/32-bit PCI bus. Du ring
run-time, the DSCC4 operates mostly as a PCI Master. However it may also be
accessed by the host processor as a PCI Slave for register read/write or access to
peripherals connected to the LBI.
Device configuration is performed via slave transactions to DSCC4 on-chip registers or
to the PCI Configuration Space.
The DSCC4 is compliant with PCI specification 2.1 at up to 33 MHz.
In addition, the DSCC4 supports little/big endian byte swapping from/to the serial
channels, and una ligned-byte access es for transmit data section s.
4.1.1 Supported PCI Transactions
Memory accesses as a PCI Master: The DSCC4 supports both the PCI Memory Write
and PCI Memory Read commands. For the PCI Memory Write command, it writes to an
agent mapped in the me mory acc ess spac e, whil e for the PC I Re ad co mma nd, it reads
from an agent mapped in the memory address space.
I/O accesses as a PCI Master: The DSCC4 does not support the PCI I/O Write nor PCI
I/O Read commands.
Memory accesse s as a PCI Slave: The D SCC4 su pports both the PCI M emory Write
and PCI Memory Read commands. For the PCI Memory Write command, the DSCC4 is
written to as an agent mapped in the memory address space, while for the PCI Memory
Read com mand, th e DSCC4 is read from as a n agen t mapped in the memory address
space.
I/O accesses as a PCI Slave: The D SCC4 does n ot s upp ort the PCI I/O Write nor PCI
I/O Read commands.
DEMUX connected to VSS: PCI operation mode
connected to VDD3: De-multiplexed bus interface mode
PEB 20534
Microprocessor Bus Interface
Semiconductor Group 52 Data Sheet 09.98
Burst Capability: The DSCC4 supports bursts of up to 3 DWORDs for reading transmit
and receive descriptors and bursts of up to 15 DWORDs for reading/writing data.
4.1.2 PCI Configuration Space Register Overv iew
The PCI Configuration Space Registers of the DSCC4 are listed in table 6.
For a detailed description of the registers see “PCI Configuration Space - Detailed
Register Description” on page 222 and “Configuration Space“ described in chapter
6 of the PCI Specification Rev. 2.1.
Table 6 PCI Configuration Space Registers
Register Name Short Name Access
(Read/
Write)
Absolute
Address Reset Value
Device ID / Vendor ID DID / VID R 00H2102/110AH
Status / Command STA / CMD R/W 04H0000/0000H
Class Code / Revision ID V1.1 CC / RID R 08H029000/20H
Builtin Self Test / Header Type /
Latency Timer /
Cache Line Size
BIST/HEAD/
LATIM/
CLSIZ
R/W 0CH00000000H
Base Address 1 BAR1 R/W 10H00000000H
Base Address 2 BAR2 R/W 14H00000000H
Base Address not used BARX R/W 18H-24H00000000H
Cardbus CIS Pointer CISP R 28H00000000H
Subsystem ID /
Subsystem Vendor ID SSID /
SSVID R2C
H00000000H
Expansion ROM Base Address ERBAD R/W 30H00000000H
Reserved RES34 R/W 34H00000000H
Reserved RES38 R/W 38H00000000H
Maximum Latency /
Minimum Grant /
Interrupt Pin /
Interrupt Line
MAXLAT /
MINGNT /
INTPIN /
INTLIN
R/W 3CH00000000H
REG40 40H00000000H
REG44 44H00000000H
REG48 48H00000000H
REG4C 4CH00000000H
PEB 20534
Microprocessor Bus Interface
Semiconductor Group 53 Data Sheet 09.98
4.2 De-multiplexed Bus Interface Extension
The DSCC4 m ay be configured f or 33 MH z/ 32-bi t D e-mu ltip lex ed bus for con nec tion to
non PCI systems with de-multiplexed processors such as the i960Hx or MC68EC0x0.
The de-multip lex ed bus interface is a syn chro nous interface very si milar to the PCI b us
with the following exceptions:
1. The W/R input/output sign al replaces the fun ction of the PCI comman d nibble in the
C/BE(3:0) bit field.
2. The transaction address is driven or read from the additional address bus A[31..2]
3. The parity signal PAR is not generated as a master or evaluated as a slave
Beside these exceptions all control signals and timings are equal to PCI bus interface
mode. Also the PCI Configuration Space must be programmed during configuration.
Note: In DEMUX mode as in PCI mode, the DSCC4 provides only the first address of a
master burst read or wri te transac tion. Addres s increme ntation m ust be provi ded
extern ally if required by the target or DSCC4 burst capability is disabled (defa ult
value) for DEMUX mode.
Note: Because the PCI command nibble is replaced by a W/R control signal only IDSEL
distingu ishes betw een PCI Confi guration Spac e and slav e register ac cess. Thus
IDSEL must be treated as a Configuration Space chipselect and remain
deasserted during all other slave register accesses.
In DEMUX bus mode the burst capability is limited to 4-dwords and must be enabled via
the ’DBE’ (Demux Burst Enable) bit in the Global Mode Register GMODE.
Even in the case that burst capability has been enabled, the target can request the
DSCC4 to stop the current transaction by asserting the STOP signal as in PCI operation.
The following diagrams illustrate the functional timing waveforms for both single and
burst transactions.
Table 7 Non-PCI Signal Extension in the De-multiplexed Bus Interface Mode
Pin
description
table
reference
Symbol Input (I)
Output (O) Function
table 1 DEMUX I DEMUX = VSS selects PCI mode,
DEMUX = VDD3 selects De-multiplexed Bus
Interface mode.
table 4 A(31:2) I/O De-multiplexed Address Bus
table 1 W/R I/O Write/Read Control signal
PEB 20534
Microprocessor Bus Interface
Semiconductor Group 54 Data Sheet 09.98
Figure 9 Master Single READ Transaction followed by a Master Single WRITE
Transaction in De-multiplexed Configuration
Table 8 DEMUX Mode Related Register and Bit-Fields
Offset
Addr. Access
Type Controlled
by Reset
Value Register Name
0008Hr/w CPU 00000000HGMODE:
Global Mode Register
Bit-Fields
Pos. Name Default Description
1 DBE 0 Demux Bus Mode Burst Enable
DBE = ’0’ Burst capability disabled, only
single DWOR D tr ans fers are performe d;
DBE = ’1’ Burst capability enabled, burst
length is limited to 4 DWORDs.
Note: Only valid in De-multiplexed bus
mode.
BE(3:0)
ADDR,
ADDR
DATA
READ Access
DATAADDR,
ADDR
BE(3:0)
WRITE Access
0ns 100ns 200ns
CLK
FRAME
D(31:0)
A(31:2)
BE(3:0)
W/R
TRDY
don’t care don’t care
don’t care don’t care
PEB 20534
Microprocessor Bus Interface
Semiconductor Group 55 Data Sheet 09.98
Figure 10 Master Burst WRITE/READ Transaction in De-multiplexed Configu-
ration
When in De-multiplexed bus configuration, the DSCC4 adheres the PCI bus protocol and
timing specificaton, except for the address and command handling. In this mode, the
addresses are provided on a separate address bus A(31:2) to eliminate the need for
external de-multiplexing buffers. The address lines A(31:2) correspond to the address
lines AD(31:2) in PCI mode. The address becomes valid with the falling edge of FRAME
and stays valid for the standard PCI address phase, the turn-around cycle and the entire
data phase. In burst mode the addresses have to be incremented externally for each
single transfer.
Moreover, in De-multiplexed mode the command signals are not used. Instead of the
command s ignals a separate pin W/R (I/O) provide s the Write/Read strobe sig nal. The
Write/Read becomes valid with the falling edge of FRAME and stays valid for the
standard PCI address phase, the turn-around cycle and the entire data phase.
BE(3:0)
ADDR,
ADDR
DATA1
WRITE/READ Access
BE(3:0) BE(3:0)
DATA2 DATA3 DATA4
BE(3:0)
0ns 100ns 200ns
CLK
FRAME
D(31:0)
A(31:2)
BE(3:0)
W/R
TRDY
don’t care
don’t care
PEB 20534
Microprocessor Bus Interface
Semiconductor Group 56 Data Sheet 09.98
The following four ’commands’ are supported:
Note: When desi gni ng a de-multiplexed s ys tem w ith t he D SCC 4 in D e-m ulti plex ed PC I
mode it is the responsibility of the glue logic to meet the bus timing/protocol of the
PCI spec ification and of the me mory devices tha t are used in the system . When
the DSCC4 operates in master mode, the bus cycle, for example, can be delayed
by the TRDY signal.
Please refer to “Application Exam ple: De-multiplexed Bus Interface” on page 407
for an de-multiplexed bus application example.
Table 9 Supported Commands in De-multiplexed Bus Mode
W/R IDSEL Master Mode Slave Mode
0 0 memory read DSCC4 register read
1 0 memory write DSCC4 register write
0 1 not supported DSCC4 PCI Configuration read
1 1 not supported DSCC4 PCI Configuration write
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 57 Data Sheet 09.98
5 DMA Controller and Central FIFOs
The DSCC4 has an 8-channel flexible DMA controller to perform data transfer with
minimal host CPU intervention and high bus efficiency between the DSCC4 and
memory.
These DMA channels service receive and transmit FIFOs of the four serial
communication controllers (SCCs) transferring data into or out of the central DMA
Controller FIFOs.
On the host system side each DMA channel transfers data either from the central receive
FIFO to the shared memory (receive direction) or from the shared memory to the central
transmit FIFO (transmit direction).
8 DMA channels, eac h corres ponding to an indivi dual SCC trans mi t or receive, operate
on linked lists in the host memory.
A ninth DMA channel transfers interrupt vectors from the central interrupt FIFO to one of
10 interrupt vector queues located in the shared host memory.
Figure 11 shows the block diagram of the DMA Controller.
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 58 Data Sheet 09.98
Figure 11 DMA Controller Block Diagram
A detailed description of the DMA channels and central FIFOs is provided in the following
chapters.
128 DWORDs Central
Receive FIFO for all
channels
128 DWORDs Central
Transmit FIFO in 4 partitions
of programmable size
Receice
DMA Channels
Transmit
DMA Channels
DMACR0
DMACR1
DMACR2
DMACR3
DMACT0
DMACT1
DMACT2
DMACT3
PCI Interface
PCI Bus
or
DEMUX Bus
Internal Master Bus
SCC
Receive FIFOs
SCC
Transmit FIFOs
15+2 DWORDs
6+2 DWORDs
DMACI
Interrupt
DMA Channel 16 DWORDs Central
Interrupt Vector FIFO Central Interrupt
Controller
DMA-Controller
DEMUX
Extension
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 59 Data Sheet 09.98
5.1 DMAC Operational Description
5.1.1 DMAC Register Overview
The following table provides an overview of all DMA Controller related registers. For
detailed register description refer to Chapter 10.
A summary is provided in the following table:
Table 10 DMA Controller Register Set
Of fset Address
relative to Base Address 0
(BAR0) in PCI Configuration
Space
Register Meaning
(not DMA channel specific registers)
Global Control Registers
0000HGCMDR Global Command Register
(see page 232)
0004HGSTAR Global Status Register
(see page 237)
0008HGMODE Global Mode Register
(see page 241)
Global Interrupt Queue Control Registers
000CHIQLENR1 Interrupt Que ue Leng th Reg ist er 1
(transmit/receive interrupt queues)
(see page 247)
0010HIQLEN R2 Interrupt Que ue Leng th Register 2
(configuration and peripheral interrupt
queue)
(see page 249)
003CHIQCFGBAR Interrupt Queue Configuration Base
Address
(see page 253)
0040HIQPBAR Int errupt Que ue Pe riph eral Base
Address
(see page 254)
Global Central FIFO Control Registers
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 60 Data Sheet 09.98
0044HFIFOCR1 FIFO Control Register 1
(Transmit FIFO partition size,
dist ing uis hed for 4 transmit channels)
(see page 255)
0048HFIFOCR2 FIFO Control Register 2
(Transmit FIFO refill threshold,
dist ing uished for 4 transmit cha nne ls)
(see page 256)
004CHFIFOCR3 FIFO Control Register 3
(Receive FIFO threshold,
common to all receive channels)
(see page 258)
0034HFIFOCR4 FIFO Control Register 4
(Transmit FIFO forward threshold,
dist ing uis hed for 4 transmit channels)
(see page 260)
(DMA channel specific registers)
Ch0 Ch1 Ch2 Ch3 (i=0..3)
0014H0018H001CH0020HIQSCCiRX
BAR Interrupt Queue Receive SCCi Base
Address
(see page 251)
0024H0028H002CH0030HIQSCCiTX
BAR Interrupt Queue Transmit SCCi Base
Address
(see page 252)
0050H005CH0068H0074HCHiCFG Channel i Configuration Register
(see page 262)
0054H0060H006CH0078HCHiBRDA Channel i Base Receive Descriptor
Address
(see page 264)
0058H0064H0070H007CHCHiBTDA Channel i Base Transmit Descriptor
Address
(see page 265)
0098H009CH00A0H00A4HCHiFRDA Channel i First Receive Descriptor
Address
(see page 266)
00C8H00CCH00D0H00D4HCHiLRDA Channel i Last Receive Descriptor
Address
(see page 268)
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 62 Data Sheet 09.98
The following table provides an overview of all DMA Controller commands. For detailed
register description refer to Chapter 10.
Table 11 DMAC Commands
Offset
Addr. Access
Type Controlled
by Reset
Value Register Name
0000Hr/w CPU 00000200HGCMDR:
Global Command Register (page 232)
Bit-Fields
Pos. Name Default Description
27..24
31..28
21
20
CFGIQ-
SCCiRX
and
CFGIQ-
SCCiTX
and
CFGIQ
CFG,
CFGIQP
0 Configure Interrupt Queue:
These command bits cause the DMAC to
establish or re-configure the dedicated
interrupt queue using the values of the
corresponding base address registers
and interrupt queue length registers.
(only pe rforme d if action request bit ’AR’
is set additionally)
13..10 TXPRi 0 Transmit Poll Request Channel i:
If the DMA transmit channel is stopped
on a HOLD condition (HOLD bit
detected), this command forces a read
transaction on the transmit descriptor
verifying the HOLD condition again.
9 IM 1 Interrupt Mask:
If set to ’1’ the action request
acknowledge interrupt is supressed.
8 IADC 0 Initialize All DMA Channels:
This bit causes the DM AC to initia liz e al l
transmit and receive channels. The
channel specific commands are ignored
in this case.
0 AR 0 Action Request:
This bit caus es the D MAC to execu te all
commands s et in registers G CMDR and
CHiCFG.
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 63 Data Sheet 09.98
0050H
005CH
0068H
0074H
r/w CPU 00000000HCHiCFG (i=0..3):
Channel i Configuration Register
(page 262)
Bit-Fields
Pos. Name Default Description
27..24 Interrupt
Mask 0 This bit field is used to mask FI and ERR
interrupt indications distinguished for
transmit and receive DMA channels.
The following commands are evaluated by the DMAC on any action request
(bit ’AR’ set in register GCMDR) without IADC command set.
22 RDR 0 Reset DMA Receiver
21 RDT 0 Reset DMA Transmitter
20 IDR 0 Initialize DMA Receiver
19 IDT 0 Initialize DMA Transmitter
Table 11 DMAC Commands (cont’d)
Offset
Addr. Access
Type Controlled
by Reset
Value Register Nam e
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 64 Data Sheet 09.98
5.1.2 DMAC Control and Data Structures
Figure 12 DMA Data Flow
PCI BridgeRAMCPU
PCI Bus
Host Bus
Serial Lines 0..3
Serial Li nes 0. . 3
Host RAM
CPU
Channel0 receive descriptor list
Channel0 transmit descriptor list
Channel0 receive
interrupt queue
PCI Bus
Load
System Overview:
Data Flow:
DSCC4
Channel0 transmit
interrupt queue
Channel1
Channel2
Channel3
Channel0
DSCC4
Configuration
interrupt queue
Peripheral
interrupt queue
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 65 Data Sheet 09.98
The CPU prepares linked lists for transmit and receive channels in the shared memory.
These may be handled by dy nam ica lly allocatin g an d l inking descriptors and buffers as
needed during runtime or by static predefined memory structures e.g. ring-chained-lists
(the ’last’ descriptor points back to the first descriptor). A mix of predefined descriptor
lists but dynamically handled data buffers may also be an appropriate solution. This
strategy depends on the specific application. The DMAC provides multiple control
mechanisms supporting all of these combinations in an efficient way.
The descriptors and data buffers can be stored in separate memory spaces within the
32-bit address range allowing full scatter/gather methods of assembling and
disassembling of packets.
Each descriptor contains a ’next descriptor address’ field to realize the linked list.
Because the DMA controlle r cann ot distingu ish betwee n valid and i nvalid a ddresses, a
’Hold’ mechanism is needed to prevent the DMA controller from branching to invalid
memory locations. A ’next descriptor address’ might be invalid for several reasons:
• no further transmit transaction is requested; therefore no further transmit descriptor is
allocate d and the ’nex t descriptor address ’ field of the la st descriptor i s invalid when
read by the DMA controller;
• temporarily the software is not able to attach new receive descriptors to the list in time;
therefore no further receive descriptor is allocated and the ’next descriptor address’
field of the last descriptor is invalid when read by the DMA controller;
• the receive descriptor list is organized as a ring; the DMA channel must be prevented
from branching a descriptor which is not yet s erviced by the CPU.
Two alternative control mechanisms are provided to detect and handle descriptor list end
(Hold) conditions:
• Hold bit control mode
(See “DMAC Operation Using Hold-Bit Control Mechanism” on page 76.)
• Last descriptor address control mode
(See “DMAC Operation Using Last Descriptor Address Control Mode” on page 78.)
The Control Mod e applies to all DMA channels transm it an d receive and is sele cted via
bit ’CMODE’ in Global Mode Register GMODE.
An HDLC frame ma y be contained in one buffer connected to one descriptor or it may
be contained in several buffers each associated with linked descriptors. A ’frame end’
indication (FE bit) will be set in each descriptor which points to the last buffer of one
HDLC frame.
The ’frame end’ indications are stored in the internal FIFOs influencing the FIFO control
(threshold) mechanisms. Therefore ’frame end’ indications (FE bit) are also used in non
frame oriented protocol modes such as ASYNC mode. They are referred to as ’frame
end/block end’ indication in the following chapters.
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 66 Data Sheet 09.98
5.1.2.1 DMAC Transmit Descriptor Lists
Each transm it descriptor consists of 4 consecu tive DWORDs located DWORD aligned
in the shared memory. The first 3 DWORDs are written by the host and read by the
correspond ing DMA channel usin g a burst transaction . They provide inform ation about
the next descriptor in the linked list, the attached transmit data buffer, and its size as well
as some control bits.
The fourth DWORD is written by the DMA channel indicating that operation on this
descripto r is finis he d.
The CPU will write the address of the first descriptor of each linked list to a dedicated
Base Address Register (BRTDAi) during initialization procedure. The corresponding
DMA channel start serving the descriptor at these addresses.
Figure 13 Transmit Descriptor List Structure
FE Hold HI NO
Next Transmit Descriptor Pointer
Transmit Data Pointer
0x0000000
FE Hold HI
31 0
byte3 byte2 byte1 byte0
byte11 byte10 byte9 byte8
byte7 byte6 byte5 byte4
byte15 byte14
byte19
31 0
Transmit Descriptor:
Transmit Data Buffer:
C0 00
0x0000
(dummy)
DWORD1
DWORD2
DWORD3
DWORD4
(DWORD5)
written by
CPU
written by
DSCC4
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 67 Data Sheet 09.98
Table 12 Transmit Descriptor Bit Field Description
DWORD Bit Field Description
1
(read by
DMAC,
written by
CPU)
FE Frame End Indication:
Not evaluated by the DMAC. This bit indicates that this
des c riptor co ntains a complet e data packet or the last
part of a data packet. This indication is forwarded to the
corresponding SCC.
An ’FI’ interrupt is generated after completion of a
transmit descriptor with FE=’1’ setting.
Hold Hold Indication:
Hold=’1’ marks the end of the descriptor chain. In this
case the DMAC will not branch to the next descriptor
address. The DMAC reads and evaluates Hold bit and
next descriptor address again on transmit poll request
command.(see Chapter 5.1.2.3)
(this bit is ignored if DMAC is configured in last
descriptor address control mod e)
HI Host Initiated Interrupt:
This bit set to ’1’ causes the DMAC to generate an
interrupt after completion of the descriptor and after
transfer of the complete data section from Host memory
to the central transmit FIFO. This may be used for
software control purposes.
NO Number OF Bytes:
This bit field determines the number of valid data bytes
in the transmit data buffer and the data buffer size. The
data buffer size is always n DWORDs, whereas n
depends on NO and the byte offset address ADD (refer
to ADD description on next pages).
2
(read by
DMA C,
written by
CPU)
Next Tx Descr Ptr Next Transmit Descriptor Address:
The DMA Channel will branch to this address when
proceeding in the linke d list.
3
(read by
DMA C,
written by
CPU)
Tx Data B uffer Ptr Transmit Data Buffer Start Address:
The DMA Channel starts reading transmit data at this
address. Read access to transmit data buffer may
occur per single DWORD transfers or up to 15
DWORDs burst transfers.
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 68 Data Sheet 09.98
In detail the DMA controller reads a transmit buffer descriptor, calculates the data buffer
address and data buffer length and transfers data up to the burst size from the data
buffer into the central transmit FIFO. For a frame longer than the burst size, this
operation is repeated as long as the transmit FIFO requests for data. For more
information about FIFO control see Chapter 5.2.
After the data buffer has been transferred, the controller marks the descriptor
“completed” and branches to the next descriptor if applicable. An ’FI’ interrupt will be
generated, if the currently completed descriptor contained an ’frame end/block end’
(FE=’1’) indication.
In HDLC mode data is transm itted a s frame s. The host i ndi cates the end of a fram e by
setting ’FE’ bit in the transmit descriptor. When a frame end is detected the DMA channel
forwards this information to the SCC. The SCC then terminates the transmission by
appendin g the CRC and the closing flag sequenc e to the data. I f (FE=0 & HOLD=1) or
(FE=0 & FTDA=LTDA) an ERR interrupt is generated by the DMA controller (see
Chapter 5.1.2.3 and Chapter 5.1.2.4).
Note: In contrast to HDLC mode all other modes (ASYNC, BISYNC, Extended
Transparent Mode) are block/character oriented. Since the DMA controller does
not disting uish be tween diffe rent protoc ol mod es (HDLC, ASYNC, ...) the ’FE’ b it
in the last descriptor of the linked list might be set also for the block/character
oriented modes as a kind of block end indication or together with an end of list
condition.
Although the DSCC4 works only DWORD oriented, it is possible to begin a transmit data
section at an uneven (not DWORD aligned) address. The two least significant bits (ADD)
of the transmit data pointer determine the beginning of the data section and the number
of data bytes in the first DWORD of the data section, respectively. Nevertheless the
DSCC4 w ill alway s perform D WORD read tran sfers (all byte enables valid) on transm it
data sections marking invalid bytes internally.
4
(written
by DMAC,
read by
CPU)
C Complete Bit:
This bit is set by the DMAC after having completed the
descriptor and corresponding data section.
The software can use this indication for memory and
linked list management.
5- Dummy DWORD;
Only necessary if compatibility between transmit and
receive descriptors is needed, i.e. receive descriptors
are manipulated by the host and attached to a transmit
descriptor list. For transmit operation, DWORD5 is
neither used by the DMAC nor by the host CPU.
Table 12 Transmit Descriptor Bit Field Description (cont’d)
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 69 Data Sheet 09.98
Table 13 shows how many bytes can be valid in the first 32-bit word depending on ADD
for both little endian mode and big endian mode. Of course, the total number of valid
bytes depends on the ’NO’ bit field value in the corresponding transmit descriptor.
Table 13 Meaning of ADD in Little/Big Endian Mode
ADD Number of Little Endian (Intel) Big Endian (Motorola)
Valid Bytes 1110010011100100
00 4, if NO > 3 byte 3 byte 2 byte 1 byte 0 byte 0 byte 1 byte 2 byte 3
01 3, if NO > 2 byte 2 byte 1 byte 0 - - byte 0 byte 1 byte 2
10 2, if NO > 1 byte 1 byte 0 - - - - byte 0 byte 1
11 1, if NO > 0 byte 0 - - - - - - byte 0
Example A: NO = 3
00 3 - byte 2 byte 1 byte 0 byte 0 byte 1 byte 2 -
Example B: NO = 2
01 2 - byte 1 byte 0 - - byte 0 byte 1 -
Example C: NO = 8
10 2 byte 1 byte 0 - - - - byte 0 byte 1
4 byte 5byte 4byte 3byte 2byte 2byte 3byte 4byte 5
2 - - byte 7byte 6byte 6byte 7 - -
Example D: NO = 1
11 1 byte 0 - - - - - - byte 0
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 70 Data Sheet 09.98
The follow ing figure provid es an exampl e, how a transmi t descriptor and it s associate d
data buffer is located in the memory as a result of a memory dump.
Figure 14 Transmit Descriptor Memory Example
Note: Although transmit descriptors consist of only 4 DWORDs it might be useful to
allocate 5 DWORD structures to achieve compatibility between transmit and
receive descriptors. In this case only small CPU performed manipulations are
necessary to convert a completed receive descriptor into a transmit descriptor re-
chained to a transmit descriptor list. This is typical e.g. for frame relay or bridging
applications where received data might be sent out again on another DSCC4 port.
1 0 0 0x009
Next Transmit Descriptor Ptr.=0x10001014
Transmit Data Pointer= 0x20000020
0x40000000
FE Hold HI
31 0
04 03 02 01
00 00 00 09
08 07 06 05
31 0
CH0BTDA = 0x10001000
DSCC4 register CH0BTDA:
Base Transmit Descriptor Address Channel 0
31 0
Host Memory Dump:
Value: Address:
0x10001000
0x10001004
0x10001008
0x1000100C
0x10001010
0x10090000
0x10010014
0x20000020
0x40000000
0x20000020
0x20000024
0x20000028
0x20000030
0x04030201
0x08070605
0x00000009
(dummy) 0x00000000
0x00000000 0x10001014
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 71 Data Sheet 09.98
5.1.2.2 DMAC Receive Descriptor Lists
Each receive descriptor consists of 5 consecutive DWORDs, located DWORD alig ned
in the shared memory. The first 3 DWORDs are read by the corresponding DMA channel
using a burs t tran sac tio n an d prov ide information abo ut the buffer size as we ll as some
control bits, the next descriptor in the linked list and the attached receive data buffer.
The fourth DWORD is written by the DMA channel indicating that operation on this
descriptor is finished. The fifth DWORD is also writte n by the DMA chann el but only in
descriptors containing the first or only data section of an HDLC frame or data block. It is
a pointer to the last descriptor containing the frame or block end (’FE’ bit) allowing the
software to unchain the complete partial descriptor list containing one frame or block
without pa rsing t hrough the list for ’FE’ indication . It is wr itten after t he frame has been
completely written to the shared memory.
The CPU will write the address of the first descriptor of each linked list to a dedicated
Base Address Register during initialization procedure. The corresponding DMA
channels start operating the linked lists at these addresses.
Figure 15 Receive Descriptor List Structure
0 Hold HI NO
Next Receive Descriptor Pointer
Receive Data Pointer
BNO
FE
Hold HI
31 0
byte3 byte2 byte1 byte0
byte11 byte10 byte9 byte8
byte7 byte6 byte5 byte4
byte15 byte14
byte19
31 0
Receive Descriptor:
Receive Data Buffer:
C 0
0x0000 0
0x00STATUS
Frame End Descriptor Pointer
DWORD1
DWORD2
DWORD3
DWORD4
(DWORD5)
written by
CPU
written by
DSCC4
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 72 Data Sheet 09.98
Table 14 Receive Descriptor Bit Field Description
DWORD Bit Field Description
1
(read by
DMAC)
Hold Hold Indication:
Hold=’1’ marks the end of the descriptor chain. In this
case the DMAC will not branch to the next descriptor
address but stop DMA operation until initialized again
(see Chapter 5.1.2.3).
(this bit is ignored if DMAC is configured in last
descriptor address control mode)
HI Host Initiated Interrupt:
This bit set to ’1’ causes the DMAC to generate an
interrupt after completion of the descriptor and after
transfer of the complete data section to the Host
memory. This may be used for software control
purposes.
NO Number OF Bytes:
This bit field determines the receive data buffer size and
should be a multiple of 4.
Note: The number of received data bytes includes the
receive status byte (RSTA) which is generated by
the SCC receiver in HDLC mode or all status
bytes optionally generated in character oriented
protoc ol mode s respe ctively .
2
(read by
DMAC)
Next Rx Descr Ptr Next Receive Desc riptor Address:
The DMA Channel will branch to this address when
proceeding in the linked list.
3
(read by
DMAC)
Rx Data Buffer
Ptr Receive Data Buffer Start Address:
The DMA Channel starts writing receive data at this
address. Write access to receive data buffer may occur
per single DWORD transfers or up to 15 DWORDs
burst transfers.
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 73 Data Sheet 09.98
In details the DMA controller reads a receive buffer descriptor, calculates the maximum
data buffer size and the data buffer address and transfers data up to the burst size from
the receive FIFO into the data buffer. For a frame longer than the burst size, this
operation is repeated as long as data is available in the FIFO. For more information
about FIFO control see Chapter 5.2).
After the data buffer has been filled, the controller writes the number of stored bytes into
the descriptor, marks the descriptor “completed” and branches to the next descriptor.
When a frame end (HDLC) or block end (e.g. termination character or time out in ASYNC
mode) is detected and all data has been transferred, the DMA controller writes the
Frame End Descriptor Pointer to the descriptor containing the beginning of the frame. A
maskable interrupt status entry is written into the interrupt queue, if initiated by the host.
As a last transaction the DMA controller writes the number of bytes stored in the last
buffer as well as a DMA controller related status byte in the descriptor.
The DMA controller related status byte indicates, if the frame or block was ended
normally or by a receiver reset command or by a HOLD bit in the current descriptor.
In HDLC mode the last byte in the data section contains the status information of the
HDLC frame such as: result of CRC check, data overflow, frame aborted. This byte is
forwarded transparently from the SCC to the data buffer following the received data.
4
(written
by DMAC)
FE Frame End Indication:
This bit set to ’1’ indicates that this descriptor contains
a complete data packet or the last part of a data packet.
An ’FI’ interrupt is generated after completion of a
receive descriptor with FE=’ 1’ se tting.
C Complete Bit:
This bit is set by the DMAC after having completed the
descriptor and corresponding data section.
The software can use this indication for memory and
linked list management.
Note: The Frame End Descriptor Pointer (DWORD 5) is
written to the receive descriptor after writing the
Complete Bit indication to DWORD 4.
5
(written
by DMAC)
FE Descr Ptr Frame End Descriptor Pointer:
This address is only written to the first descriptor of a
data packet pointing to the descriptor containing the
’FE’ indication of the same packet. If the complete
packet is stored in the first and only data section this
address is equal to the descriptor address.
Table 14 Receive Descriptor Bit Field Description (cont’d)
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 74 Data Sheet 09.98
In ASYNC mode and BISYNC mode to every data b yte an a ttached status byte c an be
stored. This byte is forwarded transparently from the SCC to the data buffer and it
contains status information such as: parity, parity error, framing error.
Since the threshold of the SCC specific receive FIFOs can be set to 1, 2, 4, 16 or
24 bytes the receive data buffer DWORDs can contain less than four valid bytes (e.g. 1
or 2 bytes) In this case the data buffer contains holes of invalid data bytes. Refer to
Table 69, “CCR2: Channel Configuration Register 2,” on page 296.
Table 15 provides examples for the receive data section.
Note: In general, BNO counts the valid bytes that have been transferred into the data
buffer including RSTA (BNO <= NO). The receive status byte (RSTA) is treated
and counted as ’data’ by the DMA controller. As an example, an HDLC frame
containing 32 bytes to be transferred to the shared memory needs 33 bytes in the
receive data buffer due to the receive status byte, which is attached to the data by
the SCC. If NO =32 in this example , the receive status byte as well as the frame
end indication will be written to the next data buffer and descriptor respectively.
Table 15 Receive Data Buffer Section
Mode BNO Little Endian Big Endian
11 10 01 00 11 10 01 00
HDLC (Default) 4 RSTA byte 2 byte 1 byte 0 byte 0 byte 1 byte 2 RSTA
HDLC 7 byte 3 byte 2 byte 1 byte 0 byte 0 byte 1 byte 2 b yte 3
(Default) RSTA byte 5 byte 4 byte 4 byte 5 RSTA
ASYNC 6 status 0 data 0 data 0 status 0
(RFDF=1 status 1 data 1 data 1 status 1
RFTH=00) status 2 data 2 data 2 status 2
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 75 Data Sheet 09.98
The following figure provides an example how a receive descriptor and its associated
data buffer is located in the memory as a result of a memory dump.
Figure 16 Receive Descriptor Memory Example
HI
31 0
04 03 02 01
00 00 00 00
00 00 A2 05
31 0
CH0BRDA = 0x30001000
DSCC4 register CH0BRDA:
Base Receive Descriptor Address Channel 0
31 0
Host Memory Dump:
Value: Address:
0x30001000
0x30001004
0x30001008
0x3000100C
0x30001010
0x00100000
0x30010014
0xC0060000
0x30001000
0x40000020
0x40000024
0x40000028
0x40000030
0x04030201
0x0000A205
0x00000000
0 0 0x010
Next Receive Descriptor Ptr.=0x30001014
Receive Data Pointer=0x40000020
0x006
0x0000
0x000x00
Frame End Descriptor Ptr.=0x30001000
0Hold HI0
00 00 00 00
0x30001014
0
1 1 00
0x40000020
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 76 Data Sheet 09.98
5.1.2.3 DMAC Operation Using Hold-Bit Control Mechanism
This mode is selected by setting bit CMODE=’0’ in register GMODE (See “GMODE:
Global Mode Register” on page 241.).
The DMA controller operates on linked lists with pointer information stored in the DSCC4
internal configuration section.
The sof tware starts D MAC op eration by writing the acti on reques t bit ’AR’ in the G lobal
Command Register (GMCDR). On this command the DSCC4 first checks the ’IADC’ bit
of the GCMDR register. Depending on the value of GCMDR.IADC all channels are
initialized (IADC=’1’) or single channels are initialized/reset. In the latter case (IADC=’0’)
the DMAC checks all channel specific Configuration registers CHiCFG for initialization
command bits (’IDR’, ’IDT’). Setting all ’IDR’ and ’IDT’ commands in registers CHiCFG
is equal to setting ’IADC’ in register GCMDR.
Each DMA channel which is triggered for initialization by one of the above mentioned
commands fe tche s the Base T rans mit/Rec eiv e D es crip tor Ad dres s (BTD A/BR DA) from
its CHiBTDA/ChiBRDA register. The DMA channel continues reading the Tx/Rx
descriptors from the shared memory which in turn point to the associated data buffer and
the next descriptor address.
The external memory associated to a DMA channel is organized as a chained list of
buffers (typically 256 bytes) set up by an external host. Each chained list is composed of
descriptors and data sections. The descriptor contains the pointer to the next descriptor,
the start a ddress of the d ata buffer and th e size of a da ta section. In transmit dire ction
the data pointer is a byte address. The descriptor als o inc ludes control informati on like
frame end (frame end for HDLC modes, block end in dication for ASYNC and extended
transparent mode), transmission hold and host initiated interrupt.
Transmit:
• In transmit direction the DMA controller reads a transmit descriptor, calculates the
data address and fills the DSCC4 central transmit FIFO. When the data transfer of the
specified section is completed, the DMA controller marks the buffer as “completed”
and branches off to the next transmit descriptor. If a frame end (FE) is indicated in the
buffer descriptor (HDLC and block oriented protocol modes), a frame end indication is
forwarded to the serial channel after the data has been transferred, and the frame will
be closed correctly by the SCC. A maskable interrupt status may be generated at the
completion of transmission to be stored in the interrupt queue by the interrupt
controller. Transmission of another frame can begin immediately.
However, if the current transmit buffer descriptor has its “HOLD” bit set, the DMA
channe l does no t branch o ff to the nex t de scriptor. If no frame en d wa s encou ntere d
in the current descriptor, an active “HOLD” bit causes a transmit FIFO underrun to
occur, and a frame to be aborted by the serial channel.
Furthermore an error interrupt is generated anytime a transmit channel detects the
HOLD condition without a frame end indication asserted.
Once, the D SCC4 has sensed the H OLD =’1’ c onditio n in t rans mit directi on, the da ta
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 77 Data Sheet 09.98
transfer c an be rest arted by a si ngl e poll comma nd (GC MDR:TXPOL Li=1; i =0...3) o r
by setting the ’AR’ bit in the GCMDR register after the configuration register has been
updat ed for the c orresponding cha nnel (new channel in itializatio n). The transm it poll
command causes the chan nel to read the H OLD bit (first DWO RD of the desc riptor)
and the next descriptor address of the current descriptor again. The DMAC branches
to the next descriptor address if HOLD bit has been reset to ’0’.
There are three methods to handle the situati on in which th e nex t des cript or addre ss
pointer is no t yet valid in the current transmit descriptor . This happens if the lo cation
of the next descriptor is not known when preparing the last descriptor of the chain:
1.Updating the next descriptor address before providing transmit poll request
command.
2. Updating BTDA address in register CHiBTDA followed by a channel initialization
command (’IDT’).
3. Programming the "next transmit descriptor address" of each descriptor with HOLD
bit set pointing to a fixed dummy descriptor with byte number 0 (NO = 0). The next
descripto r address of the dummy descriptor can be programmed to point to a newly
prepare d descri ptor list be fore provi ding a tran smi t poll requ est com mand. The DMA
channel will continue branching to the dummy descriptor, completing it without data
transfer and branch to the next descriptor which itself might point to the dummy
descriptor again if HOLD bit is set.
Receive:
• If the current receive buffer descriptor has its HOLD-bit set and the current data buffer
has been filled, the DMA channel does not branch off to the next descriptor. An active
“HOLD” bit causes a receive FIFO overflow to occur. If more data are received, those
are discarded by the serial channel.
Furthermore an error interrupt is generated anytime a receive channel detects the
HOLD condition.
The Host CPU (software) is expected to prepare sufficient amount of receive
descriptors which is supported by several control mechanisms. Thus detecting a Hold
condition in the receive descriptor list is treated as an exceptional condition by the
corresponding DMA channel.
Once, the DSCC4 has detected the HOLD=’1’ condition in receive direction, an
interrupt is generated after completion of the current receive descriptor and the
corresponding DMA channel is deactivated for receive direction as long as the host
does not request an initalization via action request command in register GCMDR. The
data transf er can be restarted by sett ing the AR bit in the GCMDR register after the
channel specific configuration register has been updated (new initialization).
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 78 Data Sheet 09.98
5.1.2.4 DMAC Operation Using Last Descriptor Address Control Mode
This mode is selected by setting bit CMODE=’1’ in register GMODE (See “GMODE:
Global Mode Register” on page 241.).
The DMA controller operates on linked lists with pointer information stored in the DSCC4
internal configuration section.
The initi alization proc edure as wel l as the CPU to DSCC4 hands haking is e qual to the
HOLD-Bit control mode as described in Chapter 5.1.2.3 with the following exception:
The host CPU does not tak e care abou t the HOLD bit. The ad dress of the descript or in
the chain at which a ’Hold’ is to be exercised is written to the corresponding LTDA/LRDA
register. The DMA channel compares its current (first) descriptor address to
LTDA/LRDA. When a match occures, a ’Hold’ condition is activated. After attaching at
least one n ew des crip tor to the linked list. Also the DMA ch ann el d oes not ta ke care o n
the HOLD bit within the descriptors but compares its current descriptor address with
LTDA/LRDA register value. In case of address match this condition is equal to the
HOLD-condition.
Figure 17 Data Transfer controlled via first and last descriptor addresses
After initialization the DMAC internally starts with the Base Tx/Rx Descriptor Address
BTDA/BRDA as the “first descriptor address“ since this address points to the first
CHiLTDA
CHiBTDA
CHiFTDA*)
CHiLRDA
CHiBRDA
CHiFRDA**)
*) FTDA starts with BTDA and is updated by the DSCC4 until FTDA=LTDA
**) FRDA starts with BRDA and is updated by the DSCC4 until FRDA=LRDA
Shared MemoryDSCC4 Register
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 79 Data Sheet 09.98
descriptor of the linke d list that has to be pro cessed by the DMA channel. The current
descriptor address is stored in register FTDA/FRDA as the (current) first descriptor
address. With branching to the next descriptor the first descriptor address register is
updated. Thus the host can keep track of the DMAC’s progress by reading the
FTDA/FRDA register. Beside the base descriptor address the user provides the Last
Tx/Rx Descriptor Address in the corresponding LTDA/LRDA register. After transferring
all data from the current data section and before branching to the next descriptor, the
DMAC compares FTDA/FRDA and LTDA/LRDA. If the corresponding addresses are not
identical the DMA channel branches to the next descriptor and the FTDA/FRDA register
is updated with “next descriptor address“ to continue normal transfer operation. If a
match occurs the DMA channel is suspended until the host writes a new last
transmit/rece ive descripto r address to LTDA /LRDA registe r. After write acce ss to these
registers the D SC C4, again, compares FTD A/FRD A an d LTD A/LR D A and proce eds as
described above.
• In transmit direction the condition LTDA equal to FTDA corresponds to the transmit
HOLD c ondition in H OLD b it con troll ed mod e. In this c ase u pdating registe r L TDA is
the equivalent operation to transmit poll request command.
• In receive direction the condition LRDA equal to FRDA corresponds to the receive
HOLD condition in HO LD bit co ntrol led mode . Updati ng regi ste r LRDA will cause the
DMA channel branching to the next receive descriptor. No new initialization command
’IDR’ is necessary as in HOLD bit controlled mode.
Re-initialization might be necessary if the linked list does not continue at the "next
descriptor address" of the last descriptor but on any other address.
There are three methods to handle the situation in which the next descriptor address
pointer is not yet va lid in the current trans mit descri ptor. This happen s if the location of
the next descriptor is not known when preparing the last descriptor of the chain:
1. Programming the "next descriptor address" of each last descriptor pointing to a newly
prepared descriptor list before updating register LTDA/LRDA. After write access to
LTDA/LRDA register, the DMA controller reads the next descriptor address again and
branches to the new descriptor.
2. Updating BTDA/BRDA address in register CHiBTDA/CHiBRDA followed by a channel
initialization command (’IDT’/’IDR’).
3. Programm ing the "next descri ptor address" of ea ch las t d esc ripto r po inti ng to a fixed
dummy descriptor with byte number 0 (NO = 0). The next descriptor address of the
dummy descriptor must be programmed to point to a newly prepared descriptor list
before updating register LTDA/LRDA. The DMA channel will branch to the dummy
descriptor, completing it without data transfer, then branch to the next descriptor which
itself might point to the dummy descriptor again.
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 80 Data Sheet 09.98
.
Figure 18 Example: ’Chain Jump’ Handling per ’Dummy Descriptor’
LTDA/LRDA
1
BTDA/BRDA
first prepared list:
second added list:
third added list:
LTDA/LRDA
2
LTDA/LRDA
3
2
3
dummy descriptor,
(NO = 0, FE = 0)
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 81 Data Sheet 09.98
5.1.3 DMAC In ter rupt Controller
The DSCC4 interrupt concept is based on 32-bit interrupt vectors generated by the
different blocks. Interrupt vectors are stored in a central interrupt FIFO which is 16
DWORDs deep. The interrupt controller transfers available vectors to one of ten circular
interrupt queues located in the shared memory depending on the source ID of each
interrupt vector.
In addit ion new in terrupt vec tors are i ndi cate d in the glo bal status regist er GS TAR on a
per queue basis and selectively confirmed by writing ’1’ to the corresponding GSTAR bit
positions. The PCI interrupt signal INTA is asserted with any new interrupt event and
remains asserted until all events are confirmed.
Each interrupt queue length and memory location can be configured via specific interrupt
queue base address registers and two shared interrupt queue length registers. The
queue length is individually programmable in multiples of 32 DWORDs (see page 247).
One dedicated interrupt queue is provided per SCC channel and directio n (IQSCCiRX
and IQSCCiTX). Non channel specific interrupt vectors generated by the DMAC itself are
transferred to the configuration queue IQCFG. The peripheral interrupt queue IQP is
used for vectors generated by one of the blocks GPP, SSC or LBI.
The internal blocks provide mask registers for suppressing interrupt indications. Masked
interrupts will neither generate an interrupt vector nor an INTA signal and GSTAR
indication. (Refer to figure “DSCC4 Logical Interrupt Structure” on page 83.)
The DMA controller (interrupt controller) itself generates 6 channel specific interrupts
regarding the transmit and receive descriptor handling:
• Host Initiated interrupt (HI):
This interrupt can be forced by setting bit ’HI’ in the receive and transmit descriptor. In
this case the DMAC will generate an HI-interrupt with completion of this descriptor i.e.
when the DMAC is ready to branch to the next descriptor address. This might be used
to monitor th e progress of the corresponding DM A channel on the des criptor list. As
an example the HI interrupt can be used to dynamically request attachment of new
receive descriptors to the list if the DMA channel comes close to the list end.
• Frame Indication interrupt (FI):
This interrupt is generated with completion of any receive and transmit descriptor with
a set ’frame end/block end’ indication, i.e. FE=’1’.
• Error interrupt (ERR):
Indicates an unexpected descriptor configuration.
receive descriptor:
ERR is generated if receive data cannot be transferred to the shared memory
complete ly becau se the frame (bloc k) does not fit int o the current data section and a
HOLD condition (HOLD bit or LRDA=FRDA) prevents the DMAC from branching to
the next descript or.
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 82 Data Sheet 09.98
ERR is also generated if an already started DMA transfer is aborted by a receive DMA
reset (RDR) command.
transmit descriptor:
In transmit direction an ERR interrupt is generated if one of the following descriptor
settings is detected
- HOLD=’1’ and FE=’0’ (the already started transmit frame could not be finished)
- LTDA=FTDA and FE=’0’ (the already started transmit frame could not be finished)
- FE=’0’ and NO=’0’ (a packet of length 0 is supposed to be a ’frame’ with FE bit set)
The DMA controller will continue ’normal’ operation in case of an ERR event.
Nevertheless these cases may result in receive data overflows or transmit data
underruns.
FI and HI interrupt indications caused by one descriptor will be generated into one
interrupt vector with ’HI’ and ’FI’ bit set.
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 83 Data Sheet 09.98
Figure 19 DSCC4 Logical Interrupt Structure
According the interrupt service routine (ISR) two different solutions (or any mix of them)
are possible:
• Traditional ISR entry on interrupt event (INTA signal asserted):
On interrup t event the CP U jumps to t he ISR entry po int. Within the ISR the DSC C4
global sta tus register GSTAR is read indicat ing which interrupt queues st ore at least
one new interrupt vector. The interrupt indication is confirmed by writing a ’1’ to the
SCC0
receive
interrupts transmit
interrupts
IQSCC0RX
IQSCC0TX
Peripherals
(SSC, GPP,
LBI)
IQCFG
IQP
DMA
Controller
Logic
HOST Memory interrupt queues
DSCC4 interrupt structure block diagram
internal interrupt bus
16
DWORD
central
interrupt
FIFO GSTAR register
INTA
signal
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 84 Data Sheet 09.98
queue spec ific bit positio ns in regi ster GSTAR and all ne w vec tors are read from th e
correspo nding inte rrupt que ues af terwards . It is poss ible to confirm all i ndicat ions by
one register access writing the value 0xFFFFFFFFH to GSTAR.
• Interrupt queue polling:
The interrup t que ues a re polled periodically for new vec tors . N ei the r IN TA si gnal nor
register GSTAR is evaluated.
The two solutions can be mixed. Queues with high interrupt activity might be serviced by
periodic queue polling and other queues on INTA and GSTAR indication only. In this
case the ISR only takes care on some GSTAR bits but must confirm all indications to
reset INTA indication.
Note that any interrupt event will cause an INTA and GSTAR indication.
To distinguish between old and new interrupt vectors in the circular queues, vectors can
be overwritten by software with all zeros. Because the number of new vectors is not
known, a queue read pointe r can be incremented as long as no new vector i s found at
this location.
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 85 Data Sheet 09.98
5.2 Central FIFOs Operational Description
The large central transmit and receive FIFOs are a major factor according to the system
performance depending on the serial lines and the system bus data rates and latencies.
Programmable thresholds and partitioning of the transmit FIFO alow optimized adaption
to the needs of any application.
5.2.1 Central FIFO Register Ove rvi ew
The following table provides an overview about all central FIFO related registers. For
detailed register description refer to Chapter 10.
Table 16 Central FIFO Control Registers
Offset
Addr. Access
Type Controlled
by Reset
Value Register Name
0044Hr/w CPU 00000000HFIFOCR1:
FIFO Control Register 1
Bit-Fields
Pos. Name Default Description
31..27,
26..22,
21..17,
15..11
TFSIZE0
TFSIZE1
TFSIZE2
TFSIZE3
0,
0,
0,
0
Transmit FIFO Section Size i (i=0..3)
Transmit FIFO Section Size for the
corresponding channel i in multiples of 4
DWORDs.
Note: The entire size of all FIFO parts
must not exceed 128 DWORDs. If
the complete FIFO is assigned to
only one channel the maximum
size is limited to 4*31=124
DWORDs.
Note: The minimum FIFO section size
for active channels is 4 DWORDs
which means a “1” programmed to
the respective TFSIZEi bit field.
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 86 Data Sheet 09.98
0048Hr/w CPU 00000000HFIFOCR2:
FIFO Control Register 2
Bit-Fields
Pos. Name Default Description
31..27,
26..22,
21..17,
15..11
TFR
THRESH0,
TFR
THRESH1,
TFR
THRESH2,
TFR
THRESH3,
0,
0,
0,
0
Transmit FIFO Refill Threshold i (i=0..3)
Transmit FIFO Refill Threshold for the
corresponding channel i in number of
DWORDs multiplied by its respective
multiplier Mx_i.
This thresho ld controls DMAC operatio n
towards the Host memory. A watermark
is calculated by:
watermark = TFRTHRESHi*Mx_i+1.
As soon as the number of valid data in
the transmit FIFO section is less than the
watermark, the DMA controller requests
for new data from shared memory.
7,5,3,1 M4_i 0 Multiplier 4
6,4,2,0 M2_i 0 Multiplier 2
Table 16 Central FIFO Control Registers (cont’d)
Offset
Addr. Access
Type Controlled
by Reset
Value Register Name
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 87 Data Sheet 09.98
004CHr/w CPU 00000000HFIFOCR3:
FIFO Control Register 3
Bit-Fields
Pos. Name Default Description
6..0 RF
THRESH 0 Receive FIFO Threshold
Receive FIFO Threshold in number of
DWORDs. A watermark is calculated by:
watermark = RFTHRESH*M_x.
When more data than specified by this
watermark is available in the receive
FIFO the DMA c ontroller is re quested to
transfer data to the channel specific data
buffers in host memory until the receive
FIFO is empty.
8 M4 0 Multiplie r 4
7 M2 0 Multiplie r 2
Table 16 Central FIFO Control Registers (cont’d)
Offset
Addr. Access
Type Controlled
by Reset
Value Register Name
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 88 Data Sheet 09.98
0034Hr/w CPU 00000000HFIFOCR4:
FIFO Control Register 4
Bit-Fields
Pos. Name Default Description
31..24,
23..16,
15..8,
7..0
TFF
THRESH0,
TFF
THRESH1,
TFF
THRESH2,
TFF
THRESH3,
0,
0,
0,
0
Transmit FIFO Forward Threshold i
(i=0..3)
Transmit FIFO Forward Threshold for
corresponding channel i in DWORDs. A
waterm ark is calc ulated by:
watermark = TFFTHRESHi
As soon as the number of valid data
belongi ng to a ne w frame in the tran smi t
FIFO is greater than the watermark, the
DMAC will provide transmit data to the
corresponding SCC. Once having
started a frame the DMAC will ignore this
threshold providing all available data to
the SCC. Threshold operation starts
again with the beginning of a new frame.
Frames shorter than the threshold will be
transferred as soon as a frame end
indication is detected by the DMAC.
Note: The maximum allowed Transmit
FIFO Forward Threshold is:
TFFTHRESHi = (TFSIZEi * 4) - 1
Note: Programming TFFTHRESHi to
zero will disable the threshold
causing the DMAC to transfer all
data immediately. This may be
useful with non packet oriented
data e.g. in ASYNC protocol
mode.
Table 16 Central FIFO Control Registers (cont’d)
Offset
Addr. Access
Type Controlled
by Reset
Value Register Name
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 89 Data Sheet 09.98
5.2.2 Central Transmit FIFO (TFIFO)
The central transmit FIFO can be partitioned in 4 sections with 128 DWORDs in total.
Each section size can be programmed according to the needs of the corresponding
serial port via register FIFOCR1. Criteria for partitioning are serial line speed (nominal
data rate) and type of traffic (bursty or constant). The software has to ensure that the sum
of section sizes does not exceed the 128 DWORDs total limit. One channel can consume
only 124 DWORDs of the central transmit FIFO.
Two thresholds per TFIFO section are provided to optimize TFIFO operation to the serial
side as well as the system interface side.
These thresholds have conflicting requirements:
1. Minimizing transmit data underrun probability in case of PCI bus latencies (especially
for high speed ports).
2. Reducing bus utilization by making maximum PCI burst transfers possible for loading
of transmit data.
As a naming convention Transmit FIFO always means the Transmit FIFO section of the
dedicated channel. All considerations apply to one transmit channel.
Requirement 1) is controlled by the Transmit FIFO forward threshold (register
FIFOCR4). Transmit data is transferred to its SCC if one of the following conditions is
true:
• A complete packet is stored in the TFIFO (Frame End (FE) indication detected). In this
case data transfer to the SCC will start although the frame is smaller than the forward
watermark.
• The TFIFO is filled beyond the forward threshold.
These two con diti ons are che cke d again after every trans fer of a complete frame to the
SCC.
Consider a small frame is stored in the transmit FIFO and the beginning of a second
frame, but the total amount of data is smaller than the forward threshold (DMAC
operation may be delayed by bus latency now). The TFIFO will start transferring data to
the SCC because a frame end (FE) indication is detected. After transfer of the complete
first frame the above two conditions are checked again. Now there is no further FE
indication in the TFIFO an d t he forw ard thre sho ld is not exceede d. Thus the TFIFO will
not start transferring the second frame until additional data is loaded.
Requirement 2) is controlled by the Transmit FIFO refill threshold (register FIFOCR2). In
case of an emp ty transmit FIFO the D MAC always tries to fill the complete TFIFO w ith
data. If a TFI FO full c onditio n occurs , the DMAC ge ts stopp ed. The DMAC start s again
if the TFIFO fill level falls below the refi ll thre sho ld and tries to get new data via the PCI
bus until the TFIFO is filled again.
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 90 Data Sheet 09.98
Figure 20 Central Transmit FIFO Section Thresholds
For all operations the burst size is limited to the available TFIFO space or to 15 DWORDs
maximum (4 DWORDs in de-multiplexed interface mode).
To guarantee maximum burst length the refill threshold should be programmed such that
(TFIFO size - refill watermark) is equal to 15 DWORDs. A lower refill threshold is no
guarentee fo r limited burst size becau se due to bus latency the available TFI FO space
might be greater when transfer starts thus allowing full burst. So 15 DWORDs burst can
happen even if the (TFIFO size - refill watermark) < 15 DWORDs.
A refill thre shold such that (TFIFO size - refil l watermark) is greater than 15 DWORDs
(the maximum burst size) does not make sense.
Important:
The following condition must be met:
refill watermark > forward watermark
Consider the following situation:
A short frame is loaded and the beginning of a second frame such that the transmit FIFO
is full and DMA operation stops.
Transmit FIFO Forward Threshold:
FIFOCR4.TFFTHRESHi
watermark = TFFTHRESHi
SCCi
transmit FIFO
Transmit FIFO Refill Threshold:
FIFOCR2.TFFTHRESHi
watermark = TFRTHRESH*Mi_x+1
Transmit data transferred
from shared memory
1
TFSIZEi*4
Transmit FIFO Section Size i:
FIFOCR1.TFSIZEi
size = TFSIZEi * 4
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 91 Data Sheet 09.98
After having transferred the first frame to the SCC, the remaining TFIFO fill level is
between the two watermarks. Now DMAC operation does NOT start if the refill
watermark is not suffused and transfer to the SCC does NOT start due to neither a frame
end indication is detected nor the forward watermark is exceeded.
This is a deadlock condition!
Note: The DMAC evaluates the 2 conditions for transfer to SCC with the beginning of
each new frame.
Note: The small SC C specific FIFOs always request for da ta if space i s available ev en
if the SCC is in power down condition. If the forward threshold condition is met,
data is transferred into the SCC FIFO immediately (6 DWORDs because transfer
to the 2 DWORD shadow FIFO does not happen in power-down).
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 92 Data Sheet 09.98
5.2.3 Central Receive FIFO (RFIFO)
The central receive FIFO is sh ared by all 4 re ceive channel s a nd dynamicall y a llo cate d
to the SCCs. One central receive FIFO threshold controls data transfer to the host
memory.
Figure 21 Central Receive FIFO Threshold
Mention, that the SCC s ty pic all y fo rwa rd rec eiv e da ta in portions of 15 D WO RDs to th e
central receive FIFO. These coherent data portions are a precondition for PCI burst
transactions.
SCC0
receive FIFO
Receive FIFO Threshold:
FIFOCR3.RFTHRESH
watermark = TFRTHRESH*Mx
Receive data transferred
to shared memory
1
128
128 DWORDs
0123
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 93 Data Sheet 09.98
5.2.4 DMAC Internal Arbitration Scheme
System Interface Arbitration (DMA Arbiter):
Once the DMA controller has been granted PCI bus access, it will attempt to transfer all
of its available data in o ne bus access. However the PCI acce ss time is limited by the
PCI bus protocol (bus arbitration and latency timer operation).
There are three priority groups as defined below:
Within each group, the group priority is shared between the channels by the round robbin
rules. After each bus access the priority is re-computed for the next access based on the
round robbi n rule s. The DM A arbiter step s through th ree groups of priority in case of at
least one DMA request per group is pending.
Priority groupings:
1. The DMA channel performing the interrupt vector transfer has the highest (first)
priority to prevent over-runs and loss of interrupt vectors.
2. The DMA channels performing the data transfer in receive direction have the second
priority (receiver group).
3. The DMA channels performing the data transfer in transmit direction have the third
priority (transmitter group).
The sub-priority of the DMA channels within the receiver or transmitter group is the same
or one channel is treated as ’high priority channel’.
This exclusive priority can be enabled via bit ’SPRI’ in register GMODE whereas the
dedicated channel can be selected via bit field ’PCH’ in register GMODE. This setting
also effects the internal SCC arbitration.
Internal SCC Arbitration (SCC Arbiter):
Two independent arbiters control the service of the 4 SCC transmit and the 4 SCC
receive channels. Each arbiter either works following the round robbin scheme or
provides ’high priority’ to one dedicated channel and services the remaining channels in
a second priority group using round robbin.
This exclusive priority can be enabled via bit ’SPRI’ in register GMODE whereas the
dedicated channel can be selected via bit field ’PCH’ in register GMODE.
The high priority option is useful if one SCC is configured in high speed mode (up to
52 MBit/s ) and the o thers are op erating at s low data rate s below 10 or 2 MBit/s. In this
case data transfer from and to the corresponding SCC FIFOs as well as data transfer
from and to host memory is preferred for the high speed transmit/receive channels. In all
other cases ’round robbin’ within the described priority groups provide a balanced
arbitration solution.
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 94 Data Sheet 09.98
5.2.5 DMAC Performance
The DMAC sup ports li near b ursting of multip le DWO RDs for tra nsfers o f both da ta and
descriptors:
• On descriptor reads: 3 (Tx) / 3 (Rx) DWORDs at a time.
• On data write/read transactions for full scatter/gather sections up to 15 DWORDs at a
time.
The burst size for transmit data is determined by:
• the central transmit FIFO partition and threshold configurations (channel specific)
• the microprocessor interface arbitration and latency
The burst size for receive data is determined by:
• the central receive FIFO threshold configuration
• the SCC receive FIFO threshold
(Only DWORDs stored in consecutive sub-sections of the central receive FIFO can be
transferred to the host memory by one burst transfer. If the SCC receive FIFO
threshold is below 15 DWORDs (60 bytes) the consecutive sub-sections in the central
receive FIFO might be smaller as well depending on other channels activity. In this
case burst transfers of receive data might be limited to the SCC receive threshold
value.)
• the microprocessor interface arbitration and latency
In PCI mode up to 15 DWORDs bursts are supported. In de-multiplexed bus interface
mode the burst size is limited to 4 DWORDs or burst transactions are suppressed at all.
The DSCC4 DMAC uses PCI (fast) back-to-back transfers to achieve a maximum
throughput within one bus arbitration phase.
PEB 20534
DMA Controller and Central FIFOs
Semiconductor Group 95 Data Sheet 09.98
5.2.6 Little / Big Endian Byte Swap Convention
The DSCC4 operates per default as a little endian device. To support integration into big
endian environments the DMAC provides an internal byte swapping mechanism which
can be enabled via bit ’ENDIAN’ in register GMODE.
The big endian byte swapping applies only to the DATA sections of the receive and
transmit descriptor lists in the shared memory.
Data sections might be prepared/evaluated by software using byte pointer operations in
the shared memory. The DSCC4 will access these data sections by DWORD read/write
transfers. In case of a big endian CPU but little endian DSCC4 mode this will result in
wrong byte orders on the serial line and in the receive data sections.
Bad/Good case example:
Figure 22 Little/Big Endian Byte Swapping
All other me mory s truct ures (de scriptors , inte rrupt vectors ) as w ell a s DSCC 4 regis ters
are organized DWORD-wise and should be operated by software using 32bit operations
only. Therefore no little/big endian ordering mismatches can occur on these structures.
However byte ordering in the local memory, as it appears to the PCI bus view, depends
on the local bus (CPU, memory, PCI bridge) realization.
CPU
(big endian)
...
BYTE *txdatabuffer;
txdatabuffer = 0x10001000;
...
for (i=0; i<9, i++)
{
*txdatabuffer = (BYTE) i;
txdatabuffer++;
}
...
0x10001000
Address:
0x10001004
0x10001008
...
...
0x00 0x01 0x02 0x03
0x04 0x05 0x06 0x07
0x08 0xXX 0xXX 0xXX
RAM
MSB
31 LSB
0
DSCC4
FLAG
0x03
0x02
0x01
0x00
0x07
0x06
0x05
0x04
0xXX
CRC
CRC
FLAG
serial transmit line,
HDLC mode assumed
time
(little endian mode:
GMODE.ENDIAN = ’0’) (big endian mode:
GMODE.ENDIAN = ’1’)
FLAG
0x03
0x02
0x01
0x00
0x07
0x06
0x05
0x04
0x08
CRC
CRC
FLAG
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 96 Data Sheet 09.98
6 Multi Function Port (MFP)
The Multi Function Port consists of a set of I/O signal pins and three internal function
blocks sharing these signal pins.
The function blocks are:
• Local Bus Interfa ce (L BI)
• Synchronous Serial Communication Interface (SSC)
• General Purpose I/O Port (GPP)
The MFP is only available in PCI bus operation mode, because in de-multiplexed bus
mode, all MFP I/O signals are used for the 32-bit address bus extension.
Various configurations allow simultaneous operation of the three function blocks:
Figure 23 MFP Configurations Overview
The configuration is selected via bit field ’PERCFG’ in register GMODE. The following
table shows all supported ’PERCFG’ bit field options (also refer to “GMODE: Global
Mode Register” on page 241):
LBI
(EBC) SSCGPP
MFP
Peripheral Block
12
LBI Ctrl. 16
LD(15:0) 8
LA(15:8),
GP(15:8)
8
LA(7:0),
GP(7:0),
SSC
LBI Ctrl. LD(15:0) LA(15:8) LA(7:0)
LBI Ctrl. LD(15:0) GP(15:8) LA(7:0)
LBI Ctrl. LD/LA(15:0) GP(15:8) SSC
LBI Ctrl. LD/LA(15:0) GP(15:8) GP(7:0)
Configurations:
LBI only
LBI + 8 bit GPP
LBI + 8 bit GPP + SSC
LBI + 16 bit GPP
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 97 Data Sheet 09.98
Table 17 MFP Configuration via GMODE Register, Bit Field ’PERCFG’:
The follo wing sub-ch apters desc ribe the th ree periphe ral funct ion blocks LBI, SSC and
GPP:
PERCFG
(2:0) Peripheral Block Configuration
The peripheral block basically consists of the functions
• Local Bus Interface (LBI)
• General Purpose Port (GPP)
• Synchronous Serial Controller (SSC)
which can be operated in various combinations/configurations.
Bit field ’PERCFG’ selects the peripheral configuration and switches the
multiplex ed sig nal pins accord ing ly:
PCI Interface Mode (DEMUX pin connected to VSS):
PERCFG Signal Pin Groups
(2:0) 109, 108 119..112 143..135
101..96 128..123
’000’ LA(15..8) LA(7..0) LD(15..0)
’001’
Reserved
. Do not use.
’010’ GP(15..8) LA(7..0) LD(15..0)
’011’ GP(15..8) SSC LD(15..0)
’100’ GP(15..8) GP(7..0) LAD(15..0)
’101’,
’110’,
’111’
Reserved
. Do not use.
DEMUX Interface Mode (DEMUX pin connected to VDD3):
Bit field ’PERCFG’ is not valid. All 32 multiplexed signals are used as
DEMUX address bus A(31:0):
PERCFG Signal Pin Groups
(2:0) 109, 108 119..112 143..135
101..96 128..123
’xxx’ A(15..8) A(7..0) A(31..0)
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 98 Data Sheet 09.98
6.1 Local Bus Interface (LBI)
The DSCC4 provides capability for the PCI host system to access peripherals connected
to the Local Bus Interface (LBI).
Note: The LBI is only available when the DSCC4 is configured for the PCI bus interface
mode. When in de-multiplexed bus interface mode, the LBI address and data pins
interface to the host system address bus.
Overview of Transactions:
Standard PCI Slave transactions are used when the PCI host system communicates with
non-intelligent LBI peripherals (single address read/write operations).
For reads, a PCI Retry sequence of operations is performed, in which the DSCC4 will
immediately terminate the PCI transaction (and request a retry) until it terminates the
transaction to the LBI. The DSCC4 uses the retry procedure because the time to
complete the data phase will require more than the maximum allowed 16 PCI clocks
(from the assertion of FRAME to the completion of the first data phase). Data transfer
will be successfully completed within a PCI retry cycle. The number of necessary PCI
retry cycles depend on PCI arbitration behavior and the time it needs to terminate the
transactio n on the local bus; PCI TRDY wa it states will no t be added for the seque ntial
retry read cycles unless the LBI arbitration time is excessive.
For write transactions, the DSCC4 will store a single data DWORD and then immediately
terminate the PCI transaction successfully. It will then arbitrate the local bus and perform
the write transaction after being granted, depending on the selected number of wait
states and LRDY bus control signal.
Thus write accesses to LBI are performed as ‘posted write’ transactions from the PCI
view. A consecutive write transaction results in PCI retry cycles in the case that the
preceding write transaction is not yet finished on the local bus.
Note: Note that the DSCC4 performs sing le word PCI Slave read or write transactions
only; Slave burst transactions to LBI are not supported.
Table 18 LBI Peripheral Transaction Options
Peripheral Type Corresponding
PCI Transaction Read Write
Non-Intelligent Slave PCI Retry operation PCI posted write
operation
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 99 Data Sheet 09.98
6.1.1 LBI Bus Modes
6.1.1.1 LBI External Bus Controller (EBC)
The External Bus Controller (EBC) provides a flexible bus interface to connect a wide
range of peripherals. In normal mode this interface is a bus master (and default bus
owner) and drives peripheral devices. It provides the ability to select busses of different
configuration: 8 bit multiplexed/de-multiplexed or 16 bit multiplexed/de-multiplexed. The
configurable pins of address signals/general purpose signals provide alternate
functionality to support the LBI pins with additional I/O signals.
The EBC also supports bus arbitration supporting other bus masters connected to the
local bus . I n this case , th e DSC C 4 can be confi gured as arbitra tion m aste r (de fau lt b us
owner) or arbitration slave
The function of the EBC is control led via th e global mode regis ter GMODE and the LBI
Configuration register LCONF. It specifies the external bus cycles in terms of address
(multiplexed/de-multiplexed), data (16-bit/8-bit) and control signal length (wait states).
6.1.1.2 Multiplexed Local Bus Modes
In the 16-bit multiplexed bus mode both the address and data lines use the pins
LD(15:0). The address is time-multiplexed with the data and has to be latched externally.
The width of the required latch depends on the selected data bus width, i.e. an 8-bit data
bus require s an 8- bit latch (the address bits LD15.. .LD8 on th e LBI port do not chan ge,
while on LD7...LD0 address and data signals are multiplexed), a 16-bit data bus requires
a 16-bit latch (the least significant address line LA0 is not relevant for word accesses).
In de-multiplexed mode, the address lines are permanently output on pins LA(15:0) and
do not requir e latches.
The EBC initiates an external access by generating the Address Latch Enable signal
(LALE) and then placing an address on the bus. The falling edge of LALE triggers an
external la tch to captu re the add ress. After a p eriod of time in which t he address m ust
have been latched externally, the address is removed from the bus. The EBC now
activates the respective command signal (LRD, LWR, LRDY) and data is driven onto the
bus either by the EBC (for wri te cycles) or by t he external memo ry/peripheral (fo r read
cycles). After a period of time, which is determined by the access time of the
memory/peripheral, data becomes valid.
Read cycles: Input data is latched and the command signal is now deactivated. This
causes the accessed device to remove its data from the bus which is then tri-stated
again.
Write cycles: The command signal is now deactivated. The data remain valid on the bus
until the next external bus cycle is started.
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 100 Data Sheet 09.98
Figure 24 Multiplexed Bus Cycle
6.1.1.3 De-multiplexed Local Bus Modes
The de-multiplexed bus modes use the LBI port pins LA(15:0) for the 16-bit address and
the LBI port pins LD(15:0) for 8/16-bit data. The EBC initiates an external access by
placing an address on the address bus. The EBC then activates the respective
command signal (LRD, LWR, LBHE). Data is driven onto the data bus either by the EBC
(for write cycles) or by the external memory/peripheral (for read cycles). After a period of
time, which is determined by the access time of the memory/peripheral, data become
valid.
Read cycles: Input data is latched and the command signal is now deactivated. This
causes the accessed device to remove its data from the data bus which is then tri-stated
again.
Write cycles: The command signal is now deactivated. If a subsequent external bus cycle
is required, the EBC places the respective address on the address bus. The data remain
valid on the bus until the next external bus cycle is started.
ITD10611
LALE
Data IN
LRD
Data OUT
LWR
Address Data
Address Data
Extended ALE
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 101 Data Sheet 09.98
Figure 25 De-multiplexed Bus Cycle
ITD10612
Address
LALE
Data IN
LRD
Data OUT
LWR
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 102 Data Sheet 09.98
6.1.1.4 Programmab le Bus Characteristics
External Data Bus Width:
The EBC can operate on 8-bit or 16-bit wide external memory/peripherals. A 16-bit data
bus uses the LBI port pins LD(15:0), while an 8-bit data bus only uses LD(7:0). This
saves bus transceivers, bus routi ng and memory cost s at the expense of trans fer time.
The EBC can control byte accesses on a 16-bit data bus.
Byte accesses on a 16-bit data bus require that the upper and lower half of the memory
can be accessed individually. In this case the upper byte is selected with the LBHE
signal, while the lower byte is selected with the LA0 signal. The two bytes of the memory
can therefore be enabled independent from each other (or together when accessing
words).
Device s such as th e ESCC2 also provid e a BHE inp ut and henc e allow byte ac cesses
in 16-bit bus mode.
When reading bytes from an external 16-bit device, whole words may be read and the
EBC automatically selects the byte to be input and discards the other.
However, it mu st be taken care on devices, whic h change thei r state when being rea d,
e.g. FIFOs, interrupt status registers, etc. In this case individual bytes should be selected
using BHE and LA0.
Switching between the Bus Modes:
The EBC bus type can be switched dynamically by software between different read/write
bus transactions. However, the connected peripherals must support the selected bus
type(s) (multiplexed mode or de-multiplexed mode).
Arbitration master/slave mode according the local bus handled dynamically by the
device itself.
Programmable bus timing characteristics:
Important timing characteristics of the external bus interface have been made user
programmable to adapt to the needs of a wide range of different external bus and
memory configurations with different types of memories and/or peripherals.
The following parameters of an external bus cycle are programmable:
•Memory Cycle Time (extendable wit h 1...15 wait s tates) defines the minimum access
time by the width of LRD/LWR strobe signals;
•LRDY Control Signal extends the transcation controlled by the target peripheral.
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 103 Data Sheet 09.98
Figure 26 Memory Cycle Time
The external bus cycles of the EBC can be extended for a memory or peripheral, which
cannot keep pace with the EBC’s maximum speed, by introducing wait states during the
access (see figure above).
The minimum LRD/LWR strobe active length is 2 LBI clock cycles (with zero MCTC wait
states).
The memory cycle time wait states can be programmed in increments of one EBC
system clock (LCLKO) within a range from 0 to 15 (default value after reset) via the
’MCTC’ bit field in the LBI Configuration register LCONF. (15-<MCTC>) waitstates will
be inserted.
ITD10613
LALE
Data IN
LRD
Data OUT
LWR
Address Data
Address Data
MCTC Wait States (1...15)
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 104 Data Sheet 09.98
6.1.1.5 Ready Signal Controlled Bus Cycles
For situations, where the programmable constant number of 15 wait states is not
enough, or where the response (access) time of a peripheral is not constant, the DSCC4
EBC interface provides external bus cycles that are terminated via a LRDY input signal.
In this case the EBC fi rst ins erts a prog ramm able number of waitsta tes (0...7) and then
monitors the LRDY line to determine the actual end of the current bus cycle. The external
device dri ves LRDY low in order to indicate that data have been latched (write cycle) or
are available (read cycle).
Figure 27 LRDY Controlled Bus Cycles
The LR DY function is enabled via bit ’RDYEN’ in the LBI Configuration register LCONF.
The LRDY sign al is alwa ys syn chronized at the input port p in. An asy nchronous L RDY
signal that has been activated by an external device may be deactivated in response to
the trailing (rising) edge of the respective command (LRD or LWR).
Combining the LRDY function with predefined waitstates is advantageous in two cases.
Memory components with a fixed access time and peripherals operating with LRDY may
be grouped into the same address window. The (external) wait states control logic in this
case would activate LRDY either upon the memory’s chip select or with the peripheral’s
LRDY output. After the predefined number of waitstates the EBC will check its LRDY
line to det ermine the end of the bus cycl e. For a memory acces s it will be low al ready,
for a peripheral access it may be delayed. As memories tend to be faster than
peripherals, there should be no impact on system performance.
When using the LRDY function with ’normally-ready’ peripherals, it may lead to
erroneous bus cycles, if the LRDY line is sampled too early. These peripherals pull their
LRDY output low, while they are idle. When they are accessed, they deactivate L RDY
ITD10614
LALE
LRD/LWR
Data OUT
LRDY
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 105 Data Sheet 09.98
until the bus cycle is complete, then drive it low again. By inserting predefined waitstates
the first LRDY sample point can be shifted to a time, where the peripheral has safely
controlled the LRDY line (e.g. after 2 waitstates, see figure 27 and 28).
Figure 28 LRDY Timing
LCLKO
LD[15..0]
(read cycle)
LRD, LWR
LRDY
LD[15..0]
(write cycle)
2 cycles fix MCTC
first LRDY
evaluation last LRDY
evaluation
(LRDY active)
transaction
termination
Notes:
- MCTC wait state configuration is assumed to one cycle in this figure
- LRDY is evaluated the first time with the clock cycle following the MCTC related wait states
- Transaction is terminated one clock cycle after detecting LRDY active
- LRDY is evaluated only if LRDY-control is enabled
65 66
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 106 Data Sheet 09.98
6.1.1.6 L BI (EBC ) Configuration
The propertie s of a bus cycle u sage of s ignal LR DY, externa l bus mod e an d waitstate s
are controll ed by LBI Configura tion register LCO NF and g lobal mo de regist er GMOD E.
This allows the use of memory components or peripherals with different interfaces within
the same system, while optimizing accesses to each of them.
LCONF is describe d in “LBI Registers Description” on page 351.
EBC Idle State:
Upon reset the MFP comes up in 16-bit de-multiplexed bus LBI mode. The EBC can then
be programmed to be arbitration master or slave by software.
When the EBC bus interface is enabled in arbitration master mode, but no external
access is currently executed, the EBC is idle. During this idle state the external interface
behaves in the following way:
Data port LD(15:0) switches in high impedance state (floating);
Address port LA(15:0) drives the last used address value;
LRD/LWR signals remain inactive (high).
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 107 Data Sheet 09.98
6.1.2 LBI Bus Arbitration
In high performance systems it may be efficient to share external resources like memory
banks or periphe ral device s among mo re than one bus controlle r. The LBI’s EBC block
supports this approach with the possibility to arbitrate the access to its external bus, i.e.
to the external devices.
This bus arbi tration allows a n external master to req ues t th e EBC’ s b us vi a the LH O LD
input signal . The EBC acknowl edges this request via the LHLDA output sign al and will
float its bus signals in this case. The new master may now access the peripheral devices
or memory banks via the same interface lines as the EBC. During this time the DSCC4
can keep on executing internal processes, as long as it does not need access to the
external bus.
When the EBC needs access to its external bus while it is occupied by another bus
master, the bus is requested via the LBREQ output signal.
The external bus arbitration is enabled by setting bit ’HLDEN’ in the LBI Configuration
register LCONF to ‘1’. This bit is allowed to be cleared by software during the execution
of program sequences, where the external resources are required but cannot be shared
with other bus masters. In this case the EBC will not answer to LHOLD requests from
other external masters.
The pins LHOLD, LHLDA and LBREQ keep their function (bus arbitration) even after the
arbitration mechanism has been switched off by clearing bit ’HLDEN’.
All three pins are used for bus arbitration after bit ’HLDEN’ was set once.
Entering the Hold State:
Access to the EBC’s external bus is requested by driving its LHOLD input low. After
synchronizing this signal the EBC will complete a current external bus cycle (if any is
active), release the external bus and grant access to it by driving the LHLDA output low.
During hold state the EBC treats the external bus interface as follows:
• Address and data bus(es) float to tri-state
• Command lines become inputs (LRD, LWR, LBHE)
Should the D SCC 4 require access to its ex ternal bu s during hold m ode, it act ivates its
bus request output LBREQ to notify the arbitration circuitry. LHOLD is activated only
during hold mode. It will be inactive during normal operation.
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 108 Data Sheet 09.98
Figure 29 External Bus Arbitration (Releasing the Bus)
Note: The DSCC4 will complete the currently running bus cycle before granting bus
access as indicated by the broken lines. This may delay hold acknowledge
compared to this figure.
The figure above shows the first possibility for LBREQ to become active.
Exiting the Hold State:
The external bus master returns the access rights to the DSCC4 EBC by driving the
LHOLD input high. After s yn chron izi ng this signal the EBC will drive the L HL DA output
high, actively drive the control signals and resume executing external bus cycles if
required.
Depending on the arbitration logic, the external bus can be returned to the EBC under
two circumstances:
• The external master does not require access to the shared resources and gives up its
own access rights, or
• The DSCC4 EBC needs access to the shared resources and demands this by
activating its LBREQ output. The arbitration logic may then deactivate the other
master’s LH LDA and so fre e the externa l bus for the EBC, depend ing on the p riority
of the different masters.
ITD10615
LHOLD
(input)
LHLDA
(output)
(output)
LBREQ
LCSO
(output)
(output signals)
LRD, LWR
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 109 Data Sheet 09.98
Figure 30 External Bus Arbitration (Regaining the Bus)
The falling LBREQ edge shows the last chance for LBREQ to trigger the indicated
regain-sequence. Even if LBREQ is activated earlier the regain-sequence is initiated by
LHOLD going high. LBREQ and LHOLD are connected via an external arbitration
circuitry. Pleas e note tha t LHOLD may als o be deact ivated w ithout th e EBC request ing
the bus.
Figure 31 below shows the correct connection of the bus arbitration signals between the
master and the slave. In order to provide correct levels during initialization of the master
and the slave, two external pull-up devices are required. One is connected to the
master’s LHOLD input, t he other to the slave’s LHLDA input.
Note: For compatibility reasons with existing applications, these pull-ups can not be
integrated into the chip.
ITD10616
LHOLD
(input)
LHLDA
(output)
(output)
LBREQ
LCSO
(output)
(output signals)
LRD, LWR
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 110 Data Sheet 09.98
Figure 31 Connection of the Master and Slave Bus Arbitration Signals
Bus Arbitration Master Initialization:
After reset, the master is normally starting execution out of external memory. During
reset, the default is the arbitration slave mode. The master arbitration mode must first be
selected by setting bit LCONF.ABM=’1’. During the initialization, the ’HLDEN’ bit in
register LCONF must be set. Since the LHOLD pin is hold high through the external pull-
up, no hold requests can occur even when the slave is not initialized yet.
Note that the HLDEN bit of the master can be reset during normal operation to force the
master to ignore hold requests from the slave until HLDEN is set again. However, the
pins LHOLD, LHLDA and LBREQ are still reserved for the bus arbitration. This is
intended to have the option to disable certain critical processes against interruption
through hold requests:
ITS10617
HOLD
HLDA
BREQ
HOLD
HLDA
BREQ
MASTER SLAVE
CC
V
CC
V
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 111 Data Sheet 09.98
Bus Arbitration Slave Initialization
The slave mus t start using intern al resources only after res et. During reset, the defa ult
is the slave mode. This is also done by setting bit LCONF.ABM=’0’. This enables the
slave mode o f the bus arbitratio n signals. Afte r this, the ’HLDEN ’ bit in register LCO NF
must be set.
Note: After setting the slave’s HLDEN bit, the LBREQ output of the slave might be
activated to low for a period of 2 LBI clock periods. If the master does not
recogniz e this hold reques t (i t depen ds on the master’s transi tio n detectio n time-
slot, whether this short pulse is detected), this pulse has n o effect. If the ma ster
recognizes this pulse, it might go into hold mode for one cycle. The exact timing in
this case will be defined later.
Note: The effect of res etting b it ’HLD EN’ in slave mode (w hether the sla ve is in hol d o r
in normal mode) will be defined later. It is recommended to not reset the slave’s
’HLDEN’ bit after initialization.
Operation of the Master/Slave Bus Arbitration:
The figure below shows the sequence of the bus arbitration signals in a master/slave
system. The start-up condition is that the master is in normal mode and operating on the
external bus, while the slave is in hold mode, operating from internal memory; the slave’s
bus interfac e is trist ated. The marked time point s in the dia gram are explai ned in detail
in the following.
1) The slave detects that it has to perform an external bus access. It activates LBREQ to
low, which issues a hold request from the master.
2) The master activates LHLDA after releasing the bus. This initiates the slave’s exit from
hold sequence.
3a) When the master detects that it also has to perform external bus accesses, it
activates LBREQ to low. The e arliest time for the maste r to a ctivate LBREQ is one LBI
clock after the activation of the master’s LHLDA signal. However, the slave will ignore
this signal until it h as c omp letely taken ove r control of the e xte rnal bus. I n thi s w ay, i t is
assured that the slave will at least perform one complete external bus access.
3b) If the master can operate from internal memory while it is in hold mode, it leaves the
LBREQ signal high until it detects that an external bus access has to be performed. The
slave therefore can stay on the bus as long as the master does not request the bus
again.
4) When the master has requested the bus again through activation of its LBREQ signal,
the slave will complete the current access and go into hold mode again. After completely
tristateing its bus interface, the slave deactivates its LBREQ signal, thus releasing the
master out of hold mode.
5) The master has terminated its hold mode and deactivates its LHLDA signal again.
Now the master again controls the external bus again.
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 112 Data Sheet 09.98
6) The master deactivates its LBREQ signal again one LBI clock after deactivation of
LHLDA. From now on (and not earlier), the slave can generate a new hold request from
the master. With this procedure it is assured that the master can perform at least one
complete bus cycle before requested to go into hold mode again by the slave.
Also shown in the figure below is the s equence of the bus control between the master
and the slave.
Figure 32 Bus Arbitration Sequence
ITS10618
M-LHOLD
S-LBREQ
M-LHLDA
S-LHLDA
M-LBREQ
S-LHOLD
LBUS Master on the Bus
Slave on the Bus
Master on the Bus
3a) 3b) 6)
5)
4)
2)
1)
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 113 Data Sheet 09.98
6.1.3 PCI to Local Bus Bridge Operation
The local bus can work in 8/16-bit multiplexed/de-multiplexed mode. This 8/16-bit
organized address space can be mapped into the host memory space using the base
address register BAR1 in the PCI Configuration Space which is initialized as part of the
device configuration. Other configuration parameters define the clock speed of the local
bus, and the number of wait states on local bus transactions.
Register Write to Peripherals:
A PCI write within the local bus address space causes the address and data to be
transferred to the peripherals on the local bus.
The DSCC4 w ill sto re a sin gle dat a D WOR D (w ith corre ct b yte ena ble inf ormat ion) and
then immediately terminate the PCI transaction successfully (posted write). The write
transaction on the local bus is performed and terminated depending on the selected
number of wait states and the LRDY bus control signal.
The PCI 32-bit addresses are automatically modified to appropriate 16 or 8-bit local bus
addresses.
Thus write accesses to LBI are performed as ‘posted write’ transactions from the PCI
view. A consecutive write transaction results in PCI retry cycles in the case that the
preceding write transaction is not yet finished on the local bus.
With this approach, consecutive PCI writes are possible to the local bus address range.
Register Read from Peripherals:
The local bus address space is mapped into the shared memory space, and hence a
read operation is si mil ar to a read from memory or any memory ma pped regis ter within
the PCI address space.
A PCI Retry sequence of operations is performed, in which the DSCC4 will immediately
terminate th e PCI trans action (and request a retry) until it terminate s the trans action to
the LBI. The DSCC4 uses the retry procedure because the time to complete the data
phase will require more than the maximum allowed 16 PCI clocks (from the assertion of
FRAME to the completion of the first data phase). Data transfer will be successfully
completed within a PCI retry cycle. The number of necessary PCI retry cycles depend
on PCI arbitration behavior and the time it needs to terminate the transaction on the local
bus.
Within the local bus, the PCI read address is physically mapped into the 8/16-bit address
space of the local bus, and the read cycle is performed to the peripheral. A 8/16 bit data
read takes place at the selected local bus speed, and the 8/16 bit data is then passed on
to the PCI cycle with the correct number of C/BE (byte enable) bits set.
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 114 Data Sheet 09.98
6.1.4 LBI Interrupt Generation
The LBI block generates interrupts for transitions on input signal LINTI. (Refer to section
“LBI Registers Description” on page 351.)
Thus local bus peripherals can generate interrupt indications to the PCI host CPU.
All interrupt events result in an LBI interrupt vector which is transferred into the peripheral
interrupt queue (refer to section “LBI Interrupt Vector” on page 397).
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 115 Data Sheet 09.98
6.2 Synchronous Serial Control (SSC) Interface
6.2.1 SSC Functional Description
6.2.1.1 Overview
The Synchronous Serial Control (SSC) Interface provides flexible high-speed serial
communication between the DSCC4 and other microcontrollers or external peripherals.
The SSC supports full-duplex and half-duplex synchronous communication up to
8.25 MBaud (@ 33 MHz bus clock). The serial clock signal can be generated by the SSC
itself (master mode) or be received from an external master (slave mode). Data width,
shift directio n, clock polarity and phase are program mable. This allow s communicat ion
with SPI-compatible, or Microwire compatible devices. Transmission and reception of
data is double-buffered. A 16-bit baud rate generator provides the SSC with a separate
internally derived serial clock signal.
The high-speed synchronous serial interface can be configured very flexibly, thus it can
be used with other synchronous serial interfaces (e.g., the ASC0 in synchronous mode),
serve for master/slave or multimaster interconnections or operate compatible with the
popular SPI interface. It allows communicating with shift registers (IO expansion),
peripherals (e.g. EEPROMs) or other controllers (networking).
Data is transmitted or received on the pins MTSR (Master Transmit/Slave Receive) and
MRST (Master Re cei ve/Sl ave Trans mit). The clock signal is an ’output’ or ’in put’ on pin
MSCLK.
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 116 Data Sheet 09.98
Figure 33 Registers and Port Pins associated with the SSC
If the MFP is programmed for SSC operation (bit field PERCFG=’011’ in register
GMODE), the lower 8 signal pins of the general purpose port (GPP) provide the SSC
specific signals whereas the upper 8 signal pins are still available for general purposes.
Neverthe less for SSC operation all GPP pin s must be programmed fo r the appropria te
direction via register GPDIR.
MSCLK
MTSR
MRST
GPDIR General Purpose Bus Direction Register
SSCBR SSC Baud Rate Generator/Reload Register
SSCTB SSC Transmit Buffer Register (write only)
SSCRB SSC Receive Buffer Register (read only)
SSCIM SSC Interrupt Mask Register
GPDATA General Purpose Bus Data Register
SSCC ON SSC Control Regis ter
SSCCSE SSC Chip Select Enable Register
SSC_IV SSC Interrupt Vector
Ports & Direct ion C ontrol
Alternate F unc t ions Data Registe rs C ontrol Registers Interr upt Co nt rol
SSCBR SSCCON
GPDIR
GPDATA SSCTB
SSCRB
SSCIM
SSCCSE SSC_IV
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 117 Data Sheet 09.98
Figure 34 Synchronous Serial Channel SSC Block Diagram
The operating mode of the serial channel SSC is controlled by its bit-addressable control
register SSCCON. This register serves for two purposes:
• during programming (SSC disabled by SSCEN=’0’) it provides access to a set of
control bits;
• during operation (SSC enabled by SSCEN=’1’) it provides access to a set of status
flags.
A detailed control register description for each of the two modes is provided in “SSC
Registers Description” on page 355.
The shift register of the SSC is connected to both the transmit pin and the receive pin via
the pin control logic (see block diagram). Transmission and reception of serial data is
synchronized and takes place at the same time, i.e. the same number of transmitted bits
is also received. Transmit data is written into the Transmit Buffer SSCTB. It is moved to
the shift register as soon as this is empty. An SSC-master (SSCMS=’1’) immediately
starts transmitting, while an SSC-slave (SSCMS=’0’) waits for an active shift clock.
MCB01957
Receiv e Int. Request
Transmit Int. Request
Error Int.Request SSC Control Block
16-Bit Shi ft Register
Transmit Buffer
Register SSCTB Register SSCRB
Receive Buffer
Pin
Control
Status Control
Shift
Clock
Generator
Baud Rate Clock
Control
Slave Clock
Master Clock
CPU
Clock
In t e r n a l B u s
MTSR
MRST
SCLK
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 118 Data Sheet 09.98
When the transfe r starts, the busy flag SSCBSY is set and a trans mit interrupt request
(SSCTXI) will be generated to indicate that SSCTB may be reloaded again. When the
programmed number of bits (2 ... 16) has been transferred, the content of the shift
register is moved to the Receive Buffer SSCRB and a receive interrupt request
(SSCRXI) will be generated. If no further transfer is to take place (SSCTB is empty),
SSCBSY will be cleared at the same time. Software should not modify SSCBSY, as this
flag is hardware controlled.
Note that only one SSC can be master at a given time.
The transfer of serial data bits may be programmed in many respects:
• The data width may be selected in a range between 2 bits and 16 bits.
• Transfer may start with the LSB or the MSB.
• The shift clock may be idle low or idle high.
• Data bits may be shifted with the leading or trailing edge of the clock signal.
• The baudrate may be set from 152 Baud up to 5 MBaud (@ 20 MHz CPU clock).
• The shift clock can be either generated (master) or received (slave).
This flexible programming allows to adapt the SSC to a wide range of applications,
where serial data transfer is required.
The Data Width Selection allows to transfer frames of any length, from 2-bit ’characters’
up to 16-bi t ’characters’. St arting with the LSB (SSCHB=’0’) allo ws commu nicating e. g.
with ASC0 devices in synchronous mode (C166 family) or 8051 like serial interfaces.
Starting with the MSB (SSCHB=’1’) allows to operate compatible with the SPI interface.
Regardless which data width is selected and whether the MSB or the LSB is transmitted
first, th e transfer data is always right al igned in registers SSCTB and SSCRB, with th e
LSB of the transfer data in bit 0 of these registers. The data bits are rearranged for
transfer by the internal shift register logic. The unselected bits of SSCTB are ignored, the
unselected bits of SSCRB will be not valid and should be ignored by the receiver service
routine.
The Clock Control allows to adapt transmit and receive behaviour of the SSC to a
variety of serial interfaces. A specific clock edge (rising or falling) is used to shift out
transmit data, while the other clock edge is used to latch in receive data. Bit SSCPH
selects the leading edge or the trailing edge for each function. Bit SSCPO selects the
level of the clock line in the idle state. Hence for an idle-high clock the leading edge is a
falling one, a 1-to-0 transition. The figure below summarizes the clock control.
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 119 Data Sheet 09.98
Figure 35 Serial Clock Phase and Polarity Options
6.2.1.2 Operational Mode: Full-Duplex Operation:
The differen t devices are connecte d through three lines . The definition of thes e lines is
always determined by the master: The line connected to the master’s data output pin
MTSR is the transmit line, the receive line is connected to its data input line MRST, and
the clock line is connected to pin MSCLK. Only the device selected for master operation
generates and outputs the serial clock on pin MSCLK. All slaves receive this clock, so
their pin MSCLK must be switched to input mode (GPDIR.p=’0’). The output of the
master’s shift register is connected to the external transmit line, which in turn is
connected to the slaves’ shift register input. The output of the slaves’ shift register is
connected to th e extern al receive line in order to enable the master to rece ive the data
shifted out of the slave. The external connections are hard-wired, the function and
direction of these pins is determined by the master or slave o peration of the individual
device.
When initializing the devices in this configuration, select one device for master operation
(SSCMS=’1’), all others must be programmed for slave operation (SSCMS=’0’).
Initialization includes the operating mode of the device's SSC and also the function of
the respective port lines (refer to section ’Port Control’).
MCD01960
Last
Bit
Transmit Data
First
Bit
Latch Data
Shift Data
Pins
MTSR/MRST
Serial Clock
SCLK
SSCPOSSCPH
00
01
10
11
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 120 Data Sheet 09.98
Figure 36 SSC Full Duplex Configuration
Note: The shift direction applies to MSB- first operation as well as to LSB-first operation.
The data output pins MRST of all slave devices are connected together onto the one
receive line in this configuration. During a transfer each slave shifts out data from its shift
register. There are two ways to avoid collisions on the receive line due to different slave
data:
1. Only one slave drives the line , i.e . enables the driver of its MRST pin . All the other
slaves ha ve to program there MR ST pi ns to input. So only one sla ve c an p ut it s da ta
onto the master’ s receive line. Only receiving of data from the master is possible. The
master selec ts the slave device from whic h it expects data eith er by separa te select
lines, or by sending a special command to this slave. The selected slave then switches
its MRST line to output, until it gets a deselection signal or command.
2. The slaves use ope n dra in output on MRST. This forms a Wired -AND co nnec tio n.
The receive line needs an e xternal pullup in this case. C orruption of the d ata on the
receive line sent by the selected slave is avoided, when all slaves which are not
selected for transmission to the master only send ’on es’. Sin ce this high level is not
actively driven onto the line, but only held through the pullup device, the selected slave
can pull this line actively to a low level when transmitting a zero bit. The master selects
the slave device, from which it expects data either by separate select lines, or by
sending a special command to this slave.
MCS01963
MRST
CLK
2Device Slave
Clock
Shift Register
MTSR
Clock
1Master
Receive
Transmit
Device
Clock
MTSR
MRST
CLK
Shift Register
MTSR
Shift Register
Clock
SlaveDevice 2
CLK
MRST
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 121 Data Sheet 09.98
After performing all necessary initializations of the SSC, the serial interfaces can be
enabled. For a master device, the alternate clock line will now go to its programmed
polarity . The alternat e data l ine w ill go to eithe r ’0’ o r ’1’, until the fi rst tran sfer w ill s tart.
After a transfer the alternate data line will always remain at the logic level of the last
transmitted data bit.
When the serial interface is enabled, the master device can initiate the first data transfer
by writing the transmit data into register SSCTB. This value is copied into the shift
register (which is assumed to be empty at this time), and the selected first bit of the
transmit data will be placed onto the MTSR line on the next clock from the baudrate
generator (transmission only starts, if SSCEN=’1’). Depending on the selected clock
phase, a clock pulse will also be generated on the MSCLK line. With the opposite clock
edge the master at the same time latches and shifts in the data detected at its input line
MRST. This ’ex changes’ the trans mit data with th e receive data . Since the cloc k line is
connected to all slaves, their shift registers will be shifted synchronously with the
master's shift register, shifting out the data contained in the registers, and shifting in the
data detected at the input line. After the preprogrammed number of clock pulses (via the
data wi dth selection) th e data transmitted by the mast er is contained in all slaves ’ shift
registers, while the master's shift register holds the data of the selected slave. In the
master and all slaves the content of the shift register is copied into the receive buffer
SSCRB and the receive interrupt flag SSCRXI is set.
A slave devic e will imme diately output the selecte d first bit (MSB or LSB of the tran sfer
data) at pin MRST, when the content of the transmit buffer is copied into the slave's shift
register. It will not wait for the next clock from the baudrate generator, as the master
does. The reason is that, depending on the selected clock phase, the first clock edge
generated by the master may already be used to clock in the first data bit. Hence the
slave's first data bit must already be valid at this time.
Note: On the SSC always a transmission and a reception takes place at the same time,
regardless wheth er valid data has been tran smitted or receiv ed. This is diffe rent,
e.g., from asynchronous reception on ASC0.
The initialization of the MSCLK pin on th e ma ste r requi res s om e attention i n ord er to
avoid undesired clock transitions, which may disturb the other receivers. The state of the
internal alternate output lines is '1' as long as the SSC is disabled. This alternate output
signal is ANDed with the respective port line output latch. Enabling the SSC with an idle-
low clock (SSCPO=’0’) will drive the alternate data output and (via the AND) the port pin
MSCLK immediately low. To avoid this, use the following sequence:
• select the clock idle level (SSCPO=’x’),
• load the port output latch with the desired clock idle level (GPDATA.p=’x’),
• switch the pin to output (GPDIR.p=’1’),
• enable the SSC (SSCEN=’1’), and
• if SSCPO=’0’: enable alternate data output (GPDATA.p=’1’).
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 122 Data Sheet 09.98
The same mechanism as for selecting a slave for transmission (separate select lines or
special commands) may also be used to promote the role of the master to another device
in the network . In this case the previous mas ter and the future master (p revious slave)
will have to toggle their operating mode (SSCMS) and the direction of their port pins (see
description above).
Chip Select Control:
There are 4 chip select pins associated with the SSC port: MCS0 to MCS3. The four chip
select lines are automatically activated at the beginning of a transfer and deactivated
again after the transfer has ended. Activation of a chip enable line always begins one
half bit time before the first data bit is output at the MTSR pin, and the deactivation
(except for the continuous transfers) is performed one half bit time after the last bit of the
transfer has been transmitted/received completely.
The chip sel ect li nes are select ed by the control bi ts ASEL0 to ASEL3 of the SSC Chip
Select Enable register SSCCSE (refer to “SSC Registers Description” on page 355).
By setting any of these bits to 0, the corresponding chip select port will be asserted when
transmitting data. All other bits of the SSCCSE register have to be set to ‘0‘.
6.2.1.3 Operational Mode: Half Duplex Operation:
In a half duplex configuration only one data line is necessary for both receiving and
transmitting of data. The data exchange line is connected to both pins MTSR and MRST
of each device, the clock line is connected to the MCLK pin.
The master device controls the data transfer by generating the shift clock, while the slave
devices receive it. Due to the fact that all transmit and receive pins are connected to the
one data exchange line, serial data may be moved between arbitrary stations.
Similar to ful l duplex mode there are two ways to avoid collisions on the data e xchang e
line:
• only the transmitting device may enable its transmit pin driver
• the non-transmitting devices use open drain output and only send ones.
Since the data inputs and outputs are connected together, a transmitting device will clock
in its own data at the input pin (MRST for a master device, MTSR for a slave). This allows
to de tect a ny co rrupti ons o n the com mon d ata ex change l ine, w here th e Rx d ata is not
equal to the Tx data.
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 123 Data Sheet 09.98
Figure 37 SSC Half Duplex Configuration
Continuous Transfers:
When the transmit interrupt request flag is set, it indicates that the transmit buffer SSCTB
is empty and ready to be loaded with the next transmit data. If SSCTB has been reloaded
by the time the current transmission is finished, the data is immediately transferred to the
shift register and the next transmission will start without any additional delay. On the data
line there is no gap between the two successive frames. For example, two byte transfers
would look the same as one word transfer. This feature can be used to interface with
devices which can operate with or require more than 16 data bits per transfer. The length
of a to tal data fra me is up to the software. This opti on can also be used to interface to
byte-wide and word-wide devices on the same serial bus.
Note: This feature only applies to multiples of the selected basic data width, since it
would require disabling/enabling of the SSC to re-program the basic data width
on-the-fly.
MCS01965
Common
Transmit/
Receive
Line
MRST
CLK
2Device Slave
Clock
Shift Register
MTSR
Clock
1Master Device
Clock
MTSR
MRST
CLK
Shift Register
Shift Register
Clock
SlaveDevice 3
CLK
MTSR
MRST
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 124 Data Sheet 09.98
Port Control:
The SSC uses three pi ns to commu nicate with the external w orld. Pin MCLK serves as
the clock line, while pins MRST (Master Receive/Slave Transmit) and MTSR (Master
Transmit/Slave Receive) serve as the serial data input/output lines. The operation of
these pins depends on the selected operating mode (master or slave).
The direction of the port lines depends on the operating mode, that is selected via
SSCCON:SSCMS.
6.2.1.4 Baud Rate Generation
The SSC interface has its own dedicated 16-bit baud rate generator with 16-bit reload
capability, allowing baud rate generation independent from timers.
The baud rate generator is clocked with the CPU clock divided by 2 (10 MHz @ 20 MHz
bus clock). The timer is counting downwards and can be started or stopped through the
global enable bit SSCEN in the SSC Control register SSCCON.
The register SSCCON (refer to section “ SSC Registers Description” on page 35 5) is
the dual-function Baud Rate Generation register. Reading SSCBR, while the SSC is
enabled, returns th e co ntents of the timer. Readin g SSCBR, while the SSC is disa ble d,
returns the programmed reload value. In this mode the desired reload value can be
written to SSCBR.
The formula s below c alculat e either the resultin g baud ra te for a give n reload value, or
the required reload value for a given baudrate:
Bit field <SSCBR> represents the contents of the reload register, taken as an unsigned
16-bit integer.
The maxi mum baud rate that can be achieved when using a PCI clock of 20 MHz is 5
MBaud. The table below lists some possible baud rates together with the required reload
values and the resulting bit times, assuming a PCI clock of 20 MHz. A PCI clock of
33 MHz is also supported.
6.2.1.5 Error Detection
The SSC is able to detect four different error conditions. Receive Error and Phase Error
are detecte d in all modes, whi le Transmit Error and Bau drate Error only apply to slave
mode. When an error is detected, the respective error flag is set. When the
corresponding error enable bit is set, also an error interrupt request will be generated by
setting SSCERI (see figure 38). The error interrupt handler may then check the error
flags to determine the cause of the error interrupt. The error flags are not reset
BaudrateSSC = fCPU
2 * (<SSCBR> + 1) SSCBR = ( fCPU
2 * BaudrateSSC ) - 1
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 125 Data Sheet 09.98
automatically (like SSCERI), but rather must be cleared by software after servicing. This
allows to se rvice some error c onditions v ia interrupt, whi le the others ma y be polled by
software.
Note: The error interrupt handler must clear the associated (enabled) errorflag(s) to
prevent repeated interrupt requests.
A Receive Error (Master or Slave mode) is detected, when a new data frame is
completely received, but the previous data was not read out of the Receive Buffer
register SSCRB. This condition sets the error flag SSCRE and, when enabled via
SSCREN, the error interrupt request flag SSCERI. The old data in the receive buffer
SSCRB will be overwritten with the new value and is unretrievably lost.
A Phase Error (Master or Slave mode) is detected, when the incoming data at pin MRST
(master mode) or MTSR (slave mode), sampled with the same frequency as the CPU
clock, changes in a range between one sample before and two samples after the latching
edge of the clock signal (refer to section ’Clock Control’). This condition sets the error
flag SSCPE and, when enabled via SSCPEN, the error interrupt request flag SSCERI.
A Baud Rate Error (Slave mod e) is detect ed, when the incomi ng cl ock sign al de viates
from the programmed baud rate by more than 100%, i.e., it has a value of either more
than double or less than half of the expected baud rate. This condition sets the error flag
SSCBE and, when enabled via SSCBEN, the error interrupt request flag SSCEIR.
Using this error detection capability requires that the slave's baud rate generator is
programme d to th e same baud ra te as the mas ter dev ice . This feat ure allo ws to det ect
false additional, or missing pulses on the clock line (within a certain frame).
If this error condition occurs and bit SSCAREN=’1’, an automatic reset of the SSC will
be performed. This is done to reinitialize the SSC, when too few or too many clock pulses
have been detected.
A Transmit Error (Slave mode) is detected, when a transfer was initiated by the master
(shift clock gets active), but the transmit buffer SSCTB of the slave was not updated
since the last tra nsfer. This condit ion sets the error flag SSCTE and, when enabl ed via
SSCTEN, the error interru pt request fl ag SSCERI. If a transfer s tarts while the trans mit
buffer is not updated, the slave will shift out the 'old' contents of the shift register, which
are usually the data received during the last transfer.
This may lead to the corruption of the data on the transmit/receive line in half-duplex
mode (open drain configuration), if this slave is not selected for transmission. This mode
requires that slaves not sele cted for trans mis sio n onl y sh ift ou t ’one s’, i.e. their transmit
buffers must be loaded with 'FFFFH' prior to any transfer.
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 126 Data Sheet 09.98
Note: A slave with push/pull output drivers, which is not selected for transmission, will
normally have its output drivers switched. However, in order to avoid possible
conflicts or misinterpretations, it is recommended to always load the slave’s
transmit buffer prior to any transfer.
Figure 38 SSC Error Interrupt Control
6.2.2 SSC Interrupt (Vector)
The SSC block generates three kinds of interrupts: transmit, receive and error interrupts.
Any of these interrupts can be enabled by setting the corresponding bit of the SSC
Interrupt Mask register SSCIM (refer to section “SSC Registers Description” on
page 355) to ‘1‘. All other bits of this register have to be set to zero.
All interrupt events result in an SSC interrupt vector which is transferred into the
peripheral interrupt queue (refer to section “SSC Interrupt Vector” on page 395).
MCS01968
Error
Interrupt
SSCEINT
&
ERI
IMER
1
Error
Transmit &
&
Receive
Error SSCRE
SSCRE
SSCTE
SSCTE
SSCPE
SSCBE
Error
Baudrate &
&
Phase
Error SSCPE
SSCBE
Register SSCCON Register SSCIM
Interrupt Vector SSC_IV
PEB 20534
Multi Function Port (MFP)
Semiconductor Group 127 Data Sheet 09.98
6.3 General Purpose Port (GPP) Interface
6.3.1 GPP Functional Description
A general purpose 8-/16-bit port is provided on pins GP0...GP15. Every pin is separately
programmable via the General Purpose Port Direction Register GPDIR to operate as an
output or an input.
The number of available port pins depends on the selected MFP configuration mode (bit
field ’PERCFG’ in register GMODE).
If defined as output, the state of the pin is directly controlled via the General Purpose Port
Data Register GPDATA. Read access to this register delivers the current state of all GPP
pins (input and output signals).
If defined as input, the state of the pin is monitored. The value is readable via GPDATA.
All changes may be (if desired) indicated via interrupt. Assigned register: General
Purpose Port Interrupt Mask Register GPIM. See “GPP Registers Description” on
page 368.
6.3.2 GPP Interrupt Vector
The GPP block generates interrupts for transitions on each input (and output) signal. Any
pin can be en abled for interrup t genera tion by settin g the c orrespon ding bi t of the G PP
Interrupt Mask register GPIM (refer to section “GPP Registers Description” on
page 368) to ‘0‘.
All interrupt events result in an SSC interrupt vector which is transferred into the
peripheral interrupt queue (refer to section “GPP Interrupt Vector” on page 398).
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 128 Data Sheet 09.98
7 Serial Communication Controller (SCC) Cores
7.1 General
The Serial Communication Controller (SCC) distinguishes itself from other
communication controllers by its advanced characteristics. The most important are:
– Support of HDLC/SDLC, ASYNC, BISYNC/MONOSYNC, and point-to-point protocols
(PPP).
– Support of layer-2 functions (HDLC mode)
In addi tion to thos e bit-oriente d funct ions comm only suppo rted by HD LC controlle rs,
such as bit stuffing, CRC check, flag and address recognition, the SCC provides a
high degree of procedural support.
– In a special operating mode (auto-mode), the SCC processes the information transfer
and the procedure handshaking (I- and S-frames of HDLC protocol) autonomously.
The only restriction is that the window size (= number of outstanding unacknowledged
frames) is limited to 1, which is sufficient for many applications. The communication
procedures are mainly processed between the communication controllers and not
between the attached hosts. Thus the dynamic load on the host and the software
expense is greatly reduced.
The host is informed about the status of the procedure and has mainly to manage the
interrupt service. Receive and transmit data are managed by the on chip DMAC
autonomously. In order to maintain cost effectiveness and flexibility, the handling of
unnumbered (U) frames, and special functions such as error recovery in case of
protocol errors, are not implemented in hardware and must be done by the user’s
software.
– Extended support of different link configurations
Besides the point-to-point configurations, the SCC allows the implementation of point-
to-multipoint or multi-master configurations without additional hardware or software
expense.
In point-to-multi point configura tions, the SCC can be use d as a maste r or as a slav e
station. Even when working as slave station, the SCC can initiate the transmission of
data at any time. An internal function block provides means of idle and collision
detection and collision resolution, which are necessary if several stations start
transmitting simultaneously. Thus, a multi-master configuration is also possible.
– Telecom specific features
In a sp ecial operating mo de, the SC C ca n transm it or rece ive data pa ckets in one of
up to 128 time-slots of programmable width (clock mode 5). Furthermore, the SCC
can transmit or receive variable data portions within a defined window of one or more
clock cycles in conjunction with an external strobe signal (clock mode 1). These
features make the SCC suitable for applications using time division multiplex
methods, such as time-slot oriented PCM systems or systems designed for packet
switching.
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 129 Data Sheet 09.98
– Support of PPP Data Link Layer frame transmission
In a special HDLC sub mode, the SCC provides transmission of PPP Data Link Layer
frames in either an asy nchronous (st art/stop), bit-synchron ous or octet-synchrono us
mode. An escape mechanism is implemented to allow control data such as
XON/XOFF to be transmitted transparently via the link, and to remove spurious control
data which may be injected into the link by intervening hardware and software.
– FIFO buffers for efficient transfer of data packets
Since all SCCs are contending for the internal busses, each SCC has an eight 32-bit
word deep FIFO in transmit and an seventeen 32-bit word deep FIFO in receive
direction for temporary storage of data packets transferred between the serial
communications interface and the central FIFOs of the DSCC4. These FIFOs allow
overlapping input/output operation (dual-port behavior).
– High Data Rate (PEB 20534H-52 only)
In a speci al operating mode (clo ck mode 4, high speed mod e) any of the four SCCs
can support high data rates: e.g. 45 Mbit/s for DS3 or 52 Mbit/s for OC1. The
aggregate bandwidth supported is 108 MBit/s per direction. This allows various
configurations, for example:
- 2 ports 52 MBit/s and 2 ports 2 MBit/s
- 2 ports 45 MBit/s and 2 ports 8 MBit/s
- 4 ports 26 MBit/s
Features included by each one of the SCCs:
• Serial Interface
– On chip clock gen eration or external clock source
– On chip DPLL for clock recovery
– Baud rate generator
– Programmable time-slot capability
– NRZ, NRZI, FM0/1 and Manchester data encoding
– Optional data flow control using modem lines (RTS, CTS, CD)
– Support of bus configuration by collision detection and resolution
– Full duplex data rates of up to 10 Mbit/s sync - 2 Mbit/s with DPLL, 2 Mbit/s async
– Full dup lex data rate of up to 52 Mbit/s (HDLC address mode 0, PPP or extended
transparent mode) in clock mode 4 (external clock source and clock
gating/gapping).
• Bit Processor Functions
– HDLC/ SDL C Mode
- Automatic flag detection and transmission
- Shared opening and closing flag
- Generation of interframe-time fill ’1’ s or flags
- Detection of receive line status
- Zero bit insertion and deletion
- CRC generation and checking (CRC-CCITT or CRC-32)
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 130 Data Sheet 09.98
- Transparent CRC option per channel and/or per frame
- Programmable Preamble (8 bit) with selectable repitition rate
- Error detection (abort, overrun, underrun, CRC error, too long/short frame)
– ASYNC Mode
- Selectable character length (5 to 8 bits)
- Even, odd, forced or no parity generation/checking
- 1 or 2 stop bit insertion in transmit
- Break detecti on/g enera tion
- Flow control by XON/XOFF character
- Immediate character insertion (for insertion of control characters in data stream)
- Termination character detection for end of block identification
- Time out detection
- Error detection (parity error, framing error)
– BISYNC Mode
- Programmable 6/8 bit SYNC pattern (MONOSYNC)
- Programmable 12/16 bit SYNC pattern (BISYNC)
- Selectable character length (5 to 8 bits)
- Even, odd, forced or no parity generation/checking
- Generation of interframe-time fill ’1’ s or SYNC characters
- CRC generation (CRC-16 or CRC-CCITT)
- Transparent CRC option per channel and/or per frame
- Programmable Preamble (8 bit) with selectable repitition rate
- Termination character detection for end of block identification
- Error detec tion (parity error, CRC e rror)
• Protocol Support (provided in HDLC/SDLC Mode)
– Address mode 0
- No address recognition
– Address mode 1
- 8-bit (high byte) address recognition
– Non-auto mode
- 8-bit (low byte) or 16-bit (high and low byte) address recognition
– Auto mode
- 8-bit or 16-bit address generatio n/rec ogn itio n
- Automatic handling of S- and I-frames
- Support of LAP-B / LAP-D
- Automatic processing of control byte(s)
- Modulo 8 or modulo 128 operation
- Programmable time-out and retry conditions
- Normal Response Mode operation for slave
– Asynchronous PPP mode
- Character oriented transmission of HDLC frame (flag, data, CRC, flag)
- Start/stop bit framing of each single character
- Automatic flag detection and transmission
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 131 Data Sheet 09.98
- Shared opening and closing flag
- Generation of interframe-time fill ’1’ s or flags
- Detection of receive line status
- No zero bit insertion/deletion
- CRC generation and checking (CRC-CCITT or CRC-32)
- Transparent CRC o ption per channel and/or per frame
- Programmable Preamble (8 bit) with selectable repitition rate
- Error detection (abort, overrun, underrun, long frame, CRC error, short frames)
- Escape mechanism for control data
- Programmable character map for escape mechanism (00H..1FH selectable from
fixed char act er map, 4 additio nal progra mma ble characters)
– Bit synchronous PPP mode
- Automatic flag detection and transmission
- Shared opening and closing flag
- Generation of interframe-time fill ’1’ s or flags
- Detection of receive line status
- Zero bit insertion and deletion
- 15 bit ’1’ abort sequence
- CRC generation and checking (CRC-CCITT or CRC-32)
- Transparent CRC o ption per channel and/or per frame
- Programmable Preamble (8 bit) with selectable repetition rate
- Error detection (abort, overrun, underrun, long frame, CRC error, short frames)
– Octet synchronous PPP mode
- Automatic flag detection and transmission
- Shared opening and closing flag
- Generation of interframe-time fill ’1’ s or flags
- Detection of receive line status
- CRC generation and checking (CRC-CCITT or CRC-32)
- Transparent CRC o ption per channel and/or per frame
- Programmable Preamble (8 bit) with selectable repetition rate
- Error detection (abort, overrun, underrun, long frame, CRC error, short frames)
- Escape mechanism for control data
- Programmable character map for escape mechanism (0x00..0x1F selectable from
fixed char act er map, 4 additio nal progra mma ble characters)
– Extended transparent mode
- Bit-transparent data transmission/reception (No HDLC-framing, no bit stuffing...)
• Protocol and Mode Independent
- D ata inv ersi on
- Data over- and underflow detection
- Timer for software support
- Internal test loop capability
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 132 Data Sheet 09.98
7.2 Protocol Modes Overview
The SCC is a mu lti -protocol co mmunic ation con troller. Th ree major p rotocol block s are
implemen ted: HDLC/SDLC, ASYNC and BISYNC.
These protocol blocks provide different protocol modes which are listed in table 19.
The protocol modes are described in details in Chapter 8:
"Detailed Protocol Description"
1) Extended tr anspa rent is a fully bit-tran spare nt tran smit/ recep tion mode w hich is treated as a sub-mod e of the
HDLC/SDLC block.
Table 19 Protocol Modes
Protocol Blocks Protocol Modes
HDLC/SDLC
PPP HDLC auto mode (16-bit)
HDLC auto mode (8-bit)
HDLC non auto mode (16-bit)
HDLC non auto mode (8-bit)
HDLC address mode 1
HDLC address mode 0
Asynchronous PPP mode
Bit synchronous PPP mode
Octet synchronous PPP mode
Extended transparent mode1)
ASYNC Asynchronous mode
Isochronous mode
BISYNC Bisynchronous mode
Monosynchronous mode
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 133 Data Sheet 09.98
7.3 SCC FIFOs
Each SCC pro vides its own tran smi t and rec eive FIFOs t o ha ndle int ernal arbi tration of
the central FIFOs.
7.3.1 SCC Transmit FIFO
The SCC transmit FIFO is divided into two parts of 6 and 2 DWORDs. The interface
between the two parts provides clock synchronization between the system clock domain
and the protocol logic working with the serial transmit clock.
Figure 39 SCC Transmit FIFO
The 6 DWORDs system clocked FIFO part always requests transmit data from the
central TFIFO if at least 4 DWORDs free space is available even if the SCC is in power-
down condition (register CCR0 bit PU=’0’).
The only exception is a transmit data underrun (XDU) event . In case of an XDU event
(e.g. after excessive PCI bus latency), the FIFO will neither request more data from the
central TFIFO nor transfer another DWORD to the protocol logic.
This XDU blocking mechanism prevents unexpected serial data and must be cleared by
a transmitter reset command. In case of a transmitter reset command (register CMDR
bit XRES=’1’) the com plete SCC transmit F IFO is cleare d and will imme diately requ est
new transmit data from the central TFIFO.
Transfer of data t o the 2 DWORD shadow part only takes pla ce if the SCC is in po wer-
up condition and an appropriate transmit clock is provided depending on the selected
clock mod e.
Serial data transmission will start as soon as at least one DWORD is transferred into the
2 DWORD shado w FIFO and tran smi ssion is ena bled de pen ding on the se lect ed clo ck
mode (CTS signal active, clock strobe signal active, valid timeslot or clock gapping signal
inactive).
DWORD8
DWORD1
DWORD2
DWORD3
DWORD4
DWORD5
DWORD6
DWORD7
to protocol logic
and serial line
transmit clock
domain
system clock
domain
data requested
from
central TFIFO
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 134 Data Sheet 09.98
7.3.2 SCC Receive FIFO
The SCC receive FIFO is divided into two parts of 15 and 2 DWORDs. The interface
between the two parts provides clock synchronization between the system clock domain
and the protocol logic working with the serial receive clock.
Figure 40 SCC Receive FIFO
With standard register settings (i.e. the SCC receive FIFO threshold is not reduced, refer
to Table 69, “CCR2: Channel Configuration Register 2,” on page 296), the SCC
receive FIFO requests data transfer to the central RFIFO if the 15 DWORDs part is
completely filled or a frame end / block end condition is detected.
This SCC receive FIFO size is optimized for high speed channel configurations. The 15
DWORDs FIFO part is transferred to the central RFIFO allocating one consecutive block
of RFIFO memory. This guarantees full 15 DWORDs burst length on the PCI/De-
multiplexed system interface which can be performed on consecutive RFIFO sections
only (refer to Chapter 5.2.3 "Central Receive FIFO (RFIFO)").
Nevertheless this FIFO depth might cause too long delay in low speed channel
configurations on data transfer to the host memory (especially ASYNC/BISYNC protocol
modes). Therefore the SCC receive FIFO threshold can be lowered in some steps
downto 1 data byte causing the SCC to request data transfer to the central RFIFO as
soon as this threshold is reached. The threshold is adjusted by bit field ’RFTH’ in register
CCR2.
In addition data stored in the SCC receive FIFO can be transferred to the central RFIFO
any time on request by setting command ’RFRD’ in register CMDR. Prior to issuing a
’RFRD’ command, the "receive FIFO not empty condition" can be tested by the host CPU
by reading bit ’RFNE’ in register STAR.
Furthermore in ASYNC mode this ’RFRD’ command can be generated automatically on
a time out condition if enabled via bit ’TOIE’ in register CCR1. In ASYNC applications
characters are often send in bl ocks which means a small time gap between cha racters
of these blocks; but also single control characters may be interleaved. In thi s case the
receive FIFO threshold might be adjusted to the length of expected ASYNC character
DWORD2
DWORD1
DWORD5
DWORD3
DWORD4
DWORD17
DWORD16
from serial
line and protocol
logic
receive clock
domain
system clock
domain
to central
RFIFO
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 135 Data Sheet 09.98
blocks or higher but single characters not exceeding the threshold are also forwarded to
the central RFIFO after time out. A time out condition is detected if the line idle time
excee ds a programmable time period (see register CCR1 bit field ’TOLEN’).
Mention that any ’RFRD’ command generated by write access to register CMDR or
automatically by time out mechanism in ASYNC/BISYNC modes always forces a
’frame end/block end’ condition (FE=’1’) causing the DMA controller to finish the current
receive descriptor.
If the SCC receive FIFO is completely filled further incoming data is ignored and a
receive data overflow condition (RDO) is detected. As soon as the receive FIFO provides
empty space receive data is accepted again waiting for a frame end or frame abort
sequence. The automatically generated receive status byte (RSTA) will contain an RDO
indication in this case and the next incoming frame will be received in a normal way.
Therefore no further CPU intervention is necessary to recover the SCC from an RDO
condition.
A "frame" with RDO status might be a mixture of a frame partly received before the RDO
event occured and the rest of this frame received after the receive FIFO again accepted
data and the frame was still incoming. A quite arbitrary series of data or complete frames
might get lost in case of an RDO event. Every frame which must be completely discarded
because of an RDO condition generates an RFO interrupt.
The SCC receive FIFO can be cleared by command ’RRES’ in register CMDR. Note that
clearing the receive FIFO during operation might delete a frame end / block end
indication. A frame which was already partly transferred to the central RFIFO cannot be
"closed" in this case because the DMA controller will not get the corresponding frame
end indication. A new frame received after receiver reset command will be appended to
this "open" frame.
In ASYNC and BISYNC protocol modes, a frame end / block end indication can be forced
by command ’RFRD’ in register CMDR to avoid this unexpected behaviour.
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 136 Data Sheet 09.98
In this case a useful sequence of clearing the SCC receive FIFO during operation is:
Register Bit field Description
ASYNC/BISYNC Modes
CCR2 RAC=’0’ Switches the receive protocol logic to inactive state not
accepting any more data from the serial line.
CMDR RRES=’1’ Resets the receive protocol logic and clears the SCC
receive FIFO (self-resetting command bit).
CMDR RFRD=’1’ Generates a frame end indication which is transferred to the
central RFIFO terminating any partial frame. If no partial
frame was stored, this will result in a "frame" with byte
number zero.
CCR2 RAC=’1’ Switches the receive protocol logic to active state accepting
receive data from the serial line.
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 137 Data Sheet 09.98
7.4 Clocking System
The DSCC4 includes an internal Oscillator (OSC) as well as four independent Baud Rate
Generators (BRG) and four Digital Phase Locked Loop (DPLL) circuits.
The transmit and receive clock can be generated either
• externally, and supplied directly via the RxCLK and/or TxCLK pins
(called external clock modes)
• internally, by selecting
– the internal oscillator (OSC) and/or the channel specific baud rate generator (BRG)
– the internal DPLL, recovering the receive (and optionally transmit) clock from the
receive data stream.
(called internal clock modes)
There are a total of 13 different clocking modes programmable via bit field ’cm’ in register
CCR0, providing a wide variety of clock generation and clock pin functions, as shown in
table 20.
The transmit clock pins (TxCLK) may also be configured as output clock and control
signals in certain clock modes if enabled via bit ’TOE’ in register CCR0.
The clocking source for the DPLL’s is always the internal channel specific BRG; the
scaling factor (divider) of the BRG can be programmed through BRR register.
There are two channel specific internal operational clocks in the SCC:
One operational clock (= transmit clock) for the transmitter part and one operational clock
(= receive clock) for the receiver part of the protocol logic.
Note: The internal timers always run using the internal transmit clock.
Table 20 Overview of Clock Modes
Clock
Type Source Generation Clock Mode
Receive
Clock
RxCLK Pins External ly 0, 1, 4, 5
OSC,
DPLL,
BRG,
Internally 2, 3a, 6, 7a
3b, 7b
Transmit
Clock
TxCLK Pins,
RxCLK Pins External ly 0a, 2a, 4, 6a
1,5
OSC,
DPLL,
BRG/BCR,
BRG
Internally 3a, 7a
2b, 6b
0b, 3b, 7b
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 138 Data Sheet 09.98
The internal structure of each SCC channel consists of 3 clocking domains, transmit,
receive, and system. These three function blocks are clocked with internal transmit
frequency fTRM, internal rec eive frequ ency fREC and sy stem frequen cy fPCI, respe ctively
(system frequency fPCI only supplies the SCC receive and transmit FIFO part facing the
DMA controller). The internal FIFO interfaces are used to transfer data between the
different clock domains.
The clocks fTRM and fREC are internal clocks only and need not be identical to external
clock inputs e.g. fTRM and TxCLK input pin.
The features of the different clock modes are summarized in table 21.
Note: If ASYNC Mode is programmed, some clock mode frequencies can or must be
divided by 1 6 as s elected by the Bit C lock Rate bit CCR1:BCR:
When BCR is set to ‘16’, oversampling (3 samples) in conjunction with majority
decision is performed. BCR has no effect when using clock mode 2, 3a, 4, 5, 6,
or 7a.
Table 21 Clock Modes of the SCCs
Channel
Configuration Clock Sources Control So urces
Clock
Mode
CCR0:
CM2,
CM1,CM0 CCR0:
SSEL to
BRG to
DPLL to
REC to
TRM CD R- Str o be X- Strobe Fr a m e-
Sync
Output
via
TxCLK
(if CCR0:
TOE = ‘1’)
0a
0b
1
2a
2b
3a
3b
4
5
6a
6b
7a
7b
0
1
X
0
1
0
1
X
X
0
1
0
1
–
OSC
–
RxCLK
RxCLK
RxCLK
RxCLK
–
–
OSC
OSC
OSC
OSC
–
–
–
BRG
BRG
BRG
–
–
–
BRG
BRG
BRG
–
RxCLK
RxCLK
RxCLK
DPLL
DPLL
DPLL
BRG
RxCLK
RxCLK
DPLL
DPLL
DPLL
BRG
TxCLK
BRG
RxCLK
TxCLK
BRG/16
DPLL
BRG
TxCLK
RxCLK
TxCLK
BRG/16
DPLL
BRG
CD
CD
–
CD
CD
CD
CD
–
–
CD
CD
CD
CD
–
–
CD
–
–
–
–
RCG
(TSAR/PC
MMRX)
–
–
–
–
–
–
TxCLK
–
–
–
–
TCG
(TSAX/PC
MMTX)
–
–
–
–
–
–
–
–
–
–
–
–
FSC
–
–
–
–
–
BRG
–
–
BRG/16
DPLL
BRG
-
TS-Control
–
BRG/16
DPLL
BRG
Clock Mode fREC fTRM
0a
0b
1
3b, 7b
fRxCLK/BCR
fRxCLK/BCR
fRxCLK/BCR
fBRG/BCR
fTxCLK
fBRG
fRxCLK/BCR
fBRG/BCR
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 139 Data Sheet 09.98
Note: If one of the clo ck modes 0b, 4, 6 or 7 is s elected the in ternal oscill ator (OSC) is
enabled which allows connection of an external crystal to pins XTAL1-XTAL2. The
output signal of the OSC can be used for one serial channel, or for all serial
channels (independent baud rate generators and DPLLs). Moreover, XTAL1
alone can be used as input for an externally generated clock.
The first two columns of Table 21 list all possible clock modes configured via bit field
’CM’ and bit ’SSEL’ in register CCR0.
For example, clock mode 6b is choosen by writing a ’6’ to register CCR1.CMi and by
setting bit CCR0.SSEL equal to ’1’. The following 4 columns (grouped as ’Clock
Sources’) specify the source of the internal clocks. Columns REC and TRM correspond
to the domain clock frequencies fREC and fTRM .
The columns grouped as ’Control Sources’ cover additional clock mode dependent
control signals like strobe signals (clock mode 1), clock gating signals (clock mode 4) or
synchronization signals (clock mode 5). The last column describes the function of signal
TxCLK which in some clock modes can be enabled as output signal monitoring the
effective transmit clock or providing a time slot control signal (clock mode 5).
The following is an example of how to read Table 21:
For clock mode 6b (row ’6b’) the TRM clock (column ’TRM’) is supplied by the baudrate
generator (BRG) output divided by 16 (source BRG/16). The BRG (column ’BRG’) is
derived from the internal oscillator which is supplied by pin XTAL1 and XTAL2.
The REC c lock (column ’REC’) i s supp lied b y t he inte rnal D PLL which itself is su pplied
by the baud rate generator (column ’DPLL’) again.
Note: The REC clock is DPLL clock divided by 16.
If enabled by bit ’TOE’ in register CCR2 the resulting transmit clock can be monitored to
pin TxCLK (last column, row ’6b’).
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 140 Data Sheet 09.98
The clocking concept is illustrated in a block diagram manner in the following figure:
Additional control signals are not illustrated (please refer to the detailed clock mode
descriptions below).
Figure 41 Clock Supply Overview
Oscillator
XTAL1
XTAL2
RxD
BRG
0b
6a/b
7a/b
2a/b
3a/b
DPLL 16:1
RxCLK
TxCLK
f
DPLL
f
BRG
f
BRG/16
f
RxCLK
f
TxCLK
f
DPLL
f
BRG
f
RxCLK
f
TRM
Transmitter Receiver
f
REC
TTL
or
CRYSTAL
f
DPLL
f
BRG
f
BRG/16
f
RxCLK
f
TxCLK
3a
7a 0b
3b
7b
2b
6b 1
50a
2a
6a
4
2a/b
3a
6a/b
7a
3b
7b 0a/b
1
5
4
settings controlled by:
register CCR0, bit field ’CM’
selects the clock mode number
register CCR0, bit ’SSEL’
selects the additional a/b option
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 141 Data Sheet 09.98
7.4.1 Clo ck M ode s
7.4.1 . 1 Clock M ode 0 (0 a/0b)
Separate, externally generated receive and transmit clocks are supplied to the SCC via
their respective pins. The transmit clock may be directly supplied by pin TxCLK
(clock mode 0a) or generated by the internal baud rate generator from the clock supplied
at pin XTAL1 (clock mode 0b).
In clock mode 0b the re sul ting tra nsm it clo ck ca n be driv en out to pi n T xC LK i f e nab led
via bit ’TOE’ in register CCR2.
Figure 42 Clock Mode 0a/0b Configuration
RxCLK
CTS
, CxD, TCG
CD
, FSC, RCG
TxCLK
RTS
RxD
TxD
XTAL1
XTAL2
clock supply
1
2
RxCLK
CTS
, CxD, TCG
CD
, FSC, RCG
TxCLK
RTS
RxD
TxD
XTAL1
XTAL2
clock supply
1
or
(tx clock monitor output)
clock mode 0b
clock mode 0a
OSC
Ctrl.
Ctrl.
Ctrl.
Ctrl.
f
BRG
= f
OSC
/k
K=(n+1)/2
M
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 142 Data Sheet 09.98
7.4.1.2 Clock Mode 1
Externally generated RxCLK is supplied to both the receiver and transmitter. In addition,
a receive s trobe can be c onnected v ia CD and a tran smit stro be via TxCLK pin. These
strobe signals work on a per bit basis. This operating mode can be used in time division
multiplex applications or for adjusting disparate transmit and receive data rates.
Note: In Extended Transparent Mode (HDLC/SDLC), the above mentioned strobe
signals provide byte synchronization (byte alignment).
In ASYNC Mode, the above mentioned strobe signals provide character
synchronization (character alignment).
This means that the strobe signal needs to be detected once only to transmit or
receive a complete byte or character respectively.
Figure 43 Clock Mode 1 Configuration
RxCLK
CTS
, CxD, TCG
CD
, FSC, RCG
TxCLK
RTS
RxD
TxD
XTAL1
XTAL2
clock supply
1
clock mode 1
receive strobe
transmit strobe
RxD
CD
(rx strobe)
TxCLK
(tx strobe)
RxCLK
TxD
V
SS
(enables transmit)
Note: In extended transparent or ASYNC protocol mode the strobe signals need to be
detected once only to transmit or receive a complete byte or character respectively.
Thus byte/character alignment is provided in these modes.
Ctrl.
Ctrl.
123
123
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 143 Data Sheet 09.98
7.4.1.3 Clo ck M ode 2 ( 2 a/2b)
The BRG is drive n by an exte rnal clock (RxCLK pin ) and delivers a referen ce clock for
the DPLL which is 16 times of the resulting DPLL output frequency which in turn supplies
the internal rece ive cl ock. Dep ending on t he progra mming o f regist er CCR0 bit ’SSEL’,
the transmit cl ock will be eithe r an ext ernal input clock s ignal provided at pi n TxCLK in
clock mode 2a or the clock delivered by the BRG divided by 16 in clock mode 2b. In the
latter case, the transmit clock c an be driven out t o pin TxCLK if ena ble d v ia b it ’ TOE’ in
register CCR2.
Figure 44 Clock Mode 2a/2b Configuration
RxCLK
CTS
, CxD, TCG
CD
, FSC, RCG
TxCLK
RTS
RxD
TxD
XTAL1
XTAL2
clock supply
1
RxCLK
CTS
, CxD, TCG
CD
, FSC, RCG
TxCLK
RTS
RxD
TxD
XTAL1
XTAL2
clock supply
1
(tx clock monitor output)
clock mode 2b
clock mode 2a
BRG
DPLL 2
BRG
DPLL 16:1
Ctrl.
Ctrl.
Ctrl.
Ctrl.
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 144 Data Sheet 09.98
7.4.1.4 Clock Mode 3 (3a/3b)
The BRG is fed with an externally generated clock via pin RxCLK. Depending on the
value of bit ’SSEL’ in register CCR0 the BRG delivers either a reference clock for the
DPLL which is 16 times of the resulting DPLL output frequency or delivers directly the
receive and transmit clock (cloc k mode 3b). In the first ca se (clock mode 3a) the DPL L
output clock is used as receive and transmit clock.
Figure 45 Clock Mode 3a/3b Configuration
RxCLK
CTS
, CxD, TCG
CD
, FSC, RCG
TxCLK
RTS
RxD
TxD
XTAL1
XTAL2
clock supply
1
RxCLK
CTS
, CxD, TCG
CD
, FSC, RCG
TxCLK
RTS
RxD
TxD
XTAL1
XTAL2
clock supply
1
(tx clock monitor output)
clock mode 3b
clock mode 3a
BRG
DPLL
(tx clock monitor output)
BRG
Ctrl.
Ctrl.
Ctrl.
Ctrl.
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 145 Data Sheet 09.98
7.4.1.5 Clock Mode 4 (High Speed Interface Clock Mode)
Separate, externally gen erate d re cei ve and tran smi t c loc ks are su ppl ied vi a pins RxClk
and TxClk. In addition separate receive and transmit clock gating signals are supplied
via pins RCG and TCG. These gating signals work on a per bit basis.
Note: Clock mode 4 can only be applied in combination with High Speed Serial Mode
(register CCR0 bit HS =’1’). This setting optimizes the internal signal and clock
trees for high speed timing requirements.
Note: For correct operation only HDLC Address Mode 0, PPP and Extended
Transparent Mode should be used.
Figure 46 Clock Mode 4 (High Speed) Configuration
RxCLK
CTS, CxD,
TCG
CD, FSC,
RCG
TxCLK
RTS
RxD
TxD
XTAL1
XTAL2
clock supply
1
clock mode 4
transmit clock gate signal
receive clock gate signal
2
tx clock out signal
TxCLK
TCG
TxD
TxCLKout
signal delay
RxCLK
RCG
RxD
1 clock delay
Ctrl.
Ctrl.
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 146 Data Sheet 09.98
The transmit clock supplied at pin TxCLK can be switched out to pin RTS in phase with
the transmit data on pin TxD if bit ’TCLKO’ in register CCR1 is set. Due to internal signal
delays the transmit data output signal delay to TxCLK may be high with regard to the total
clock period of 19.2 nanoseconds which is the minimum high speed clock period.
Therefore the transmit clock is supplied to pin RTS such that TxD signal has a small
delay to this mon itor clock which might be usef ul for conn ection to external transc eiver
devices.
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 147 Data Sheet 09.98
7.4.1.6 Clock Mode 5
This operation mode has been designed for application in time-slot oriented PCM
systems.
Note: For correct operation NRZ data coding/encoding should be used.
The receive and transmit clock are common for each channel and must be supplied
externally via pin RxC LK. The SCC receiv es an d transm its onl y d uring fix ed tim e-slots.
Either one time-slot
– of programmable width (1 … 512 bit, via TSAR and TSAX registers), and
– of programmable location with respect to the frame synchronization signal (via pin
FSC)
or up to 32 time-slots
– of constant width (8 bits), and
– of programmable location with respect to the frame synchronization signal (via pin
FSC)
can be selected.
The time-slot locations can be programmed independently for receive and transmit
direction via TSAX/TSAR and PCMMTX/PCMMRX registers.
Depending on the value program med v ia th ose re gisters, th e rece ive/trans mit t ime-slot
starts with a delay of 1 (minimum delay) up to 1024 clock periods following the frame
synchronization signal.
Figure 47 shows how to select a time-slot of programmable width and location and
figure 48 shows how to select one or more time-slots of 8-bit width.
If bit ’TOE’ in register CCR0 is set, the selected transmit time-slot(s) is(are) indicated at
an output st atus sig nal v ia pin TxCLK, wh ich is driv en to ‘l ow ’ durin g the ac tiv e trans mit
window.
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 148 Data Sheet 09.98
Figure 47 Selecting one time-slot of programmable delay and width
31 24 23 16 15 8 7 0
TTSN TCS TCC0
31 24 23 16 15 8 7 0
RTSN RCS RCC0
TTSA: Transmit Time Slot Assignment Register
RTSA: Receive Time Slot Assignment Register
TEPCM = ’0’: TPCM Mask Disabled
REPCM = ’0’: RPCM Mask Disabled
TS delay (transmit):
1 + TTSN*8 + TCS
(1...1024)
TS delay (receive):
1 + RTSN*8 + RCS
(1...1024)
TS width (transmit):
TCC
(1...512 clocks)
TS width (receive):
RCC
(1..512)
FSC
RxCLK
active
time slot
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 149 Data Sheet 09.98
Note: If time-slot 0 is to be selected, the DELAY has to be as long as the PCM frame
itself to achieve synchronization (at least for the 2nd and subsequent PCM
frames): DEL AY = PCM frame length = 1 + xTSN*8 + xC S. xTSN and xCS have
to be set appropriately.
Example: Time-slot 0 in E1 (2.048 Mbit/s) system has to be selected.
PCM frame length is 256 clocks. 256 = 1+ xTSN*8 + xCS. => xTSN = 31, xCS = 7.
Note: In extended transparent mode the width xCC of the selected time-slot has to be
n
×
8 bit because of character synchronization (byte alignment). In all other modes
the width can be used to define windows down to a minimum length of one bit.
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 150 Data Sheet 09.98
Figure 48 Selecting one or more time-slots of 8-bit width
The common transmit and receive clock is supplied at pin RxCLK and the common frame
synchronisation signal at pin FSC. The "strobe signals" for active time slots are
generated in ternally by the time slot assigner block (TSA) indepen dent in transmit and
receive direction.
When th e tran sm it an d rec eiv e PC M m as ks are enabl ed, bit fiel ds ’TCC’ and ’RCC’ are
ignored because of the constant 8-bit time slot width.
31 24 23 16 15 8 7 0
TTSN TCS
TCC
1
TTSA: Transmit Time Slot Assignment Register
TEPCM = ’1’: TPCM Mask Enabled
TS delay (transmit):
1 + TTSN*8 + TCS
(1...1024)
TS delay (receive):
1 + RTSN*8 + RCS
(1...1024)
31 24 23 16 15 8 7 0
TPCMM: Transmit PCM Mask Register
...
1
3
1
17
TS0 TS1 TS2 TS3 TS4 TS5 TS16 TS17
8 bit
31 24 23 16 15 8 7 0
RTSN RCS
RCC
1
RTSA: Receive Time Slot Assignment Register
REPCM = ’1’: TPCM Mask Enabled
31 24 23 16 15 8 7 0
RPCMM: Receive PCM Mask Register
FSC
RxCLK
active
time slot
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 151 Data Sheet 09.98
Figure 49 Clock Mode 5 Configuration
Note: The transmit time slot delay and width is programmable via bit fields ’TTSN’, ’TCS’
and ’TCC’ in register TTSA.
The receive time slot delay and width is programmable via bit fields ’RTSN’, ’RCS’
and ’RCC’ in register RTSA.
RxCLK
CTS
, CxD, TCG
CD,
FSC
, RCG
TxCLK
RTS
RxD
TxD
XTAL1
XTAL2
clock supply
1
clock mode 5
time slot indicator signal
Time Slot
Assigner
(TSA)
RxCLK
FSC
internal
tx strobe
TS delay TS width
TxCLK
TS-Control
TxD
internal
rx strobe
TS delay TS width
RxD
012n ... 0n
Ctrl.
Ctrl.
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 152 Data Sheet 09.98
7.4.1.7 Clock Mode 6 (6a/6b)
This clock mode is identical to clock mode 2a/2b except that the clock source of the BRG
is supplied at pin XTAL1.
The BRG is driven by the internal oscillator and delivers a reference clock for the DPLL
which is 16 times the resulting DPLL output frequency which in turn supplies the internal
receive clock. Depending on the programming of register CCR0 bit ’SSEL’, the transmit
clock will be either a n external input clock signal provided at pi n TxCLK in clock mode
6a or the clock delivered by the BRG divided by 16 in clock mode 6b. In the latter case,
the transmit clock can be driven out to pin TxCLK if enabled via bit ’TOE’ in register
CCR2.
Figure 50 Clock Mode 6a/6b Configuration
RxCLK
CTS
, CxD, TCG
CD
, FSC, RCG
TxCLK
RTS
RxD
TxD
XTAL1
XTAL2
clock supply
1
RxCLK
CTS
, CxD, TCG
CD
, FSC, RCG
TxCLK
RTS
RxD
TxD
XTAL1
XTAL2
(tx clock monitor output)
clock mode 6b
clock mode 6a
BRG
DPLL
BRG
DPLL 16:1
or
V
SS
V
SS
or
OSC
OSC
Ctrl.
Ctrl.
Ctrl.
Ctrl.
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 153 Data Sheet 09.98
7.4.1.8 Clo ck M ode 7 ( 7 a/7b)
This clock mode is identical to clock mode 3a/3b except that the clock source of the BRG
is supplied at pin XTAL1.
The BRG is driven by the internal oscillator. Depending on the value of bit ’SSEL’ in
register CCR0 the BRG delivers either a reference clock for the DPLL which is 16 times
the resulting DPLL output frequency (clock mode 7a) or delivers directly the receive and
transmit clock (clock mode 7b). In clock mode 7a the DPLL output clocks receive and
transmit data.
Figure 51 Clock Mode 7a/7b Configuration
RxCLK
CTS
, CxD, TCG
CD
, FSC, RCG
TxCLK
RTS
RxD
TxD
XTAL1
XTAL2
RxCLK
CTS
, CxD, TCG
CD
, FSC, RCG
TxCLK
RTS
RxD
TxD
XTAL1
XTAL2
(tx clock monitor output)
clock mode 7b
clock mode 7a
BRG
DPLL (tx clock monitor output)
BRG
or
V
SS
V
SS
OSC
OSC
or
Ctrl.
Ctrl.
Ctrl.
Ctrl.
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 154 Data Sheet 09.98
7.4.2 Baud Rate Generator (BRG)
Each serial channel provides a baud rate generator (BRG) whose division factor is
controlled by register BRR. The function of the BRG depends on the selected clock
mode.
The clock division factor k is calculated by:
Table 22 BRR Register and Bit-Fields
Offset
Addr. Access
Type Controlled
by Reset
Value Register Name
012CH
01ACH
022CH
02ACH
r/w CPU 00000000HBRR:
Baud Rate Register
Bit-Fields
Pos. Name Default Description
11..8 BRM 0 Baud Rate Factor M,
range M = 0..15
5..0 BRN 0 Baud Rate Factor N
range N = 0..63
kN1+()2M
×=
fBRG fin k⁄=
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 155 Data Sheet 09.98
7.4.3 Clock Recovery (DPLL)
The SCC o ffers the a dvantage o f recoverin g the recei ved clock f rom the received d ata
by means of internal DPLL circuitry, thus eliminating the need to transfer additional clock
information vi a a separa te serial cloc k line. For this purp ose, the DPLL i s supplied with
a ‘referen ce clock ’ from the BRG w hich is 16 times the expec ted data c lock rate (clock
mode 2, 3a, 6, 7a). The transmit clock may be obtained by dividing the output of the BRG
by a constant factor of 16 (clock mode 2b, 6b; bit ’SSEL’ in register CCR0 set) or also
directly from the DPLL (clock mode 3a, 7a).
The main task of the DPLL is to derive a receive clock and to adjust its phase to the
incoming data stream in order to enable optimal bit sampling.
The mechani sm for c lock recov ery depend s on the se lected data encod ing (see “Data
Encoding” on page 162).
The following functions have been implemented to facilitate a fast and reliable
synchronization:
Interference Rejection and Spike Filtering
Two or more edges in the same directional data stream within a time period of 16
reference clocks are considered to be interference and consequently no additional clock
adjustment is performed.
Phase Adjustment (PA)
Referring to figures 52, 53 and 54, in the case where an edge appears in the data stream
within the PA fields of the time window, the phase will be adjusted by 1/16 of the data.
Phase Shift (PS) (NRZ, NRZI only)
Referring to fig ure 52 in the case where an edge appea rs in the dat a stream within the
PS field of the time window, a second sampling of the bit is forced and the phase is
shifted by 180 degre es.
Note: Edges in all other parts of the time window will be ignored.
This operation facilitates a fast and reliable synchronization for most common
applications. Above all, it implies a very fast synchronization because of the phase shift
feature: one edge on the re ceived data st ream is enough for the DPLL to sy nchronize,
thereby eliminating the need for synchronization patterns, sometimes called preambles.
However, in case of extremely high jitter of the incoming data stream the reliability of the
clock recovery cannot be guaranteed.
The SCC offers the option to disable the Phase Shift function for NRZ and NRZI
encodings by setting bit ’PSD’ in register CCR0. In this case, the PA fields are extended
as shown in figure 53.
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 156 Data Sheet 09.98
Now, the DPLL is more insensitive to high jitter amplitudes but needs more time to reach
the optimal sampling position. To ensure correct data sampling, preambles should
precede the data information.
Figures 52, 53 and 54 explain the DPLL algorithms used for the different data
encodings.
Figure 52 DPLL Algorithm for NRZ and NRZI
Coding with Phase Shift Enabled ( C CR0:PSD = ‘ 0’)
Figure 53 DPLL Algorithm for NRZ and NRZI
Encoding with Phase Shift Disabled (CCR0:PSD = ‘1’)
012345678910 11 12 13 14 15
0 +PA PS -PA 0
Bit Cell
DPLL
Count
Output
DPLL
Correction
ITD01806
0123456789101112131415
0 +PA -PA 0
Bit Cell
DPLL
Count
Output
DPLL
Correction
ITD04820
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 157 Data Sheet 09.98
Figure 54 DPLL Algorithm for FM0, FM1 and Manchester Coding
To supervise correct function when using bi-phase encoding, a status flag and a
maskable interrupt inform about synchronous/asynchronous state of the DPLL.
0123456789101112131415
0 +PA - ignore - -PA 0
Bit Cell (FM Coding)
DPLL
Count
Clock
Transmit
Correction
ITD01807
76543210
Bit Cell (Manchester Coding)
+PA - ignore -
Receive
Clock
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 158 Data Sheet 09.98
7.5 SCC Interrupt Interface
Special events in the SCC are indicated by interrupts either for receive direction or for
transmit di rection. Accordingl y individual transmit and receive interrupt queues located
in the shared memory are provided for each of the 4 SCCs. The interrupts generated by
the SCC are forwarded as interrupt vectors to the channel specific and direction specific
interrupt queues. See “Interrupt Interface” on page 71 and “Interrupt Vector
Definitions” on page 280 .
Each interrupt indicated by the interrupt vector or interrupt status register (ISR) can
selectively be maske d (disabled) by setting the corresponding bit in the interrupt mask
register IMR.
Note: A dedicated transmit interrupt vector ma y cont ain recei ve interrupt indicati ons. In
that case the receive interrupt indications can be ignored, since a dedicated
receive interrupt vector is generated, additionally.
Note: A dedi cated rece ive interru pt vector ma y contai n transmit in terrupt indi cati ons. In
that case the transmit interrupt indications can be ignored, since a dedicated
transmit interrupt vector is generated, additionally.
In interrup t visible mode (C CR0:VIS = ’1’), masked i nterrupt stat us bits neither act ivate
the interrupt s ignal nor genera te an interrupt ve ctor, but are indi cated in the respec tive
interru pt status register ISR.
This mode is useful when some interrupt st atus bits are to ge nerate an interrup t vector
and other status bits are to be polled in the individual interrupt status register.
When using visible mode, only unmasked interrupt status bits are reset when the
interrupt status register is read.
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 159 Data Sheet 09.98
7.6 High Speed Channel Operation (PEB 20534H-52 only)
Any channel of the DSCC4 may be configured for operation at data rates up to 52 MBit/s.
To configure o ne SC C for hig h s peed ope ratio n, bi t ’H S’ in re gis ter C CR0 has to be set
(allows data rates above 10 MBit/s) and clock mode 4 has to be selected to configure the
internal clock unit appropriately.
The protocols supported at this high data rate are limited to HDLC address mode 0, PPP
and extended transparent mode (PPP modes up to 45 MBit/s).
The high speed mode will operate with clocks provided on TxCLK (transmitter) and
RxCLK (receiver). Clock gating is supplied on TCG and RCG pins to allow external
transceivers (HSSI, DS3, etc.) to enable/disable frame/block bursts. When the clock
gating signals are inactive, the SCC will not latch data from pin RxD nor will it output
data on pin TxD (idle ’1’ is transmitted when TCG is inactive).
Beside clock gating via signals TCG/RCG, clock gapping is also supported in high speed
mode.
In transmit direction one clock delay is inserted between detecting an active transmit
gate signal TCG and the first transmit data bit driven to the line.
A transmit clock which is in phase to the data bits on pin TxD can be provided on pin RTS
by setting bit ’TxCLKO’ in register CCR1.
Depending on the expected data traffic the user can select appropriate depth for the
central transmit FIFO. In addition a second transmit FIFO threshold (register FIFOCR4)
controls data transfer to the SCC to prevent data underrun conditions when a new frame
is started.
The central rec eive FIFO is alloc ated dy namica lly by th e SCC as the data are rec eived
on the serial line.
7.7 Serial Bus Configuration Mode
Beside the poin t-to-point configuratio n, the SCC effectively supports point-to-multipo int
(pt-mpt, or bus) configurations by means of internal idle and collision detection/collision
resolution methods.
In a pt-mpt configuration, comprising a central station (master) and several peripheral
stations (slaves), or in a multimaster configuration, data transmission can be initiated by
each station over a common transmit line (bus). In case more than one station attempts
to transmit data simultaneously (collision), the bus has to be assigned to only one
station.
– In HDLC/SDLC mode, a collision-resolution procedure is implemented by the SCC.
Bus assignment is bas ed on a priority me ch anis m with rotating p ri oriti es. This allows
each station a bus access within a predetermined maximum time delay (deterministic
CSMA/CD), no matter how many transmitters are connected to the serial bus.
– In BISYNC mode, the collision-resolution is implemented by the microprocessor.
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 160 Data Sheet 09.98
– In ASYNC mode, a bus configuration is not recommended.
Note: In high speed operation the usage of bus configuration is not supported.
Prerequisites for bus operation are:
• NRZ encoding
• ‘OR’ing of data from every transmitter on the bus (this can be realized as a wired-OR,
using the TxD open drain capability)
• Feedback of bus information (CxD input).
The bus configuration is selected via register CCR0.
Note: Central clock supply for each station is not necessary if both the receive and
transmit clock is recovered by the DPLL (clock modes 3a, 7a). This minimizes the
phase shi ft between the individual transmit clock s.
The bus configuration mode operates independently of the clock mode, e.g. also
together with clock mode 1 (receive and transmit strobe operation).
7.7.1 Serial Bus Access Procedure
The idle st ate of the bus is identified by eigh t or more con secutive ‘1’s. When a devic e
starts tr ans mis sio n of a frame , the bus is re cog niz ed to be b usy by the other de vic es at
the moment the first ‘zero’ is transmitted (e.g. first ‘zero’ of the opening flag in
HDLC mode).
After the frame has been transmitted, the bus becomes available again (idle).
Note: If the bus is o ccupied by other transmitters and /or there is no tran smit request in
the SCC, logical ‘1’ will be continuously transmitted on TxD.
7.7.2 Serial Bus Collisions and Recovery
During the transmission, the data transmitted on TxD is compared with the data on CxD.
In case of a mismatch (‘1’ sent and ‘0’ detected, or vice versa) data transmission is
immediately aborted, and idle (logical ‘1’) is transmitted.
HDLC/SDLC: Transmission will be initiated again by the SCC as soon as possible if the
first part of the frame is still present in the SCC transmit FIFO. If not, an XMR interrupt is
generated.
Since a ‘zero’ (‘low’) on the bus prevails over a ‘1’ (high impedance) if a wired-OR
connection is implemented, and since the address fields of the HDLC frames sent by
different stations norma lly differ from o ne another, t he fac t t hat a collision h as oc curred
will be detected prior to or at the latest within the address field. The frame of the
transmitter with the highest temporary priority (determined by the address field) is not
affected and is transmitted successfully. All other stations cease transmission
immediately and return to bus monitoring state.
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 161 Data Sheet 09.98
BISYNC: Transmitter and SCC transmit FIFO are reset and pin TxD goes to ‘1’. The
XMR interrupt is provided which requests the microprocessor to repeat the whole
message or block of characters.
ASYNC: Bus configuration is not recommended.
Note: If a wired-OR connection has been realized by an external pull-up resistor without
decoupling, the data output (TxD) can be used as an open drain output and
connected directly to the CxD input.
For correct identification as to which frame is aborted and thus has to be repeated
after an XMR interrupt has occurred, th e contents of SC C transmit FIFO have to
be unique, i.e. SCC transmit FIFO as well as the central transmit FIFO should not
contain d ata of more than o ne frame. For this p urpose new d ata my be provi ded
to the DMA controller only after ’ALLS’ interrupt status is detected.
7.7.3 Serial Bus Access Priority Scheme
To ensure that all competing stations are given a fair access to the transmission medium.
Once a station has successfully completed the transmission of a frame, it is given a lower
level of priority . This priority mec han ism is bas ed o n the req uirem ent th at a sta t ion may
attempt transmitting only when a determined number of consecutive ‘1’s are detected on
the bus.
Normall y, a transmi ssion ca n star t when eigh t consec utive ‘1’ s on the b us are detec ted
(through pin CxD). When an HDLC frame has been successfully transmitted, the internal
priority class is decreased. Thus, in order for the same station to be able to transmit
another frame , ten c onsecutiv e ‘1’s on the bus must be d etected . This gua rantees that
the transmission requests of other stations are satisfied before the same station is
allowed a second bus access. When ten consecutive ‘1’s have been detected,
transmission is allowed again and the priority class (of all stations) is increased (to eight
‘1’s).
Inside a priority class, the order of transmission (individual priority) is based on the HDLC
address, as explained in the preceding paragraph. Thus, when a collision occurs, it is
always the station transmitting the only ‘zero’ (i.e. all other stations transmit a ‘one’) in a
bit position of the address field that wins, all other stations cease transmission
immediately.
7.7.4 Serial Bus Configuration Timing Modes
If a bus con figu ratio n has be en sele cted , the SC C provi des two tim ing mod es, diffe ring
in the time interval between sending data and evaluation of the transmitted data for
collision dete cti on.
• Timing mode 1 (CCR0:SC1, SC0 = ‘01’)
Data is output with the rising edge of the transmit clock via the TxD pin, and evaluated
1/2 a clock period later at the CxD pin with the falling clock edge.
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 162 Data Sheet 09.98
• Timing mode 2 (CCR0:SC1, SC0 = ‘11’)
Data is output with the falling clock edge and evaluated with the next falling clock
edge. Th us one co mplete clock period i s a vai labl e between da ta output an d c olli sion
detection.
7.7.5 Functions Of Signal RTS in HDLC Mode
In clock modes 0 and 1, the RTS output can be programmed via regist er CCR1 (SOC
bits) to be active when data (frame or character) is being transmitted. This signal is
delayed by one clock period with respect to the data output TxD, and marks all data bits
that could be transmitted without collision (see figure 55). In this way a configuration
may be implemented in which the bus access is resolved on a local basis (collision bus)
and where the data are sent one clock period later on a separate transmission line.
Figure 55 Request-to-Send in Bus Operation
Note: For details on the functions of the RTS pin refer to “Modem Control Signals
(RTS, CTS, CD)” on page 164.
7.8 Data Encoding
The SCC supports the following coding schemes for serial data:
– Non-Return-To-Zero (NRZ)
– Non-Return -To-Zero-Inverted (NR ZI)
– FM0 (also known as Bi-Phase Space)
– FM1 (also known as Bi-Phase Mark)
– Manchester (also known as Bi-Phase)
7.8.1 NRZ and NRZI Encoding
NRZ: The signal level co rre sp onds to the valu e of t he d ata bit. By prog ra mmi ng b i t DI V
(CCR1 register), the SCC may invert the transmission and reception of data.
NRZI: A logical ‘0’ is indicated by a transition and a logical ‘1’ by no transition at the
beginning of the bit cell.
ITT00242
Collision
TxD
CxD
RTS
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 163 Data Sheet 09.98
Figure 56 NRZ and NRZI Data Encoding
7.8.2 FM0 and FM1 Encoding
FM0: An edge occurs at the beginning of every bit cell. A logical ‘0’ has an additional
edge in the center of the bit cell, whereas a logical ‘1’ has none. The transmit clock
precedes the receive clock by 90°.
FM1: An edge occurs at the beginning of every bit cell. A logical ‘1’ has an additional
edge in the center of the bit cell, a logical ‘0’ has none. The transmit clock precedes the
receive clock by 90°.
Figure 57 FM0 and FM1 Data Encoding
0110010
ITD05313
Transmit/
Receive Clock
NRZ
NRZI
110010
ITD01809
Receive
Clock
FM0
FM1
Transmit
Clock
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 164 Data Sheet 09.98
7.8.3 Manchester Encoding
Manchester: In the first half of the bit cell, th e physical sign al level correspon ds to the
logical value of the data bit. At the center of the bit cell this level is inverted. The transmit
clock precedes the receive clock by 90°. The bit cell is shifted by 180° in comparison with
FM coding.
Figure 58 Manchester Data Encoding
7.9 Modem Control Signals (RTS, CTS, CD)
7.9.1 RTS/CTS Handshaking
The SCC provides two pins (RTS, CTS) per serial channel supporting the standard
request-to-send modem handshaking procedure for transmission control.
A transmit request will be indicated by outputting logical ‘0’ on the request-to-send output
(RTS). It is also possible to control the RTS output by software. After having received the
permission to transmit (CTS) the SCC starts data transmission.
HDLC/SDLC and BISYNC: In the case where permission to transmit is withdrawn in the
course of transmission, the frame is aborted and IDLE is sent. After transmission is
enabled again by re-activation of CTS, and if the beginning of the frame is still available
in the SCC, the frame will be re-tran smitted (se lf-recovery). However, if th e permission
to transmit is withdrawn aft er th e d ata av ailable in th e s hadow part o f th e SC C transmit
FIFO has been completely transmitted and the pool is released, the transmitter and the
SCC transmit FIFO are reset, the RTS output is deactivated and an interrupt (XMR) is
generated.
Note: For correct identification as to which frame is aborted and thus has to be repeated
after an XMR interrupt has occurred, the contents of
SCC transmit FIFO
have to
be unique, i.e. SCC transmit FIFO as well as central transmit FIFO should not
contain data of more than one frame, which could happen if transmission of a new
110010
ITD01810
Receive
Clock
Manchester
Transmit
Clock
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 165 Data Sheet 09.98
frame is started by providing new data to the DMA controller too early. For this
purpose the All Sent interrupt (ISR.ALLS) ha s to be w aited for b efore forwarding
new data to the DMA controller.
ASYNC: In the case where permission to transmit is withdrawn, transmission of the
current character is completed. After that, IDLE is sent. After transmission is enabled
again by re-activation of CTS, the next available character is sent out.
Note: In the case where permission to transmit is not required, the CTS input can be
connected directly to
VSS
.
Additiona lly, any transition on the CTS in put pin w ill generat e an interru pt indicate d via
register ISR, if this function is enabled by setting the ’CSC’ bit in register IMR to ’0’.
Figure 59 RTS/CTS Handshaking
Beyond this standard RTS function, signifying a transmission request of a frame
(Request To Send), in HDLC mode the RTS output may be programmed for a special
function via SOC1, SOC0 bits in the CCR1 register. This is only available if the serial
channel is operating in a bus configuration mode in clock mode 0 or 1.
• If SOC1, SOC0 bits are set to ‘11’, the RTS output is active (= low) during the
reception of a frame.
• If SOC1, SOC0 bits are s et to ‘10’, t he RTS output fun ction is di sabled and the RT S
pin remains always high.
ITT00244
Sampling
CTS
TxCLK
TxD
RTS
~
~~
~~
~~
~
PEB 20534
Serial Communication Controller (SCC) Cores
Semiconductor Group 166 Data Sheet 09.98
7.9.2 Carrier Detect (CD) Receiver Control
Similar to the RTS/CTS control for the tr ansmitter, the SCC supports the carrier detec t
modem control function for the serial receiver if the Carrier Detect Auto Start (CAS)
function is programmed by setting th e ’CAS’ bit in registe r CCR1. This function is always
available in clock modes 0, 2, 3, 6, 7 via the CD pin. In clock mode 1 the CD function is
not supported. See table 21 for an overview.
If the CAS function is selected, the receiver is enabled and data reception is started when
the CD input is detected to be high. If CD input is set to ‘low’, reception of the current
character (byte) is still completed.
7.10 L ocal Loop Test Mode
To provide fast and efficient testing, the SCC can be operated in a test mode by setting
the ’TLP’ bit in register CCR1. The on-chip serial data input and output signals
(TxD,RxD) are connected, generating a local loopback. As a result, the user can perform
a self-test of the SCC.
Figure 60 SCC Test Loop
Note: The transmit data is not disconnected from pin TxD during test loop operation, i.e.
transmit data is always provided to pin TxD.
Note: An sufficient clock mode must be used for test loop operation such tha t rec eiver
and transmitter operate with the same frequencies depending on the clock supply
(e.g. clock mode 2b or 6b).
Note: Test loop operation is not supported in high speed channel operation
(clock mode 4).
SCC transmit
logic
SCC receive
logic
TLP=’0’
TLP=’1’
RxD
TxD
PEB 20534
Detailed Protocol Description
Semiconductor Group 167 Data Sheet 09.98
8 Detailed Protocol Description
The following table 23 prov ides an overview of all su pported protocol mod es and their
assignment to three major protocol engines HDLC, ASYNC, BISYNC.The protocol
engine of each SCC is selected via bit field ’SM’ in register CCR0. The HDLC Sub Modes
are selected via addition al bit fields in registers CCR0 an d CCR1.
PEB 20534
Detailed Protocol Description
Semiconductor Group 168 Data Sheet 09.98
1) Extended transparent is a fully bit-transparent transmit/reception mode which is treated as a sub-mode of the
HDLC/SDLC block.
All modes are discussed in details in this chapter.
Table 23 Protocol Mode Overview
Protocol
Engine: Protocol Mode:
Register
CCR0
Setting:
Register CCR1 Setting:
bit fields
MDS, ADM bit field
PPPM
HDLC/
SDLC
SM = ’00’
HDLC auto mode 16 bit MDS = ’00’
ADM = ’1’ PPPM = ’00’
8 bit MDS = ’00’
ADM = ’0’
HDLC non auto mode 16 bit MDS = ’01’
ADM = ’1’
8 bit MDS = ’01’
ADM = ’0’
HDLC address mode 1 MDS = ’10’
ADM = ’1’
HDLC address mode 0 MDS = ’10’
ADM = ’0’
PPP mode asynchronous MDS=’xx’
ADM=’x’ PPPM = ’10’
bit synchronous PPPM = ’11’
octet synchronous PPPM = ’01’
Extended transparent mode1) MDS = ’11’
ADM=’x’ PPPM = ’00’
Register CCR0 Setting
(bit BCR):
ASYNC
SM = ’11’ Asynchronous mode BCR = ’1’
Isochronous mode BCR = ’0’
Register CCR1 Setting
(bit EBIM):
BISYNC
SM = ’10’ Bisynchronous mode EBIM = ’1’
Monosynchronous mode EBIM = ’0’
PEB 20534
Detailed Protocol Description
Semiconductor Group 169 Data Sheet 09.98
8.1 HDLC/SDLC Protocol Modes
The HDLC controller of each serial channel (SCC) can be programmed to operate in
various modes, which are different in the treatment of the HDLC frame in receive
direction. Thus, the receive data flow and the address recognition features can be
performed in a very flexible way satisfying almost any application specific requirements.
There are 4 di fferen t H DL C ope ra tin g modes whic h c an b e selected vi a reg ister CCR1.
Two more protocol modes (PPP, Extended Transparent Mode) are treated as sub-
modes of the HDLC controller part.
8.1.1 HDLC
The following table provides an overview of the different address comparison
mechanisms in HDLC operating modes:
Table 24 Address Comparison Overview
Mode Address Field High Address Byte Low Address Byte
Auto Mode,
Non-Auto Mode
16 bit FEH
1111 11 C/R 02
RAL1
RAL2
FCH
1111 11 C/R 02
RAL1
RAL2
RAH1 RAL1
RAH2 RAL2
8 bit RAL1 -
RAL2 -
Address
Mode 1
8 bit FEH-
FCH-
RAH1 -
RAH2 -
Address
Mode 0
None - -
PEB 20534
Detailed Protocol Description
Semiconductor Group 170 Data Sheet 09.98
8.1.1.1 Auto Mode
Characteristics: Window size 1, random message length, address recognition.
The SCC processes autonomously all numbered frames (S-, I-frames) of an HDLC
protocol. The HDLC control field, I-field data of the frames and an additional status byte
are temporarily stored in the SCC receive FIFO.
Depending on the selected address mode, the SCC can perform a 2-byte or 1-byte
address recognition. If a 2-byte address field is selected, the high address byte is
compared with the fixed value FEH or FCH (group address) as well as with
two individually programmable values in RAH1 and RAH2 registers. According to the
ISDN LAPD protocol, bit 1 of the high byte address will be interpreted as
COMMAND/RESPONSE bit (C/R), dependent on the setting of the CRI bit in RAH1, and
will be excluded from the address comparison.
Simi larly , two c omp arison valu es can be pr ogramm ed in speci al re gist ers (RA L1, RAL 2)
for the low address byte. A valid address will be recognized in case the high and low byte
of the address field correspond to one of the compare values. Thus, the SCC can be
called (addressed) with 6 different address combinations, however, only the logical
connection identified through the address combination RAH1, RAL1 will be processed in
the auto-mo de, all others in the non auto-mode. HDLC frames with address fields th at
do not match any of the address combinations, are ignored by the SCC.
In the case of a 1 -byte ad dress, RAL1 and RAL2 w ill be used as c omparison v alues in
the RADR register. According to the X.25 LAPB protocol, the value in RAL1 will be
interpreted as COMMAND and the value in RAL2 as RESPONSE.
The address bytes can be masked to allow sel ective broadc ast frame recognition. For
further information see “Receive Address Handling” on page 185.
8.1.1.2 Non Auto Mode
Characteristics: address recognition, arbitrary window size.
All frames with valid addresses (address recognition identical to auto-mode) are
forwarded directly to the system memory.
The HDLC con trol f ield, I-field data an d an add itio nal status byte are tem pora ri ly sto re d
in the SCC receive FIFO.
In non-auto-mode, all frames with a valid address are treated similarly.
The address bytes can be masked to allow selective broadcast frame recognition.
PEB 20534
Detailed Protocol Description
Semiconductor Group 171 Data Sheet 09.98
8.1.1.3 Address Mode 1
Characteristics: address recognition high byte.
Only the high byte of a 2-byte address field will be compared. The address byte is
compared with the fixed value FEH or FCH (group address) as well as with two
individually programmable values RAH1 and RAH2 in the RADR register. The whole
frame excluding the first address byte will be stored in the SCC receive FIFO.
The address bytes can be masked to allow selective broadcast frame recognition.
8.1.1.4 Address Mode 0
Characteristics: no address recognition
No address recognition is performed and each complete frame will be stored in the SCC
receive FIFO.
8.1.2 HDLC/PPP Protocol Mode
PPP (as described in RFC1662) can work over 3 modes: asynchronous HDLC,
synchronous HDLC, and octet synchronous. The DSCC4 supports asynchronous HDLC
PPP over ISDN or DDS circuits as well as synchronous HDLC PPP for use over dial -
up connections. The octet synchronous mode of PPP protocol (RFC 1662) supports PPP
over SONET applications.
Both the asynchronous HDLC PPP mode, as well as the synchronous HDLC PPP
modes, are submodes of the HDLC mode. Either mode is selected by configuring the
DSCC4 for the standard HDLC mode. In addition the appropriate PPP mode is selected
via bit field ’PPPM’ in register CCR2.
The DSCC4 prov ides logic to c onvert an HDLC frame t o an ASYNC charac ter stream
with the spe cified m apping f unction s. Layer 3 PPP funct ions are n ormally i mplemen ted
in software.
The PPP-support hardware allows software to perform segmentation and reassembly of
PPP payloads, a nd allows the DSCC4 to pe rform the asynchronous HDLC PPP or the
synchronous HDLC PPP protocol conversions as required for the network interface.
8.1.2.1 Bit Synchronous PPP
The DSCC4 transfers a data block from the shared memory, inserts HDLC Header
(Opening Flag), and appends the HDLC Trailer (CRC, Ending Flag). Zero-bit stuffing
algorithm is also performed. No character mapping is performed. The bit-synchronous
PPP mode differs from the HDLC mode (address mode 0) only in the abort sequence:
HDLC requires at least 7 ’1’ bit whereas PPP requires at least 15 ’1’ bit abort sequence.
PEB 20534
Detailed Protocol Description
Semiconductor Group 172 Data Sheet 09.98
For receive operation the DSCC4 monitors the incoming data stream for the Opening
Flag (7E Hex) to identify the beginning of a HDLC packet. Subsequent bytes are part of
data and are processed as normal HDLC packet including checking of CRC.
8.1.2.2 Octet Synchronous PPP
The DSCC4 transfers a data block from the shared memory, inserts HDLC Header
(Opening Flag), and appends the HDLC Trailer (CRC, Ending Flag). Beside this
standard HDLC operation, zero-bit stuffing is not performed, but character mapping is
performed.
For receive operation the DSCC4 monitors the incoming data stream for the Opening
Flag (7E Hex) to identify the beginning of a HDLC packet. Subsequent bytes are part of
data and are processed as normal HDLC packet including checking of CRC. Received
mapped characters are unmapped.
8.1.2.3 Aynchronous PPP
For transmit operation, the DSCC4 reads a data block from host memory, inserts the
HDLC header (Opening Flag), and appends the HDLC trailer (CRC, Ending Flag). Each
octet (including HDLC framing flags and idle flags) is converted into async character
format (1 start, 8 data bits, 1 stop bit) and then transmitted using the asynchronous
character formatter block.
In receive direction any async character is transferred into the DSCC4’s ASYNC
Character De-Formatting logic block, where it is translated back into the original
information octet. Th e information octets are then transferred to ho st the memory as in
HDLC address mode 0 operation.
8.1.3 Extended Transparent Mode
Characteristics: fully transparent
In extended transparent mode, fully transparent data transmission/reception without
HDLC framing is performed, i.e. without FLAG generation/recognition, CRC
generation/check, or bit stuffing. This allows user specific protocol variations.
In clock mode 1 and clock mode 5 byte alignment is provided.
8.1.4 HDLC Receive Data Processing Overview
The following two figures give an overview about the management of the received
frames in the different HDLC operating modes.
PEB 20534
Detailed Protocol Description
Semiconductor Group 173 Data Sheet 09.98
Figure 61 SCC Receive Data Flow (HDLC Modes) part a)
CRC16
FLAGFLAG (high) (low)
16 bit ADDR
CTRL I (data)
/32
RFIFO
RAH1,2 RAL1,2
optional 1) optional 2)
RSTA
RSTA
registers
involved
(address compare)
CRC16
FLAGFLAG (low)
8 bit
ADDR
CTRL I (data)
/32
RFIFO
RAL1,2
opt. 1) optional 2)
RSTA
RSTA
registers
involved
(address
compare)
CRC16
FLAGFLAG (high) (low)
16 bit ADDR
data
/32
RFIFO
RAH1,2 RAL1,2
optional 1) optional 2)
RSTA
RSTA
registers
involved
(address compare)
CRC16
FLAGFLAG (low)
8 bit
ADDR
data
/32
RFIFO
RAL1,2
opt. 1) optional 2)
RSTA
RSTA
registers
involved
(address
compare)
Automode 16Automode 8Non-Automode 16Non-Automode 8
option1)
The 8 or 16 bit address field can optionally be forwarded to the RFIFO (bit ’RADD’ in
register CCR2)
option 2)
The 16 or 32 bit CRC field can optionally be forwarded to the RFIFO (bit ’RCRC’ in register
CCR2)
PEB 20534
Detailed Protocol Description
Semiconductor Group 174 Data Sheet 09.98
Figure 62 SCC Receive Data Flow (HDLC Modes) part b)
8.1.5 HDLC Transmit Data Processing Overview
Two different types of frames can be transmitted:
– I-frames and
– transparent frames
as shown below.
CRC16
FLAGFLAG (low)
8 bit
ADDR
data
/32
RFIFO
RAH1,2
opt. 1) optional 2)
RSTA
RSTA
registers
involved
(address
compare)
CRC16
FLAG
FLAG data
/32
RFIFO
optional 2)
RSTA
RSTA
registers
involved
Address Mode 1Address Mode 0
option1)
The 8 or 16 bit address field can optionally be forwarded to the RFIFO (bit ’RADD’ in
register CCR2)
option 2)
The 16 or 32 bit CRC field can optionally be forwarded to the RFIFO (bit ’RCRC’ in
register CCR2)
PEB 20534
Detailed Protocol Description
Semiconductor Group 175 Data Sheet 09.98
Figure 63 SCC Transmit Data Flow (HDLC Modes)
For transmiss ion of I-frame s (s elected via bi t ’SXIF’ in regi ste r CC R2), the address and
control fields are generated autonomously by the SCC and the data in the corresponding
transmit data buffer is entered into the information field of the frame. This is possible only
if the SCC is operated in auto mode.
For (address) transparent frames, the address and the control fields have to be entered
in the transmit data buffer by software. This is possible in all operating modes and used
also in auto-mode for sending U-frames.
If bit ’X CRC’ in register CCR2 is set, the CRC c hecksum will n ot be genera ted internally.
The checksum has to be provided via the transmit data buffer as the last two or four bytes
by software. The transmitted frame will be closed automatically only with a (closing) flag.
CRC16
FLAG
FLAG
8 bit
ADDR
data
/32
TFIFO
XAD1
optional 2)
registers
involved
All Frames with automatic 8 or 16 bit Address and Control Byte Processing
(Auto Mode, Non-Auto Mode):
option 2)
Generation of the 16 or 32 bit CRC field can optionally be disabled by setting
bit ’XCRC’ in register CCR2, in which case the CRC must be calculated and
written into the last 2 or 4 bytes of the transmit FIFO, to immediately proceed
closing flag.
16 bitADDR
XAD2
CRC16
FLAG
FLAG data
/32
TFIFO
optional 2)
Frames without automatic Address and Control Byte Processing
(Address Mode 0):
CTRL
internally
generated
PEB 20534
Detailed Protocol Description
Semiconductor Group 176 Data Sheet 09.98
Note: The SCC does not check whether the length of the frame, i.e. the number of bytes,
to be transmitted makes sense according the HDLC protocol or not.
8.1.6 Procedural Support (Layer-2 Functions)
When ope rating in th e auto mod e, the SCC offers a hig h degre e of pro tocol su pport . In
addition to address recognition, the SCC autonomously processes all (numbered) S- and
I-frames (window size 1 only) with either normal or extended control field format
(modulo-8 or modulo-128 sequence numbers – selectable via register CCR1 bit ’MCS’).
The following functions will be performed:
– updating of transmit and receive counter
– evaluation of transmit and receive counter
– processin g of S commands
– flow co ntrol with RR/RNR
– generation of responses
– recognition of protocol errors
– transmission of S commands, if acknowledgement is not received
– continuous status query of remote station after RNR has been received
– programmable timer/ repeater f unctions .
In additio n, all unnu mbered f rames are forwarde d direct ly to the p rocessor. The logical
link can be initialized by software at any time (Reset HDLC Receiver by RRES command
in register CMDR).
Additional logical connections can be operated in parallel by software.
PEB 20534
Detailed Protocol Description
Semiconductor Group 177 Data Sheet 09.98
8.1.7 Full-Duplex LAPB/LAPD Operation
Initially (i.e. after RESET), the LAP controllers of the two serial channels are configured
to function as a combined (primary/secondary) station, where they autonomously
perform a subset of the balanced X.25 LAPB/ISDN LAPD protocol.
Reception of Frames:
The logical processing of received S-frames is performed by the SCC without
interrupti ng the host. The host is merely informed by interrupt of status changes in the
remote station (receiver ready / receiver not ready) and protocol errors (unacceptable
N(R), or S-frame with I field).
I-frames are also processed autonomously and checked for protocol errors. The I-frame
will not be accepted in the case of sequence errors (no interrupt is forwarded to the host),
but is immediat ely con firme d by an S-re spo nse . If th e ho st sets the SCC int o a ‘ receive
not ready’ status, an I-frame will not be accepted (no interrupt) and an RNR response is
transmitted. U-frames are always stored in the RFIFO and forwarded directly to the host.
The logical sequence and the reception of a frame in auto mode is illustrated in
figure 64.
Note: The state variables N(S), N(R) are evaluated within the window size 1, i.e. the
SCC checks only the least significant bit of the receive and transmit counter
regardless of the selected modulo count.
PEB 20534
Detailed Protocol Description
Semiconductor Group 178 Data Sheet 09.98
Figure 64 Processing of Received Frames in Auto Mode
ITD00230
Command
with p=1
?
Y
?
Ready
Rec. N
f=p
Trm RR
ActivRec.
Set RRNR
Response
PCE
Int :
1
Response
Trm RNR
f=p
?
Overflow
Data
N
:Int RME
Set RDO
Response
Trm RR
f=p
RMEInt :
N
Y
Y
N
=V
Y
N(S) (R)+1
Rec. Ready :Int RME
N
Set R D O
Data
Overflow
?
N
Y
ALLSInt :
Acknowledge
RESET Wait for
+1
(S) =
Y
=VN(R) (S)+1
Y
?
Acknowledge
Wait for
N
Y
:Int XMR
RES ET Wa i t for
Acknowledge
Acknowledge
RESET Wait for
ALLSInt :
Response
f=1
?
N
(S)+1N(R)=V
N
Y
Wait for
Acknowledge
?
NN
Y
?
CRC Error
Set CRCE
N
N
Set RAB
Aborted
?
Y
U Frame
1
YProt. Error
?
N
PCEInt :
:Int RME
Set CRCE N
?
CRC Error
Y
Set RA B
Aborted
?
N
Y
I Fram e
N
1
YProt. Error
?
N
or Abort
CRC Error
Y
RNR
?
1
RESET RRNR
?
,
YCRC Error
or Abort
N
?
Prot. Error
Y
N
:Int PCE
SREJREJRR,
1
Y
(R)+1(R) =VV
V(S)
V
V(S)
V=
(S) +1
Y
?
?
ALLS:Int
PEB 20534
Detailed Protocol Description
Semiconductor Group 179 Data Sheet 09.98
Transmission of Frames:
The SCC autonomously transmits S commands and S responses in the auto mode.
Either transparent or I-frames can be transmitted by the user. The software timer has to
be operated in the internal timer mode to transmit I-frames. After the frame has been
transmitted, the timer is self-started, the XFIFO is inhibited, and the SCC waits for the
arrival of a positive acknowledgement. This acknowledgement can be provided by
means of an S- or I-frame.
If no positive acknowledgement is received during time t1, the SCC transmits an S-
command (p = ‘1’), which must be answered by an S-response (f = ‘1’). If the S-response
is not received, the process is performed n1 times (in HDLC known as N2, refer to
register TIMR).
Upon the arrival of an acknowledgement or after the completion of this poll procedure
the XFIFO is enabled and an inte rrupt is g enerated. Inte rrupts ma y be triggered by the
following:
• message has been positively acknowledged (ALLS interrupt)
• message must be repeated (XMR interrupt)
• response has not been received (TIN interrupt).
In automode only when the ALLS interrupt has been issued data of a new frame
may be provided to the DMA controller!
Upon arrival of an RNR frame, the software timer is started and the status of the remote
station is polled periodically after expiration of t1, until the status ‘receive ready’ has been
detected. Th e user is info rmed via the a ppropriate interrupt . If no resp onse is received
after n1 times, a TIN interrupt, and t1 clock periods thereafter an ALLS interrupt is
generated and the process is terminated.
Note: The internal timer mode should only be used in the auto mode.
Transparent frames can be transmitted in all operating modes.
PEB 20534
Detailed Protocol Description
Semiconductor Group 180 Data Sheet 09.98
Figure 65 Timer Procedure/Poll Cycle
ITD00231
Wait for
Acknowledge
Set
?
?
2
?
with f=1
Response
Wait for
Acknowledge
?
RRNR
N
NN
Y Y
N
2
YY
=
?
(R) (S)+1VN
Y
N
1tLoad
Rec.RNRRec.RRIRec. Frame
T Proc. Activ
TINInt :
1
Load t1
Y
?
Ready
Rec. N
Command p=1
Trm RR
,,
Trm RNR
Command p=1
n1 n1-1=
N
?
Y
Y
?
N
n1 = 7
n1 = 0
Run Out
1 2
2
Load t 1
Trm RR/RNR
CMDR ; STI
Command p=1
Trm I Fram e
Acknowledge
Set wait for
InactivT Proc. 1
RNR
Set RRNR
Rec.
1t
Load n1
Load n1
11.06.1996 B/R
PEB 20534
Detailed Protocol Description
Semiconductor Group 181 Data Sheet 09.98
Examples
The interaction between SCC and the host during transmission and reception of I-frames
is illustrated in the following two figures. The flow control with RR/RNR of I-frames during
transmission/reception is illustrated in figure 66. Both, the sequence of the poll cycle
and protocol errors are shown in figure 67.
Figure 66 Transmission/Reception of I-Frames and Flow Control
Figure 67 Flow Control: Reception of S-Commands and Protocol Errors
RME
RME
WFA
Transmit with
FrameI
Confirm I Frame
ALLS
ALLS WFA
Reception FrameI
Transmit FrameI
RR(1)
(0.0)
I
RR(1)
I(0.1)
(1.1)
I
(1.2)
I
RR(2)
RR(0)f=1
RNR
RSC(RNR)
RSC(RR)
XMR
WFA
t1
t1 RR(0)p=1
RNR(0)f=1
RNR(0)
I(0.0)
RR(0)p=1
ALLS
RNR
RME
XRNR
RNR(0)
RR
(0.0)
I
RR(0)p=1
RR(0)f=1
RR(0)p=1
RR(0)f=1
I(0.0)
RR(1)
t1
t1
t1
RRp=1
Poll Cycle
Protocol Error
I
RR(0)p=1
RR(1)
RR(2)
ALLS
PCE
TIN
WFA
ALLS
RR(0)
WFA
RRp=1
(0.0)
PEB 20534
Detailed Protocol Description
Semiconductor Group 182 Data Sheet 09.98
Protocol Error Handling:
Depending on the error type, erroneous frames are handled according to table 25.
Note: The station variables ( V(S), V(R) ) are not changed.
8.1.8 Half-Duplex SDLC-NRM Operation
The LAP controllers of the two serial channels can be configured to function in a half-
duplex Normal Response Mode (NRM), where they operate as a slave (secondary)
station, by setting the NRM bit in the CCR1 register of the corresponding channel.
In contrast to the full-duplex LAP B/LAP D operation, where the combined
(primary + secondary) station transmits both commands and responses and may
transmit d ata at an y time, the NRM mod e allows only respo nses to be transmit ted and
the secondary station may transmit only when instructed to do so by the master (primary)
station. The SCC gets the permission to transmit from the primary station via an S-, or I-
frame with the poll bit (p) set.
The NRM mode can be profitably used in a point-to-multipoint configuration with a fixed
master-slave relationship, which guarantees the absence of collisions on the common
transmit line. It is the responsibility of the master station to poll the slaves periodically
and to handle error situations.
Prerequisite for NRM operation is:
– auto mode with 8-bit address field selected
Register CCR1 bit fields MDS1, MDS0, ADM = ‘000’
– external timer mode
Register TIMR bit TMD = ‘0’
– same transmit and receive addresses, since only responses can be transmitted, i.e.
Register XADR bit fields XAD1 = XAD2 and register RADR bit fields RAL1 = RAL2
(address of secondary).
Table 25 Error Handling
Frame Type Error Type Generated
Response Generated
Interrupt Rec. Status
I CRC error
aborted
unexpec. N(S)
unexp ec. N(R)
–
–
S-frame
–
FI (from DMAC)
FI (from DMAC)
–
PCE
CRC error
abort
–
–
S CRC error
aborted
unexp ec. N(R)
with I-field
–
–
–
–
–
–
PCE
PCE
–
–
–
–
PEB 20534
Detailed Protocol Description
Semiconductor Group 183 Data Sheet 09.98
Note: The broadcast address may be programmed in bit field RAL2 if broadcasting is
required.
In this case bit fields RAL1 and RAL2 are not equal.
The primary station has to operate in transparent HDLC mode.
Reception of Frames:
The reception of frames functions similarly to the LAPB/LAPD operation (see “Full-
Duplex LAPB/LAPD Operation” on page 177).
Transmission of Frames:
The SCC does not transmit S-, or I-frames if not instructed to do so by the primary station
via an S-, or I-frame with the poll bit set.
The SCC can be prepared to send an I-frame by the host by setting bit ’SXIF’ in register
CCR2. The transmission of the frame, however, will not be initiated by the SCC until
reception of either an
• RR, or
•I-frame
with poll bit set (p = ‘1’).
After the frame has been transmitted (with the final bit set), the host has to wait for an
ALLS or XMR interrupt (positiv e ackn owledgement).
Since the on-chip timer of the SCC must be operated in the external timer mode
(a secon dary does not poll th e primary for acknowledge ments), time r supervi sion must
be done by the primary station.
Upon the arrival of an acknowledgement the SCC transmit FIFO is enabled and an
interrupt is forwarded to the host, either the
– message has been positively acknowledged (ALLS interrupt), or the
– message must be repeated (XMR interrupt).
Additionally, the timer can be used under host control to provide timer recovery of the
secondary if no acknowledgements are received at all.
Note: A secondary will transmit transparent frames only if the permission to send is
given by receiving an S-frame or I-frame with poll bit set (p = ‘1’).
Examples:
A few examples of SCC/host interaction in the case of normal response mode (NRM)
mode are shown in figure 68 and 68.
PEB 20534
Detailed Protocol Description
Semiconductor Group 184 Data Sheet 09.98
Figure 68 No Data to Send: Data Reception/Transmission
Figure 69 Data Transmission (without error), Data Transmission (with error)
RR(0)f=1
RR(0)p=1
DSCC4 Secondary Primary
ITD01800
(0,1)f=1
(0,0)p=1
ITD00237
(1,1)p=1
RR(2)f=1
ALLS
RME
XIF I
I
I
(0,0)f=1
RR(0)p=1
ITD00238
RR(1)p=0
ALLS
XIF
I
(0,0)f=1
RR(0)p=1
ITD01801
RR(0)p=1
XMR
XIF
I
RR(0)f=1
t
PEB 20534
Detailed Protocol Description
Semiconductor Group 185 Data Sheet 09.98
8.1.9 S pecial Functio ns
8.1.9.1 Extended Transparent Transmission and Reception
When programmed in the extended transparent mode via the MODE register
(MODE:MDS1, MDS0, ADM = ‘111’), the SCC performs fully transparent data
transmission and reception without HDLC framing, i.e. without
• FLAG insertion and deletion
• CRC generation and checking
• bit stuffing.
This feature can be profitably used e.g. for:
• user specific protocol variations
• line state monitoring, or
• test purposes, in particular for monitoring or intentionally generating HDLC protocol
rule violations (e.g. wrong CRC)
Character or octet boundary synchronization can be achieved by using clock mode 1
with an external receive strobe input to pin CD.
8.1.9.2 Receive Address Handling
The Receive Address Low/High Bytes (RAL1/RAH1 and RAL2/RAH2) in register RADR
can be masked on a p er bit b asi s by setting the corres pon din g bits in the mask registe r
RAMR. This allows extended broadcast address recognition. Masked bit positions
always match in comparison of the received frame address with the respective address
fields in register RADR.
This feature is applicable to all HDLC protocol modes with address recognition (auto
mode, non-auto mode and address mode 1). It is disabled if all bits of mask bit fields AML
and AMH are set to ‘zero’ (which is the RESET value).
The function of RADR:RAL2/RADR:RAH2 and detection of the fixed group address FEH
or FCH if applicable to the selected operating mode remains unchanged.
As an option in the auto mode, non-auto mode and address mode 1, the 8/16 bit address
field of rec eived frames ca n b e p ush ed to the rece ive data buffe r (firs t o ne/t wo by tes of
the frame). This function is especially useful in conjunction with the extended broadcast
address recognition. It is enabled by setting control bit ’RADD’ in register CCR2.
8.1.9.3 Shared Flags
If the ‘Shared Flag’ feature is enabled by setting bit ’SFLG’ in register CCR1 the closing
flag of a previously transmitted frame simultaneously becomes the opening flag of the
following frame if there is one already available in the SCC transmit FIFO.
In receive direction the SCC always expects and handles ’Shared Flags’. ’Shared
Zeroes’ of consec utive flags are also supported.
PEB 20534
Detailed Protocol Description
Semiconductor Group 186 Data Sheet 09.98
8.1.9.4 One Bit Insertion
Similar to the zero bit insertion (bit stuffing) mechanism, as defined by the HDLC
protocol, the SCC offers a completely new feature of inserting/deleting a ’one’ after
seven c onsecutive ‘z eros’ into the tra nsmit/receive data stream, if th e serial chann el is
operating in bus configuration mode. This method is useful if clock recovery is performed
by DPLL.
Since only NRZ data encoding is supported in a bus configuration, there are possibly
long sequences without edges in the receive data stream in case of successive ‘0’s
received, and the DPLL may lose synchronization.
Enabling the one bit insertion feature by setting bit ’OIN’ in register CCR2, it is
guaranteed that at least after
– 5 consecutive ‘1’s a ‘0’ will appear (bit stuffing), and after
– 7 consecutive ‘0’s a ‘1’ will appear (one insertion)
and thus a correct function of the DPLL is ensured.
Note: As with the b it stuffing, the ‘one i nsertion’ is fully trans parent to th e user, but it is
not in accordance with the HDLC protocol, i.e. it can only be applied in proprietary
systems using circuits that also implement this function, such as the
SAB 82525/SAB 82526.
8.1.9.5 Preamble Transmission
If enabled via bit ’EPT’ in register CCR2, a programmable 8-bit pattern is transmitted with
a selec tabl e numb er of rep etition s after Interframe Timefil l t ransmis sion i s stop ped an d
a new frame is ready to be sen t out. The 8 b it preamb le pa ttern can be program med in
bit field ’PRE’ and the repetition time in bit field ’PREREP’ of register CCR2.
Note: Zero Bit Insertion is disabled during preamble transmission. To guarantee correct
function the programmed preamble value should be different from Receive
Address Byte values defined for any of the connected stations.
8.1.9.6 CRC Generation and Checking
In HDLC/SDLC mode, error protection is done by CRC generation and checking.
In standard applications, CRC-CCITT algorithm is used. The Frame Check Sequence at
the end of each frame consists of two bytes of CRC checksum.
If required, the CRC-CCITT algorithm can be replaced by the CRC-32 algorithm,
enabled via bit ’C32’ in register CCR1. In this case the Frame Check Sequence consists
of four bytes.
As an option in non-auto mode or address mode 0, the internal handling of received and
transmitted CRC checksum can be influenced via control bits ’RCRC’ and ’XCRC’ in
register CCR2.
PEB 20534
Detailed Protocol Description
Semiconductor Group 187 Data Sheet 09.98
Receive direction:
The received CRC checksum is always assumed to be in the 2 (CRC-CCITT) or 4 ( CRC-
32) last bytes of a frame, immediate ly preceding a closing flag. If bit ’RCRC’ is set, the
received CRC checksum is treated as data and will be written to the receive data buffer
in the shared memory where it precedes the frame status byte. Nevertheless the
received CRC checksum is additionally checked for correctness. If non-auto mode is
selected, the limits for ‘Valid Frame’ check are modified (refer to description of the
Receive Status Byte (RSTA)).
Transmit direction:
If bit ’XCRC’ is set, the CRC checksum is not generated internally. The checksum has to
be provided v ia the tran smit dat a buf fer by soft ware. The transmitt ed fra me will onl y be
closed automatically with a (closing) flag.
Note: The SCC does not check whether the length of the frame, i.e. the number of bytes,
to be transmitted makes sense or not according the HDLC protocol.
PEB 20534
Detailed Protocol Description
Semiconductor Group 188 Data Sheet 09.98
8.1.9.7 Data Transparency in PPP Mode
When transporting bit-files (as opposed to text files), or compressed files, the characters
could easily represent MODEM control characters (such as CTRL-Q, CTRL-S) which the
MODEM would not pass through. The DSCC4 maintains an Async Control Character
Map (ACCM) for characters 00-1F Hex. Whenever there is a mapped character in the
data stream , the transm itter pr ecedes that cha racter wit h a control -escape char acter of
7DH. After the control-escape, the character itself is transmitted with bit 5 inverted.
character e.g. 13H is mapped to 7DH, 33H).
At the receive end, a 7DH character is discarded and the following character is modified
by inverting bit 5 (e.g. if 7DH, 33H is received, the 7DH is discarded and the 33H is
changed to 13 H the original character).
In addition to the ACCM, 4 user programmable characters (especially outside the range
00-1F Hex) can also be mapped using the control-escape sequence described above.
These characters are specified in register UADC.
The recei ver discards all characters whic h are received un mapped, but ex pected to be
mappe d because o f ACCM and UD AC regis ter conte nts. If this oc curs withi n an HDLC
frame, the unexpected characters are discarded before forwarded to the receive CRC
checking unit.
7DH (control-escape) and 7EH (flag) octets in the data stream are mapped in general.
The sequence of mapping control logic is:
1. 7DH and 7EH octets,
2. ACCM,
3. UDAC.
The 32 lookup octet values (00H-1FH) are stored within the device. One dword
programmable register is used to select which of the 32 fixed characters have to be
mapped using the control-escape sequence. This is maintained by register ACCM.
Register UDAC provides the additional 4 user programmable characters to be mapped.
PEB 20534
Detailed Protocol Description
Semiconductor Group 189 Data Sheet 09.98
This mechanism is applied to asynchronous HDLC PPP mode as well as to octet
synchronous HDLC PPP mode.
Figure 70 PPP Mapping/Unmapping Example
31 24 23 16 15 8 7 0
ACCM: Async Control Character Map Register
1F
0
00
0
19
1
131E
0
15
0
14
0
...
...
...
...
12
0
11
0
31 24 23 16 15 8 7 0
UDAC: User Defined Async Control Character Map Register
7Eh 7Eh 7Eh
20h
13
H
20
H
01
H
02
H
data in
transmit
FIFO:
HDLC
framing: 13
H
20
H
01
H
02
H
7E
H
7E
H
33
H
00
H
01
H
02
H
7E
H
7D
H
7D
H
7E
H
PPP
mapping:
33
H
00
H
01
H
02
H
7E
H
7D
H
7D
H
7E
H
received
character:
13
H
20
H
01
H
02
H
7E
H
7E
H
PPP
unmapping:
13
H
20
H
01
H
02
H
data in
receive
FIFO:
serial
line
Note: CRC generation/checking is assumed to be disabled in this example; according the PPP
mapping/unmapping, CRC characters are treated as ’data’ characters being mapped/unmapped if
necessary .
PEB 20534
Detailed Protocol Description
Semiconductor Group 190 Data Sheet 09.98
8.1.9.8 Receive Length Check Feature
The SCC offers the possibility to supervise the maximum length of received frames and
to terminate data reception in the case that this length is exceeded.
This feature is controlled via the special Receive Length Check Register RLCR.
The function is enabled by setting bit ’RC’ (Receive Check) and the maximum frame
length to be checked is programmed via bit field ’RL’. The maximum receive length can
be determined as a multiple of 32-byte blocks as follows:
MAX_LENGTH = (RL + 1) × 32 ,
MAX_LENGTH = 32 ... 65536
where RL is the value written to bit field ’RL’.
All frames excee ding this length are treated as if they had been abort ed by the remote
station, i.e. the CPU is informed via
– a FLEX interrupt is generated by the SCC, and
– the receive abort indication ’RAB’ in the Receive Status Byte (RSTA) is set.
Receive operation continues with the beginning of the next receive frame.
PEB 20534
Detailed Protocol Description
Semiconductor Group 191 Data Sheet 09.98
8.2 Asynchronous (A SYNC ) Protocol Mode
8.2.1 Character Framing
Character framing is achieved by start and stop bits. Each data character is preceded by
one Start bit a nd term inated by one or two stop bits. The chara cter len gth is s electa ble
from 5 up to 8 bits. Optionally, a parity bit can be added which complements the number
of ones to an even or odd quantity (even/odd parity). The parity bit can also be
programmed to ha ve a fixed val ue (Ma rk or Spac e). The chara cter fo rmat c onfi gurat ion
is performed via appropriate bit fields in register CCR2. Figure 71 shows the
asynchronous character format.
Figure 71 Asynchronous Character Frame
8.2.2 Data Reception
The SCC offers the flexibility to combine clock modes, data encoding and data sampling
in many different ways. However, only definite combinations make sense and are
recommended for correct operation:
8.2.2.1 Asynchronous Mode
Prerequisites:
• Bit clock rate 16 selected (register CCR0, bit BCR = ‘1’)
• Clock mode 0, 1, 3b, 4, or 7b selected (register CCR0, bit field ’CM’)
• NRZ data encoding selected (register CCR0, bit field ’SC’)
The receiver w hi ch opera tes wi th a clo ck rate e qua l to 16 times the no mina l (ex pec ted )
data bit rate, synchronizes itself to each character by detecting and verifying the start bit.
Since character length, parity and stop bit length is known, the ensuing valid bits are
D1 D2 D3 D4 D5 D6 D7 Parity
Par. Par. Par.
D0
(LSB)
1
Start
Bit
5 to 8 Data Bits
(6 to 9 Bits with Parity)
1 or 2
Stop
Bits
Character Frame
ITD01804
PEB 20534
Detailed Protocol Description
Semiconductor Group 192 Data Sheet 09.98
sampled. Oversampling (3 samples) around the nominal bit center in conjunction with
majority decision is provided for every received bit (including start bit).
The sync hroni za tion las t s for o ne cha racte r, the next incomin g c hara cter cau ses a new
synchronization to be performed. As a result, the demand for high clock accuracy is
reduced. Two communication stations using the asynchronous procedure are clocked
independently, their clocks need not be in phase or locked to exactly the same frequency
but, in fact, may differ from one another within a certain range.
8.2.2.2 Isochronous Mode
Prerequisites:
• Bit clock rate 1 selected (register CCR0 bit BCR = ‘0’)
• Clock mode 2, 3a, 6, or 7a (DPLL mode) has to be used in conjunction with FM0, FM1
or Manchester encoding (register CCR0 bit fields ’CM’ and ’SC’).
The isochronous mode uses the asynchronous character format. However, each data bit
is only sampled once (no oversampling).
In clock modes 0 and 1, the input clock has to be externally phase locked to the data
stream. This mode allows much higher transfer rates. Clock modes 3b, 4 and 7b are not
recommended due to difficulties with bit synchronization when using the internal baud
rate generator.
In clock modes 2, 3a, 6, and 7a, clock recovery is provided by the internal DPLL. Correct
synchronization of the DPLL is achieved if there are enough edges within the data
stream, which is generally ensured only if Bi-Phase encoding (FM0, FM1 or Manchester)
is used.
8.2.2.3 Storage of Receive Data
If the receiver is ena bled, received data is stored in the SCC re ceive FIFO (the LSB is
received first). Moreover, the CD input may be used to control data reception. Character
length, number of stop bits and the optional parity bit are checked. Storage of parity bits
can be disabled. Erro rs are indicated via interrupts. Additionally, the character specific
error status (framing and parity) can optionally be stored in the SCC receive FIFO.
Filling of the the SCC receive FIFO is controlled by
• a programmable threshold level (bit field ’RFTH’ in register CCR2),
• the selected data format (bit ’RFDF’ in register CCR2),
• the parity storage selection (bit ’DPS’ in register CCR2),
• detection of the programmable Termination Character (bit ’TCDE’ and bit field ’TC’ in
register TCR).
Additionally, the time-out event inte rrupt as an opti onal status information indicates that
a certain time (refer to register CCR1) has elapsed since the reception of the last
character.
PEB 20534
Detailed Protocol Description
Semiconductor Group 193 Data Sheet 09.98
8.2.3 Dat a Tr ansmi ss i on
The selection of asynchronous or isochronous operation has no further influence on the
transmitter. The bit clock rate is solely a dividing factor for the selected clock source.
Transmission of the contents of the SCC transmit FIFO starts after providing data to the
DMA controller. The character frame for each character, consisting of start bit, the
character itself with defined character length, optionally generated parity bit and stop
bit(s) is assembled.
After finishing transmission (indicated by the ‘ALLS’ interrupt), IDLE sequence (logical
‘1’) is transmitted on transmit pin TxD.
Additionally, the CTS signal may be used to control data transmission.
8.2.4 S pecial Functio ns
8.2.4.1 Break Detection/Generation
Break generation:
On issuing the transmit break command (bit ’XBRK’ in register CCR2), the TxD pin is
immediately forced to physical ‘0’ level with the next following clock edge, and released
with the first clock edge after this command is reset again by software.
Break detection:
The SCC recogni zes the break co ndition upon rec eiving consec utive (physical) ‘0’ s for
the defined character length, the optional parity and the selected number of stop bits
(‘zero’ character and framing error). The ‘zero’ character is not pushed to RFIFO. If
enabled, the ’Break’ interrupt (BRK) is generated.
The break condition will be present until a ‘1’ is received which is indicated by the ‘Break
Terminated’ interrupt (BRKT).
8.2.4. 2 In-band Flow Control by XON/XOFF Characters
Programmable XON and XOFF characters:
The XNXF register contains the programmable values for XON and XOFF characters.
The number of sig nificant bit s in a regist er is dete rmined b y the pro grammed c haracte r
length via bit field ’CHL’ in register CCR2.
Additionally, two programmable eight-bit values ’MXN’ and ’MXF’ serve as masks for the
characters XON and XOFF, respectively:
A ‘1’ in any mask bit position has the effect that no comparison is performed between the
correspondin g bits in the re ceived c haracters (‘don’ t cares’) a nd the XO N/XOFF val ue.
At RESET, the masks are ‘zero’ed, i.e. all bit positions will be compared.
A received character is considered to be recognized as a valid XON or XOFF character
PEB 20534
Detailed Protocol Description
Semiconductor Group 194 Data Sheet 09.98
– if it is correctly framed (correct length),
– if its bits match the ones in the XON or XOFF registers over the programmed character
length,
– if it has correct parity (if applicable).
Received XON and XOFF characters are stored in the SCC receive FIFO, as any other
characters, when bit DXS is set to ’0’ in register CCR2. Otherwise they are not stored in
the receive FIFO.
In-Band Flow Control of Transmitted Characters:
Recognition of an XON or XOFF character causes always a corresponding maskable
interrupt status to be generated.
Further action depends on the setting of control bit ’FLON’ (Flow Control On) in register
CCR2:
0: No further action is automatically taken by the SCC.
1: The reception of an XOFF character automatically turns off the transmitter after the
currently transmitted character (if any) has been shifted out completely (entering
XOFF state). The reception of an XON character automatically makes the transmitter
resume transmit ting (enter ing XON state).
After hardware RESET, bit CCR2:FLON is ‘0’.
When bit CCR2:FLON is programmed from ‘0’ to ‘1’, the transmitter is first in the
‘XON state’, until an XOFF character is received.
When bit CC R2:FLON is programme d from ‘1’ to ‘0’, the transmit ter al ways g oes in th e
‘XON state’, and transmission is only controlled by the user and by the CTS signal input.
The in-band flow c ontrol of the tr ansmitte r via rec eived XON and XOFF c haracte rs ca n
be combined with control via CTS pin, i.e. the effect of the CTS pin is independent of
whether in-band control is used or not. The transmitter is enabled only if CTS is ‘low’ and
XON state has been reached.
Transmitter Status Bit:
The status bit ‘Flow Control Status’ (bit ’FCS’ in register STAR) indicates the current
state of the transmitter, as follows:
0: if the transmitter is in XON state,
1: if the transmitter is in XOFF state.
Note: The transm itter ca nnot b e t urned o ff by softw are wi thout d isruptin g data poss ibly
remaining in the transmit FIFO.
Flow Control for Received Data:
After writing a character value to register TICR (Transmit Immediate Character) its
character contents is inserted into the outgoing character stream
• immediately upon writing this register by the microprocessor if the transmitter is in
IDLE state. If no further characters (transmit FIFO empty) are to be transmitted, i.e.
PEB 20534
Detailed Protocol Description
Semiconductor Group 195 Data Sheet 09.98
the transmitter returns to IDLE state after transmission of the TIC and an ALLS (All
Sent) interrupt will be generated.
• after the end of a character currently being transmitted if the transmitter is not in IDLE
state. This does not affect the contents of the transmit FIFO. Transmission of
characters from transmit FIFO is resumed after the TIC is send out.
Transmission via this register is possible even when the transmitter is in XOFF state
(however, CTS must be ‘low’).
The TIC valu e is an eight-b it value. The number of sign ificant bits is determine d by the
programmed asynch character length. Parity value (if programmed) and selected
number of stop bits are automatically appended, equal to the characters provided via the
transmit data buffer. The usage of TIC is independent of in-band flow control
mechanism, i.e. is not affected by bit ’FLON’ in register CCR2 anyway.
To control multiple accesses to register TICR, an additional status bit STAR:TEC (TIC
Executing) is provided which signals that the transmission command of currently
programmed TIC is accepted but not yet completely executed. Further access to register
TIC is only allowed if bit STAR:TEC is ‘0’ again.
8.2.4.3 Out-of-band Flow Control
Transmitter:
The transmitter output is enabled if CTS sig nal is ‘LOW’ AN D the XON state is reac hed
in case of in-band flow control is enabled. If the in-band flow control is disabled
(CCR2:FLON = ‘0’), the transmitter is only controlled by the CTS signal.
Nevertheless setting bit CCR1:FCTS = ‘1’ allows the transmitter to send data
independent of the condition of the CTS signal, the in-band flow control (XON/XOFF)
mechanism would still be operational if enabled via bit CCR2:FLON = ‘1’.
Receiver:
For some applications it is desirable to provide means of out-of-band flow control to
indicate to the far end transmitter that the local receiver’s buffer is getting full.
This flow control can be used between two DTEs as shown in figure 72 and between a
DTE and a DCE (MODEM) as shown in figure 73 that supports this kind of bi-directional
flow control.
Setting bit CCR1:FRTS = ‘1’ and CCR1:RTS = ‘0’ invokes this out-of-band flo w control
for the receiver. When the shadow part of the SCC receive FIFO has reached a pre-
defined threshold of 20 bytes, the RTS signal is forced inactive (HIGH). When the
shadow part of the receive FIFO is empty, the RTS is re-asserted (‘LOW’). Note that the
data is immediately transferred from the shadow receive FIFO to the DMA accessible
FIFO (as long as there is space available). Thus when the shadow FIFO reaches the
20 bytes threshold, there are 4 more bytes storage available before an overflow can
occur. This provides s ufficient time for the fa r end transmit ter to react to the chan ge in
the RTS signal and stop sending more data.
PEB 20534
Detailed Protocol Description
Semiconductor Group 196 Data Sheet 09.98
Figure 72 shows the connection between two SCC devices as DTEs. The RTS of DTE-
A (SCC) feeds the CTS input of the second DTE-B (another SCC). For example while
DTE-A is receiving data and its receive FIFO threshold is reached, the RTS signal goes
in-active ’HIGH’ forcing the CTS of DTE-B to become in-active indicating that
transmiss ion has to stop after fin ishing the curre nt charac ter. Both DTE dev ices s hould
also be using the CTS signal to flow control their transmitters. When the shadow receive
FIFO in DTE-A is cleared its RTS goes active ‘LOW’ and this signals the far end DTE-B
to resume transmission. Data flow control from DTE,-B to DTE-A works in the same way.
Figure 72 Out-of-Band DTE-DTE Bi-directional Flow Control
ITS08517
TxD
RxD
CTS
RTS
DSCC4
RTS
CTS
RxD
TxD
DSCC4
RS232c Signals
(drivers not shown)
DTE ABDTE
PEB 20534
Detailed Protocol Description
Semiconductor Group 197 Data Sheet 09.98
Figure 73 shows an SCC as a DTE connected to a DCE (MODEM equipment).
The RTSA feeds the RTSB input of the DCE (MODEM equipment) that supports bi-
directional flow control. So when the DTE-A’s receiver threshold is reached, the RTSA
signal goes active ’HIGH’ which is sensed by the DCE and it stops transmitting. Similarly
if the DCE’s receiver threshold is reached, it deactivates the CTSB (’HIGH’) and causes
the DTE to sto p transmission. These types of D CEs hav e fairly deep buf fers to ensure
that it can continue to receive data from the line even though it is unable to pass the data
to the DTE for short periods of time. Note that a SCC can also be used in the DCE
equipment as shown. Exchange of signals (e .g. RTS to CTS) is necessarily inside the
DCE equipment.
Figure 73 Out-of-Band DTE-DCE Bi-directional Flow Control
RTS and CTS are used to indicate when the local receiver’s buffer is nearly full. This
alerts the far end transmitter to stop transmission.
The combination of transmitter and receiver out-of-band control features mentioned
above enables data to be exchanged between two devices without software intervention
for flow control.
TxD
RxD
CTS
RTS
DSCC4
RTS
CTS
RxD
TxD
DSCC4
RS232c Signals
(drivers not shown)
DTE A BDCE MODEM 1)
RTS
CTS
RxD
TxD
1) Some of the newer MODEMs support bi-directional flow control.
PEB 20534
Detailed Protocol Description
Semiconductor Group 198 Data Sheet 09.98
8.3 Character Oriented Synchronous (BISYNC) Protocol Mode
8.3.1 Character Framing
Character oriented protocols achieve synchronization between transmitting and
receiving station by means of special SYN characters. Two examples are the
MONOSYNC and IBM’s BISYNC procedures. BISYNC has two starting SYN characters
while MONOSYNC uses only one SYN. Figure 74 gives an example of the message
format.
Figure 74 BISYNC Message Format
The SYN character, its length, the length of data characters and additional parity
generation are programmable:
• 1 SYN character with 6 or 8 bit length (MONOSYNC), programmable via register
SYNCR.
• 2 SYN characters w ith 6 or 8 bit length each (BISYNC), program mable via reg isters
SYNCR.
• Data character length may vary from 5 to 8 bits (bit field ’CHL’ in register CCR2).
• P arity informat ion (even/odd parity, mark, s pace) may be appended to the ch aracter
(bit ’PARE’ and bit field ’PAR’ in register CCR2).
8.3.2 Data Reception
The receiver is generally activated by setting bit ’RAC’ in register CCR2. Additionally, the
CD signal may be used to control data reception depending on the selected clock mode.
After issuing the HUNT command, the receiver monitors the incoming data stream for
the presenc e of speci fied SYN charac ter(s). Howev er, data re ceptio n is still disabled . If
synchronization is gained by detecting the SYN character(s), an SCD interrupt is
genera ted and all follo wing data is pus hed to the receive FIFO, i.e. con trol sequences ,
data characters and optional CRC frame checking sequence (the LSB is received first).
SYN
(SYNL) (SYNH)
SYN SOH Header STX Text ETX CRC(Data)
2 Leading
SYN
Characters
Start
of
Header Text
of
Start End
of
Text
Frame
Checking
Sequence
ITD01805
PEB 20534
Detailed Protocol Description
Semiconductor Group 199 Data Sheet 09.98
In normal operation, SYN characters are excluded from storage to receive FIFO. SYN
character length can be specified independently of the selected data character length. If
required, the character parity bit and/or parity status is stored together with each data
byte in the receive FIFO.
As an option, the loading of SYN characters in receive FIFO may be enabled by setting
the bit ’SLOAD’ in register CCR2. Note that in this case SYN characters are treated as
data. Consequently, for correct operation it must be guaranteed that SYN character
length equal s the character length + optional pa rity bit. This is the user’s responsibility
by appropriate software settings.
Filling of the receive FIFO is controlled by a programmable threshold level.
Reception is stopped if
1. the receiver is deactivated by resetting the RAC bit, or
2. the CD signal goes inactive (if Carrier Detect Auto Start is enabled), or
3. the HUNT command is issued again, or
4. the Receiver Reset command (RRES) is issued, or
5. a programmed Termination Character has been found (optional).
On actions 1. and 2., reception remains disabled until the receiver is activated again.
After this is done, and generally in cases 3. and 4., the receiver returns to the (non-
synch roniz ed) Hun t state. In cas e 5. a HUN T com mand ha s to be is sued. R ecepti on of
data is internally disabled until synchronization is regained.
Note: Further checking of frame length, extraction of text or data information and
verifying the Frame Checking Sequence (e.g. CRC) has to be done by the
microprocessor.
8.3.3 Dat a Tr ansmi ss i on
Transmission of data provided in the shared memory is started after the DMA controller
forwards the first data bytes to the SCC transmit FIFO (the LSB is sent out first).
Additionally, the CTS signal may be used to control data transmission. The message
frame is as sem ble d b y a ppen din g a ll data char act ers t o th e s pec ifie d SYN c harac ter(s )
until Transmit Message End condition is detected (FE indication via DMAC). Internally
generated parity information may be added to each character (SYN, CRC and Preamble
ch aracters are excluded) .
If enabled via CRC Append bit (bit ’CAPP’ in register CCR2), the internally calculated
CRC checksum (16 bit) is added to the message frame. Selection between CRC-16 and
CRC-CCITT algorithms is provided.
Note: - Internally generated SYN characters are always excluded from CRC calculation,
- CRC checksum (2 bytes) is sent without parity.
The internal CRC generator is automatically initialized before transmission of a new
frame starts. The initialization value is selectable.
PEB 20534
Detailed Protocol Description
Semiconductor Group 200 Data Sheet 09.98
After finishing data transmission, Interframe Timefill (SYN characters or IDLE) is
automati cally sent.
8.3.4 Special Functions
8.3.4.1 Preamble Transmission
If enabled via register CCR2, a programmable 8-bit pattern (bit field ’PRE’) is transmitted
with a sel ect able number of repetiti ons af ter In terfram e Time fill transmission is stopped
and a new frame is ready to be sent out.
Note: If the preamble pattern equals the SYN pattern, reception is triggered by the
preamble.
8.3.4.2 CRC Parity Inhibit
If the internal CRC generator is not used for calculation of the Frame Check Sequence,
an externally calculated checksum (16 bits) can be appended to the message frame
without internally generated parity information, although parity is enabled for data
characters.
Prerequisites are:
• CRC generator disabled (CAPP = ‘0’),
• Frame/Block End indication has to be issued with t he checksum p rovided in shared
memory.
The programmed character length has no influence on this function.
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 201 Data Sheet 09.98
9 Reset and Initialization Procedure
9.1 Reset and Power-On
The DSCC 4 offers se vera l re set funct ions . An external l ow s ignal on signal RST resets
the DSCC4. It immediately drives all PCI outputs to their benign state, pin REQ is driven
to
tristate
. The TxDn output signals are driven to high impedance, the RTSn output
signals are driven to inactive. All bi-directional signals (e.g. Multi Function Port (M FP))
are switched to input function.
All registers and functions are initialized to known states (RESET values).
After RST is deasserted, the functional blocks PCI, SCCs, LBI, SSC and GPP are in
reset state or standby mode.
Table 26 Status after Hardware Reset
Block Reset Status
PCI interface Standby
PCI Config Space registers accessible.
Global Registers Standby
Slave registers accessible;
Host Bus Interface enabled in PCI mode!
DMAC Standby
Slave registers accessible.
SCC Standby
Slave registers accessible
Reset Status :
– power-down mode
– HDLC mode
– NRZ coding
–…
Refer to reset values in “SCC Registers - Detailed
Register Description” on page 272.
LBI Reset
Slave register LCONF has reset value.
SSC Standby
Slave registers accessible.
GPP Standby
Slave registers accessible.
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 202 Data Sheet 09.98
Software Reset
Software reset control bits are also available to reset single sections of the DSCC4.
Reset of
– DMA receiver is performed via CHiCFG.RDR bit (channel specific)
– DMA transmitter is performed via CHiCFG.RDT bit (channel specific)
– transmitter in SCC is performed via CMDR.XRES bit (channel specific)
– receiver in SCC is performed via CMDR.RRES bit (channel specific)
– EBC (LBI, SSC, GPP) block is performed via LCONF.EBCRES bit.
Note: The software reset only affects the internal state machines. The registers are not
reset.
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 203 Data Sheet 09.98
9.2 Initializatio n Example
The first st ep of initializati on is done alre ady via hardware configuration. If D EMUX pin
is connected to VDD3, the DSCC4 is configured in de-multiplexed bus interface mode.
Otherwise (DEMUX connected to VSS) the DSCC4 is configured in PCI mode and the
MFP is available as LBI port (Reset state).
After hardware reset, the host has to write a minimum set of registers to initialize the
functional blocks. The following tables provide initialization sequences assuming that in
parallel the ho st will reserve/prepare m emo ry spa ce fo r inte rru pt ring buffers and linked
list data structures before activating the DSCC4’s DMAC by setting GCMDR.AR (See
“Start of Operation” on page 207). During the initialization phase the DSCC4 is
operating in slave mode.
Table 27 Global Configuration of D SCC4 and Initialization of DMAC (Interrupt
Channel)
Step Action
1 Select Little-/Big-Endian mode via bit GMODE.ENDIAN.
2 Select Burst-/No-Burst mode via bit GMODE.BURST.
(Valid only in DEMUX mode)
3 Select Priority Scheme via bit GMODE.SPRI and
GMODE.CHN.
4 Configure MFP via bit fields GMODE.PERCFG and
GMODE.LCD as:
- LD15..0, LA15...0 (LBI mux/demux)
- LD15..0, GP 15...8, LA7...0 (GPP + LBI)
- LD15..0, GP15...8, SSC (GPP + SSC + LBI)
- LD15..0, GP15...0 (GPP + LBI)
5Set interrupt queue base addresses in registers
IQSCCnRXBAR, IQSCCnTXBAR, IQPSCCRXBAR,
IQPSCCTXBAR, IQCFGBAR, IQPBAR
In parallel
reserve memory
space for
interrupt queues
(ring buffers) in
shared memory
6 Set interrupt queue lengths in registers
IQLENR1, IQLENR2
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 204 Data Sheet 09.98
For the SCC, first, the serial mode, the configuration of the serial port and the clock mode
have t o be defined. The host may s witch the SC C between po wer-up an d power-d own
mode. This has no influence upon the contents of the registers, i.e. the internal state
remains stored. In power-down mode, however, all internal clocks are disabled, no
interrupts are forwarded to the host. This state can be used as a standby mode, when
the SCC is temporarily not used, thus substantially reducing power consumption.
Table 28 Initialization of DMAC (Data Channels)
Step Action
1 Select Control Mode of the DMA controller via
GMODE.CMODE bi t.
In parallel
prepare linked
list(s) in shared
memory.
GMODE.CMODE = ’0’
(causes the DMAC to check
HOLD bit before branching
to next descriptor).
GMODE.CMODE = ’1’
(causes the DMAC to
compare FTDA/FRDA and
LTDA/LRDA before branch-
ing to next descriptor).
2 Select FIFO size and thresholds: FIFOCR1...4
3 Set base descriptor addresses: BRDA, BTDA.
4Verify/set HOLD=’1’ in last
element of linked list. Write last transmit/receive
descriptor address to
LTDA/LRDA register.
5 Configure channels: CHiCFG
- Interrupt Mask (RFI, TFI, RERR, TERR)
- DMAC Command (RDR, RDT, IDR, IDT)
Table 29 Initialization of the SCC(s)
Step Action
1 Set clock mode specific features: CCR0, BRR, TTSA, RTSA, TPCMM,
RPCMM.
2 Set serial mode (HDLC, ASYNC, BISYNC): CCR0
3 Set serial port configuration (Encoding, output driver select, handshaking
mechansi m): CCR0, CCR1
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 205 Data Sheet 09.98
In PCI mode the MFP can be used to access external periphe rals like ESCC2, HSCX,
FALC54, when LBI configuration is selected. In LBI configuration even data can be
exchanged with a local microprocessor. Moreover, a General Purpose Port (GPP) can
be used for con trol p urposes. Altern atively , seri al c ommuni cation c an be perfo rmed via
the SSC port.
Depending on the configuration selected in GMODE register (bit field
GMODE.PERCFG), the appropriate register set of the MFP has to be initialized, too (LBI,
SSC, GPP regist ers).
4 Set serial mode specific features for
HDLC
Auto Mode, NRM Mode: CCR1, TIMR, XADR, RADR, RAMR
Non Auto Mod e: CCR1, CCR2, RADR, RAMR
Transparen t Mode: CCR1, CCR2 , RADR, RAMR
Extended Transparent Mode: CCR1
Asynchronous PPP: CCR1, CCR2, UDAC, ACCM
Synchronous PPP: CCR1, CCR2, UDAC, ACCM
General:
Shared flags, CRC reset level: CCR1
Preamble, ITF/OIN, SCC RFIFO configuration: CCR2
Receive Length Check: RLCR
ASYNC
Bit clock rate: CCR0
Data format, SCC R FIFO configuration, Flow control: CCR2, XNXF
Character Insertion: TICR
Termination Character: TCR
BISYNC
Bisync/Monosync: CCR1
Sync character: SYNCR
Data format, SCC RFIFO configuration, Preamble, CRC: CCR2
Termination character: TCR
5 Set interrupt mask: IMR
Table 29 Initialization of the SCC(s) (cont’d)
Step Action
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 206 Data Sheet 09.98
Table 30 Initialization of the MFP
Step Action
1 When the MFP has been configured as
– LBI port: configure LCONF
– GP port: configure GPDIR, GPIM
– SSC port: configure SSCCON, SSCBR, SSCCSE, SSCIM.
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 207 Data Sheet 09.98
9.3 Start of Ope ratio n
After having p erformed the i nitializati on, the host reques ts the acti vation of th e DSCC4
by enabling the SCC(s) and setting the corresponding bits in the GCMDR.
A correct sequence is:
• activation of the SCC
• activation of the DMAC (receive and transmit)
• enabling the SCC receiver
Table 31 Activation of DMAC and SCC
Step Action by Host Action by DSCC4
1 Set appropriate command bits for
interrupt queue initialization and
Action Request (AR) bit in GCMDR.
2 The DMAC sets up the interrupt
queues. When the configuration was
successful, the DMAC sets
GSTAR.ARACK=’1’.
If enable d, INT signal is activa ted and
the DMAC stores the corresponding
configuration interrupt vector in the
interrupt queue IQCFG located in the
shared memory.
3 Serve interrupts.
4 Set the SCC to power-up mode via
CCR0.PU.
The SCC receiver should remain
disab led (CC R2.R AC= ’0’).
5 Reset SCC transmitter.
6 The SCC requests for data from the
central TFIFO and, if data are
available, data transmission is started.
7 Receive/transmit interrupts caused by
the SCC are forwarded as interupt
vectors through the central interrupt
queue to the appropriate interrupt
buffers in shared memory.
8 Serve interrupts.
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 208 Data Sheet 09.98
9 Set appropriate channel configuration
bits (receive and transmit) and Action
Request (AR) bit in GCMDR.
10 After the DSCC4 has become bus
master the DMAC the internal DMA
channels for data transfer. When the
configuration was successful, the
DMAC sets GSTAR.ARACK=’1’.
If e nabled, INT signal is activate d and
the DMAC stores the corresponding
configuration interrupt vector in the
interrupt queue IQCFG located in the
shared memory. Moreover the DMAC
branches to the base receive/transmit
descriptor (referenced by BRDA/
BTDA) of the linke d list.
In transmit direction it starts trans-
ferring data - if available - from shared
memory into the on chip central
transmit FIFO. Since the SCCs
receivers are not activated so far, no
data are received from SCCs.
11 Read GSTAR for interrupt
information and acknowledge
interrupts by writing a ’1’ back to the
bits, which indicate an interrupt
12 Enable the SCC receiver
(CCR2.RAC=’1’). SCC starts data reception. The
received data are transferred to the
central RFIFO. As soon as the
threshold of the RFIFO has been
reached, the DMAC transfers the data
into the shared memory.
Table 31 Activation of DMAC and SCC (cont’d)
Step Action by Host Action by DSCC4
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 209 Data Sheet 09.98
After start procedure the continuous operation of data transfer is essentially controlled
by the host and the DMAC sharing the data structures and interrupt queues located in
the memory.
Table 32 Continuous Operation of Data Transfer
Step Action by Host Action by DSCC4 (DMAC)
HOLD bit ctrld. LTDx/FTDx ctrld
.
HOLD bit ctrld. LTDx/FTDx ctrld.
1 Add list elements to linked list, if
available. The DMAC
transfers data
between shared
memory and on
chip FIFOs.
It branches to the
next descriptor as
long as
HOLD = ’0’.
The DMAC
transfers data
between shared
memory and on
chip FIFOs.
It branches to the
next descriptor as
long as FxDA is
not equal to LxDA.
Set HOLD=’1’ in
new last element
of linked list and
reset HOLD bit in
the previously last
descriptor.
Write new last
transmit/receive
descriptor
address in
LTDA/LRDA
register.
2 Set appropriate
poll bit in GCMDR
register
GCMDR.TXPRi.
3 If HOLD=1 has
been sensed
before, the DMAC
reads current
descriptor again
and branches
(HOLD=0) to next
descriptor.
Otherwise the poll
request is ignored.
If FTDA/FRDA =
LTDA/LRDA has
been sensed
before, the DMAC
compares
FTDA/FRDA to
updated
LTDA/LRDA and
branches to next
descriptor.
4 Generally, host initiated interrupts
might be generated.
5 Serve interrupts appropriately.
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 210 Data Sheet 09.98
The sequence of functional steps shown in table 32 are repeated as long as data
transmission is required and no error does occur.
The procedures to stop data transmission and/or reception are shown in table 33 and
table 34.
Table 33 Stop Data Transmission
Step Action by Host Action by DSCC4
HOLD bit ctrld. LTDx/FTDx ctrld
.
2Write address of the “buffer-stop“ to
BTDA.
Initi aliz e CHiC F G regis ter
- Interrupt Mask:
(RFI, TFI, RERR, TERR)
- DMAC Command (RDT)
3 Set GCMDR.AR.
4 The DMAC transmit channel discards
the transmit data s tored internally and
returns to its reset state. No additi onal
data are read from the shared
memory.
5 Reset transmitter in SCC.
6 Data stored internally in SCC are
discarded and transmission stops.
Table 34 Stop Data Reception
Step Action by Host Action by DSCC4
1 Disable receiver in SCC via
CCR2.RAC.
2 SCC stops data reception. The
received data stored internally are
transferred to the shared memory.
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 211 Data Sheet 09.98
Beside normal operation, exceptions might happen such as:
– no memory avai lab le for re ceived data (HOLD =’1’ or FRD A=L RD A in c urren t re ce ive
descriptor), which leads to a receive data overflow
– no data available for transmission (HOLD=’1’, FE=’0’ or FTDA=LTDA, FE=’0’ in
current transmit descriptor ), which leads to transmit data underrun
– general failure, which can cause a software restart.
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 212 Data Sheet 09.98
Table 35 Exceptional handling in Case of Receive Data Overflow
Step Action by Host Action by DSCC4
HOLD bit ctrld. LTDx/FTDx ctrld
.
No Host action has to be performed in case of a receive data overflow event. The
DSCC4 marks the receive descriptor (data section) containing incomplete data with an
’RDO’ indication in the receive status byte (RSTA). The DMA controller proceeds with
the next receive descriptor; no difference to handling of a non-exceptional frame.
Table 36 Exceptional handling in Case of Transmit Data Underrun
Step Action by Host Action by DSCC4
HOLD bit ctrld. LTDx/FTDx ctrld
.
1 Prepare linked list for future data
transmission in shared memory.
Set HOLD=’1’ in
last element of
linked lis t.
-
2 Initialize CHiCFG register
- Interrupt Mask:
(RFI, TFI, RERR, TERR)
- DMAC Command (RDT)
3 Reset transmitter in SCC via
CMDR.XRES. SCC starts requesting for transmit
data from central TFIFO.
4 - Write last transmit
descriptor
address in LTDA
register.
Update BTDA.
Initi aliz e CHiC F G regis ter
- Interrupt Mask:
(RFI, TFI, RERR, TERR)
- DMAC Command (IDT and AR)
5 SCC gets new data (if available) from
the central TFIFO and Data
transmission starts.
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 213 Data Sheet 09.98
In case of a general res tart the initia lization a nd start sequ ences hav e to be perfo rmed
as described in tables 31 to 32.
7 Set GCMD R .AR.
8 After the DSCC4 has become bus
master the DMAC sets up the internal
DMA channels for data transmission. If
availablle, data are transferred from
shared memory to on chip TFIFO.
Table 36 Exceptional handling in Case of Transmit Data Underrun (cont’d)
Step Action by Host Action by DSCC4
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 214 Data Sheet 09.98
9.4 Initializ ation Exam ple
9.4.1 Test Loop For Data Transfer in HDLC Address Mode 0
The data in the transmit data buffer referenced by Tx Data Pointer are transmitted via
SCC1’s tes t loop and stored afte r rec eption in the receive data buffer refe renc ed by Rx
Data Buffer.
Figure 75 Overview of Data Stuctures in shared Memory before Transmission
IQSCC1TX
ITD10619
Address
0000 6000 0000 4000
To be set by DSCC4
Set by host
~
~
IQCFG
~
~
Address
IQSCC1RX
~
~~
~
0000 2000
Address
~
~~
~
0001 0010
0001 0000
0001 0004
Address
0001 000C
0001 0008
Address
Transmit Descriptor List:
0001 100C
0001 1008
0001 1004
0001 1000
CFG: HOLD = 1, HI = 1, NO = 32
Next Rx Descriptor Pointer
CFG: FE = 1, HOLD = 1, NO = 16
Next Tx Descriptor Pointer
Receive Descriptor List:
0001 0100
0040 0000
6020 0000
0040 0000
0001 1100
C010 0000
BNO, Status
Rx Data Pointer
Status
Tx Data Pointer 0000 0000
0000 0003
0000 0002
0000 0001
FE Descriptor Pointer
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 215 Data Sheet 09.98
Table 37 Register Initialization for HDLC Transparent Mode 0, Test Loop
Register Access
<= (write)
=> (read)
Value Meaning
GMODE <= 0000 0000 RESET Value:
- DMAC is controlled by HOLD bit
- Little Endian
- Default Priority Scheme
- MFP configured as LBI (not needed in
this example)
IQLENR1 <= 0000 0000 RESET Value:
Size of ring buffers: 32 entries
IQLENR2 <= 0000 0000 RESET Value:
Size of ring buffers: 32 entries
IQSCC1RXBAR <= 0000 2000 IQ Base Address for SCC1,RX
IQSCC1TXBAR <= 0000 4000 IQ Base Address for SCC1,TX
IQCFGBAR <= 0000 6000 IQ Base Address for CFG
FIFOCR1 <= 07C0 0000 max. possible buffer of TFIFO reserved
for SCC1: 124 32-bi t words
FIFOCR2 <= 0040 0000 Watermark of TFIFO (SCC1 portion) is
set to 2 (example).
(As soon as less than two DWORDs are
in the central TFIFO buffer, the TFIFO
requests for more data.)
FIFOCR3 <= 0000 0000 RESET Value:
Watermark of RFIFO is set to one.
(As soon as one 32-bit word is stored in
the RFIFO, the RFIFO requests for data
transfer to shared memory.
FIFOCR4 <= 0000 0000 RESET Value:
Watermark of TFIFO forward threshold
(SCC1 portion) is set to one.
(As soon as at least one 32-bit word is in
the central TFIFO, the TFIFO transfers
data to SCC1 transmit FIFO.)
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 216 Data Sheet 09.98
GCMDR <= 2220 0001 Command Bits:
- Configure IQ SCC1 RX
- Configure IQ SCC1 TX
- Configure IQ CFG
- Action Request
DSCC4 performs the configuration and requests for the bus to transfer a CFG
interrupt vector to the IQCFG in shared memory
GSTAR => 0020 0001 Indication Bits:
- CFG interrupt indicated
- Action Request Acknowledge indicated
GSTAR <= 0020 0001 Acknowledge the interrupt indications:
- CFG interrupt
- Action Request Acknowledge
1st entry of
IQCFG in
shared memory
=> A000 0001 Indication Bits:
- CFG interrupt queue ID
- Action Request Acknowledge indicated
CCR0 (SCC1) <= 8000 0016 Power Up
NRZ
HDLC
Clock Mode 6b, assuming that a clock is
provided on XTAL1
CCR1 (SCC1) <= 0204 8100 TxD output driver select
HDLC Transparent Mode 0
Test Loop
FCTS=’1’
CCR2 (SCC1) <= 0803 0008 Receiver active
Continuous FLAG sequences as
interframe time fill
RFTH=’011’ (default value)
IMR (SCC1) <= FFFA EF3D Interrupts are enabled as follows:
ALLS, XDU, XPR, RDO, RFS, RFO
CMDR (SCC1) <= 0101 0000 Commands:
- Transmitter Reset
- Receiver Reset
Table 37 Register Initialization for HDLC Transparent Mode 0, Test Loop
Register Access
<= (write)
=> (read)
Value Meaning
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 217 Data Sheet 09.98
SCC1 resets transmitter and receiver. After transmitter reset an XPR interrupt is
generated. A corresponding interrupt vector is forwarded through the interrupt
queue to the appropriate interrupt ring buffer in shared memory.
GSTAR => 0200 0000 Indication Bit:
- SCC1 TX interrupt indicated
GSTAR <= 0200 0000 Indicated interrupt is acknowledged.
1st entry of
IQSCC1TX in
shared memory
=> 5200 1000 Indication Bits:
- SCC1 TX interrupt queue ID
- Caused by SCC
- XPR interrupt indicated
CH1CFG <= 0018 0000 Configuration of DMAC channels:
- Enable all interrupts: RFI, TFI, RERR,
TERR
-Set commands: IDR, IDT
CH1BRDA <= 0001 0000 Set base receive descriptor address.
(See figure 75.)
CH1BTDA <= 0001 1000 Set base transmit descriptor address.
(See figure 75.)
GCMDR <= 0000 0001 Command Bit:
- Action Request
DSCC4 checks the CHiCFG registers and performs the configuration of the DMA
channels as required. After configuration an appropriate interrupt vector is
generated and forwarded to IQCFG in shared memory
After configuration the DMAC transfers transmit data from the tx data buffer to the
central TFIFO. These data are forwarded to the SCC1, which loops back the data
at the serial port . The received data are forwarded to the central RFIFO. Then the
DMAC transfers the receive data to the rx data buffer in shared memory.
The data transmission is completed with appropriate interrupts: ALLS, RFS, HI, FI.
GSTAR => 2220 0001 Indication Bits:
- SCC1 RX interrupt indicated
- SCC1 TX interrupt indicated
- CFG interrupt indicated
- Action Request Acknowledge indicated
GSTAR <= 2220 0001 Indicated interrupts are acknowledged.
Table 37 Register Initialization for HDLC Transparent Mode 0, Test Loop
Register Access
<= (write)
=> (read)
Value Meaning
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 218 Data Sheet 09.98
See figure 76 for data structure in shared memory after data transmission.
2nd entry of
IQCFG in
shared memory
=> A000 0001 Indication Bits:
- CFG interrupt queue ID
- Action Request Acknowledge indicated
2nd entry of
IQSCC1TX in
shared memory
=> 5002 0000 Indication Bits:
- DMA(SCC) TX interrupt queue ID
- Caused by DMA
- FI interrupt indicated
3rd entry of
IQSCC1TX in
shared memory
=> 5204 0000 Indication Bits:
- SCC1 TX interrupt queue ID
- Caused by SCC
- ALLS interrupt indicated
1st entry of
IQSCC1RX in
shared memory
=> 1200 0040 Indication Bits:
- SCC1 RX interrupt queue ID
- Caused by SCC
- RFS interrupt indicated
2nd entry of
IQSCC1RX in
shared memory
=> 1006 0000 Indication Bits:
- SCC1 RX interrupt queue ID
- Caused by DMAC
- HI, FI interrupt indicated
Table 37 Register Initialization for HDLC Transparent Mode 0, Test Loop
Register Access
<= (write)
=> (read)
Value Meaning
PEB 20534
Reset and Initialization Procedure
Semiconductor Group 219 Data Sheet 09.98
Figure 76 Overview of Data Stuctures in shared Memory after Transmission
IQSCC1RX
~
~
CFG: HOLD = 1, HI = 1, NO = 32
Next Rx Descriptor Pointer
CFG: FE = 1, HOLD = 1, NO = 16
Next Tx Descriptor Pointer
FE Descriptor Pointer
Address
0000 6000
Set by DSCC4
Set by host
IQCFG
~
~~
~
A000 0001
A000 0001
0000 4000
Address
0001 0010
0001 0000
0001 0004
Address
0001 000C
0001 0008
Address
Transmit Descriptor List:
0001 100C
0001 1008
0001 1004
0001 1000
Receive Descriptor List:
Status: Completed.
Tx Data Pointer
0001 0100
0001 0000
C011 0000
0040 0000
6020 0000
4000 0000
0040 0000
0001 1100
C010 0000
BNO = 17, Status = 0
Rx Data Pointer
= Don’t care in transparent mode 0
IQSCC1TX
"Ax" = Receive Status Byte: "Ax" means
ITD10620
5210 1000
5002 0000
5204 0000
Valid Frame, CRC OK;
~
~
1006 0000
1200 0040
Ax
"x"
~
~
0000 2000
Address
~
~
0000 0001
0000 0003
0000 0002
0000 0000
0000 0003
0000 0002
0000 0001
0000 0000
PEB 20534
Detailed Register Description
Semiconductor Group 220 Data Sheet 09.98
10 Detailed Register Description
The DSCC4 has a set of PCI Configuration Space registers and several PCI memory
mapped on-chip registers which allow configuration and control of the different functions
within the DSCC4. Additionally, receive/transmit descriptors and data sections as well as
interrupt status queues are located in the shared memory. The on-chip registers as well
as the data structures and inte rrupt qu eues in the share d memory a re des cribed i n this
chapter.
PEB 20534
Detailed Register Description
Semiconductor Group 221 Data Sheet 09.98
10.1 Register Range Overview and Address Mapping
Table 38 Register Range and Address Mapping
Base
Address Range
Offset
Address
Register/Address Range Description
Total Range Number of
used DWORD
registers
None,
selected via IDSEL
signal (and bus
commands in PCI
interface mode)
2 KByte
000H...7FFH
16
000H...03CH
PCI Configuration Space
Register Set
BAR0 0000H256 Byte
0000H...00FFH
59
0000H...00ECH
Global Registers
0100H128 Byte
0100H...017FH
23
0100H...0158H
SCC0 Registers
0180H128 Byte
0180H...01FFH
23
0180H...02D8H
SCC1 Registers
0200H128 Byte
0200H...027FH
23
0200H...0258H
SCC2 Registers
0280H128 Byte
0280H...02FFH
23
0280H...02D8H
SCC3 Registers
0300H128 Byte
0300H...037FH
1
0300H
LBI Control Register
0380H128 Byte
0380H...03FFH
6
0380H...0394H
SSC Control Registers
0400H128 Byte
0400H...047FH
3
0400H...0408H
GPP Control Registers
0480H896 Byte
0480H...07FFH
0 (unused)
BAR1 0000H64 KByte
0000H...FFFFH
(16384) Local Bus address range
mapped into PCI (HOST)
memory address space
PEB 20534
Detailed Register Description
Semiconductor Group 222 Data Sheet 09.98
10.2 PCI Configuration Space - Detailed Register Description
According to the PCI Specification V2.1 the DSCC4 supports the register layout of the
predefined header region of the PCI Configuration Space.
Table 39 DSCC4: PCI Configuration Space Register Set
The predefined header region has a size of 64 bytes and consists of fields that uniquely
identify the device and allow the device to be controlled.
The DSCC4 supports the 64-byte header portion of the configuration space of PCI
Specifica tion Rev 2.1 to id entify the dev ice upon initial ins tallati on and power-up. None
31 1615 0
Device ID (=2102H) Vendor ID(=110AH)00H
Status (=0280H) Command (=0000H)04H
Class Code (=028000H) Revision ID (=20H)08H
BIST
(=00H)Header Type
(=00H)Latency Timer
(=00H)Cache Line Size
(=00H)0CH
BAR0 (Base Address Register 0): base address of DSCC4 on-chip registers 10H
BAR1 (Base Address Register 1): base address of Local Bus Interface 14H
Base Address Register (not used) 18H
Base Address Register (not used) 1CH
Base Address Register (not used) 20H
Base Address Register (not used) 24H
Reserved 28H
Reserved 2CH
Expansion ROM Base Address 30H
Reserved 34H
Reserved 38H
Max_Lat (=0AH) Min_Gnt (=03H) Interrupt Pin (=01H) Interrrupt Line (=00H)3CH
Predefined header region of the PCI Configuration Space
PEB 20534
Detailed Register Description
Semiconductor Group 223 Data Sheet 09.98
of the additional locations 64-255 are used, nevertheless the DSCC4 responds to
configuration read/write cycles within the address range 00H to FCH.
These configuration registers are addressed only using the PCI Configuration read/write
cycles and using the IDSEL/DEVSEL handshake signals.
The PCI Configuration Space is also valid in de-multiplexed bus interface mode, i.e. pin
DEMUX connected to VDD3. In this case only signal IDSEL is used to select the PCI
Configuration register set on read/write transactions.
The Base Address fields in th e con figu r atio n spa ce define the memory base addresses
and the corresponding address range the DSCC4 will respond to. The first Base
Address is the base address of the DSCC4’s on-chip register range (Global control
registers, SCC registers, LBI control registers, GPP Control registers, SSC registers).
The second Base Address is the base address of the memory mapped LBI address
space.
According to the PCI Specification, mapped address ranges are evaluated by writing all
ones to the base address registers and read ing back the value. The number of leading
zeros determine the supported address range:
The status/command register (offset address 04H) of the PCI Configuration Space
describes an d determines the DSCC4 PCI syst em behavior and is des cribed in deta ils
in Table 39:
Table 40 PCI Base Address Ranges
Offset
Address Register Register Read Value
(after writing 0xFFFFFFFFH)Supported Address Range
10HBAR0 0xFFFFF800H2 KByte
(DSCC4 on chip registers)
14HBAR1 0xFFFF0000H64 KByte
(Local bus address range mapped
to PCI (HOST) address space)
PEB 20534
Detailed Register Description
Semiconductor Group 224 Data Sheet 09.98
Note: Bit locations containing a ’0’ or ’1’ are hardwired status and configuration settings
specifying a fixed device behavior. These bit locations are also described in Table
39.
Table 41 PCI Configuration Space: Status/Command Register
CPU Accessibility: read/write (PCI Configuration Cycles)
Reset Value: 0280 000H
Offset Address: 04H (PCI Configuration Space Offset Address)
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
PCI Status Information
DPE
SSE
RMA
RTA
001
DPED
10000000
Bit1514131211109876543210
PCI Command Bits
000000
FBBE
SERRE
0
PER
00
SC
BM
MS
IOS
PEB 20534
Detailed Register Description
Semiconductor Group 225 Data Sheet 09.98
Table 42 Status and Comma nd regis ter bits
Bit
Location Symbol Description
31 DPE Detected Parity Error
This bit is set by the device whenever it detects a parity error,
even if parity error handling is disabled (as controlled by bit 6
in the Command register).
30 SSE Signaled System Error
This bit will be set when
•the SERR
(SERRE) Enable bit is set in the Command
register
and
one of the following events occured:
1. A transaction in which the DSCC4 acts as a master is
terminated with master abort.
2. A transaction in which the DSCC4 acts as a master is
terminated with target abort by the involved target.
3. The transaction has an address parity error and the Parity
Error Response bit is set.
29 RMA Received Master Abort
This bit is set whenever the DSCC4 aborts a transaction with
master abort. This occurs when no device responds.
28 RTA Received Target Abort
This bit is set whenever a device responds to a master
transaction of the DSCC4 with a target abort.
27 0BSignaled Target Abort
The DSCC4 will never signal “Target Abort”.
26..25 01BDEVSEL Timing
The DSCC4 is a medium device.
24 DPED Data Parity Error Detected
This bit is set when the following three conditions are met:
1. the device asserted PERR itself or observed PERR
asserted;
2. the device setting the bit acted as the bus master for the
transaction in which the error occurred;
3. and the Parity Error Response bit is set in the Command
register.
23 1BFast Back-to-Back Capable
The DSCC4 is fast Back-to-Back capable.
PEB 20534
Detailed Register Description
Semiconductor Group 226 Data Sheet 09.98
22 0BUDF Supported
No UDFs are supported by the DSCC4.
21 0B66 MHz Capable
The DSCC4 is not 66 MHz capable.
20...16 00000BReserved
15...10 000000BReserved
9 FBBE Fast Back-to-Back enable
A value of ’1’ means the DSCC4 is allowed to generate fast
Back-to-Back transactions to different agents.
A value of ’0’ means the DSCC4 is only allowed to generate
fast Back-to-Back transaction to the same agent.
8 SERRE SERR Enable
A value of ’1’ enables the SERR driver.
A value of ’0’ disables the SERR driver .
70
BWait Cycle Control
The DSCC4 does never perform address/data stepping.
6 PER Parity Error Response
When this bit is set the DSCC4 will take its normal action when
a parity error is detected. When this bit is ’0’ the DSCC4
ignores any parity errors that it detects and continues normal
operation.
50
BVGA Palette Snoop
The DSCC4 is no VGA-Device.
40
BMemory Wri te and Inva lidate Enable
The “Invalidate” command is not supported by the DSCC4.
3SCSpecial Cycles
All special cycles are ignored.
Note: Although this bit can be set it has no effect.
2BMBus Master
A value of ’1’ enables the bus master capability.
Note: Before giving the first action request it is necessary to
set this bit.
1MSMemory Space
A value of ’1’ allows the DSCC4 to respond to Memory Space
Addresses.
Note: This bit must be set before the first read/write
transactions to the DSCC4 will be started.
Table 42 Status and Command register bits (cont’d)
PEB 20534
Detailed Register Description
Semiconductor Group 227 Data Sheet 09.98
0IOSIO Space
I/O Space accesses to the DSCC4 are not supported.
Note: Although this bit can be set it has no effect.
Table 42 Status and Comma nd regis ter bits (cont’d)
PEB 20534
Detailed Register Description
Semiconductor Group 228 Data Sheet 09.98
10.3 On-Ch ip Registers Description
10.3.1 Global Registers - Detailed Register Description
10.3.1.1 Global Registers Overview
The DSCC4 global registers are used to configure and control the DMA controller,
central FIFOs and genral device functions.
The full 32 bit address location of each global register consists of:
• Base Address Register 0 (PCI Configuration Space, address location 10H)
• Register address offset, which is in the range 0000H ...00FCH
All registers are 32-bit organized registers.
Table 43 provides an overview about all global registers:
Table 43 DSCC4 Global Register Overview
Offset Register Meaning
General registers:
0000HGCMDR Global Command Register
0004HGSTAR Global Status Register
0008HGMODE Global Mode Register
Interrupt Queue (IQ) specific registers (and FIFO Control 4 register):
000CHIQLENR0 IQ Length Register 0
0010HIQLENR1 IQ Length Register 1
0014HIQSCC0RXBAR IQ SCC0 RX Base Address Reg.
0018HIQSCC1RXBAR IQ SCC1 RX Base Address Reg.
001CHIQSCC2RXBAR IQ SCC2 RX Base Address Reg.
0020HIQSCC3RXBAR IQ SCC3 RX Base Address Reg.
0024HIQSCC0TXBAR IQ SCC0 TX Base Address Reg.
0028HIQSCC1TXBAR IQ SCC1 TX Base Address Reg.
002CHIQSCC2TXBAR IQ SCC2 TX Base Address Reg.
0030HIQSCC3TXBAR IQ SCC3 TX Base Address Reg.
0034HFIFOCR4 FIFO Control Register 4
0038H
RESERVED
-
PEB 20534
Detailed Register Description
Semiconductor Group 229 Data Sheet 09.98
003CHIQCFGBAR IQ CFG Base Addre ss Reg.
0040HIQPBAR IQ Peripheral Base Address Reg.
DMA Controller (DMAC) specific registers:
0044HFIFOCR1 FIFO Control Regis ter 1
0048HFIFOCR2 FIFO Control Regis ter 2
004CHFIFOCR3 FIFO Control Register 3
0050HCH0CFG Channel 0 Configuration Register
0054HCH0BRDA Channel 0 Base Rx Descr Address
0058HCH0BTDA Channel 1 Base Tx Descr Address
005CHCH1CFG Channel 1 Configuration Register
0060HCH1BRDA Channel 1 Base Rx Descr Address
0064HCH1BTDA Channel 1 Base Tx Descr Address
0068HCH2CFG Channel 2 Configuration Register
006CHCH2BRDA Channel 2 Base Rx Descr Address
0070HCH2BTDA Channel 2 Base Tx Descr Address
0074HCH3CFG Channel 3 Configuration Register
0078HCH3BRDA Channel 3 Base Rx Descr Address
007CHCH3BTDA Channel 3 Base Tx Descr Address
0080H
...
RESERVED
-
0094H
0098HCH0FRDA Channel 0 First Rx Descr Address
009CHCH1FRDA Channel 1 First Rx Descr Address
00A0HCH2FRDA Channel 2 First Rx Descr Address
00A4HCH3FRDA Channel 3 First Rx Descr Address
00A8H
RESERVED
-
00ACH
RESERVED
-
00B0HCH0FTDA Channel 0 First Tx Descr Address
00B4HCH1FTDA Channel 1 First Tx Descr Address
00B8HCH2FTDA Channel 2 First Tx Descr Address
Table 43 DSCC4 Global Register Overview (cont’d)
Offset Register Meaning
PEB 20534
Detailed Register Description
Semiconductor Group 230 Data Sheet 09.98
00BCHCH3FTDA Channel 3 First Tx Descr Address
00C0H
RESERVED
-
00C4H
RESERVED
-
00C8HCH0LRDA Channel 0 Last Rx Descr Address
00CCHCH1LRDA Channel 1 Last Rx Descr Address
00D0HCH2LRDA Channel 2 Last Rx Descr Address
00D4HCH3LRDA Channel 3 Last Rx Descr Address
00D8H
RESERVED
-
00DCH
RESERVED
-
00E0HCH0LTDA Channel 0 Last Tx Descr Address
00E4HCH1LTDA Channel 1 Last Tx Descr Address
00E8HCH2LTDA Channel 2 Last Tx Descr Address
00ECHCH3LTDA Channel 3 Last Tx Descr Address
00F0H
...
RESERVED
-
00FCH
Table 43 DSCC4 Global Register Overview (cont’d)
Offset Register Meaning
PEB 20534
Detailed Register Description
Semiconductor Group 231 Data Sheet 09.98
10.3.1.2 Global Regist ers Description
Each register description is organized in three parts:
• a head with general information about reset value, access type (read/write), offset
address and usual handling;
• a table containing the bit information (name of bit positions);
• a table containing the detailed description of each bit.
PEB 20534
Detailed Register Description
Semiconductor Group 232 Data Sheet 09.98
Table 44 GCMDR: Global Command Register
CPU Accessibility: read/write
Reset Value: 0000 0200H
Offset Address: 0000H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
Interrupt Queue Configuration Commands
CFG
IQ
SCC
3
RX
CFG
IQ
SCC
2
RX
CFG
IQ
SCC
1
RX
CFG
IQ
SCC
0
RX
CFG
IQ
SCC
3
TX
CFG
IQ
SCC
2
TX
CFG
IQ
SCC
1
TX
CFG
IQ
SCC
0
TX
00CFG
IQ
CFG
CFG
IQ
P
0000
Bit1514131211109876543210
Transmit Poll Requests / Interrupt Mask AR Action Request (AR)
00
TXPR3
TXPR2
TXPR1
TXPR0
IMAR
00000000AR
PEB 20534
Detailed Register Description
Semiconductor Group 233 Data Sheet 09.98
CFGIQ
SCC3RX Configure Interrupt Queue SCC3 Receive (Channel RX 3)
CFGIQ
SCC2RX Configure Interrupt Queue SCC2 Receive (Channel RX 2)
CFGIQ
SCC1RX Configure Interrupt Queue SCC1 Receive (Channel RX 1)
CFGIQ
SCC0RX Configure Interrupt Queue SCC0 Receive (Channel RX 0)
Only valid, if action request bit ’AR’ is set.
The DSCC4 DMA (interrupt) controller will transfer interrupt vectors
generated by the dedicated SCC receiver (3..0) to the corresponding
interrupt queue which must be configured via ’CFGIQSCCiRX’
command bits :
bit=’0’ The DSCC4 DMA (interrupt) controller does NOT
configure/re-configure the corresponding interrupt queue,
if action request bit ’AR’ is set to ’1’.
bit=’1’ Causes the DSCC4 DMA (interrupt) controller to
configure/re-configure the corresponding interrupt queue,
if action request bit ’AR’ is set to ’1’.
On action request, the DMA (interrupt) controller will
evaluate the corresponding interrupt queue base address
and length registers which must have been programmed
by software before.
PEB 20534
Detailed Register Description
Semiconductor Group 234 Data Sheet 09.98
CFGIQ
SCC3TX Configure Interrupt Queue SCC3 Transmit (Chann el TX 3)
CFGIQ
SCC2TX Configure Interrupt Queue SCC2 Transmit (Chann el TX 2)
CFGIQ
SCC1TX Configure Interrupt Queue SCC1 Transmit (Chann el TX 1)
CFGIQ
SCC0TX Configure Interrupt Queue SCC0 Transmit (Chann el TX 0)
Only valid, if action request bit ’AR’ is set.
The DSCC4 DMA (interrupt) controller will transfer interrupt vectors
generated by the dedicated SCC transmitter (3..0) to the corresponding
interrupt queue which must be configured via ’CFGIQSCCiTX’ command
bits:
bit=’0’ The DSCC4 DMA (interrupt) controller does NOT
configure/re-configure the corresponding interrupt queue,
if action request bit ’AR’ is set to ’1’.
bit=’1’ Causes the DSCC4 DMA (interrupt) controller to configure
the corresponding interrupt queue, if action request bit
’AR’ is set to ’1’.
On action request, the DMA (interrupt) controller will
evaluate the corresponding interrupt queue base address
and length registers which must have been programmed
by software before.
CFGIQCFG Configure Interrupt Queue Configuration (-)
Only valid, if action request bit ’AR’ is set.
The DSCC4 DMA (interrupt) controller will transfer action request
acknowledge/failure interrupt vectors to the configuration interrupt
queue which must be configured via ’CFGIQCFG’ command bits:
bit=’0’ The DSCC4 DMA (interrupt) controller does NOT
configure/re-configure the configuration interrupt queue, if
action request bit ’AR’ is set to ’1’.
bit=’1’ Causes the DSCC4 DMA (interrupt) controller to configure
the configuration interrupt queue, if action request bit ’AR’
is set to ’1’.
On action request, the DMA (interrupt) controller will
evaluate the configuration interrupt queue base address
and length registers which must have been programmed
by software before.
PEB 20534
Detailed Register Description
Semiconductor Group 235 Data Sheet 09.98
CFGIQP Configure Interrupt Queue Peripheral (-)
Only valid, if action request bit ’AR’ is set.
The DSCC4 DMA (interrupt) controller will transfer interrupt vectors
generated by the Loccal Bus Interface (LBI) to the peripheral interrupt
queue which must be configured via ’CFGIQP’ command bits:
bit=’0’ The DSCC4 DMA (interrupt) controller does NOT
configure/re-configure the peripheral interrupt queue, if
action request bit ’AR’ is set to ’1’.
bit=’1’ Causes the DSCC4 DMA (interrupt) controller to configure
the peripheral interrupt queue, if action request bit ’AR’ is
set to ’1’.
On action request, the DMA (interrupt) controller will
evaluate the peripheral interrupt queue base address and
length registers which must have been programmed by
software before.
TXPRi
(i=3...0) Transmit Poll Request Channel i (Channel TX 3...0)
Self-clearing command bit, only valid in ’HOLD’ bit controlled DMA
controller mode (bit CMODE = ’0’ in register GMODE):
TXPRi=’0’ No Transmit Poll Request is performed. The
corresponding DMA controller transmit channel is
stopped when HOLD=’1’ has been detected in the current
transmit descrip tor.
TXPRi=’1’ Setting this bit to ’1’, when HOLD=’1’ has been detected
in the current transmit descriptor, will cause the DSCC4 to
poll the ’HOLD’ bit in the current transmit descriptor, i.e.
the DSCC4 reads the configuration word (DWORD 0) and
next descriptor address (DWORD 1) of the current
descriptor again. If the ’HOLD’ bit is detected cleared (’0’),
the DMA controller will branch to the next descriptor.
When the DMA controller is not in ’HOLD’ state, this
command is discarded.
PEB 20534
Detailed Register Description
Semiconductor Group 236 Data Sheet 09.98
IMAR Interrupt Mask Action Request (-)
On any action request, the DSCC4 will generate either an ’action request
acknowledge’ or an ’action request failed’ interrupt vector which is
transferred into the configuration interrupt queue. These interrupts can
be masked via bit ’IMAR’:
IMAR=’0’ ’action request acknowledge’ and ’action request failed’
interrupt vectors respectively are generated and
transferred into the configuration interrupt queue.
IMAR=’1’ (Reset value)
’action request acknowledge’ and ’action request failed’
interrupt vectors respectively are NOT generated (and
thus NOT transferred into the configuration interrupt
queue).
AR Action Request (-)
Self-clearing command bit:
AR=’0’ No action request is performed.
AR=’1’ If this bit is set to ’1’, the DMA controller will evaluate:
• register GCMDR for all interrupt queue configuration
commands;
• all DMA channel specific configuration registers
(CHiCFG, i=3...0) for reset and initialization commands.
Any command (command bit set to ’1’) will cause the
corresponding configuration process to start.
A ’action request acknowledge’ or ’action request failed’
interrupt is generated after completion of all configuration
processes and a corresponding status bit is set in register
GSTAR.
PEB 20534
Detailed Register Description
Semiconductor Group 237 Data Sheet 09.98
The Global Status Register indicates whether an action request was executed
successfully or not. It also gives information about the interrupt source and which
interrupt queue has been written to when INTA is activated.
Ten interrupt queues are provided:
– four queues for receive interrupt vectors of the SCCs (SCCi, i=0...3)
– four queues for transmit interrupt vectors of the SCCs (SCCi, i=0...3)
– one queue for configuration interrupt vectors (action request acknowledge/failed)
– one queue for interrupts of the internal peripherals (SSC, LBI, and GPP).
To clear any bit in the status register, the host CPU must set the corresponding bit to “1“
by register write access. Signal INTA will be deasserted by the DSCC4 if ALL GSTAR
indications are cleared.
Table 45 GSTAR: Global Status Register
CPU Accessibility: read/write
Reset Value: 0000 000H
Offset Address: 0004H
typical usage: written by DSCC4 as interrupt indication
evaluated by CPU and written as interrupt confirmation
Bit31302928272625242322212019181716
Queue Specific Interrupt Indication
II
SCC
3
RX
II
SCC
2
RX
II
SCC
1
RX
II
SCC
0
RX
II
SCC
3
TX
II
SCC
2
TX
II
SCC
1
TX
II
SCC
0
TX
00II
CFG 0II
P
SSC
II
P
LBI
0II
P
GPP
Bit1514131211109876543210
Action Request Result Status
00000000000000
ARF
ARACK
PEB 20534
Detailed Register Description
Semiconductor Group 238 Data Sheet 09.98
IISCC3RX Interrupt Indication Queue SCC3 Receive (Channel RX 3)
IISCC2RX Interrupt Indication Queue SCC2 Receive (Channel RX 2)
IISCC1RX Interrupt Indication Queue SCC1 Receive (Channel RX 1)
IISCC0RX Interrupt Indication Queue SCC0 Receive (Channel RX 0)
These bits indicate whether at least one new interrupt vector was
transferred into the corresponding receive interrupt queue:
bit=’0’ No new interrupt vector was transferred into the
corresp ond ing que ue.
bit=’1’ At least one new interrupt vector was transferred into the
corresp ond ing que ue.
IISCC3TX Interrupt Indication Queue SCC3 Transmit (Chann el TX 3)
IISCC2TX Interrupt Indication Queue SCC2 Transmit (Chann el TX 2)
IISCC1TX Interrupt Indication Queue SCC1 Transmit (Chann el TX 1)
IISCC0TX Interrupt Indication Queue SCC0 Transmit (Chann el TX 0)
These bits indicate whether at least one new interrupt vector was
transferred into the corresponding transmit interrupt queue:
bit=’0’ No new interrupt vector was transferred into the
corresp ond ing que ue.
bit=’1’ At least one new interrupt vector was transferred into the
corresp ond ing que ue.
IICFG Interrupt Indication Configuration Queue (-)
This bit indicates whether at least one new interrupt vector was
transferred into the configuration interrupt queue:
bit=’0’ No new interrupt vector was transferred into the
configuration interrupt queue.
bit=’1’ At least one new interrupt vector was transferred into the
configuration interrupt queue.
PEB 20534
Detailed Register Description
Semiconductor Group 239 Data Sheet 09.98
IIPSSC Interrupt Indication Peripheral Queue (SSC Interrupt) (-)
This bit indicates whether at least one new SSC interrupt vector was
transferred into the peripheral interrupt queue:
bit=’0’ No new SSC interrupt vector was transferred into the
peripheral interrupt queue.
bit=’1’ At least one new SSC interrupt vector was transferred into
the peripheral interrupt queue.
IIPLBI Interrupt Indication Peripheral Queue (LBI Interrupt) (-)
This bit indicates whether at least one new LBI interrupt vector was
transferred into the peripheral interrupt queue:
bit=’0’ No new LBI interrupt vector was transferred into the
peripheral interrupt queue.
bit=’1’ At least one new LBI interrupt vector was transferred into
the peripheral interrupt queue.
IIPGPP Interrupt Indication Peripheral Queue (GPP Interrupt) (-)
This bit indicates whether at least one new GPP interrupt vector was
transferred into the peripheral interrupt queue:
bit=’0’ No new GPP interrupt vector was transferred into the
peripheral interrupt queue.
bit=’1’ At least one new GPP interrupt vector was transferred into
the peripheral interrupt queue.
ARF Action Request Failed Status (-)
This bit indicates that an action request command was completed with
an ’action request failed’ condition:
bit=’0’ No action request was performed or no ’action request
failed’ condition occured completing an action request.
bit=’1’ The last action request command was completed with an
’action request failed’ condition.
PEB 20534
Detailed Register Description
Semiconductor Group 240 Data Sheet 09.98
ARACK Action Request Acknowledge Status (-)
This bit indicates that an action request command was completed
successfully:
bit=’0’ No action request was performed or completed
successfully.
bit=’1’ The last action request command was completed
successfully.
PEB 20534
Detailed Register Description
Semiconductor Group 241 Data Sheet 09.98
Table 46 GMODE: Global Mode Register
CPU Accessibility: read/write
Reset Value: 0000 000H
Offset Address: 0008H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
General Configuration
0000000000OSC
PD LCD(1:0) PERCFG(2:0)
Bit1514131211109876543210
General Configuration
SPRI
CHN(1:0)0000000000
ENDIAN
BURST
CMODE
PEB 20534
Detailed Register Description
Semiconductor Group 242 Data Sheet 09.98
OSCPD Oscillator Power Down (-)
This bit switches the internal oscillator (used if a crystal is connected to
pins XTAL1 and XTAL2) in power-down (stand-by) mode:
OSCPD=’0’ Normal operation. The internal oscillator works, if a crystal
is connected to pins XTAL1 and XTAL2.
OSCPD=’1’ The internal oscillator is in power-down mode.
LCD(1:0) LBI Clock Division (-)
The internal LBI operating clock (which is monitored on output pin
LCLKO) is internally derived from the PCI clock input pin CLK and a
clock division unit. The division factor can be selected via this bit field:
LCD = ’00’
Reserved
, do not use.
LCD = ’01’ LCLK = CLK / 2
LCD = ’10’ LCLK = CLK / 4
LCD = ’11’ LCLK = CLK / 16
Note: The LBI clock signal monitored on pin LCLKO is an asymmetric
clock signal. The LBI clock high phase time is always equal the
PCI clock high phase time (typical 15 nano seconds). The LBI
clock low phase time is extended respectively.
PEB 20534
Detailed Register Description
Semiconductor Group 243 Data Sheet 09.98
PERCFG
(2:0) Peripheral Block Configuration (-)
The peripheral block basically consists of the functions
• Local Bus Interface (LBI)
• General Purpose Port (GPP)
• Synchronous Serial Controller (SSC)
which can be operated in various combinations/configurations.
Bit field ’PERCFG’ selects the peripheral configuration and switches the
multiplex ed sig nal pins accord ing ly:
PCI Interface Mode (DEMUX pin connected to VSS):
PERCFG Signal Pin Groups
(2:0) 109, 108 119..112 143..135
101..96 128..123
’000’ LA(15..8) LA(7..0) LD(15..0)
’001’
Reserved
. Do not use.
’010’ GP(15..8) LA(7..0) LD(15..0)
’011’ GP(15..8) SSC LD(15..0)
’100’ GP(15..8) GP(7..0) LAD(15..0)
’101’,
’110’,
’111’
Reserved
. Do not use.
DEMUX Interface Mode (DEMUX pin connected to VDD3):
Bit field ’PERCFG’ is not valid. All 32 multiplexed signals are used as
DEMUX address bus A(31:0):
PERCFG Signal Pin Groups
(2:0) 109, 108 119..112 143..135
101..96 128..123
’xxx’ A(15..8) A(7..0) A(31..0)
PEB 20534
Detailed Register Description
Semiconductor Group 244 Data Sheet 09.98
SPRI Select Priority (-)
This bit selects one DMA channel (transmit and receive) as a high priority
chann el. The dedicated ch annel 3..0 is selected via b it field ’CHN’.
If enabled, the selected serial channel (SCCi) is serviced by the DMA
controller with high priority, whereas the remaining channels are in a
second round-robbin priority group:
SPRI=’0’ No SCC is selected as high priority channel. All SCCs are
serviced in a round-robbin manner.
Bit field ’CHN’ is don’t care.
SPRI=’1’ The SCC selected via bit field ’CHN’ is serviced as high
priority channel, i.e. requests to central receive and
transmit FIFOs are priorized against the other channels.
CHN(1:0) Channel Number (for highest priority) (-)
This bit field is only valid if bit ’SPRI’ is set to ’1’ and specifies the serial
channel which is serviced with highest priority by the DMA controller:
CHN=’00’ Channel 0 (transmit and receive) is selected as high
priority channel.
CHN=’01’ Channel 1 (transmit and receive) is selected as high
priority channel.
CHN=’10’ Channel 2 (transmit and receive) is selected as high
priority channel.
CHN=’11’ Channel 3 (transmit and receive) is selected as high
priority channel.
PEB 20534
Detailed Register Description
Semiconductor Group 245 Data Sheet 09.98
ENDIAN Endian Selection (-)
This bit selects whether receive and transmit data buffers are handled
with Intel or Motorola like byte ordering:
ENDIAN=’0’ The DWORDs of receive and transmit data buffers are
evaluated based on a little endian (Intel like) byte
ordering.
ENDIAN=’1’ The DWORDs of receive and transmit data buffers are
evaluated based on a big endian (Motorola like) byte
ordering. Therefore the byte ordering is automatically
swapped by the DMA controller.
Note: The little/big endian selection (byte-swapping) effects only DSCC4
operation on receive and transmit data buffer sections. Descriptor
reads and writes as well as register access is not effected anyway.
DBE DEMUX Bur st Enable (-)
This bit is only valid if the DSCC4 is running in de-multiplexed bus
interface mode, i.e. pin DEMUX connected to VDD3.
By default value, the burst functionality is disabled in DEMUX mode and
can be enabled via setting this bit. However burst length is limited to 4
DWORDs in DEMUX mode (15 DWORDs in PCI mode):
DBE=’0’ Burst functionality is disabled. The DSCC4 will perform all
transactions to the host memory using single DWORD
read/write bus transfers.
DBE=’1’ Burst functionality is enabled. The DSCC4 performs burst
transfers for operation on descriptors and data sections
(like in PCI mode). Burst length is limited to 4 DWORDs
maximum.
CMODE DMA Control Mode (-)
This bit selects between the two global DMA controller mechanisms for
handling descriptor chain end conditions:
CMODE=’0’ ’HOLD’ bit control mode.
The descriptor chain end condition is controlled via the
’HOLD’ bit in each receive/transmit descriptor.
CMODE=’1’ Last Receive/Transmit Descriptor Address mode.
The descriptor chain end condition is controlled via
registers LRDA/LTDA.
PEB 20534
Detailed Register Description
Semiconductor Group 246 Data Sheet 09.98
PEB 20534
Detailed Register Description
Semiconductor Group 247 Data Sheet 09.98
Table 47 IQLENR1: Interrupt Queue Length Register 1
CPU Accessibility: read/write
Reset Value: 0000 000H
Offset Address: 000CH
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
Interrupt Queues Length Configuration
IQSCC0RXLEN IQSCC1RXLEN IQSCC2RXLEN IQSCC3RXLEN
Bit1514131211109876543210
Interrupt Queues Length Configuration
IQSCC0TXLEN IQSCC1TXLEN IQSCC2TXLEN IQSCC3TXLEN
PEB 20534
Detailed Register Description
Semiconductor Group 248 Data Sheet 09.98
IQSCC3
RXLEN Interrupt Queue SCC3 Receive Length (Channel RX 3)
IQSCC2
RXLEN Interrupt Queue SCC2 Receive Length (Channel RX 2)
IQSCC1
RXLEN Interrupt Queue SCC1 Receive Length (Channel RX 1)
IQSCC0
RXLEN Interrupt Queue SCC0 Receive Length (Channel RX 0)
These bit fields determine the length of the corresponding receive
interrupt queue (related to the respective SCC receive channel):
Queue
length: Queue Length = (1 + ’IQSCCiRXLEN’) * 32 DWORDS,
’IQSCCiRXLEN’ = 0...15
IQSCC3
TXLEN Interrupt Queue SCC3 Transmit Length (Channel TX 3)
IQSCC2
TXLEN Interrupt Queue SCC2 Transmit Length (Channel TX 2)
IQSCC1
TXLEN Interrupt Queue SCC1 Transmit Length (Channel TX 1)
IQSCC0
TXLEN Interrupt Queue SCC0 Transmit Length (Channel TX 0)
These bit fields determine the length of the corresponding transmit
interrupt queue (related to the respective SCC transmit channel):
Queue
length: Queue Length = (1 + ’IQSCCiTXLEN’) * 32 DWORDS,
’IQSCCiTXLEN’ = 0...15
PEB 20534
Detailed Register Description
Semiconductor Group 249 Data Sheet 09.98
Table 48 IQLENR2: Interrupt Queue Length Register 2
CPU Accessibility: read/write
Reset Value: 0000 000H
Offset Address: 0010H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
Interrupt Queues Length Configuration
00000000 IQCFGLEN IQPLEN
Bit1514131211109876543210
(unused)
0000000000000000
PEB 20534
Detailed Register Description
Semiconductor Group 250 Data Sheet 09.98
IQCFGLEN Interrupt Queue Configuration Length (-)
These bit field determins the length of the configuration interrupt queue:
Queue
length: Queue Length = (1 + ’IQCFGLEN’) * 32 DWORDS,
’IQCFGLEN’ = 0...15
IQPLEN Interrupt Queue Peripheral Length (-)
These bit field determins the length of the peripheral interrupt queue:
Queue
length: Queue Length = (1 + ’IQPLEN’) * 32 DWORDS,
’IQPLEN’ = 0...15
PEB 20534
Detailed Register Description
Semiconductor Group 251 Data Sheet 09.98
Table 49 IQSCCiRXBAR:
Interrupt Queue SCCi Receiver Base Address Register (i=0...3)
CPU Accessibility: read/write
Reset Value: 0000 000H
Channel 0 Channel 1 Channel 2 Channel 3
Offset Address: 0014H0018H001CH0020H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
IQSCCiRXBAR(31:16)
Bit1514131211109876543210
IQSCCiRXBAR(15:2) 0 0
IQSCCi
RXBAR i = 3...0 (RX Channel 3...0)
These registers determine the base address of the respective receive
interrupt queue and can be located anywhere in the 32 bit address
range.
Note: The interrupt queue base addresses must be 32-bit aligned, i.e. bit
1 and 0 must be set to ’0’.
PEB 20534
Detailed Register Description
Semiconductor Group 252 Data Sheet 09.98
Table 50 IQSCCiTXBAR:
Interrupt Queue SCCi Receiver Base Address Register (i=0...3)
CPU Accessibility: read/write
Reset Value: 0000 000H
Channel 0 Channel 1 Channel 2 Channel 3
Offset Address: 0024H0028H002CH0030H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
IQSCCiTXBAR(31:16)
Bit1514131211109876543210
IQSCCiTXBAR(15:2) 0 0
IQSCCi
TXBAR i = 3...0 (TX Channel 3...0)
These registers determine the base address of the respective transmit
interrupt queue and can be located anywhere in the 32 bit address
range.
Note: The interrupt queue base addresses must be 32-bit aligned, i.e. bit
1 and 0 must be set to ’0’.
PEB 20534
Detailed Register Description
Semiconductor Group 253 Data Sheet 09.98
Table 51 IQCFGBAR:
Interrupt Queue Configuration Base Address Register
CPU Accessibility: read/write
Reset Value: 0000 000H
Offset Address: 003CH
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
IQCFGBAR(31:16)
Bit1514131211109876543210
IQCFGBAR(15:2) 0 0
IQCFGBAR (-)
This register determins the base address of the configuration interrupt
queue and can be located anywhere in the 32 bit address range.
Note: The interrupt queue base address must be 32-bit aligned, i.e. bit 1
and 0 must be set to ’0’.
PEB 20534
Detailed Register Description
Semiconductor Group 254 Data Sheet 09.98
Table 52 IQPBAR:
Interrupt Queue Peripheral Base Address Register
CPU Accessibility: read/write
Reset Value: 0000 000H
Offset Address: 0040H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
IQPBAR(31:16)
Bit1514131211109876543210
IQPBAR(15:2) 0 0
IQPBAR (-)
This register determins the base address of the peripheral interrupt
queue and can be located anywhere in the 32 bit address range.
Note: The interrupt queue base address must be 32-bit aligned, i.e. bit 1
and 0 must be set to ’0’.
PEB 20534
Detailed Register Description
Semiconductor Group 255 Data Sheet 09.98
Table 53 FIFOCR1: FIFO Control Register 1
CPU Accessibility: read/write
Reset Value: 0000 000H
Offset Address: 0044H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
Central Transmit FIFO Partition Size
TFSIZE0 TFSIZE1 TFSIZE2 0
Bit1514131211109876543210
Central Transmit FIFO Partition Size
TFSIZE3 00000000000
TFSIZEi
(i = 0...3) Transmit FIFO Size (Partition Channel i) (TX Channel i)
These bit fields determine the channel specific partition size of the
central transmit FIFO in multiples of 4 DWORDs:
partition size i = ’TFSIZEi’ * 4 DWORDs,
range 0...124 DWORDs
Note: The entire size of all FIFO partitions must not exceed the total
FIFO size of 128 DWORDs which is the responsibility of the
software.
Note: If the complete central transmit FIFO is assigned to only one
channel, the maximum partition size is 124 DWORDs.
Note: The minimum allowed partition size for used channels is 4
DWORDs, i.e. ’TFZIZEi’ = 1.
PEB 20534
Detailed Register Description
Semiconductor Group 256 Data Sheet 09.98
Table 54 FIFOCR2: FIFO Control Register 2
CPU Accessibility: read/write
Reset Value: 0000 000H
Offset Address: 0048H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
Central Transmit FIFO Refill Threshold
TFRTHRES0 TFRTHRES1 TFRTHRES2 0
Bit1514131211109876543210
Central Transmit FIFO Refill Threshold Multipliers
TFRTHRES3 0 0 0 M4_
0M2_
0M4_
1M2_
1M4_
2M2_
2M4_
3M2_
4
PEB 20534
Detailed Register Description
Semiconductor Group 257 Data Sheet 09.98
TFRTHRESi
(i = 0...3) Transmit FIFO Refill Threshold Channel i (TX Channel i)
These bit fields determine the channel specific Transmit FIFO Refill
Threshold for the corresponding channel i in number of DWORDs
multiplied by its respective multiplier Mi_x.
This threshold controls DMAC operation towards the Host memory.
A watermark is calculated by:
watermark = TFRTHRESH*Mi_x+1
.
As soon as the number of valid data in the transmit FIFO is less than
watermark
the DMA controller requests for new data from shared
memory.
After initialization the DMAC fills the complete transmit FIFO; when the
number of data in the transmit FIFO decreases below the
watermark
the
transmit FIFO requests for refill. The
number
of data to be transferred
into the transmit FIFO is calculated by:
number=TFSIZEi-watermark
.
The transmit FIFO cannot be filled when the DMA channel is in internal
HOLD state, i.e. ’HOLD’ bit has been detected (GMODE.CMODE=’0’)
or Last Descriptor Adress matches the current descriptor address
FTDA = LTDA (GMODE.CMODE=’1’).
Note: The watermark has to be equal or lower than the specified size of
the corresponding central transmit FIFO partition.
Note: The minimum allowed TFRTHRESHi value for used channels is
“1”.
Mx_i
(i = 0...3) Multiplier 2 or 4 (TX Channel i)
These bits enable a multiplier 2 or 4 respectively for the corresponding
’TFRTHRESi’ value:
M2_i = ’0’ The multiplier ’by 2’ is disabled
M2_i = ’1’ The ’TFRTHRESi’ bit field value is multiplied by 2.
M4_i = ’0’ The multiplier ’by 4’ is disabled
M4_i = ’1’ The ’TFRTHRESi’ bit field value is multiplied by 4.
Note: It is recommended not to set both multiplier enable bits of one
channel to ’1’.
PEB 20534
Detailed Register Description
Semiconductor Group 258 Data Sheet 09.98
Table 55 FIFOCR3: FIFO Control Register 3
CPU Accessibility: read/write
Reset Value: 0000 000H
Offset Address: 004CH
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
(unused)
0000000000000000
Bit1514131211109876543210
(unused) Multipliers Central Receive FIFO Threshold
0000000M4M2 RFTHRES
PEB 20534
Detailed Register Description
Semiconductor Group 259 Data Sheet 09.98
RFRTHRES Receive FIFO Threshold (-)
This bit field determins the central Receive FIFO Threshold in number of
DWORDs multiplied by its respective multiplier ’M2’ or ’M4’.
This threshold controls DMAC operation towards the Host memory.
A watermark is calculated by:
watermark = RFRTHRESH*Mx
.
When more data than specified by this watermark are available in the
receive FIFO the DMA controller is requested to transfer the received
data to the channel specific data buffers in the host memory until the
central receive FIFO is empty. If no host memory buffer is available for
a channel, since the internal HOLD state is reached, i.e. ’HOLD’ bit has
been detected (GMODE.CMODE=’0’) or Last Descriptor Adress mat-
ches the current descriptor address FRDA = LRDA
(GMODE.CMODE=’1’), no data can be transferred.
Note: The watermark has to be lower than the maximum central receive
FIFO size of 128 DWORDs .
Note: The minimum allowed RFRTHRESH value is “1”.
Mx Multiplier 2 or 4 (-)
These bits enable a multiplier 2 or 4 respectively for the ’RFRTHRES’
value:
M2 = ’0’ The multiplier ’by 2’ is disabled
M2 = ’1’ The ’RFRTHRES’ bit field value is multiplied by 2.
M4 = ’0’ The multiplier ’by 4’ is disabled
M4 = ’1’ The ’RFRTHRES’ bit field value is multiplied by 4.
Note: It is recommended not to set both multiplier enable bits to ’1’.
PEB 20534
Detailed Register Description
Semiconductor Group 260 Data Sheet 09.98
Table 56 FIFOCR4: FIFO Control Register 4
CPU Accessibility: read/write
Reset Value: 0000 000H
Offset Address: 0034H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
Central Transmit FIFO Forward Thresholds
TFFTHRES3 TFFTHRES2
Bit1514131211109876543210
Central Transmit FIFO Forward Thresholds
TFFTHRES1 TFFTHRES0
PEB 20534
Detailed Register Description
Semiconductor Group 261 Data Sheet 09.98
TFFTHRESi
(i = 0...3) Transmit FIFO Forward Threshold Channel i (TX Channel i)
These bit fields determine the channel specific Transmit FIFO Forward
Threshold for the corresponding channel i in number of DWORDs.
This threshold controls DMAC operation towards the serial channels
(SCCi).
A watermark is calculated by:
watermark = TFFTHRESH
.
As soon as the number of valid data belonging to a new frame in the
central transmit FIFO is greater than the watermark, the DMAC will
provide transmit data to the corresponding SCC. Once having started
one frame, the DMAC will ignore this threshold providing all available
data of the current frame to the SCC. Threshold operation starts again
with the beginning of a new frame. Frames shorter than the threshold will
be transferred as soon as a frame end indication is detected by the
DMAC.
Note: The maximum allowed Transmit FIFO Forward Threshold is:
TFFTHRESHi = (TFSIZEi * 4) - 1 DWORDs,
whereas ’TFSIZEi’ is the channel specific central transmit FIFO
partition size programmed in register FIFOCR1
Note: Programming TFFTHRESHi to zero will disable the threshold
causing the DMAC to transfer all data immediately. This may be
useful for not frame/packet oriented data transmission, e.g. in
ASYNC protocol mode.
PEB 20534
Detailed Register Description
Semiconductor Group 262 Data Sheet 09.98
Table 57 CHiCFG: Channel i Configuration Register (i=3...0)
CPU Accessibility: read/write
Reset Value: 0000 000H
Channel 0 Channel 1 Channel 2 Channel 3
Offset Address: 0050H005CH0068H0074H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
DMA Channel Interrupt Masks DMA Channel Commands
0000
MRFI
MTFI
MRERR
MTERR
0
RDR
RDT
IDR
IDT
000
Bit1514131211109876543210
(unused)
0000000000000000
PEB 20534
Detailed Register Description
Semiconductor Group 263 Data Sheet 09.98
MRFI Mask Receive FI-Interrupt (Channel i) (RX Channel)
This bit enables/disables the receive FI-interrupt indication for the DMA
channel, the register is dedicated to (i=3..0):
MRFI=’0’ FI-interrupt generation is enabled for the dedicated DMA
recei ve cha nnel.
MRFI=’1’ FI-interrupt generation is disabled for the dedicated DMA
recei ve cha nnel.
MTFI Mask Transmit FI-Interrupt (Channel i) (TX Channel)
This bit enables/disables the transmit FI-interrupt indication for the DMA
channel, the register is dedicated to (i=3..0):
MTFI=’0’ FI-interrupt generation is enabled for the dedicated DMA
transmit channe l.
MTFI=’1’ FI-interrupt generation is disabled for the dedicated DMA
transmit channe l.
MRERR Mask Receive ERR-Interrupt (Channel i) (RX Channel)
This bit enables/disables the receive ERR-interrupt indication for the
DMA channel, the register is dedicated to (i=3..0):
MRERR=’0’ ERR-interrupt generation is enabled for the dedicated
DMA receive channel.
MRERR=’1’ ERR-interrupt generation is disabled for the dedicated
DMA receive channel.
MTERR Mask Transmit ERR-Interrupt (Channel i) (TX Channel)
This bit enables/disables the transmit ERR-interrupt indication for the
DMA channel, the register is dedicated to (i=3..0):
MTERR=’0’ ERR-interrupt generation is enabled for the dedicated
DMA transmit channel.
MTERR=’1’ ERR-interrupt generation is disabled for the dedicated
DMA transmit channel.
PEB 20534
Detailed Register Description
Semiconductor Group 264 Data Sheet 09.98
Table 58 CHiBRDA:
Channel i Base Receive Descriptor Address Register (i=3...0)
CPU Accessibility: read/write
Reset Value: 0000 000H
Channel 0 Channel 1 Channel 2 Channel 3
Offset Address: 0054H0060H006CH0078H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
CHiBRDA(31:16)
Bit1514131211109876543210
CHiBRDA(15:2) 0 0
CHiBRDA i = 3...0 (RX Channel 3...0)
These registers determine the base address of the channel specific
receive descriptor chain and can be located anywhere in the 32 bit
address range.
Note: The descriptor base addresses must be 32-bit aligned, i.e. bit 1
and 0 must be set to ’0’.
PEB 20534
Detailed Register Description
Semiconductor Group 265 Data Sheet 09.98
Table 59 CHiBTDA:
Channel i Base Transmit Descriptor Address Register (i=3...0)
CPU Accessibility: read/write
Reset Value: 0000 000H
Channel 0 Channel 1 Channel 2 Channel 3
Offset Address: 0058H0064H0070H007CH
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
CHiBTDA(31:16)
Bit1514131211109876543210
CHiBTDA(15:2) 0 0
CHiBTDA i = 3...0 (TX Channel 3...0)
These registers determine the base address of the channel specific
transmit descriptor chain and can be located anywhere in the 32 bit
address range.
Note: The descriptor base addresses must be 32-bit aligned, i.e. bit 1
and 0 must be set to ’0’.
PEB 20534
Detailed Register Description
Semiconductor Group 266 Data Sheet 09.98
Table 60 CHiFRDA:
Channel i First (Current) Receive Descriptor Address Register
(i=3...0)
CPU Accessibility: read/write
Reset Value: 0000 000H
Channel 0 Channel 1 Channel 2 Channel 3
Offset Address: 0098H009CH00A0H00A4H
typical usage: written by DSCC4
evaluated by CPU
Bit31302928272625242322212019181716
CHiFRDA(31:16)
Bit1514131211109876543210
CHiFRDA(15:2) 0 0
CHiFRDA i = 3...0 (RX Channel 3...0)
The DMA controller writes the first/current address of the channel
specific receive descriptor chain to these registers, i.e. the address of the
receive descriptor, the DMA receive channel i is currently working on.
These registers are only valid, if the DMA controller is operating in Last
Descriptor Address Mode (bit CMODE set to ’1’ in register GMODE).
PEB 20534
Detailed Register Description
Semiconductor Group 267 Data Sheet 09.98
Table 61 CHiFTDA:
Channel i First (Current) Transmit Descriptor Address Register
(i=3...0)
CPU Accessibility: read/write
Reset Value: 0000 000H
Channel 0 Channel 1 Channel 2 Channel 3
Offset Address: 00B0H00B4H00B8H00BCH
typical usage: written by DSCC4
evaluated by CPU
Bit31302928272625242322212019181716
CHiFTDA(31:16)
Bit1514131211109876543210
CHiFTDA(15:2) 0 0
CHiFTDA i = 3...0 (TX Channel 3...0)
The DMA controller writes the first/current address of the channel
specific transmit descriptor chain to these registers, i.e. the address of
the transmit descriptor, the DMA transmit channel i is currently working
on.
These registers are only valid, if the DMA controller is operating in Last
Descriptor Address Mode (bit CMODE set to ’1’ in register GMODE).
PEB 20534
Detailed Register Description
Semiconductor Group 268 Data Sheet 09.98
Table 62 CHiLRDA:
Channel i Last Receive Descriptor Address Register (i=3...0)
CPU Accessibility: read/write
Reset Value: 0000 000H
Channel 0 Channel 1 Channel 2 Channel 3
Offset Address: 00C8H00CCH00D0H00D4H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
CHiLRDA(31:16)
Bit1514131211109876543210
CHiLRDA(15:2) 0 0
PEB 20534
Detailed Register Description
Semiconductor Group 269 Data Sheet 09.98
CHiLRDA i = 3...0 (RX Channel 3...0)
These registers determine the last descriptor address of the channel
specific receive descriptor chain and can be located anywhere in the 32
bit address range. The last descriptor address is written by the CPU and
marks the corresponding descriptor as the last descriptor in the receive
desc ri ptor chain.
After write access to one of these registers, the DMA controller again
compares the first (current) receive descriptor address (register
CHiFRDA) with the last descriptor address (register CHiLRDA), if the
corresponding DMA controller channel was in internal HOLD state. If
these addresses do not match any more, the DMA channel leaves the
internal HOLD state, re-reads the next descriptor address of the current
receive descriptor and continues operation.
These registers are only valid, if the DMA controller is operating in Last
Descriptor Address Mode (bit CMODE set to ’1’ in register GMODE).
Note: The last descriptor addresses must be 32-bit aligned, i.e. bit 1 and
0 must be set to ’0’.
PEB 20534
Detailed Register Description
Semiconductor Group 270 Data Sheet 09.98
Table 63 CHiLTDA:
Channel i Last Transmit Descriptor Address Register (i=3...0)
CPU Accessibility: read/write
Reset Value: 0000 000H
Channel 0 Channel 1 Channel 2 Channel 3
Offset Address: 00E0H00E4H00E8H00ECH
typical usage: written by DSCC4
evaluated by CPU
Bit31302928272625242322212019181716
CHiLTDA(31:16)
Bit1514131211109876543210
CHiLTDA(15:2) 0 0
PEB 20534
Detailed Register Description
Semiconductor Group 271 Data Sheet 09.98
CHiFTDA i = 3...0 (TX Channel 3...0)
These registers determine the last descriptor address of the channel
specific transmit descriptor chain and can be located anywhere in the 32
bit address range. The last descriptor address marks the corresponding
descriptor as the last descriptor in the transmit descriptor chain.
After write access to one of these registers, the DMA controller again
compares the first (current) transmit descriptor address (register
CHiFTDA) with the last descriptor address (register CHiLTDA), if the
corresponding DMA controller channel was in internal HOLD state.If
these addresses do not match any more, the DMA channel leaves the
internal HOLD state, re-reads the next descriptor address of the current
transmit descriptor and continues operation.
These registers are only valid, if the DMA controller is operating in Last
Descriptor Address Mode (bit CMODE set to ’1’ in register GMODE).
Note: The last descriptor addresses must be 32-bit aligned, i.e. bit 1 and
0 must be set to ’0’.
PEB 20534
Detailed Register Description
Semiconductor Group 272 Data Sheet 09.98
10.3.2 SCC Registers - Detailed Register Description
10.3.2.1 SCC Registers Overview
The SCC registers are used to configure and control each of the four Serial
Communi cation Contro ller (SCC). The compl ete SCC register se t exists four times, i.e.
once for each SCC, distinguished by a SCC specific offset address.
The full 32 bit address location of each SCC register consists of:
• Base Address Register 0 (PCI Configuration Space, address location 10H)
• SCC specific offset address:
SCC0: 0100H
SCC1: 0180H
SCC2: 0200H
SCC3: 0280H
• Register address offset, which is in the range 00H ...58 H
Most regi sters and registe r bit positio ns are shared by all SCC protocol mode s (HDLC,
ASYNC, BISYNC). Neverthe less the meaning (and name) of single bit positions might
defer between different protocol modes. All registers are 32-bit organized registers.
Table 64 provides an overview about all SCC registers:
PEB 20534
Detailed Register Description
Semiconductor Group 273 Data Sheet 09.98
Table 64 SCC Register Overview
Register
Offset Access
Type Register Valid in
Mode(s)
00Hw CMDR Command Register H/A/B
04Hr STAR Status Register H/A/B
08Hr/w CCR0 Channel Configuration Register 0 H/A/B
0CHr/w CCR1 Channel Configuration Register 1 H/A/B
10Hr/w CCR2 Channel Configuration Register 2 H/A/B
14Hr/w ACCM ASYNC Control Character Map H (PPP)
18Hr/w UDAC User Defined ASYNC Character H (PPP)
1CHr/w TTSA Transmit Time Slot Assignment Register H/A/B
20Hr/w RTSA Receive Time Slot Assignment Register H/A/B
24Hr/w PCMMTX PCM Mask for Transmit H/A/B
28Hr/w PCMMRX PCM Mask for Receive H/A/B
2CHr/w BR R Baud Rate Register H/A/B
30Hr/w TIMR Timer Register H/A/B
34Hr/w XADR Transmit Address Register H
38Hr/w RADR Receive Address Reg iste r H
3CHr/w RAMR Receive Address Mask Register H
40Hr/w RLC R Re cei ve Leng th Che ck Re gis ter H
44Hr/w XNXFR XON/XOFF Register A
48Hr/w TCR Termination Character Register A/B
4CHr/w TICR Transmit Immediate Character Register A
50Hr/w SYNCR Synchronization Character Register B
54Hr/w IMR Interrupt Mask Register H/A/B
58Hr/w ISR Interrupt Status Register H/A/B
PEB 20534
Detailed Register Description
Semiconductor Group 274 Data Sheet 09.98
10.3.2.2 SCC Registers Description
Each register description is organized in three parts:
• a head with ge neral in forma tion a bout res et val ue, acc ess t ype (rea d/write), channe l
specific offset addresses and usual handling;
• a table containing the bit information (name of bit positions) distinguished for the three
major protocol modes HDLC (H), ASYNC (A) and BISYNC (B);
• a tab le containing the detail ed descriptio n of each bit; the correspo nding modes, the
bit is valid for, are marked again by a bracket term right beside the full bit name.
PEB 20534
Detailed Register Description
Semiconductor Group 275 Data Sheet 09.98
Table 65 CMDR: Command Register
CPU Accessibility: write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 0100H0180H0200H0280H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
Transmitter Commands Receiver Commands
H0000000
XRES
0000000
RRES
A0000000
XRES
000000
RFRD
RRES
B0000000
XRES
00000
HUNT
RFRD
RRES
Bit1514131211109876543210
Mode
Internal Commands Remote Control
H000000STI00000000RNR
A000000STI000000000
B000000STI000000000
PEB 20534
Detailed Register Description
Semiconductor Group 276 Data Sheet 09.98
XRES Transmitter Reset Command (all modes)
Self-clearing command bit:
XRES=’1’ The SCC transmit FIFO is cleared and the transmitter
protocol engines are reset to their initial state.
The SCC transmit FIFO requests new transmit data from
the central TFIFO immediately after transmitter reset
procedure.
A transmitter reset command is recommended after all
changes in protocol mode configurations (switching
between the protoc ol eng ines HDLC/ ASYNC /BISYN C or
sub-modes of HDLC).
HUNT Enter Hunt State Command (bisync mode)
Self-clearing command bit:
HUNT=’1’ This command forces the receiver to enter its ’HUNT’
state immediately. Thus synchronization is ’lost’ and the
receiver starts searching for new SYNC characters.
RFRD Receive FIFO Read Enable Command (async/bisync mode)
Self-clearing command bit:
RFRD=’1’ This command forces insertion of a ’block end’ indication
in the SCC receive FIFO. If the receive FIFO is not empty
(bit ’RFNE’ set in register STAR) and data was not
transferred to the central RFIFO because neither the
receive threshold is exceeded nor a block end indication
is stored, this command forces data transfer to the central
RFIFO.
Note: This command always generates a ’block-end’
indication. If the receive FIFO was empty, the DMA
Controller will finish the current receive descriptor
with receive byte number zero and frame-
end/blo ck -e nd indi cation bit set.
PEB 20534
Detailed Register Description
Semiconductor Group 277 Data Sheet 09.98
RRES Receiver Reset Command (all modes)
Self-clearing command bit:
RRES=’1’ The SCC receive FIFO is cleared and the receiver
protocol engines are reset to their initial state.
The SCC receive FIFO accepts new receive data from the
protocol engine immediately after receiver reset
procedure.
It is recommended to disable data reception before
issuing a receiver reset command by setting bit
CCR2.RAC = ’0’ and enabling data reception afterwards.
A ’receiver reset command’ is recommended after all
changes in protocol mode configurations (switching
between protocol the engines HDLC/ASYNC/BISYNC or
sub-modes of HDLC).
STI Start Timer Command (all modes)
Self-clearing command bit:
HDLC Automde:
In HDLC Automode the timer is operating in ’internal timer mode’. The
timer is started automatically by the SCC when an I-Frame is sent out
and needs to be acknowledged.
If the ’STI’ command is issued by software:
STI=’1’ An S-Frame with poll bit set is sent out and the internal
timer is started expecting an acknowledge from the
remote station via an I- or S-Frame.
The timer is stopped after receiving an acknowledge
otherwise the timer expires generating a timer interrupt.
All protocol modes except HDLC Automode:
In these modes the timer is operating in ’external timer mode’.
STI=’1’ This commands starts timer operation in ’external timer
mode’.
The timer can be stopped by rewriting register TIMR. If the
timer expires a timer interrupt is generated.
Note: Internal/External timer operation mode must be selected by bit
’TMD’ in register TIMR.
PEB 20534
Detailed Register Description
Semiconductor Group 278 Data Sheet 09.98
RNR Receiver Not Ready Command (hdlc mode)
NON self-clearing command bit:
This command bit is significant in Automode only.
RNR=’0’ Forces the receiver to enter its ’receiver-ready’ state. The
receiver acknowledges received poll or I-Frames with a
’receiver-ready’ indication.
RNR=’1’ Forces the receiver to enter its ’receiver-not-ready’ state.
The receiver acknowledges received poll or I-Frames with
a ’receiver-not-ready’ indication.
PEB 20534
Detailed Register Description
Semiconductor Group 279 Data Sheet 09.98
Table 66 STAR: Status Register
CPU Accessibility: read
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 0104H0184H0204H0284H
typica l usage : upda ted by DSC C 4
read and evaluated by CPU
Bit31302928272625242322212019181716
Mode
Command Status Transmitter Status Receiver Status Automode
Status
H000
CEC
000
CTS
00
CD
RLI
DPLA
WFA
XRNR
RRNR
A00
TEC
CEC
FCS
00
CTS
0
RFNE
CD
0
DPLA
000
B000
CEC
000
CTS
SYNC
RFNE
CD
0
DPLA
000
Bit1514131211109876543210
Mode
unused
H0000000000000000
A0000000000000000
B0000000000000000
PEB 20534
Detailed Register Description
Semiconductor Group 280 Data Sheet 09.98
TEC TIC Executing (async mode)
TIC=’0’ No TIC (transmit immediate character) is currently in
transmission. Access to register TICR is allowed to initiate
a TIC transmissio n.
TIC=’1’ A TIC command (write access to register TICR) is
accepted but not completely executed. No further write
acc ess to registe r TICR is a llowed until ’ TIC’ bi t is cl eared
by the DS CC4.
CEC Command Executing (all modes)
CEC=’0’ No command is currently in execution. The command
register CMDR can be written by CPU.
CEC=’1’ A command (writ ten previ ous ly to regist er CMDR) is
currently in execution. No further command can be written
to register CMDR by CPU.
Note: CEC will be active at most 2.5 receive or transmit clock cycles
(depending on whether a receiver or transmitter related command
is executed).
CEC will stay active if the SCC is in power-down mode or if no
serial clock, needed for command execution, is available.
FCS Flow Control Status (async mode)
If (in-band) flow control mechanism is enabled via bit ’FLON’ in register
CCR2 this bit indicates the current state of transmitter:
FCS=’0’ Transmitter is ready (always after transmitter reset
command or XON-character detected).
FCS=’1’ Transmitter is stopped (XOFF-character detected).
CTS Clear To Send Input Signal State (all modes)
CTS=’0’ CTS input signal is inactive (high level)
CTS=’1’ CTS input signal is active (low level)
PEB 20534
Detailed Register Description
Semiconductor Group 281 Data Sheet 09.98
SYNC Synchronization Status (bisync mode)
Indicates whether the receiver is in synchron ized state. After a ’HUNT’
command ’SYNC’ bit is cleared and the receiver starts searching for a
SYNC character. When found the ’SYNC’ status bit is set immediately,
an SCD-interrupt is generated (if enabled) and receive data is forwarded
to the receive FIFO.
SYNC=’0’ Synchronization is lost or not yet achieved.
(after reset, after new ’HUNT’ command has been issued
before SYNC character is found)
SYNC=’1’ The receiver is in synchronized state.
RFNE (SCC) Receive FIFO Not Empty (async/bisync modes)
This status bit is set if the SCC receive FIFO stores at least one valid
byte which might be not transferred to the central RFIFO because the
programmed threshold level is not exceeded and no frame end/ block
end condition is generated.
RFNE=’0’ SCC rece ive FIFO is empty.
RFNE=’1’ SCC receive FIFO is not empty.
CD CD (Carrier Detect) Input Signal State (all modes)
CD=’0’ CD input signal is inactive (low level)
CD=’1’ CD input signal is active (high level)
RLI Receive Line Inactive (hdlc mode)
This bit indicates that neither flags as interframe time fill nor data are
received via the receive line.
RLI=’0’ Receive line is active, no constant high level is detected.
RLI=’1’ Receive line is inactive, i.e. more than 7 consecutive ’1’
are detected on the line.
PEB 20534
Detailed Register Description
Semiconductor Group 282 Data Sheet 09.98
DPLA DPLL Asynchronous (all modes)
This bit is only valid if the receive clock is recovered by the DPLL and
FM0, FM1 or Manchester data encoding is selected. It is set when the
DPLL has lost synchronization. In this case reception is disabled
(receive abort condition) until synchronization has been regained. In
addition transmission is interrupted in all cases where transmit clock is
derived from the DPLL (clock mode 3a, 7a). Interruption of transmission
is performed the same way as on deactivation of the CTS signal.
DPLA=’0’ DPLL is synchronized.
DPLA=’1’ DPLL is asynchronous (re-synchronization process is
started automatically).
WFA Wait For Acknowledgement (hdlc mode)
This status bit is significant in Automode only. It indicates whether the
Automode state machine expects an acknowledging I- or S-Frame for a
previously sent I-Frame.
WFA=’0’ No acknowledge I/S-Frame is expected.
WFA=’1’ The Automode state machine is waiting for an
achnowledging S- or I-Frame.
XRNR Transmit RNR Status (hdlc mode)
This status bit is significant in Automode only. It indicates the receiver
status of the local station (SCC).
XRNR=’0’ The receiver is ready and will automatically answer poll-
frames with a S-Frame with ’receiver-ready’ indication.
XRNR=’1’ The receiver is NOT ready and will automatically answer
poll-frames with a S-Frame with a ’receiver-not-ready’
indication.
RRNR Received RNR Status (hdlc mode)
This status bit is significant in Automode only. It indicates the receiver
status of the remote station.
RRNR=’0’ The remote station receiver is ready.
RRNR=’1’ The remote receiver is NOT ready.
(A ’receiver-not-read y’ indication wa s received from the
remote station)
PEB 20534
Detailed Register Description
Semiconductor Group 283 Data Sheet 09.98
Table 67 CCR0: Channel Configuration Register 0
CPU Accessibility: read/write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 0108H0188H0208H0288H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
Power Mode - Serial Port
Configurationt Protocol Engine
Selection
H
PU
00000000
SC2
SC1
SC0
00
SM1
SM0
A
PU
00000000
SC2
SC1
SC0
00
SM1
SM0
B
PU
00000000
SC2
SC1
SC0
00
SM1
SM0
Bit1514131211109876543210
Mode
Interrupts DPLL Clock Source Clock Mode
H000
VIS
000
PSD
BCR
0
TOE
SSEL
HS
CM2
CM1
CM0
A000
VIS
000
PSD
BCR
0
TOE
SSEL
0
CM2
CM1
CM0
B000
VIS
000
PSD
BCR
0
TOE
SSEL
0
CM2
CM1
CM0
PEB 20534
Detailed Register Description
Semiconductor Group 284 Data Sheet 09.98
PU Power Up (all modes)
PU=’0’ The SCC is in ’power-down’ mode. The protocol engines
are switched off (standby) and no operation is performed.
This may be used to save power when SCC is not in use.
Note: The DMA Controller accessible part of the SCC
transmit FIFO requests for transmit data even in
’power-down’ mode .
PU=’1’ The SCC is in ’power-up’ mode.
SC(2..0) Serial Port Configuration (all modes)
This bit field selects the line coding of the serial port.
Note, that special operation modes and settings may require or exclude
operation in special line coding modes. Refer to the ’prerequisites’ in the
dedicated mode descriptions.
SC = ’000’ NRZ data encoding
SC = ’001’ Bus configuration, timing mode 1;
NRZ data encoding
SC = ’010’ NRZI data encoding
SC = ’011’ Bus configuration, timing mode 2;
NRZI data encoding
SC = ’100’ FM0 data encoding
SC = ’101’ FM1 data encoding
SC = ’110’ Mancheste r data encoding
SC = ’111’
Reserved
(do not use)
Note: If bus configuration mode is selected, only NRZ/NRZI data
encoding is supported.
PEB 20534
Detailed Register Description
Semiconductor Group 285 Data Sheet 09.98
SM(1..0) Serial Port Mode (all modes)
This bit field selects one of the three protocol engines.
Depending on the selected protocol engine the SCC related registers
change or special bit positions within the registers change their meaning.
SM = ’00’ HDLC/SDLC protocol engine
SM = ’01’
Reserved
(do not use)
SM = ’10’ B ISYNC proto col engine
SM = ’11’ ASYNC protocol engine
VIS Masked Interrupts Visible (all modes)
VIS=’0’ Masked interrupt status bits are visible on interrupt status
register (ISR) read accesses but cannot be cleared by
these read accesses.
VIS=’1’ Masked interrupt status bits are visible and automatically
cleared after interrupt status register (ISR) read access.
Note: Masked interrupts will not generate an interrupt vector to the
interrupt cont roll er.
PSD DPLL Phase Shift Disable (all modes)
This option is only applicable in the case of NRZ or NRZI line encoding
is selected.
PSD=’0’ Normal DPLL operation.
PSD=’1’ The phase shift function of the DPLL is disabled. The
windows for phase adjustment are extended.
PEB 20534
Detailed Register Description
Semiconductor Group 286 Data Sheet 09.98
BCR Bit Clock Rate (all modes)
This bit is only valid in asynchronous HDLC/PPP, ASYNC and BISYNC
protocol mode and only in clock modes not using the DPLL (0, 1, 3b, 7b).
It is also invalid in high speed operation clock mode 4.
BCR=’0’ Selects isochronous operation with bit clock rate 1. Data
bits are sampled once.
BCR=’1’ Selects standard asynchronous operation with bit clock
rate 16. Using 16 samples per bit, data bits are sampled 3
times around the nominal bit center. The resulting bit
value is determined by majority decision of the 3 samples.
For correct operation NRZ data encoding has to be
selected.
TOE Transmit Clock Out Enable (all modes)
For clock modes 0b, 2b, 3a, 3b, 6b, 7a and 7b, the internal transmit clock
can be monitored on pin TxCLK as an output signal. In clock mode 5, a
time slot control signal marking the active transmit time slot is output on
pin TxCLK.
Bit ’TOE’ is invalid for all other clock modes.
TOE=’0’ TxCLK pin is input.
TOE=’1’ TxCLK pin is switched to output function if applicable for
the selected clock mode.
SSEL Clock Source Select (all modes)
Distinguishes between the ’a’ and ’b’ option of clock modes 0, 2, 3, 6 and
7.
SSEL=’0’ Option ’a’ is selected.
SSEL=’1’ Option ’b’ is selected.
PEB 20534
Detailed Register Description
Semiconductor Group 287 Data Sheet 09.98
HS High Speed (PEB 20534H-52 only) (hdlc mode)
Configures the SCC for High Speed operation.
HS=’0’ Normal, non high speed operation.
HS=’1’ Switches the internal clocking and data paths for high
speed operation up to 52 MBit/s and enables the clock
gating mechanism.
Note: Bit ’HS’ should be set only in conjunction with clock mode 4
selection.
CM(2..0) Clock Mode (all modes)
This bit field selects one of main clock modes 0..7. The internal
CM = ’000’ clock mode 0
CM = ’001’ clock mode 1
CM = ’010’ clock mode 2
CM = ’011’ clock mode 3
CM = ’100’ clock mode 4 (high speed operation clock mode)
(PEB 20534H-52 only)
SC = ’101’ clock mode 5 (time slot oriented clocking mode)
SC = ’110’ clock mode 6
SC = ’111’
clock mod e 7
PEB 20534
Detailed Register Description
Semiconductor Group 288 Data Sheet 09.98
Table 68 CCR1: Channel Configuration Register 1
CPU Accessibility: read/write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 010CH018CH020CH028CH
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
Output Control
H00
SOC1
SOC0
0
DIV
ODS
00
ICD
TCLKO
RTS
FCTS
CAS
0
A00000
DIV
ODS
00
ICD
0
RTS
FRTS
FCTS
CAS
0
B00000
DIV
ODS
00
ICD
0
RTS
FRTS
FCTS
CAS
0
Bit1514131211109876543210
Mode
Transmitter/Receiver Configuration
H
MDS1
MDS0
ADM
NRM
PPM1
PPM0
MCS
TLP
SFLG
00000
CRL
C32
A0000000
TLP
TOIE
TOLEN
B0000
SLEN
BISNC
0
TLP
00000000
PEB 20534
Detailed Register Description
Semiconductor Group 289 Data Sheet 09.98
SOC(1..0) Serial Output Control (hdlc mode)
This bit field selects the RTS signal output function.
(This bit field is only valid in bus configuration modes selected via bit field
SC(2:0) in register CCR0).
SOC = ’0X’ RTS ouput signal is active during transmission of a frame
(active low).
SOC = ’10’ RTS ouput signal is always inactive (high).
SOC = ’11’ RTS ouput signal is active during reception of a frame
(active low).
DIV Data Inversion (all modes)
This bit is only valid if NRZ data encoding is selected via bit field SC(2:0)
in register CCR0.
DIV=’0’ No Data Inversion.
DIV=’1’ Data is transmitted/received inverted (on a per bit basis).
In HDLC and HDLC Synchronous PPP modes the
continuous ’1’ idle sequence is NOT inverted.
Interframe time fill flag transmission is inverted.
Note: It is recommended not to use DIV=’1’ in combination with high
speed operation, i.e. clock mode 4 and CCR0.HS=’1’.
ODS Output Driver Select (all modes)
The transmit data output pin TxD can be configured as push/pull or open
drain output chracte ristic.
ODS=’0’ TxD pin is open drain output.
ODS=’1’ TxD pin is push/pull output.
ICD Invert Carrier D etect Pin Polarity (all modes)
ICD=’0’ Carrier Detect (CD) input pin is active high.
ICD=’1’ Carrier Detect (CD) input pin is active low.
PEB 20534
Detailed Register Description
Semiconductor Group 290 Data Sheet 09.98
TCLKO Transmit Clock Output (hdlc mode)
This bit is only valid in high speed operation mode, i.e. clock mode 4 and
CCR0.HS = ’1’.
In high speed mode the internal transmit clock is supplied to pin RTS
as
an output.
TCLKO=’0’ Pin RTS is an unused input signal in high speed operation
mode.
It should be connected to a known level via a pull-up/down
resistor.
TCLKO=’1’ The internal transmit clock is supplied as an output signal
to pin RTS.
RTS Request To Send pin control (all modes)
The request to send pin RTS can be controlled by the DSCC4 as an
ou tput autonomously or via setting/clearing bit ’RTS’ .
This bit is not valid in high speed operation mode.
RTS=’0’ RTS (output) pin is controlled by the DSCC4
autonomously.
• In HDLC modes RTS is activated during transmission.
• In ASYNC and BISYNC mode, the function depends on
bit ’FRTS’ in register CCR1.
• In bus configuration mode the functionality depends on
bit field ’SOC’ setting.
RTS=’1’ • In HDLC modes RTS (output) is activated (low) until
this bit is cleared by software again.
• In ASYNC and BISYNC mode, the function depends on
bit ’FRTS’ in register CCR1.
PEB 20534
Detailed Register Description
Semiconductor Group 291 Data Sheet 09.98
FRTS Flow Control (using signal RTS)(async/bisync modes)
In ASYNC and BISYNC modes, bit ’FRTS’ together with bit ’RTS’
determine the function of signal RTS:
FRTS, RTS
0, 0 Pin RTS is controlled by the DSCC4 autonomously.
RTS is activated (low) as soon as transmit data is
available within the SCC transmit FIFO.
1, 0 Pin RTS is controlled by the DSCC4 autonomously
supporting bi-directional data flow control.
RTS is activated (low) if the shadow part of the SCC
receive FIFO is empty and de-activated (high) when the
SCC receive FIFO fill level reaches its receive FIFO
threshold.
0, 1 Forces pin RTS to active state (low).
1, 1 Forces pin RTS to inactive state (high).
FCTS Flow Control (using signal CTS)(all modes)
This bit controls the function of pin CTS.
This bit is not valid in high speed operation mode.
FCTS = ’0’ The transmitter is stopped if CTS input signal is inactive
(high) and enabled if active (low).
Note: In character oriented protocol modes (ASYNC,
BISYNC, asynchronous PPP), the current byte is
completely sent even if CTS becomes inactive
during transmission.
FCTS = ’1’ The transmitter is enabled disregarding CTS input signal.
Note: In ASYNC mode the transmitter is additionally controlled by in-
band flow control mechanism (if enabled).
PEB 20534
Detailed Register Description
Semiconductor Group 292 Data Sheet 09.98
CAS Carrier Detect Auto Start (all modes)
CAS = ’0’ The CD pin is used as general input.
In clock mode 1 and 4, clock mode specific control signals
must be provided at this pin (receive strobe, receive
gating).
In clock mode 5, this input signal is unused.
A pull-up/down resistor is recommended if unused.
CAS = ’1’ The CD pin enables/disables the receiver for data
reception. (Polarity of CD pin can be configured via bit
’ICD’.)
Note: In ASYNC mode the transmitter is additionally controlled by in-
band flow control mechanism (if enabled).
MDS(1..0) Mode Select (hdlc modes)
This bit field selects the HDLC protocol sub-mode including the
’extended transparent mode’. ’MDS’ selection is invalid if any PPP mode
is enabled via bit field ’PPPM’.
MDS = ’00’ Automode.
MDS = ’01’ Non-Automode.
MDS = ’10’ Address Mode 0/1.
(Option ’0’ or ’1’ is selected via bit ’ADM’.)
MDS = ’11’ Extended transparent mode (bit transparent
transmission/reception).
ADM Address Mode Select (hdlc mode)
The meaning of this bit depends on the selected protocol sub-mode:
Automode, Non-Automode:
Determines the address field length of a HDLC frame.
ADM = ’0’ 8-bit address field.
ADM = ’1’ 16-bit address field.
Address mode 0/1:
Determines whether address mode 0 or 1 is selected.
ADM = ’0’ Address Mode 0 (no address recognition).
ADM = ’1’ Address Mode 1 (high byte address recognition).
PEB 20534
Detailed Register Description
Semiconductor Group 293 Data Sheet 09.98
NRM Normal Response Mode (hdlc mode)
This bit is valid in HDLC Automode operation only and determines the
function of the Automode LAP-Controller:
NRM = ’0’ Full-duplex LAP-B / LAP-D operation.
NRM = ’1’ Half-duplex normal response mode (NRM) operation.
PPPM(1..0) PPP Mode Select (hdlc mode)
This bit field enables and selects the HDLC PPP protocol modes:
PPPM = ’00’ No PPP protocol operation. The HDLC sub-mode is
determined by bit field ’MDS’.
PPPM = ’01’ Octet synchronous PPP protocol operation.
PPPM = ’10’ Asynchronous PPP protocol operation.
Note: Bit ’BCR’ in register CCR0 must be set to ensure
proper asynchronous reception.
PPPM = ’11’ Bit synchronous PPP protocol operation.
SLEN SYNC Character Length (bisync mode)
This bit selects the SYNC character length in BISYNC/MONOSYNC
operation mode:
SLEN = ’0’ 6 bit (MONOSYNC), 12 bit (BISYNC).
SLEN = ’1’ 8 bit (MONOSYNC), 16 bit (BISYNC).
BISNC Enable BISYNC Mode (bisync mode)
This bit is selects BISYNC/MONOSYNC operation:
BISNC = ’0’ 6 bit (MONOSYNC), 12 bit (BISYNC).
BISNC = ’1’ 8 bit (MONOSYNC), 16 bit (BISYNC).
MCS Modulo Count Select (hdlc mode)
This bit is valid in HDLC Automode operation only and determines the
control field format:
MCS = ’0’ Basic operation, one byte control field (modulo 8 counter
operation).
MCS = ’1’ Extended operation, two bytes control field (modulo 128
co unter operati on).
PEB 20534
Detailed Register Description
Semiconductor Group 294 Data Sheet 09.98
TLP Test Loop (all modes)
This bit controls the internal test loop between transmit and receive data
signals. The test loop is closed at the far end of serial transmit and
receive line just before the respective TxD and RxD pins:
TLP = ’0’ Test loop disabled.
TLP = ’1’ Test loop enabled.
The software is responsible to select a clock mode which
allows correct reception of transmit data depending on the
external clock supply. Transmit data is also sent out via
pin TxD, but receive input pin RxD is internally
disconnected during test loop operation.
Note: It is recommended not to use the test loop in high
speed opera tion mode (clock mode 4). A non high
speed clock mode s hould be sel ected for te st loop
operation.
SFLAG Shared Flags Transmission (hdlc mode)
This bit enables ’shared flag transmission’ in HDLC protocol mode. If
another transmit frame begin is stored in the SCC transmit FIFO, the
closing flag of the preceding frame becomes the opening flag of the next
frame (shared flags):
SFLAG = ’0’ Shared flag transmission disabled.
SFLAG = ’1’ Shared flag transmission enabled.
Note: The receiver always supports shared flags and shared zeros of
consecutive flags.
TOIE Time Out Indication Enable (async mode)
If this bit is selected in ASYNC mode, any time out event will
automatically generate a ’RFRD’ command thus inserting a ’frame end/
block end’ indication into the receive FIFO. This causes the SCC receive
FIFO to forward received data to the DMA Controller even if the receive
FIFO threshold is not exceeded. The DMA controller is forced to finish
the current receive descriptor with an ’frame end / block - end’ indication:
TOIE = ’0’ Automatic Time Out processing disabled.
TOIE = ’1’ Automatic Time Out processing enabled.
PEB 20534
Detailed Register Description
Semiconductor Group 295 Data Sheet 09.98
TOLEN
(6..0) Time Out Length (async mode)
This bit field determines the time out period. If there is no receive line
activity for the configured period of time, a time out indication is
generated if enabled via bit ’TOIE’.
The period of time is programmable in multiples of character frame
length (CFL) time equivalents including start, parity and stop bits:
TOLEN T = ((TOLEN + 1) * 4) * CFL
CRL CRC Reset Value (hdlc mode)
This bit defines the initial value of the internal transmit/receive CRC
generators:
CRL=’0’ Initial value is 0xFFFFH (16 bit CRC), 0xFFFFFFFFH
(32 bit CRC).
Default value for most HDLC/SDLC applications.
CRL=’1’ Initial value is 0x0000H (16 bit CRC), 0x00000000H
(32 bit CRC).
C32 CRC 32 Select (hdlc mode)
This bit enables 32-bit CRC operation for transmit and receive.
C32=’0’ 16-bit CRC-CCITT generation/checking.
C32=’1’ 32-bit CRC generation/checking.
Note: The internal ’valid frame’ criteria is updated depending on the
selected number of CRC-bytes.
PEB 20534
Detailed Register Description
Semiconductor Group 296 Data Sheet 09.98
Table 69 CCR2: Channel Configuration Register 2
CPU Accessibility: read/write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 0110H0190H0210H0290H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
Receiver Configuration
H0000
RAC
0000
DRCRC
RCRC
RADD
RFDF
RFTH(2:0)
A00
CHL1
CHL0
RAC
DXS
XBRK
STOP
PAR1
PAR0
PARE
DPS
RFDF
RFTH(2:0)
B00
CHL1
CHL0
RAC
00
SLOAD
PAR1
PAR0
PARE
DPS
RFDF
RFTH(2:0)
Bit1514131211109876543210
Mode
Transmitter/Receiver Configuration
HPRE(7:0)
EPT
0
PRE1
PRE0
ITF
SXIF
OIN
XRCR
A000000000000000
FLON
BPRE(7:0)
EPT
0
PRE1
PRE0
ITF
CRL
CAPP
CRCM
PEB 20534
Detailed Register Description
Semiconductor Group 297 Data Sheet 09.98
CHL(1..0) Character Length (async/bisync modes)
This bit field selects the number of data bits within a character:
CHL = ’00’ 8-bit data.
CHL = ’01’ 7-bit data.
CHL = ’10’ 6-bit data.
CHL = ’11’ 5-bit data.
RAC Receiver active (all modes)
Switches the receiver between operational/inoperational states:
RAC=’0’ Receiver inactive, receive line is ignored.
RAC=’1’ Receiver ac tive.
DXS Disable Storage of XON/XOFF Characters (async mode)
In ASYNC mode, XON/XOFF characters might be filtered out or stored
to the SCC receive FIFO:
DXS=’0’ All received characters including XON/XOFF characters
are stored in the receive FIFO.
DXS=’1’ XON/XOFF characters are filtered out and not stored in
the receive FIFO.
XBRK Transmit Break (async mode)
XBRK=’0’ Norm al transmit operation.
XBRK=’1’ Forces the TxD pin to ’low’ level immediately (break
condition), regardless of any character being currently
transmitted. This command is executed immediately with
the next rising edge of the transmit clock and further
transmission is disabled. The currently sent character is
lost.
Data stored in the SCC transmit FIFO will be sent as soon
as the break condition is cleared (XBRK=’0’). A transmit
reset command (bit ’XRES’ in register CMDR) does NOT
clear the break condition automatically.
PEB 20534
Detailed Register Description
Semiconductor Group 298 Data Sheet 09.98
STOP Stop Bit number (async mode)
This bit selects the number of stop bits per ASYNC character:
STOP=’0’ 1 stop bit per character.
STOP=’1’ 2 stop bits per character.
SLOAD Enable SYNC Character Load (bisync mode)
In BISYNC mode, SYNC characters might be filtered out or stored to the
SCC receive FIFO:
SLOAD=’0’ All received characters including SYNC characters are
stored in the receive FIFO.
SLOAD=’1’ SYNC chara cters are filtered out and not stored in the
receive FIFO.
PAR(1..0) Parity Format (async/bisync modes)
This bit field selects the parity generation/checking mode:
PAR = ’00’ SPACE (’0’), a constant ’0’ is inserted as parity bit.
PA R = ’01’ Odd pari ty.
PAR = ’10’ Even parity.
PAR = ’11’ MARK (’1’), a constant ’1’ is inserted as parity bit.
The received parity bit is stored in the SCC receive FIFO depending on
the selected character format:
• as lead ing bit immediately preceding the data bits if character length
is 5, 6 or 7 bits and bit ’DPS’ in register CCR2 is cleared (’0’).
• as LSB of the status byte belonging to the character if character length
is 8 bi ts and the corresponding receive FIFO da ta format is s elected
(RFDF = ’1’).
A parity error is indicated in the MSB of the status byte belonging to each
character if enabled. In addition, a parity error interrupt can be
generated.
PARE Parity Enable (async/bisync modes)
PARE=’0’ Parity generation/checking is disabled.
PARE=’1’ Parity generation/checking is enabled.
PEB 20534
Detailed Register Description
Semiconductor Group 299 Data Sheet 09.98
DPS Data Parity Storage (async/bisync modes)
Only valid if parity generation/checking is enabled via bit ’PARE’:
DPS= ’0’ The parit y bit is stored.
DPS=’1’ The parity bit is not stored in the data byte containing
ch aracter d ata.
The parity bit is always stored in the status byte.
DRCRC Disable Receive CRC Che cking (hdlc mode)
DPS=’0’ The receiver expects a 16 or 32 bit CRC within a HDLC
frame. CRC processing depends on the setting of bit
’RCRC’.
Frames shorter than expected are marked ’invalid’ or are
discarded (refer to RSTA description).
DPS=’1’ The receiver does not expect any CRC within a HDLC
frame. The criteria for ’valid frame’ indication is updated
accordingly (refer to RSTA description).
Bit ’RCRC’ is ignored.
RCRC Receive CRC Checking Mode (hdlc mode)
This bit is only valid in Non-Automode and Address Mode 0:
RCRC=’0’ The received checksum is evaluated, but not forwarded to
the receive FIFO.
RCRC=’1’ The received checksum (2 or 4 bytes) is evaluated and
forwarded to the receive FIFO as data. In Non-Automode
the criteria for ’valid frame’ is updated (refer to RSTA
description).
RADD Receive Address Pushed to RFIFO (hdlc mode)
This bit is only valid if a HDLC sub-mode with address field support is
selected (Automode, Non-Automode, Address Mode 1):
RADD=’0’ The received HDLC address field (either 8 or 16 bit
depending on bit ’ADM’) is evaluated, but NOT forwarded
to the receive FIFO.
RADD=’1’ The received HDLC address field (either 8 or 16 bit
depending on bit ’ADM’) is evaluated and forwarded to the
receive FIFO.
PEB 20534
Detailed Register Description
Semiconductor Group 300 Data Sheet 09.98
RFDF Receive FIFO Data Format (all modes)
HDLC modes:
This bit is only valid if the minimum receive FIFO threshold is selected
(bit field RFTH(2..0) = ’000’):
RFDF=’0’ A minimum of one byte is stored in the receive FIFO
before forwarded to the central RFIFO.
RFDF=’1’ A minimum of two bytes is stored in the receive FIFO
before forwarded to the central RFIFO.
ASYNC/BISYNC modes:
In ASYNC/BISYNC modes, the character format is determined as
follows:
Char5
P
data byte:
0457
Char6
P
0
657
Char7
P
067
Char8
07
RFDF=’0’
(no parity bit stored)
Char5
P
data byte (d):
0457
Char6
P
0
657
Char7
P
067
Char8
07
RFDF=’1’
(no parity bit stored)
status byte (s):
PE
067
FE P
PE
067
FE P
PE
067
FE P
PE
0
6
7
FE P
P
: Parity bit stored in data byte, can be disabled via bit ’DPS’
PE: Parity Error
FE: Frame Error
P: Parity bit stored in second data byte (= status byte)
PEB 20534
Detailed Register Description
Semiconductor Group 301 Data Sheet 09.98
RFTH(2..0) Receive FIFO Threshold (all modes)
This bit field defines the level up to which the SCC receive FIFO is filled
with valid data before transfer to the central RFIFO is requested.
(Transfer always starts immediately in case of a ’frame end / block end’
condition.)
The meaning depends on the selected protocol engine:
HDLC Modes:
RFTH(2..0) Threshold level in number of data (d) and status (s) bytes
depending on bit ’RFDF’:
RFDF = ’0’ RFDF = ’1’
’000’ 1d 2d
’001’ 4d don’t care
’010’ 16d don’t care
’011’ 24d don’t care
’100’ 32d don’t care
’101’ 60d don’t care
’110’ 1d 2d
’111’ 1d 2d
ASYNC/BISYNC Modes:
RFTH(2..0) Threshold level in number of data (d) and status (s) bytes
depending on bit ’RFDF’:
RFDF = ’0’ RFDF = ’1’
’000’ 1d 1d+1s
’001’ 4d 2d+2s
’010’ 16d 8d+8s
’011’ 24d 12d+12s
’100’ 32d 16d+16s
’101’ 60d 30d+30s
’110’ 1d 1d+1s
’111’ 1d 1d+1s
PEB 20534
Detailed Register Description
Semiconductor Group 302 Data Sheet 09.98
PRE Preamble (hdlc/bisync modes)
This bit field determines the preamble pattern which is send out during
preamble transmission.
Note: In HDLC-mode, zero-bit insertion is disabled during preamble
transmission.
EPT Enable Preamble Transmission (hdlc/bisync modes)
This bit enables preamble transmission. The preamble is started after
interframe timefill (ITF) transmission is stopped because a new frame is
ready to be transmitted. The preamble pattern consists of 8 bits defined
in bit field ’PRE(7..0)’ which is send repetitively. The number of
repetitions is determined by bit field ’PRE(1..0)’:
EPT=’0’ Preamble transmission is disabled.
EPT=’1’ Preamble transmission is enabled.
Note: Preamble operation does NOT influence HDLC shared flag
transmission if enabled.
PRE(1..0) Number of Preamble Repetitions (hdlc/bisync modes)
This bit field determines the number of preambles transmitted:
PRE = ’00’ 1 preamble.
PRE = ’01’ 2 preambles.
PRE = ’10’ 4 preambles.
PRE = ’11’ 8 preambles.
ITF Interframe Time Fill (hdlc/bisync modes)
This bit selects the idle state of the transmit pin TxD:
ITF=’0’ Continuous logical ’1’ is send during idle phase.
ITF=’1’ Continuous flag sequences are send (’01111110’ flag
pattern).
Note: It is recommended to clear bit ’ITF’ in bus configuration modes, i.e.
conti nuo us o nes ar e sen d as idle sequence an d data enc odin g is
NRZ.
PEB 20534
Detailed Register Description
Semiconductor Group 303 Data Sheet 09.98
SXIF Selects Transmission of I-Frames (hdlc mode)
This bit is valid in HDLC-Automode only:
SXIF=’0’ Causes the SCC in Automode to transmit a HDLC frame
as an I-frame. Additionally to the opening flag sequence,
the address and control field of the frame is automatically
added by the SCC. An all-sent (ALLS) interrupt is
generated after receiving the corresponding
acknowledgement
SXIF=’1’ Continuous flag sequences are send (’01111110’ flag
pattern).
CRL CRC Reset Value (bisync mode)
This bit defines the initial value of the internal transmit/receive CRC
generators:
CRL=’0’ Initial value is 0xFFFFH (16 bit CRC), 0xFFFFFFFFH
(32 bit CRC).
CRL=’1’ Initial value is 0x0000H (16 bit CRC), 0x00000000H
(32 bit CRC).
OIN One Insertion (hdlc mode)
In HDLC mode a one-insertion mechanism similar to the zero-insertion
can be activated:
OIN=’0’ The ’1’ insertion mechanism is disabled.
OIN=’1’ In transmit direction a logical ’1’ is inserted to the serial
data stream after 7 consecutive zeros.
In receive direction a ’1’ is deleted from the receive data
stream after receiving 7 consecutive zeros.
This enables clock information to be recovered from the
receive data stream by means of a DPLL, even in the case
of NRZ data encoding, because a transition at bit cell
boundary occurs at least every 7 bits.
PEB 20534
Detailed Register Description
Semiconductor Group 304 Data Sheet 09.98
CAPP CRC Append (bisync mode)
In BISYNC mode the CRC generator can be activated:
CAPP=’0’ No CRC generation/checking is active in BISYNC mode.
CAPP=’1’ The CRC generator is activated:
1. The CRC generator is initialized every time the
transmission of a new ’frame’ starts. The CRC
initialization value can be selected via bit ’CRL’ in
register CCR2 (for BISYNC operat ion).
2. The CRC is automatically appended to the last
transmitted data of a ’frame’.
CRCM CRC Mode Select (bisync mode)
In BISYNC mode the CRC generator can be configured for two different
generator polynom s:
CRCM=’0’ CRC-16:
The polynominal is x16+x15+x2+1.
CRCM=’1’ CRC-CCITT:
The polynominal is x16+x12+x5+1.
XCRC Transmit CRC Checking Mode (hdlc mode)
XCRC=’0’ The transmit checksum (2 or 4 bytes) is generated and
appended to the transmit data automatically.
XCRC=’1’ The transmit checksum is not generated automatically.
The checksum is expected to be provided by software as
the last 2 or 4 bytes in the transmit data buffer.
Note: The transmitter does NOT check whether the
number of data bytes makes sense, i.e. a valid
frame length.
PEB 20534
Detailed Register Description
Semiconductor Group 305 Data Sheet 09.98
FLON Flow Control Enable (async mode)
In ASYNC mode, in-band flow control is supported:
FLON=’0’ No automatic in-band flow-control is performed. However
recognition of a flow control character (XON/XOFF)
causes always a maskable interrupt event.
FLON=’1’ Automatic in-band flow-control is performed.
Reception of a XOFF character (defined in register XNXF)
turns off the transmitter after the currently transmitted
character has been shifted out completely (XOFF state).
Reception of a XON character (defined in register XNXF)
resumes the transmitter from XOFF into XON state ready
to send available trans mit data bytes.
The current flow control state is indicated via bit ’FCS’ in
register Star.
Any transmitter reset switches the flow-control logic to
XON state.
PEB 20534
Detailed Register Description
Semiconductor Group 306 Data Sheet 09.98
Table 70 ACCM: PPP ASYNC Control Character Map
CPU Accessibility: read/write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 0114H0194H0214H0294H
typical usage: written by CPU, valid in HDLC PPP protocol mode only
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
ASYNC Character Control Map (high)
H1F 1E 1D 1C 1B 1A 19 18 17 16 15 14 13 12 11 10
A0000000000000000
B0000000000000000
Bit1514131211109876543210
Mode
ASYNC Character Control Map (low)
H0F 0E 0D 0C 0B 0A 09 08 07 06 05 04 03 02 01 00
A0000000000000000
B0000000000000000
PEB 20534
Detailed Register Description
Semiconductor Group 307 Data Sheet 09.98
ACCM ASYNC Character Control Map (hdlc mode)
This bit field is valid in HDLC asynchronous and octet-synchronous PPP
mode only:
Each bit selects the corresponding character (indicated as hex value
1FH..00H in the register description table) as control character which has
to be mapped into the transmit data stream.
PEB 20534
Detailed Register Description
Semiconductor Group 308 Data Sheet 09.98
Table 71 UDAC: User Defined PPP ASYNC Control Character Map
CPU Accessibility: read/write
Reset Value: 7E7E 7E7EH
SCC0 SCC1 SCC2 SCC3
Offset Address: 0118H0198H0218H0298H
typical usage: written by CPU, valid in HDLC PPP protocol mode only
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
ASYNC Character 3 ASYNC Character 2
HAC3 AC2
A0000000000000000
B0000000000000000
Bit1514131211109876543210
Mode
ASYNC Character 1 ASYNC Character 0
HAC1 AC0
A0000000000000000
B0000000000000000
PEB 20534
Detailed Register Description
Semiconductor Group 309 Data Sheet 09.98
AC3..0 User Defined ASYNC Character Control Map (hdlc mode)
This bit field is valid in HDLC asynchronous and octet-synchronous PPP
mode only:
These bit fields define user determined characters as control characters
which have to be mapped into the transmit data stream.
In register ACCM only characters 00H..1FH can be selected as control
characters. Register UDAC allows to specify any four characters in the
range 00H..FFH .
The default value is a 7EH flag which must be always mapped. Thus no
additional character is mapped if 7EH ’s are programed to bit fields
AC3...0 (reset value).
(7EH is mapped automatically, even if not defined via a AC bit field.)
PEB 20534
Detailed Register Description
Semiconductor Group 310 Data Sheet 09.98
Table 72 TTSA: Transmit Time Slot Assignment Register
CPU Accessibility: read/write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 011CH019CH021CH029CH
typical usage: written by CPU, valid in HDLC clock mode 5 only
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
Tx Time Slot Number Tx Clock Shift
H0 TTSN(6:0) 00000 TCS(2:0)
A0 TTSN(6:0) 00000 TCS(2:0)
B0 TTSN(6:0) 00000 TCS(2:0)
Bit1514131211109876543210
Mode
Transmit Time Slot Control Transmit Channel Capacity
H
TEPCM
000000 TCC(8:0)
A
TEPCM
000000 TCC(8:0)
B
TEPCM
000000 TCC(8:0)
PEB 20534
Detailed Register Description
Semiconductor Group 311 Data Sheet 09.98
TTSN(6:0) Transmit Time Slot Number (all modes)
This bit field selects one of 128 possible time-slots in which data is
allowed to be transmitted. The number of bits per time-slot can be
programmed via bit field ’TCC’.
TCS(2:0) Transmit Clock Shift (all modes)
This bit field determines the transmit clock shift.
TEPCM Enable PCM Mask Transmit (all modes)
This bit selects the additional Transmit PCM Mask (refer to register
PCMMTX):
TEPCM=’0’ Standard time-slot configuration.
TEPCM=’1’ The time-slot width is constant 8 bit, bit fields ’TTSN’ and
’TCS’ determine the offset of the PCM mask and ’TCC’ is
ignored. Each time-slot selected via register PCMMTX is
an active transmit timeslot.
TCC(8:0) Transmit Channel Capacity (all modes)
This bit field determines the transmit time-slot width in standard time-slot
configuration (bit TEPCM=’0’):
Number of bits = TCC + 1, (1...512 bits/time-slot)
PEB 20534
Detailed Register Description
Semiconductor Group 312 Data Sheet 09.98
Table 73 RTSA: Receive Time Slot Assignment Register
CPU Accessibility: read/write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 0120H01A0H0220H02A0H
typical usage: written by CPU, valid in HDLC clock mode 5 only
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
Rx Time Slot Number Rx Clock Shift
H0 RTSN(6:0) 00000 RCS(2:0)
A0 RTSN(6:0) 00000 RCS(2:0)
B0 RTSN(6:0) 00000 RCS(2:0)
Bit1514131211109876543210
Mode
Receive Time Slot Control Receive Channel Capacity
H
REPCM
000000 RCC(8:0)
A
REPCM
000000 RCC(8:0)
B
REPCM
000000 RCC(8:0)
PEB 20534
Detailed Register Description
Semiconductor Group 313 Data Sheet 09.98
RTSN(6:0) Receive Time Slot Number (all modes)
This bit field selects one of 128 possible time-slots in which data is
received. The number of bits per time-slot can be programmed via bit
field ’RCC ’.
RCS(2:0) Receive Clock Shift (all modes)
This bit field determines the receive clock shift.
REPCM Enable PCM Mask Receive (all modes)
This bit selects the additional Receive PCM Mask (refer to register
PCMMRX):
REPCM=’0’ Standard time-slot configuration.
REPCM=’1’ The time-slot width is constant 8 bit, bit fields ’RTSN’ and
’RCS’ determine the offset of the PCM mask and ’RCC’ is
ignored. Each time-slot selected via register PCMMRX is
an active receive timeslot.
RCC(8:0) Receive Channel Capacity (all modes)
This bit field determines the receive time-slot width in standard time-slot
configuration (bit REPCM=’0’):
Number of bits = RCC + 1, (1...512 bits/time-slot)
PEB 20534
Detailed Register Description
Semiconductor Group 314 Data Sheet 09.98
Table 74 PCMMTX: PCM Mask for Transmit Direction
CPU Accessibility: read/write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 0124H01A4H0224H02A4H
typical usage: written by CPU, valid in HDLC clock mode 5 only
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
PCM Mask for Transmit Direction (high)
HT31 T30 T29 T28 T27 T26 T25 T24 T23 T22 T21 T20 T19 T18 T17 T16
AT31 T30 T29 T28 T27 T26 T25 T24 T23 T22 T21 T20 T19 T18 T17 T16
BT31 T30 T29 T28 T27 T26 T25 T24 T23 T22 T21 T20 T19 T18 T17 T16
Bit1514131211109876543210
Mode
PCM Mask for Transmit Direction (low)
HT15 T14 T13 T12 T11 T10 T09 T08 T07 T06 T05 T04 T03 T02 T01 T00
AT15 T14 T13 T12 T11 T10 T09 T08 T07 T06 T05 T04 T03 T02 T01 T00
BT15 T14 T13 T12 T11 T10 T09 T08 T07 T06 T05 T04 T03 T02 T01 T00
PEB 20534
Detailed Register Description
Semiconductor Group 315 Data Sheet 09.98
PCMMTX PCM Mask for Tran smit Direc tion (hdlc mode)
This bit field is valid in HDLC clock mode 5 only and the PCM mask must
be enabled via bit ’TEPCM’ in register TTSA.
Each bit selects one of 32 (8-bit) transmit time-slots. The offset of time-
slot zero to the frame sync pulse can be programmed via register TTSA
bit field ’TTSN’.
PEB 20534
Detailed Register Description
Semiconductor Group 316 Data Sheet 09.98
Table 75 PCMMRX: PCM Mask for Receive Direction
CPU Accessibility: read/write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 0128H01A8H0228H02A8H
typical usage: written by CPU, valid in HDLC clock mode 5 only
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
PCM Mask for Receive Direction (high)
HT31 T30 T29 T28 T27 T26 T25 T24 T23 T22 T21 T20 T19 T18 T17 T16
AT31 T30 T29 T28 T27 T26 T25 T24 T23 T22 T21 T20 T19 T18 T17 T16
BT31 T30 T29 T28 T27 T26 T25 T24 T23 T22 T21 T20 T19 T18 T17 T16
Bit1514131211109876543210
Mode
PCM Mask for Receive Direction (low)
HT15 T14 T13 T12 T11 T10 T09 T08 T07 T06 T05 T04 T03 T02 T01 T00
AT15 T14 T13 T12 T11 T10 T09 T08 T07 T06 T05 T04 T03 T02 T01 T00
BT15 T14 T13 T12 T11 T10 T09 T08 T07 T06 T05 T04 T03 T02 T01 T00
PEB 20534
Detailed Register Description
Semiconductor Group 317 Data Sheet 09.98
PCMMRX PCM Mask for Re ceiv e Direction (hdlc mode)
This bit field is valid in HDLC clock mode 5 only and the PCM mask must
be enabled via bit ’REPCM’ in register RTSA.
Each bit selects one of 32 (8-bit) receive time-slots. The offset of time-
slot zero to the frame sync pulse can be programmed via register RTSA
bit field ’RTSN’.
PEB 20534
Detailed Register Description
Semiconductor Group 318 Data Sheet 09.98
Table 76 BRR: Baud Rate Register
CPU Accessibility: read/write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 012CH01ACH022CH02ACH
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
(unused)
H0000000000000000
A0000000000000000
B0000000000000000
Bit1514131211109876543210
Mode
Baud Rate Generator Factors
H0000 BRM(3:0) 00 BRN(5:0)
A0000 BRM(3:0) 00 BRN(5:0)
B0000 BRM(3:0) 00 BRN(5:0)
PEB 20534
Detailed Register Description
Semiconductor Group 319 Data Sheet 09.98
BRM(3:0) Baud Rate Factor ’M’ (all modes)
BRN(5:0 ) Baud Rate Factor ’N’ (all modes)
This bit fields determine the division factor of the internal baud rate
generator. The baud rate generator input clock and the usage of baud
rate generator output depends on the selected clock mode.
The division factor k is calculated by:
with M=0..15 and N=0..63.
kN1+()2M
×=
fBRG fin k⁄=
PEB 20534
Detailed Register Description
Semiconductor Group 320 Data Sheet 09.98
Table 77 TIMR: Timer Register
CPU Accessibility: read/write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 0130H01B0H0230H02B0H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
Timer Configuration Timer Value
HSRC 0 0 TMD 0 CNT(3:0) TVALUE(23:16)
ASRC 0 0 TMD 0 CNT(3:0) TVALUE(23:16)
BSRC 0 0 TMD 0 CNT(3:0) TVALUE(23:16)
Bit1514131211109876543210
Mode
Timer Value
HTVALUE(15:0)
ATVALUE(15:0)
BTVALUE(15:0)
PEB 20534
Detailed Register Description
Semiconductor Group 321 Data Sheet 09.98
SRC Clock Source (all modes)
This bit selects the clock source of the internal timer:
SRC = ’0’ The timer is clocked by the effective transmit clock.
SRC = ’1’ The timer is clocked by the frame-sync synchronization
signal supplied via the CD pin in clock mode 5.
(Valid in clock mode 5 only.)
TMD Timer Mode (all modes)
This bit selects between internal and external timer operation mode:
TMD=’0’ External timer mode:
The timer is controlled by the CPU via access to registers
CMDR and TIMR.
The timer can be started any time by setting bit ’STI’ in
register CMDR. The timer stops automatically after it has
expired and generates a timer interrupt. The timer can be
stopped any time by writing zero to the value bit field.
TMD=’1’ Internal timer mode: (valid in HDLC Automode only)
The timer is used by the DSCC4 for protocol specific time-
out and retry transactions.
PEB 20534
Detailed Register Description
Semiconductor Group 322 Data Sheet 09.98
CNT(2..0) Counter (all modes)
The meaning of this bit field depends on the timer mode.
In ’internal timer mode’ (HDLC Automode and bit TMD=’1’):
• Retry Counter (in HDLC protocol known as ’N2’):
Bit fiel d ’CNT’ indic ates the num ber of S-Command f rames (with p oll
bit set) which are transmitted autonomously by the DSCC4 after every
expi ration of the time out period ’t’ (d etermined b y ’VALUE’), in c ase
an I-Frame gets not acknowledged by the opposite station. The
maximum value is 6 S-command frames. If ’CNT’ is set to ’7’, the
number of S-commands is unlimited in case of no acknowledgement.
In external timer mode (bit TMD=’0’):
• Restart Counter :
Bit field ’CNT’ indicates the number of automatic restarts which are
performed by the DSCC4 after every expiration of the time out period
’t’, in case the timer is not stopped by writing a zero to bit field
’TVALUE’.The maximum value is 6 restarts. If ’CNT’ is set to ’7’, a
timer interrupt is generated periodically with time period ’t’ determined
by bit field ’TVALUE’.
TVALUE
(23:0) Timer Expiration Value (all modes)
This bit field determines the timer expiration period ’t’:
(’CP’ is the clock period depending on bit ’SRC’.)
t TVALUE 1+()CP⋅=
PEB 20534
Detailed Register Description
Semiconductor Group 323 Data Sheet 09.98
Table 78 XADR: Transmit Address Register
CPU Accessibility: read/write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 0134H01B4H0234H02B4H
typical usage: written by CPU, valid in HDLC mode only
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
(unused)
H0000000000000000
A0000000000000000
B0000000000000000
Bit1514131211109876543210
Mode
Transmit Address (low) Transmit Address (high)
HXAD2 (low byte) XAD1 (high byte) 0 XAD
1_0
or XAD2 (RESPON SE) or XAD1 (COMM A ND)
A0000000000000000
B0000000000000000
PEB 20534
Detailed Register Description
Semiconductor Group 324 Data Sheet 09.98
XAD1 and XAD2 bit fields are valid in HDLC modes with automatic address field
handling only (Automode, Address Mode 1, Non-Automode). They can be programmed
with with one individual address byte which is inserted automatically into the address
field (8 or 16 bit) of a HDLC transmit frame. The function depends on the selected
protocol mode and address field size (bit ’ADM’ in register CCR1).
XAD2 Transmit Address 2 (hdlc mode)
2-byte address field:
Bit field XAD2 constitutes the low byte of the 2-byte address field.
(In ISDN LAP-D, the low byte is known as ’TEI’.)
1-byte address field:
According to the X.25 LAP-B protocol, XAD2 is the address of a
’RESPONSE’ frame.
XAD1 Transmit Address 1 (hdlc mode)
2-byte address field:
Bit field XAD1 constitutes the high byte of the 2-byte address field. Bit 1
must be set to ’0’. According to the ISDN LAP-D protocol, bit 1 is
interpreted as the C/R (COMMAND/RES PONSE) bit. This bit is
manipulated automatically by the DSCC4 according to the setting of bit
’CRI’ in register RADR:
bit 1 (C/R)
Commands Transmit 1 0
Responses Transmit 0 1
CRI=1 CRI=0
(In ISDN LAP-D, the low byte is known as ’SAPI’.)
1-byte address field:
According to the X.25 LAP-B protocol, XAD2 is the address of a
’COMMAND’ frame .
PEB 20534
Detailed Register Description
Semiconductor Group 325 Data Sheet 09.98
Table 79 RADR: Receive Address Register
CPU Accessibility: read/write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 0138H01B8H0238H02B8H
typical usage: written by CPU, valid in HDLC mode only
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
Receive Address 1 (high) Receive Address 1 (low)
HRAH1 CRI RAH
1_0 RAL1
or RAH1 RAL1
A0000000000000000
B0000000000000000
Bit1514131211109876543210
Mode
Receive Address 2 (high) Receive Address 2 (low)
HRAH2 RAL2
A0000000000000000
B0000000000000000
PEB 20534
Detailed Register Description
Semiconductor Group 326 Data Sheet 09.98
In operating modes that provide address recognition, the high/low byte of the received
address is compared with the individually programmable values in register RADR.
This addresses can be masked on a per bit basis by setting the corresponding bits in
register RAMR to allow extended broadcast address recognition. This feature is
applicable to all HDLC sub-modes with address recognition.
RAH1 Receive Address 1 Byte High (hdlc mode)
In HDLC Automode bit ’1’ is reserved for ’CRI’ (Command Response
Indication). In all other modes RAH1 is an 8 bit address.
CRI Command/Response Indication
The setting of this bit effects the meaning of the ’C/R’ bit in the receive
status byte (RSTA):
C/R meaning C/R Value
Command received 0 1
Response received 1 0
CRI=1 CRI=0
Note: If 1-byte address field is selected in HDLC Automode, RAH1 must
be set to 0x00
H
.
RAL1 Receive Address 1 Byte Low (hdlc mode)
The general function whether it must be written or read by the CPU and
its meaning depends on the selected operating mode:
•Auto- / Non-Automode (16-bit address)
RAL1 can be programmed with the value of the first individual low
address byte.
•Auto- / Non-Automode (8-bit address)
According to X.25 LAP-B protocol, the address in RAL1 is considered
as the address of a ’COMMAND’ frame.
RAH2 Receive Address 2 Byte High (hdlc mode)
RAL2 Receive Address 2 Byte Low (hdlc mode)
Value of the second individually programmable high/low address byte. If
a 1-byte address field is selected, RAL2 is considered as the address of
a ’RESPONSE’ frame according to X.25 LAP-B protocol.
PEB 20534
Detailed Register Description
Semiconductor Group 327 Data Sheet 09.98
Table 80 RAMR: Receive Address Mask Register
CPU Accessibility: read/write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 013CH01BCH023CH02BCH
typical usage: written by CPU, valid in HDLC mode only
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
Receive Mask Address 2 (high) Receive Mask Address 2 (low)
HAMRAH2 AMRAL2
A0000000000000000
B0000000000000000
Bit1514131211109876543210
Mode
Receive Mask Address 1 (high) Receive Mask Address 1 (low)
HAMRAH1 AMRAL1
A0000000000000000
B0000000000000000
PEB 20534
Detailed Register Description
Semiconductor Group 328 Data Sheet 09.98
AMRAH2 Receive Mask Address 2 Byte High (hdlc mode)
AMRAL2 Receive Mask Address 2 Byte Low (hdlc mode)
AMRAH1 Receive Mask Address 1 Byte High (hdlc mode)
AMRAL1 Receive Mask Address 1 Byte Low (hdlc mode)
Setting a bit in this bit field to ’1’ masks the corresponding bit in bit field
{’RAH2’/’RAL2’/’RAH1’/’RAL1’} of register RADR. A masked bit position
always matches when comparing the received frame address with bit
field {’RAH2’/’RAL2’/’RAH1’/’RAL1’} allowing extended broadcast
mechanism.
bit = ’0’ The dedicated bit position is NOT masked. This bit
position in the received address must match with the
corresponding bit position in bit field
{’RAH2’/’RAL2’/’RAH1’/’RAL1’} to accept the frame.
bit = ’1’ The dedicated bit position is masked. This bit position in
the received address NEED NOT match with the
corresponding bit position in bit field
{’RAH2’/’RAL2’/’RAH1’/’RAL1’} to accept the frame.
PEB 20534
Detailed Register Description
Semiconductor Group 329 Data Sheet 09.98
Table 81 RLCR: Receive Length Check Register
CPU Accessibility: read/write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 0140H01C0H0240H02C0H
typical usage: written by CPU, valid in HDLC mode only
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
(unused)
H0000000000000000
A0000000000000000
B0000000000000000
Bit1514131211109876543210
Mode
Receive Length Control Receive Length Limit
HRCE 0 0 0 0 RL(10:0)
A0000000000000000
B0000000000000000
PEB 20534
Detailed Register Description
Semiconductor Group 330 Data Sheet 09.98
RCE Receive Length Check Enable (hdlc mode)
This bit is valid in HDLC mode only and enables/disables the receive
length check function:
RCE = ’0’ No receive length check on received HDLC frames is
performed.
RCE = ’1’ The receive length check is enabled. All bytes of a HDLC
frame which are transferred to the receive FIFO
(depending on the selected protocol sub-mode and
receive CRC handling) are counted and checked against
the maximum length check limit which is programmed in
bit field ’RL’.
A frame exceeding the maximum length is treated as if it
were aborted on the receive line (receive abort ’RAB’
interrupt and status indication).
In addition a ’FLEX’ interrupt is generated if enabled.
Note: The Receive Status Byte (RSTA) is part of the
frame length checking.
Thus it is guaranteed, that the number of bytes
transferred to the host memory for one frame never
exceeds the value programmed to bit field ’RL’.
RL(10:0) Receive Length Check Limit (hdlc mode)
This bit-field defines the receive length check limit if checking is enabled
via bit ’RCE’:
RL(10:0) The receive length limit is calculated by:
Limit RL 1+()32⋅=
PEB 20534
Detailed Register Description
Semiconductor Group 331 Data Sheet 09.98
Table 82 XNXF: XON/XOFF In-Band Flow Control Character Register
CPU Accessibility: read/write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 0144H01C4H0244H02C4H
typical usage: written by CPU, valid in ASYNC mode only
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
XON Character) XOF Character
H0000000000000000
AXON XOFF
B0000000000000000
Bit1514131211109876543210
Mode
XON Character Mask XOFF Character Mask
H0000000000000000
AMXON MXOFF
B0000000000000000
PEB 20534
Detailed Register Description
Semiconductor Group 332 Data Sheet 09.98
XON XON Character (async mode)
This bit field specifies the XON character for in-band flow control in
ASYNC protocol mode. The number of significant bits starting with the
LSB depends on the character length (5..8 bits) selected via bit field
’CHL’ in register CCR2.
A received character is recognized as a valid XON-character, if
• th e ch arac ter was c orrect ly fra med (c haracter length as p ro gram med
and correct parity if checking is enabled)
• each bit position of the received character which is not masked via bit
field ’MXON’ matches with the corresponding bit in bit field ’XON’.
Received characters recognized as ’XON’ character are always stored
in the receive FIFO as normal receive data, but generate an appropriate
XON interrupt if enabled and switch the transmitter into ’XON’ state if in-
band flow control is enabled via bit ’FLON’ in register CCR2.
XOFF XOFF Character (async mode)
This bit field specifies the XOFF character for in-band flow control in
ASYNC protocol mode. The number of significant bits starting with the
LSB depends on the character length (5..8 bits) selected via bit field
’CHL’ in register CCR2.
A received character is recognized as a valid XOFF-character, if
• th e ch arac ter was c orrect ly fra med (c haracter length as p ro gram med
and correct parity if checking is enabled)
• each bit position of the received character which is not masked via bit
field ’MXOFF’ matches with the corresponding bit in bit field ’XOFF’.
Received characters recognized as ’XOFF’ character are always stored
in the receive FIFO as normal receive data, but generate an appropriate
XOFF interrupt if e nab led a nd sw it ch the transmitte r i nto ’XO FF’ s tate if
in-band flow control is enabled via bit ’FLON’ in register CCR2.
PEB 20534
Detailed Register Description
Semiconductor Group 333 Data Sheet 09.98
MXON XON Character Mask (async mode)
Setting a bit in this bit field to ’1’ masks the corresponding bit in bit field
’XON’ of register XNXF. A masked bit position always matches when
comparing the received character with bit field ’XON’.
bit = ’0’ The dedicated bit position is NOT masked. This bit
position in the received character must match with the
corresponding bit position in bit field ’XON’ to recognize
the received character as a ’XON’ character.
bit = ’1’ The dedicated bit position is masked. This bit position in
the received address NEED NOT match with the
corresponding bit position in bit field ’XON’ to recognize
the received character as a ’XON’ character.
MXOFF XOFF Character Mask (async mode)
Setting a bit in this bit field to ’1’ masks the corresponding bit in bit field
’XOFF’ of register XNXF. A masked bit position always matches when
comparing the received character with bit field ’XOFF’.
bit = ’0’ The dedicated bit position is NOT masked. This bit
position in the received character must match with the
corresponding bit position in bit field ’XOFF’ to recognize
the received character as a ’XOFF’ character.
bit = ’1’ The dedicated bit position is masked. This bit position in
the received address NEED NOT match with the
corresponding bit position in bit field ’XOFF’ to recognize
the received character as a ’XOFF’ character.
PEB 20534
Detailed Register Description
Semiconductor Group 334 Data Sheet 09.98
Table 83 TCR: Termination Character Register
CPU Accessibility: read/write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 0148H01C8H0248H02C8H
typical usage: written by CPU, valid in ASYNC/BISYNC modes only
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
(unused)
H0000000000000000
A0000000000000000
B0000000000000000
Bit1514131211109876543210
Mode
Termination Character Control Termination Character
H0000000000000000
A
TCDE
0000000 TC(7:0)
B
TCDE
0000000 TC(7:0)
PEB 20534
Detailed Register Description
Semiconductor Group 335 Data Sheet 09.98
TCDE Termination Character Detection Enable (async/bisync modes)
This bit is valid in ASYNC/BISYNC modes only and enables/disables the
termination character detection mechanism:
TCDE = ’0’ No receive termination character detection is performed.
TCDE = ’1’ The termination character detection is enabled. The
receive data stream is monitored for the occurence of a
termination character (TC) programmed via bit field ’TC’.
When this character is detected, an internal ’frame end /
block end’ indication is generated. This causes the DMA
controller to complete the current receive descriptor and
branch to the next receive descriptor address.
Note: If the programmed character length (bit field ’CHL’
in register CCR2) is less than 8 bits, the most
significant unused bits of bit field ’TC’ must be set to
’0’. Otherwise no termination character will be
detected.
TC(7:0) Termination Character (async/bisync modes)
This bit-field defines the termination character which is monitored on the
receive data stream if enabled via bit ’TCDE’.
PEB 20534
Detailed Register Description
Semiconductor Group 336 Data Sheet 09.98
Table 84 TICR: Transmit Immediate Character Register
CPU Accessibility: read/write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 014CH01CCH024CH02CCH
typical usage: written by CPU, valid in ASYNC mode only
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
(unused)
H0000000000000000
A0000000000000000
B0000000000000000
Bit1514131211109876543210
Mode
(unused) Transmit Immediate Character
H0000000000000000
A00000000 TIC(7:0)
B0000000000000000
PEB 20534
Detailed Register Description
Semiconductor Group 337 Data Sheet 09.98
TIC Transmit Immediate Character (async mode)
On write access to this register the ASYNC protocol engine will
automatically insert the character defined by bit field ’TIC’ into the
transmit data stream.
This happens
• imm ediatel y a fter writ e acce ss to re gister TI CR i f the tran smitter is in
IDLE state (no other character is currently transmitted). The
transmitter returns to IDLE state after transmission of the TIC.
• immediately after the character which is currently in transmission is
completed. After transmission of the TIC, the transmitter continues
with transmission of characters which are still stored in the transmit
FIFO. Thus the TIC is inserted into the data stream between the
characters provided via the transmit FIFO.
The TIC transmission is independent off in-band flow control. Thus the
TIC is send out even if the transmitter is in XOFF-state. However the
transmitter must be enabl ed vi a sig nal C T S (depe nding on bit ’FCT S’ in
register CCR1).
The number of significant bits (starting with the LSB) depends on the
character length programmed in bit field ’CHL’ in register CCR2. All
character framing related settings in register CCR2 (start bit, parity
generation, number of stop bits) also apply to the TIC character framing.
As long as the TIC character is not completely send, status bit TIC
Execution (’TEC’) in status register STAR is set to ’1’ by the DSCC4. No
further write access to register TICR is allowed until ’TEC’ status
indication is cleared by the DSCC4.
PEB 20534
Detailed Register Description
Semiconductor Group 338 Data Sheet 09.98
Table 85 SYNCR: Synchronization Character Register
CPU Accessibility: read/write
Reset Value: 0000 0000H
SCC0 SCC1 SCC2 SCC3
Offset Address: 0150H01D0H0250H02D0H
typical usage: written by CPU, valid in BISYNC mode only
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
(unused)
H0000000000000000
A0000000000000000
B0000000000000000
Bit1514131211109876543210
Mode
Synchronization Character(s)
H0000000000000000
A0000000000000000
BSYNCH(7:0) SYNCL(7:0)
PEB 20534
Detailed Register Description
Semiconductor Group 339 Data Sheet 09.98
SYNCH(7:0) Synchronization Character (high) (bisync mode)
SYNCL(7:0) Sync hroniza tion Character (low) (bisync mode)
This register is only valid in BISYNC protocol mode.
The synchronization (SYNC) character format depends on the setting of
bit ’BISNC’ and ’SLEN’ in register CCR1:
• MONOSYNC Mode (CCR1.BISNC = ’0’)
The SYNC character is defined by bit field ’SYNCL’:
a) SLEN = ’0’: the 6 bit SYNC character is specified by bits (5..0)
b) SLEN = ’1’: the 8 bit SYNC character is specified by bits (7..0).
• BISYNC Mode (CCR1.BISNC = ’1’)
The SYNC character is defined by bit fields ’SYNCL’ and ’SYNCH’:
a) SLEN = ’0’: the 12 bit SYNC character is specified by bits (5..0) of
each bit field, i.e. SYNC(11..0) = SYNCH(5..0), SYNCL(5..0)
b) SLEN = ’1’: the 16 bit SYNC character is specified by bits (7..0) of
each bit field, i.e. SYNC(15..0) = SYNCH(7..0), SYNCL(7..0).
In transmit di recti on the SYNC cha racter is se nt continuou sly if no da ta
has to be transmitted and interframe timefill control is enabled by setting
bit ’ITF’ to ’1’ in register CCR2.
In receive direction the receiver monitors the data stream for occurence
of the specified SYNC pattern if operating in ’HUNT’ mode (bit ’HUNT’ in
register CMDR).
PEB 20534
Detailed Register Description
Semiconductor Group 340 Data Sheet 09.98
Table 86 IMR: Interrupt Mask Register
CPU Accessibility: read/write
Reset Value: FFFF FFFFH
SCC0 SCC1 SCC2 SCC3
Offset Address: 0154H01D4H0254H02D4H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
Mode
Output Control
H1111111111111
ALLS
1
XDU
A1111111111111
ALLS
1
XOFF
B1111111111111
ALLS
1
XDU
Bit1514131211109876543210
Mode
Transmitter/Receiver Configuration
H
TIN
CSC
XMR
XPR
1111
RDO
RFS
RSC
PCE
PLLA
CDSC
RFO
FLEX
A
TIN
CSC
XOFF
XPR
11
BRK
BRKT
TCD
TIME
PERR
FERR
PLLA
CDSC
RFO
1
B
TIN
CSC
XMR
XPR
1111
TCD
1
PERR
SCD
PLLA
CDSC
RFO
1
PEB 20534
Detailed Register Description
Semiconductor Group 341 Data Sheet 09.98
(31:0) Interrupt Mask Bits (all modes)
Each SCC interrupt event can generate an interrupt vector as well as an
interrupt signal indication to pin INTA. Each bit position of register IMR
is a mask for the corresponding interrupt event in the interrupt status
register ISR. Masked interrupt events neither generate an interrupt
vector nor an interrupt indication to via pin INTA.
bit = ’0’ The corresponding interrupt event is NOT masked and will
generate an interrupt vector as well as an interrupt
indication via pin INTA.
bit = ’1’ The corresponding interrupt event is masked and will
NEITHER generate an interrupt vector NOR an interrupt
indication via pin INTA.
Moreover, masked interrupt events are:
• not di splaye d in the interrupt status regis ter ISR if bit ’ VIS’ i n regis ter
CCR0 is programmed to ’0’.
• are displayed in interrupt status register ISR if bit ’VIS’ in register
CCR0 is programmed to ’1’.
Note: After RESET, all interrupt events are masked.
For detailed interrupt event description refer to the corresponding bit
position in register ISR.
PEB 20534
Detailed Register Description
Semiconductor Group 342 Data Sheet 09.98
Table 87 ISR: Interrupt Status Register
CPU Accessibility: read
Reset Value: FFFF FFFFH
SCC0 SCC1 SCC2 SCC3
Offset Address: 0158H01D8H0258H02D8H
typical usage: written by DSCC4
evaluated by CPU
Bit31302928272625242322212019181716
Mode
Output Control
H1111111111111
ALLS
1
XDU
A1111111111111
ALLS
1
XOFF
B1111111111111
ALLS
1
XDU
Bit1514131211109876543210
Mode
Transmitter/Receiver Configuration
H
TIN
CSC
XMR
XPR
1111
RDO
RFS
RSC
PCE
PLLA
CDSC
RFO
FLEX
A
TIN
CSC
XON
XPR
11
BRK
BRKT
TCD
TIME
PERR
FERR
PLLA
CDSC
RFO
1
B
TIN
CSC
XMR
XPR
1111
TCD
1
PERR
SCD
PLLA
CDSC
RFO
1
PEB 20534
Detailed Register Description
Semiconductor Group 343 Data Sheet 09.98
ALLS ALL Sent Interrupt (all modes)
HDLC Mode:
This bit is set to ’1’:
• if the last bit of the current HDLC frame is sent out via pin TxD and no
further frame is stored in the SCC transmit FIFO, i.e. the transmit FIFO
is empty (Address Mode 0/1, Non-Automode);
• if an I-frame is sent out completely via pin TxD and either a valid
acknowledge S-frame has been received or a time-out condition
occured because no valid acknowledge S-frame has been received in
time (Automode).
ASYNC/BISYNC Mode:
This bit is set to ’1’, if the last character is completely sent via pin TxD
and no further data is stored in the SCC transmit FIFO, i.e. the transmit
FIFO is empty.
XDU Transmit Data Underrun Interrupt (hdlc/bisync mode)
HDLC Mode:
This bit is set to ’1’, if the current frame was terminated by the SCC with
an abort sequence, because neither a ’frame end / block end’ indication
was detected in the FIFO (to complete the current frame) nor more data
is available in the SCC transmit FIFO.
Note: The transmitter is stopped if this condition occurs and needs to be
reset via command bit ’XRES’ in register CMDR. Furthermore the
XDU interrupt indication MUST be cleared by generating an
interrupt vec tor, thus bi t ’XDU’ sh ould not be masked v ia register
IMR.
BISYNC Mode:
This bit is set to ’1’, if the current transmission was terminated with IDLE
sequence because no more data is available in the SCC transmit FIFO.
Note: The transmitter is stopped if this condition occurs and needs to be
reset via command bit ’XRES’ in register CMDR. Furthermore the
XDU interrupt indication MUST be cleared by generating an
interrupt vec tor, thus bi t ’XDU’ sh ould not be masked v ia register
IMR.
PEB 20534
Detailed Register Description
Semiconductor Group 344 Data Sheet 09.98
XOFF XOFF Character Detected Interrupt (async mode)
ASYNC Mode:
This bit is set to ’1’, if the currently received character matched the XOFF
character programmed in bit field ’XOFF’ in register XNXF and indicates,
that the transmitter is switched to XOFF-state if in-band flow control is
enabled via bit ’FLON’ in register CCR2.
TIN Timer Interrupt (all modes)
This bit is set to ’1’, if the internal timer was activated and has expired
(refer also to description of timer register TIMR).
CSC CTS Status Change (all modes)
This bit is set to ’1’, if a transition occurs on signal CTS. The current state
of signal CTS is monitored by status bit ’CTS’ in status register STAR.
XMR Transmit Mes sage Re peat (hdlc/bisync modes)
HDLC Mode:
This bit i s set to ’1’ , if transm ission of the last frame has to be repeated
(by software), because
• the SCC has received a negative acknowledge to an I-frame in HDLC
Automode operation;
• a collision occured after at least four bytes of data has been
completely sent out, i.e. automatic re-transmission cannot be
performed by the SCC;
•CTS
signa l was deasserted after a t lea st four bytes of data has b een
completely sent out.
Note: For easy recovery from a collision event (in bus configuration
only), the SCC transmit FIFO should not contain more than one
complete frame. This can be achieved by using the ’ALLS’
interrupt to control the corresponding DMA controller transmit
channel forwarding a new frame on all sent (ALLS) event only.
BISYNC Mode:
This bit is se t to ’1 ’, if tra nsm iss ion of the last block of ch aracters has to
be repeated (by software), because CTS signal was deasserted after at
least four bytes of data has been completely sent out.
PEB 20534
Detailed Register Description
Semiconductor Group 345 Data Sheet 09.98
XON XON Character Detected Interrupt (async mode)
ASYNC Mode:
This bit is set to ’1’, if the currently received character matched the XON
character programmed in bit field ’XON’ in register XNXF and indicates,
that the transmitter is switched to XON-state if in-band flow control is
enabled via bit ’FLON’ in register CCR2.
XPR Transmit Pool Ready Interrupt (all modes)
This bit is set to ’1’, if a transmitter reset command was executed
success ful ly (com mand bit ’XR ES’ i n re gis ter C MDR ) and transmit data
can be written to the FIFO by the DMA controller.
A ’XPR’ interrupt is not generated, if no sufficient transmit clock is
available (depending on the selected clock mode).
BRK Break Interrupt (async mode)
This bit is set to ’1’, if a break condition was detected on the receive line,
i.e. a low level for a time equal to (character length + parity bit + stop
bit(s)) bits depending on the selected ASYNC character format.
BRKT Break Terminated Interrupt (async mode)
This bit is set to ’1’, if a previously detected break condition on the
receive line is terminated by a low to high transition.
RDO Receive Data Overflow Interrupt (hdlc mode)
This bit is set to ’1’, if receive data of the current frame got lost because
of a SCC receive FIFO full condition. However the rest of the frame is
received a nd disc arded as lon g as the re ceive FIFO re mains f ull and is
stored as soon as FIFO space is available again. The receive status byte
(RSTA) of such a frame contains an ’RDO’ indication.
TCD Termination Character Detected Interrupt (async/bisync modes)
This bit is set to ’1’, if a terminatio n characte r is detec ted in the receive
data stre am. The SCC will in sert a ’frame end / bloc k end’ indicatio n to
the SCC receive FIFO which causes the DMAC to finish the current
receive descriptor.
PEB 20534
Detailed Register Description
Semiconductor Group 346 Data Sheet 09.98
RFS Receive Frame Start Interrupt (hdlc mode)
This bit is set to ’1’, if the beginning of a valid frame is detected by the
receiver. A valid frame is detected either if a valid address field is
recognized (in all operating modes with address recognition) or if a start
flag is recognized (in all operating modes with no address recognition).
TIME Time Out Interrupt (async mode)
This bit is set to ’1’, if the time out limit is exceeded, i.e. no new character
was receiv ed in a programmable perio d of time (refer to register CC R1
bit fields ’TOIE’ and ’TOLEN’ for more information).
RSC Receive Status Change Interrupt (hdlc mode)
This bit is valid in HDLC Automode only.
It is set to ’1’, if a status change of the remote station receiver has been
detected by receiving a S-frame with receiver ready (RR) or receiver not
ready (RNR) indication. Because only a status change is indicated via
this interrupt, the curren t status can be evaluated by reading bit ’RRNR’
in status register STAR.
PERR Parity Error Interrupt (async/bisync modes)
This bit is only valid if parity checking/generation is enabled via bit
’PARE’ in register CCR2.
It is set to ’1’, if a character with wrong parity has been received. If
enabled via bit ’RFDF’, this error status is additionally stored in the
receive status byte generated for each receive character.
PCE Protocol Error Interrupt (hdlc mode)
This bit is valid in HDLC Automode only.
It is set to ’1’, if the receiver has detected a protocol error, i.e. one of the
following events occured:
• an S- or I-frame was received with wrong N(R) counter value;
• an S-frame containing an I-control field was received.
PEB 20534
Detailed Register Description
Semiconductor Group 347 Data Sheet 09.98
FERR Framing Error Interrupt (async mode)
This bit is set to ’1’, if a character framing error is detected, i.e. a ’0’ was
ampled at a position where a stop bit ’1’ was expected due to the
selected character format.
SCD SYN Character Detected Interrupt (bisync mode)
This bit is set to ’1’, if a synchronization character (SYNC) was detected
after the rece iver was switche d to HUNT-mode ( by command b it ’HUNT’
in register CMDR).
PLLA DPLL Asynchronous Interrupt (all modes)
This bit is only valid, if the receive clock is derived from the internal DPLL
and FM0 , FM1 or Manc hester dat a encodi ng is select ed (de pending on
the selected clock mode and dat a encoding mode). It i s set to ’1’ if the
DPLL has lost synchronization. Reception is disabled until
synchronization has been regained again. If the transmitter is supplied
with a clock derived from the DPLL, transmission is also interrupted.
CDSC Carrier Detect Status Change Interrupt (all modes)
This b it i s s et to ’1’, if a state trans itio n h as been detect ed at s ign al CD .
Because only a state transition is indicated via this interrupt, the current
status can be evaluated by reading bit ’CD’ in status register STAR.
RFO Receive FIFO Overflow In terru pt (all modes)
HDLC Mode:
This bit is set to ’1’, if the SCC receive FIFO is full and a complete frame
must be discarded. This interrupt can be used for statistical purposes
and might indicate that the DMAC was not able to service the SCC
receive FIFO quickly enough, e.g. PCI bus latencies are too bad.
ASYNC/BISYNC Mode:
This bit is set to ’1’, if the SCC receive FIFO is full and another received
character must be discarded. This interrupt can be used for statistical
purposes a nd m igh t indicate that th e D MAC was not a ble to s ervi ce th e
SCC receive FIFO quickly enough, e.g. PCI bus latencies are too bad.
PEB 20534
Detailed Register Description
Semiconductor Group 348 Data Sheet 09.98
FLEX Frame Length Exceeded Interrupt (hdlc mode)
This bit is set to ’1’, if the frame length check feature is enabled and the
current received frame is aborted because the programmed frame length
limit was exceeded (refer to register RLCR for detailed description).
PEB 20534
Detailed Register Description
Semiconductor Group 349 Data Sheet 09.98
10.3.3 Peripheral Registers - Detailed Register Description
10.3.3.1 Peripheral Registers Overview
The DSCC4 Peripheral registers are used to configure and control the function blocks
LBI, SSC and GPP.
The full 32 bit address location of each global register consists of:
• Base Address Register 0 (PCI Configuration Space, address location 10H)
• Register address offset, which is in the range 0300H ...07FFH
All registers are 32-bit organized registers.
Table 88 provides an overview about all global registers:
PEB 20534
Detailed Register Description
Semiconductor Group 350 Data Sheet 09.98
Table 88 DSCC4 Peripheral Register Overview
Offset Register Meaning
LBI block specific registers:
0300HLCONF LBI Configuration Register
0304H
...
Reserved
-
037CH
SSC block specific registers:
0380HSSCCON SSC Control Register
0384HSSCBR SSC Baud Rate Generator Register
0388HSSCTB SSC Transmit Buffer
038CHSSCRB SSC Receive Buffer
0390HSSCCSE SSC Chip Select Enable Register
0394HSSCIM SSC Interrupt Mask Register
0398H
...
Reserved
-
03FCH
GPP block specific registers:
0400HGPDIR GPP Direction Configuration Register
0404HGPDATA GPP Data I/O Register
0408HGPOD GPP Open Drain Function Register
040CH
...
Reserved
-
07FCH
PEB 20534
Detailed Register Description
Semiconductor Group 351 Data Sheet 09.98
10.3.3.2 LBI Registers Description
Table 89 LCONF: LBI Configuration Register
CPU Accessibility: read/write
Reset Value: 0000 000H
Offset Address: 0300H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
LBI General Configuration
LINTIC
00000000
EBCRES
000000
Bit1514131211109876543210
LBI General Configuration
000000
EALE
HDEN
BTYP(1:0)
RDEN
ABM
MCTC(3:0)
PEB 20534
Detailed Register Description
Semiconductor Group 352 Data Sheet 09.98
LINTIC LBI Interrupt Input Control (-)
This bit selects whether the LBI interrupt input pin LINTI is a low or high
active input signal:
LINTIC=’0’ LINTI input pin ia a low active input signal, i.e. LINTI=’0’
generates an interrupt indication.
LINTIC=’1’ LINTI input pin ia a high active input signal, i.e. LINTI=’1’
generates an interrupt indication.
EBCRES LBI External Bus Controller Reset (-)
Via this bit the complete LBI block can be reset (disabled) or enabled:
EBCRES
=’0’ The LBIl block is forced into its reset state. Also all
dedicated pins are in reset state (same as hardware
reset).
EBCRES
=’1’ The LBI block isenabled. The function depends to the
selected configuration.
Note: This Reset control bit is not self clearing. The LBI block remains in
its reset state, until a ’1’ is written to bit ’EBCRES’.
EALE LBI Extended ALE (-)
This bit selects whether the LBI ALE output signal is generated for one
LBI clock period or an extended period:
EALE=’0’ ALE signal high time is 1 LBI clock period.
EALE=’1’ ALE signal high time is 1 LBI clock period + 1 PCI clock
high time.
PEB 20534
Detailed Register Description
Semiconductor Group 353 Data Sheet 09.98
HDEN LBI HOLD Enable (-)
This bit selects whether the LBI bus arbitration interface (pins LHOL D,
LHDLA, LBREQ) is enabled or disabled:
HDEN=’0’ The LBI bus arbitration interface is disabled. The DSCC4
(LBI) is always active bus master.
HDEN=’1’ The LBI bus arbitration interface is enabled. The DSCC4
(LBI) shares bus mastership with one or more other bus
masters.
The DSCC4 can be default arbitration master or
arbitration slave depending on the setting of bit ’ABM’.
BTYP(1:0) LBI Bus Type (-)
The Local Bus Interface (LBI) supports 4 different bus configurations
which are selected via this bit field:
BTYP = ’00’ 8 bit address/data de-multiplexed bus
BTYP = ’01’ 8 bit address/data multiplexed bus
BTYP = ’10’ 16 bit address/data de-multiplexed bus
BTYP = ’11’ 16 bit address/data multiplexed bus
Note: The Peripheral Configuration must be selected accordingly (bit
field ’PERCFG’ in register GMODE).
RDEN LBI LRDY Enable (-)
This bit selects whether the LRDY control input signal is evaluated or
ignored by the LBI:
RDEN=’0’ Input signal LRDY is ignored (but should be connected to
a defined level). The bus cycle depends only on the
selected number of wait states (bit field ’MCTC’).
RDEN=’1’ Input signal LRDY is evaluated after the number of
selected wait states have been inserted. The bus
transaction is terminated after the first detection of
LRDY=’0’ (active).
PEB 20534
Detailed Register Description
Semiconductor Group 354 Data Sheet 09.98
ABM LBI Arbitration Master (-)
The DSCC4 (LBI) is always master (initiator) of bus transactions on the
local bus. Nevertheless the local bus can be shared with other master
peripherals. In this case the bus arbitration interface must be enabled by
setting bit ’HDEN’ to ’1’.
Bit ’ABM’ selects whether the DSCC4 (LBI) is arbitration default master
or arbitration slave, i.e. another peripheral is arbitration default bus
master.
ABM=’0’ The DSCC4 (LBI) is arbitration slave. Pin LHDLA of the
bus arbitration interface is an input signal.
ABM=’1’ The DSCC4 (LBI) is arbitration default master. Pin LHDLA
of the bus arbitration interface is an output signal.
MCTC(3:0) LBI Memory Cycle Time Control (-)
Via this bit field, a constant number of wait states can be selected for
each LBI bus cycle (read and write). The wait states are inserted into the
read and write strobe signal (LRD, LWR) active time (based on LBI clock
cycles):
MCTC Wait States:
’0000’ 15
’0001’ 14
... ...
’000’ 0
Note: The minimum active time of read and write strobe signals is 2 LBI
clocks. MCTC wait states are additional. If LRDY control is
enabled, further wait states may be inserted depending on the
LRDY input signal which is generated by the connected
peripherals.
PEB 20534
Detailed Register Description
Semiconductor Group 355 Data Sheet 09.98
10.3.3.3 SSC Registers Description
Table 90 SSCCON: SSC Control Register
CPU Accessibility: read/(write)
(Do not write in operating mode,
i.e. write access with SSCEN = ’1’.)
Reset Value: 0000 000H
Offset Address: 0380H
typical usage: written by CPU for control (configuration mode),
read by CPU for status information (operating mode)
evaluated/updated by DSCC4
Bit31302928272625242322212019181716
Mode
(unused)
Control
0000000000000000
Status
0000000000000000
Bit1514131211109876543210
Mode
SSC General Configuration And Status
Control
SSCEN=’0’
SSCMS
0
SSCREN
SSCBEN
SSCPEN
SSCREN
SSCTEN
0
SSCPO
SSCPH
SSCHB
SSCBM(3:0)
Status
SSCEN=’1’
SSCMS
0
SSCBSY
SSCBE
SSCPE
SSCRE
SSCTE
0000 SSCBC(3:0)
PEB 20534
Detailed Register Description
Semiconductor Group 356 Data Sheet 09.98
SSCEN SSC Enable (-)
This bit selects whether the SSC is in ’configuration mode’ or in normal
’operation mode’. The meaning of bits 14..0 depends on the setting of
this bit:
SSCEN=’0’ The SSC is in configuration mode. Bits 14..0 provide
con trol bits for SSC configuration.
SSCEN=’1’ The SSC is in normal operation mode. Bits 14..0 provide
status information bits.
SSCMS SSC Master Select (configuration/operation mode)
This bit selects whether the SSC is operating in master or in slave mode.
Bit ’SSCMS’ is valid in configuration and operation mode of register
SSCCON:
SSCMS=’0’ The SSC is slave. Operation is performed by the shift
clock, supplied at pin MCLK (input).
SSCEN=’1’ The SSC is master. Operation is performed by the
internally generated shift clock which is monitored at pin
MCLK (outpu t).
SSCREN SSC Receive Error Enable (configuration mode)
This bit selects whether the SSC ignores or checks receive errors:
SSCREN
=’0’ The SSC ignores receive errors.
SSCREN
=’1’ The SSC checks receive errors.
SSCBSY SSC Busy Status Flag (operation mode)
This bit indicates that a transfer is currently in process:
SSCBSY
=’0’ No tr ansfer is in process.
SSCBSY
=’1’ A transfer is currently in process.
PEB 20534
Detailed Register Description
Semiconductor Group 357 Data Sheet 09.98
SSCBEN SSC Baud Rate Error Enable (configuration mode)
This bit selects whether the SSC ignores or checks baud rate errors:
SSCREN
=’0’ The SSC ignores baud rate errors.
SSCREN
=’1’ The SSC checks baud rates errors.
SSCBE SSC Baud Rate Status Flag (operation mode)
This bit indicates a baud rate mismatch:
SSCBE=’0’ No baud rate mismatch is detected.
SSCBE=’1’ A baud rate mismatch is detected, i.e. the slaves baudrate
deviates from the expected baud rate more than a factor
of 2 or 0.5.
SSCPEN SSC Phase Error Enable (configuration mode)
This bit selects whether the SSC ignores or checks phase errors:
SSCPEN
=’0’ The SSC ignores phase errors.
SSCPEN
=’1’ The SSC checks phase errors.
SSCPE SSC Baud Rate Status Flag (operation mode)
This bit indicates a phase error:
SSCPE=’0’ No phase error is detected.
SSCPE=’1’ A phase error is detected, i.e. a transition occured on the
receive data signal within a guard window around the
sampling clock edge.
PEB 20534
Detailed Register Description
Semiconductor Group 358 Data Sheet 09.98
SSCREN SSC Receive Error Enable (configuration mode)
This bit selects whether the SSC ignores or checks receive errors:
SSCREN
=’0’ The SSC ignores receive errors.
SSCREN
=’1’ The SSC checks receive errors.
SSCRE SSC Receive Status Flag (operation mode)
This bit indicates a receive error:
SSCRE=’0’ No receive error is detected.
SSCRE=’1’ A receive error is detected, i.e. reception is completed
before the receive buffer was read by the CPU.
SSCTEN SSC Transmit Error Enable (configuration mode)
This bit selects whether the SSC ignores or checks transmit errors:
SSCTEN
=’0’ The SSC ignores transmit errors.
SSCTEN
=’1’ The SSC checks transmit errors.
SSCTE SSC Transmit Status Flag (operation mode)
This bit indicates a transmit error:
SSCTE=’0’ No transmit error is detected.
SSCTE=’1’ A transmit error is detected, i.e. transmission starts before
the transmit buffer has been updated by the CPU.
SSCPO SSC Polarity Control (configuration mode)
This bit selects the polarity of the clock:
SSCPO=’0’ The idle clock line is ’low’. Leading clock edge is a low-to-
high transition.
SSCPO=’1’ The idle clock line is ’high’. Leading clock edge is a high-
to-low transition.
PEB 20534
Detailed Register Description
Semiconductor Group 359 Data Sheet 09.98
SSCPH SSC Clock Phase Control (configuration mode)
This bit selects the active clock phase for operation. The definition of the
leading clock edge depends on the setting of bit ’SSCPO’:
SSCPH=’0’ Transmit data is shifted with the leading clock edge,
receive data is latched with the trailing clock edge.
SSCPH=’1’ Transmit data is shifted with the trailing clock edge,
receive data is latched with the leading clock edge.
SSCHB SSC Heading (Bit Order) Control (configuration mode)
This bit selects if LSB or MSB is transmitted/received first:
SSCHB=’0’ LSB first operation on transmit and receive.
SSCHB=’1’ MSB first operation on transmit and receive.
SSCBM
(3:0) SSC Data Width Control (configuration mode)
Via this bit field, the data width (active part of the transmit and receive
buffers) can be selected in the range 2 to 16 bit:
SSCBM Data width:
’0000’
Reserved
. Do not use.
’0001’ Data width is 2 bit.
’0010’ Data width is 3 bit.
... ...
’111’ Data width is 16 bit.
SSCBC
(3:0) SSC Shift Counter (operation mode)
This bit field is used by the SCC as shift counter and is updated with
every bit shift operation.
PEB 20534
Detailed Register Description
Semiconductor Group 360 Data Sheet 09.98
Table 91 SSCBR: SSC Baud Rate Register
CPU Accessibility: read/write
Reset Value: 0000 000H
Offset Address: 0384H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
(unused)
0000000000000000
Bit1514131211109876543210
SSC Baud Rate Generator
SSCBR(15:0)
PEB 20534
Detailed Register Description
Semiconductor Group 361 Data Sheet 09.98
These bits define the baud rate used for data transfer via the SSC interface. Reading
SSCBR (while SSC is enabled via bit ’SSCEN’ in register SSCCON) returns the timer
value. Reading SSCBR (while SSC is disabled) returns the programmed reload value.
The desired reload value of the baud rate can be written to SSCBR when the SSC
interface is disabled.
Note: The contents of SSCBR must always be > 0.
Never write to SSCBR, while the SSC is enabled.
Table 92 SSC Baud Rate Values
SSCBR(15:0) Baud Rate Bit Time
0000HReserved. Use a reload
value > 0. ---
0001H5 MBaud 200 ns
0002H3.3 MBaud 300 ns
0003H2.5 MBaud 400 ns
0004H2.0 MBaud 500 ns
0009H1.0 MBaud 1 µs
0063H100 KBaud 10 µs
03E7H10 KBaud 100 µs
270FH1.0 KBaud 1 ms
FFFFH152.6 Baud 6.6 ms
PEB 20534
Detailed Register Description
Semiconductor Group 362 Data Sheet 09.98
Bit field ’SSCTB’ is written by the CPU and contains the transmit data to be transmitted.
The number of valid bits depend on bit field ’SSCBM’ in register SSCCON (configuration
mode).
Table 93 SSCTB: SSC Transmit Buffer Register
CPU Accessibility: write
Reset Value: 0000 000H
Offset Address: 0388H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
(unused)
0000000000000000
Bit1514131211109876543210
SSC Transmit Buffer
SSCTB(15:0)
PEB 20534
Detailed Register Description
Semiconductor Group 363 Data Sheet 09.98
Bit field ’SSCRB’ is read by the CPU and contains the receive data. The number of valid
bits depend on bit field ’SSCBM’ in register SSCCON (configuration mode).
Table 94 SSCRB: SSC Receive Buffer Register
CPU Accessibility: read
Reset Value: 0000 000H
Offset Address: 038CH
typical usage: written by DSCC4
evaluated by CPU
Bit31302928272625242322212019181716
(unused)
0000000000000000
Bit1514131211109876543210
SSC Receive Buffer
SSCRB(15:0)
PEB 20534
Detailed Register Description
Semiconductor Group 364 Data Sheet 09.98
Table 95 SSCCSE: SSC Chip Select Enable Register
CPU Accessibility: read/write
Reset Value: 0000 000H
Offset Address: 0390H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
(unused)
0000000000000000
Bit1514131211109876543210
SSC Chip Select Control
00000000
ASEL3
ASEL2
ASEL1
ASEL0
0000
PEB 20534
Detailed Register Description
Semiconductor Group 365 Data Sheet 09.98
ASEL3 SSC Chipselect 3 (-)
ASEL2 SSC Chipselect 2 (-)
ASEL1 SSC Chipselect 1 (-)
ASEL0 SSC Chipselect 0 (-)
These bits determine the function of the chipselect signals MCSi (i=3..0):
ASELi=’0’ The MCSi chipselect pin is active low (constant ’0’).
ASELi=’1’ The MCSi chipselect pin is controlled automatically by the
SSC transmitter.
PEB 20534
Detailed Register Description
Semiconductor Group 366 Data Sheet 09.98
Table 96 SSCIM: SSC Interrupt Mask Register
CPU Accessibility: read/write
Reset Value: 0000 000H
Offset Address: 0394H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
(unused)
0000000000000000
Bit1514131211109876543210
SSC Interrupt Mask Control
0000000000000
IMRX
IMER
IMTX
PEB 20534
Detailed Register Description
Semiconductor Group 367 Data Sheet 09.98
Note: The transmit interrupt notifies the CPU about the start of a transmission.
The receive interrupt notifies transfer of the received data to the shared memory.
The error interrupt notifies the CPU about different error conditions related to data
transmission and reception. To further specify what sort of error interrupt the user
wants to trace, the corresponding bits of the SSC control register SSCCON has to
be set. The SSC error conditions that can be checked are transmit errors
(SSCCON(bit 8) =‘1‘), phase errors (SSCCON(bit 10)=‘1‘) and baud rate errors
(SSCCON(bit 11)=‘1‘). If any of these error conditions shall not be checked, the
corresponding enable bit has to be set to ‘0‘.
Example
To check for transmit errors only:
SSCIM(bit 1)=‘1‘, SSCCON(bit 8)=‘1‘, SSCCON(bit 10)=‘0‘,SSCCON(bit 11)=‘0‘
IMRX SSC Receive Interrupt Mask (-)
This bit enables/disables receive interrupt indications:
IMRX=’0’ SSC receive interrupts are disabled.
IMRX=’1’ SSC receive interrupts are enabled.
IMER SSC Error Interrupt Mask (-)
This bit enables/disables error interrupt indications:
IMRX=’0’ SSC error interrupts are disabled.
IMRX=’1’ SSC error interrupts are enabled.
IMTX SSC Transmit Interrupt Mask (-)
This bit enables/disables transmit interrupt indications:
IMRX=’0’ SSC transmit interrupts are disabled.
IMRX=’1’ SSC transmit interrupts are enabled.
PEB 20534
Detailed Register Description
Semiconductor Group 368 Data Sheet 09.98
10.3.3.4 GPP Registers Description
Table 97 GPDIR: GPP Direction Register
CPU Accessibility: read/write
Reset Value: 0000 000H
Offset Address: 0400H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
(unused)
0000000000000000
Bit1514131211109876543210
GPP I/O Signal Direction Control
GPDIR(15:0)
PEB 20534
Detailed Register Description
Semiconductor Group 369 Data Sheet 09.98
GPDIR
(15:0) GPP I/O Signal Direction Control (-)
Each bit of this bit field controls the direction (input/output) of the
corresponding General Purpose Pin. E.g. GPDIR bit 8 determins the I/O
characteristic of pin GP8.
Note: Even if not configured for GPP operation (bit field ’PERCFG’ in
register GMODE), this register must be programmed appropriately
for correct SSC or LBI operation.
(For detailed information refer to the chapters describing SSC and
LBI operation.)
GPDIR(i)=’0’ Pin GPi is configured as output pin.
GPDIR(i)=’1’ Pin GPi is configured as input pin.
PEB 20534
Detailed Register Description
Semiconductor Group 370 Data Sheet 09.98
Table 98 GPDATA: GPP Data Register
CPU Accessibility: read/write
Reset Value: 0000 000H
Offset Address: 0404H
typical usage: written by CPU for input signals,
read by CPU for output signals;
evaluated by DSCC4 for output signals,
written by DSCC4 for input signals;
Bit31302928272625242322212019181716
(unused)
0000000000000000
Bit1514131211109876543210
GPP I/O Data Control
GPDATA(15:0)
PEB 20534
Detailed Register Description
Semiconductor Group 371 Data Sheet 09.98
GPDATA
(15:0) GPP I/O Data Control (-)
Each bit of this bit field is related to the corresponding GPP signal pin.
The function of each bit depends on the I/O configuration of the
dedicated GPP pin:
Output pins (GPDIR(i)=’1’ in register GPDIR):
Write access:
GPDATA(i)
=’0’ Pin GPi output pin is set to ’0’ (low).
GPDATA(i)
=’1’ Pin GPi output pin is set to ’1’ (high).
Read access:
GPDATA(i)
=’0’ Current level of GPi output pin is ’0’ (low).
GPDATA(i)
=’1’ Current level of GPi output pin is ’1’ (high).
Input pins (GPDIR(i)=’0’ in register GPDIR):
Write access:
Write access to GPDATA bit locations related to input signals is ignored.
Read access:
GPDATA(i)
=’0’ Current level of GPi input pin is ’0’ (low).
GPDATA(i)
=’1’ Current level of GPi input pin is ’1’ (high).
PEB 20534
Detailed Register Description
Semiconductor Group 372 Data Sheet 09.98
Table 99 GPIM: GPP Interrupt Mask Register
CPU Accessibility: read/write
Reset Value: 0000 000H
Offset Address: 0408H
typical usage: written by CPU
evaluated by DSCC4
Bit31302928272625242322212019181716
(unused)
0000000000000000
Bit1514131211109876543210
GPP Interrupt Mask Control
GPIM(15:0)
PEB 20534
Detailed Register Description
Semiconductor Group 373 Data Sheet 09.98
GPIM(15:0) GPP Interrupt Mask Control (-)
Each bit of this bit field enables/disables interrupt generation in case of
transitions on the corresponding GPP pins. Even if not configured for
GPP operation (bit field ’PERCFG’ in register GMODE), this register
should be programmed appropriately for correct SSC or LBI operation
(masking interrupt generation).
Interrupt generation should be disabled for GPP pins, configured as
output pins via bit field ’GPDIR’ in register GPDIR.
GPIM(i)=’0’ Pin GPi interrupt generation is disabled.
GPIM(i)=’1’ Pin GPi interrupt generation is enabled.
PEB 20534
Host Memory Organization
Semiconductor Group 374 Data Sheet 09.98
11 Host Memory Organization
11.1 Linked List Structure
11.1.1 Transmit Descriptor Lists
Each transm it descriptor consists of 4 consecu tive DWORDs located DWORD aligned
in the shared memory. The first 3 DWORDs are read by the corresponding DMA channel
using a burst transaction and provide information about the next descriptor in the linked
list, the attached transmit data buffer and its size as well as some control bits.
The fourth DWORD is written by the DMA channel indicating that operation on this
descripto r is finis he d.
The CPU will write the address of the first descriptor of each linked list to a dedicated
Base Address Register during initialization procedure. The corresponding DMA
channels start operating the linked lists at these addresses.
Figure 77 Transmit Descriptor List Structure
11.1.1.1 Transmit Descriptor
The transmit descriptor lists are prepared by the host within the shared memory and read
by DSCC 4 DM A control ler, whe n requested to do by th e host either v ia an ’ AR’ (Actio n
Request) command or an transmit poll command or after branching from previous
transmit descriptors. The handling of transmit descriptor lists is described in details in
FE Hold HI NO
Next Transmit Descriptor Pointer
Transmit Data Pointer
0x0000000
FE Hold HI
31 0
byte3 byte2 byte1 byte0
byte11 byte10 byte9 byte8
byte7 byte6 byte5 byte4
byte15 byte14
byte19
31 0
Transmit Descriptor:
Transmit Data Buffer:
C0 00
0x0000
(dummy)
DWORD1
DWORD2
DWORD3
DWORD4
(DWORD5)
written by
CPU
written by
DSCC4
PEB 20534
Host Memory Organization
Semiconductor Group 375 Data Sheet 09.98
chapter “DMAC Transmit Descriptor Lists” on page 66. The transmit descriptor contains
4 DWORDs which are described in the following table:
Table 100 Transmit Descriptor
DWORD
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
0
FE
HOLD
HI
NO
1 Next Transmit D es cr ipt or Pointer
2 Transm it Dat a Pointer
30C00000000000000
DWORD
15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0
00000000000000000
1 Next Transmit Descriptor Pointer
2 Transmit Data Point er
30000000000000000
PEB 20534
Host Memory Organization
Semiconductor Group 376 Data Sheet 09.98
FE: Frame End, set by the host
It indicates that the current transmit data section (addressed by Transmit
Data Pointer) contains the end of a frame (HDLC, PPP) or the end of data
block (ASYNC, BISYNC). When transferring the last data from this transmit
data section into the internal FIFO the DMAC marks these data with an
’frame end / block end’ indication bit.
GMODE.CMOD=’0’:
After that it checks the HOLD bit stored in the on-chip memory. If HOLD=’0’,
it branches to the next transmit descriptor. Otherwise the corresponding
DMAC transmit channel is deactivated as long as the host CPU does not
request reactivation via the GCMDR register (either transmit poll request or
action request with ’IDT’ command).
GMODE.CMOD=’1’:
After that it checks if the first (current) transmit descriptor address (LTDA) is
equal to the last transmit descriptor address (LTDA) stored in the
corresponding channel specific on-chip registers. When both addresses
differ, it branches to the next transmit descriptor. Otherwise the
corresponding DMAC transmit channel is deactivated as long as the host
CPU does not write a new LTDA value to LTDA register or provides an
action request with ’IDT’ command.
HOLD: Hold (only valid when GMODE.CMODE=0)
It indicates whether the current descriptor is the last element of a linked list
or not:
HOLD=’0’: A next descriptor is available in the shared memory; after
checking the HOLD bit stored in the on-chip memory the
DMAC branches to next transmit descriptor
HOLD=’1’: The current descriptor is the last one that is available for the
DMAC. The corresponding DMAC channel is deactivated for
transmit direction as long as the microprocessor does not
request an activation via the CMDR register.
NO: Byte Number
This byte number defines the number of bytes stored in the data section to
be transmitted. Thus the maximum length of data buffer is 8191 bytes (i.e.
NO = 1FFFH). A transmit descriptor and the corresponding data section
must contain at least either one data byte or a frame end indication. Other-
wise an DMA controller interrupt with ’ERR’ bit set is generated.
PEB 20534
Host Memory Organization
Semiconductor Group 377 Data Sheet 09.98
HI: Host Initiated Interrupt
If the HI bit is set, the corresponding DMAC generates an interrupt with set
HI bit after transferring all data bytes of the current data section.
Next
Transmit
Descriptor
Pointer:
This 32-bit pointer contains the start address of the next transmit descrip-
tor. After sending the indicated number of data bytes, DSCC4 branches to
the next transmit descriptor to continue transmission. The transmit descrip-
tor is read entirely at the beginning of transmission and stored in on-chip
memory. Therefore all information in the next descriptor must be valid
when the DSCC4 branches to this descriptor.
This pointer is not used if a transmitter reset or initialization channel com-
mand is detected while the DSCC4 still reads data from the current trans-
mit descriptor. In this case BTDA value in the BTDA register is used as a
pointer for the next transmit descriptor to be branched to.
Transmit
Data
Pointer:
This 32-bit pointer contains the start address of the transmit data section.
Although DSCC4 works long word oriented, it is possible to begin transmit
data section at byte addresses.
C: Complete
This bit is set by the DSCC4 if
- it completes reading data section normally
- it was aborted by a transmitter reset command.
PEB 20534
Host Memory Organization
Semiconductor Group 378 Data Sheet 09.98
11.1.2 Receive Descriptor Lists
Each receive descriptor consists of 5 consecutive DWORDs located DWORD aligned in
the shared mem ory. The first 3 DWORDs are read by the correspon ding DMA channel
using a burst transaction and provide information about the next descriptor in the linked
list, the attached receive data buffer and its size as well as some control bits.
The fourth DWORD is written by the DMA channel indicating that operation on this
descripto r is finishe d. The fifth DWORD is also writ ten by the DMA channel but on ly in
descriptors containing the first or only data section of an HDLC frame or data block. It is
a pointer to the last descriptor containing th e frame or block end (’FE’ bit) allowing the
software to unchain the complete partial descriptor list containing one frame or block
without parsing through the list for ’FE’ indication.
The CPU will write the address of the first descriptor of each linked list to a dedicated
Base Address Register during initialization procedure. The corresponding DMA
channels start operating the linked lists at these addresses.
Figure 78 Receive Descriptor List Structure
0 Hold HI NO
Next Receive Descriptor Pointer
Receive Data Pointer
BNO
FE
Hold HI
31 0
byte3 byte2 byte1 byte0
byte11 byte10 byte9 byte8
byte7 byte6 byte5 byte4
byte15 byte14
byte19
31 0
Receive Descriptor:
Receive Data Buffer:
C 0
0x0000 0
0x00STATUS
Frame End Descriptor Pointer
DWORD1
DWORD2
DWORD3
DWORD4
(DWORD5)
written by
CPU
written by
DSCC4
PEB 20534
Host Memory Organization
Semiconductor Group 379 Data Sheet 09.98
11.1.2.1 Receive Descriptor
The receive descriptor lists are prepared by the host within the shared memory and read
by DSCC 4 DMA cont rolle r, when request ed to do by the host e ither via an ’AR’ (Act ion
Request) c ommand or a fter branchin g from previo us receive descriptors . The handl ing
of receive descriptor lists is described in details in chapter “DMAC Receive Descriptor
Lists” on page 71. The receive descriptor contains 5 DWORDs which are described in
the following table:
Table 101 Receive Descriptor
DWORD
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
00
HOLD
HI
NO
1 N ex t Re ce iv e De scriptor Pointer
2 Receive Data Po int er
3FEC0BNO
4 Frame End D es cr ipt or Pointer
DWORD
1514131211109876543210
00000000000000000
1 N ex t Re ce iv e De scriptor Pointe r
2 Receive Data Po int er
3 STATUS 00000000
4 Frame En d D es criptor Pointer
PEB 20534
Host Memory Organization
Semiconductor Group 380 Data Sheet 09.98
HOLD: Hold (only valid when GMODE.CMODE=0)
It indicates whether the current descriptor is the last element of a linked list
or not:
HOLD=’0’: A next descriptor is available in the shared memory; after
checking the HOLD bit stored in the on-chip memory the
DMAC branches to next receive descriptor
HOLD=’1’: The current descriptor is the last one that is available for the
DMAC. After completion of the current receive descriptor an
interrupt is generated and the corresponding DMAC channel
is deactivated for receive direction as long as the micropro-
cessor does not request an activation via the CMDR register.
HI: Host Initiated Interrupt
If the HI bit is set, the corresponding DMAC generates an interrupt with set
HI bit after transferring all data bytes into the current data section.
NO: Byte Number
This byte number defines the size of the receive data section allocated by
the host. It has to be a multiple of 4 bytes which is responsibility of the soft-
ware. The maximum buffer length is 8188 bytes (i.e. NO = 1FFCH).
Next
Receive
Descriptor
Pointer:
This 32-bit pointer contains the start address of the next receive descrip-
tor . After completion of the current receive descriptor the DSCC4 branches
to the next receive descriptor to continue reception. The receive descriptor
is read entirely at the beginning of reception and stored in on-chip
memory. Therefore all information in the next descriptor must be valid
when the DSCC4 branches to this descriptor.
This pointer is not used if a receiver reset command is detected while the
DSCC4 still writes data to the current receive descriptor. In this case
BRDA is used as a pointer for the next receive descriptor to be branched
to.
Receive
Data
Pointer:
This 32-bit pointer contains the start address of the receive data section.
The start address must be DWORD aligned.
PEB 20534
Host Memory Organization
Semiconductor Group 381 Data Sheet 09.98
FE: Frame End
It indicates that the current receive data section (addressed by Receive
Data Pointer) contains the end of a frame (HDLC, PPP) or the end of data
block (ASYNC, BISYNC, XTRANS). This bit is set by the DMAC after
transferring the last data from the internal FIFO (indicated by the END bit)
into the receive data section. Moreover the BNO and STATUS is updated
and the ’C’ bit is set by the DMAC.
GMODE.CMODE=’0’:
After that it checks the HOLD bit stored in the on-chip memory. If
HOLD=’0’, it branches to the next receive descriptor. Otherwise the corre-
sponding DMAC receive channel is deactivated as long as the host CPU
does not request reactivation via the GCMDR register (action request with
’IDR’ command).
GMODE.CMODE=’1’:
After that it checks if the first (current) receive descriptor address (LRDA) is
equal to the last receive descriptor address (LRDA) stored in the
corresponding channel specific on-chip registers. When both addresses
differ, it branches to the next receive descriptor. Otherwise the
corresponding DMAC receive channel is deactivated as long as the host
CPU does not write a new LRDA value to LRDA register or provides an
action request with ’IDR’ command.
C: Complete
This bit is set by the DSCC4 if
- it completed filling data section normally
- it was aborted by a receiver reset command
- end of frame (HDLC, PPP) or end of block (ASYNC, BISYNC, BTRANS)
was stored in the receive data section.
BNO: Byte Number of Received Data
DSCC4 writes the number of data bytes it has stored in the current data
secti on int o BNO
Frame
End
Descriptor
Pointer:
This 32-bit pointer is valid only in the descriptor, that contains the data
pointer to the first data section of an HDLC frame or
ASYNC/BISYNC/BTRANS block. This pointer is updated by the DSCC4
with the address of the descriptor that contains the data pointer to the last
section (FE) of the HDLC frame or ASYNC/BISYNC/BTRANS block.
PEB 20534
Host Memory Organization
Semiconductor Group 382 Data Sheet 09.98
Receive descriptor STATUS bit field:
76543210
000000RA 0
RA: Receive Abort
This bit indicates that the reception of a frame (HDLC, PPP) or block
(ASYNC, BISYNC) was ended by a DMA receiver reset command or by a
HOLD bit in the current receive descriptor or by a FRDA=LRDA.condition.
PEB 20534
Host Memory Organization
Semiconductor Group 383 Data Sheet 09.98
11.1.2.2 Receive Data Section Status Byte (HDLC Mode)
In HDLC protocol mode, the last byte of a frame (Receive Status Byte, RSTA) - located
in the data section - contains error indications caused by the SCC (e.g. CRC, receive
abort, …).
The contents of the RSTA byte relates to the received HDLC frame and is generated
when end-of-frame is recognized at the serial receive interface.
RSTA
76543210
VFR RDO CRC RAB HA1 HA0 C/R LA
PEB 20534
Host Memory Organization
Semiconductor Group 384 Data Sheet 09.98
VFR… Valid Frame
Determines whether a valid frame has been received.
1…valid
0…invalid
An invalid frame is either a frame which is not an integer number of 8 bits
(n * 8 bits) in length (e.g. 25 bits), or a frame which is too short taking into
account the operation mode selected via CCR1 (MDS1, MDS0, ADM)
and the selected CRC algorithm (CCR1:C32) as follows:
for CCR2:DRCRC = ’0’: (CCR2:RCRC has no affect)
• auto-/non-auto mode (16-bit address)
4 bytes (CRC-CCITT) or 6 (CRC-32)
• auto-/non-auto mode (8-bit address)
3 bytes (CRC-CCITT) or 5 (CRC-32)
• transparent mode 1:
3 bytes (CRC-CCITT) or 5 (CRC-32)
• transparent mode 0:
2 bytes (CRC-CCITT) or 4 (CRC-32)
Shorter frames are not reported anyway.
for CCR2:DRCRC = ’1’: (CCR2:RCRC has no affect)
• auto-/non-auto mode (16-bit address):
2bytes
• auto-/non-auto mode (8-bit address):
1byte
• transparent mode 1:
1byte
• transparent mode 0:
1byte
Shorter frames are not reported anyway.
RDO… Receive Data Overflow
A data overflow has occurred during reception of the frame. Additionally,
an interrupt can be generated (refer to ISR:RDO/IMR:RDO).
CRC… CRC Compare /Check
0…CRC check failed, received frame contains errors.
1…CRC check OK, received frame is error-free.
RAB… Receive Message Aborted
The received frame was aborted from the transmitting station. According
to the HDLC protocol, this frame must be discarded by the receiver
station.
PEB 20534
Host Memory Organization
Semiconductor Group 385 Data Sheet 09.98
HA1…
HA0… High Byte Address Compare
Significant only if 2-byte address mode has been selected. In operating
modes which provide high byte address recognition, the DSCC4
compares the high byte of a 2-byte address with the contents of two
individually programmable addresses (RADR:RAH1, RADR:RAH2) and
the fixed values FEH and FCH (broadcast address). Dependent on the
result of this comparison, the following bit combinations are possible:
10…RAH1 has been recognized.
00…RAH2 has been recognized.
01…broadcast address has been recognized.
If RAH1, RAH2 contain identical values, a match is indicated by ‘10’.
C/R… Command/Response
Significant only if 2-byte address mode has been selected. Value of the
C/R bit (bit in high address byte) in the received frame. The interpretation
depends on the setting of the CRI bit in the RADR register. Refer also to
the description of RADR register.
LA… Low Byte Address Compare
Not significant in transparent and extended transparent operating mode.
the below byte address of a 2-byte address field, or the single address
byte of a 1-byte address field is compared with two addresses
(RADR:RAL1, RADR:RAL 2).
0…RAL2 has been recognized.
1…RAL1 has been recognized.
According to the X.25 LAPB protocol, RAL1 is interpreted as the address
of a COMMAND frame and RAL2 is interpreted as the address of a
RESPONSE frame.
PEB 20534
Host Memory Organization
Semiconductor Group 386 Data Sheet 09.98
11.1.2.3 Receive Data Secti on Status Byte (ASYNC/BISYNC Modes)
In ASYNC/BISYNC protocol mode additionally to every data byte, an attached status
byte can be stored (CCR2:RFDF=’1’).
The data character and status character format is determined as follows:
Figure 79 ASYNC/BISYNC Receive Status Character Format
Note: The ’Frame Error’ (FE) status bit is only valid in ASYNC protocol mode.
Char5
P
data byte:
0457
Char6
P
0
657
Char7
P
067
Char8
07
RFDF=’0’
(no parity bit stored)
Char5
P
data byte (d):
0457
Char6
P
0
657
Char7
P
067
Char8
07
RFDF=’1’
(no parity bit stored)
status byte (s):
PE
067
FE P
PE
067
FE P
PE
067
FE P
PE
0
6
7
FE P
P
: Parity bit stored in data byte, can be disabled via bit ’DPS’
PE: Parity Error
FE: Frame Error
P: Parity bit stored in second data byte (= status byte)
PEB 20534
Host Memory Organization
Semiconductor Group 387 Data Sheet 09.98
11.2 Interrupt Qu eue Struct ure
11.2 .1 Interr upt Queue O ve rview
The DSCC4 interrupt concept is based on 32-bit interrupt vectors generated by the
different blocks. Interrupt vectors are stored in a central interrupt FIFO which is 16
DWORDs deep. The interrupt controller transfers available vectors in one of ten circular
interrupt queues located in the shared memory. Each queue is dedicated to the interrupt
source.
In addit ion new in terrupt vectors are indicate d in the glo bal stat us register GSTAR on a
per queue basis and selectively confirmed by writing ’1’ to the corresponding GSTAR bit
positions. The PCI interrupt signal INTA is asserted with any new interrupt event and
remains asserted until all events are confirmed.
(For more detailed information refer to chapter “DMAC Interrupt Controller” on page 81.)
PEB 20534
Host Memory Organization
Semiconductor Group 388 Data Sheet 09.98
Figure 80 DSCC4 Logical Interrupt Structure
SCC0
receive
interrupts transmit
interrupts
IQSCC0RX
IQSCC0TX
Peripherals
(SSC, GPP,
LBI)
IQCFG
IQP
DMA
Controller
Logic
HOST Memory interrupt queues
DSCC4 interrupt structure block diagram
internal interrupt bus
16
DWORD
central
interrupt
FIFO GSTAR register
INTA
signal
PEB 20534
Host Memory Organization
Semiconductor Group 389 Data Sheet 09.98
11.2.2 Interrupt Vector Overview
Figure 81provides an overview about all DSCC4 interrupt vectors. The different interrupt
sources (types) are distinguished by the most significant 8 bit of the 32-bit interrupt
vector. The dedicated host memory interrupt queues, the vectors are transferred to, are
referenced in brackets under each interrupt vector name:
Figure 81 Interrupt Vector Overview
1
CFG IV
(IQCFG)
0x00h
31 24 23 16 15 8 7 0
0 000000
0
DMAC IV
(IQSCCi)
0x0000hCh ID
0
1
000000
0010 000 000000
ARF
ARACK
HI
FI
ERR
0
SCC IV
(IQSCCi)
Interrupt Status Register (ISR) bit field 18..0Ch ID
0
0
000000
SSC IV
(IQP)
D/E=1: SSCCON Register (Status) bit field 16..0
D/T=0, R/T=1: Receive Buffer Register SSCRB bit field 16..0
D/T=0, R/T=0: 0000h
00 00001100
R/T
D/E
ERR
RXI
TXI
LBI IV
(IQP)
00 00001 1 0 1 0 0 0x0000h000
GPP IV
(IQP)
00 00001 1 1 1 0 0 GPP Data Register (GPDATA) bit field 15..0000
Source Coding
Ch ID (receive):
000: SCC0RX
001: SCC1RX
010: SCC2RX
011: SCC3RX
Ch ID (transmit):
100: SCC0TX
101: SCC1TX
110: SCC2TX
111: SCC3TX
PEB 20534
Host Memory Organization
Semiconductor Group 390 Data Sheet 09.98
11.2.2.1 Configuration Interrup t Vector
Configuration interrupt vectors are transferred to Configuration Interrupt Queue ’IQCFG’.
Table 102 CFGIV: Configuration Interrupt Vectori
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
Source Id=1010 000000000000
1514131211109876543210
00000000000000
ARF
ARACK
Source ID 10102Configuration Interrupt Vector (IQCFG)
ARF Action Request Failed Interrupt
This bit indicates that an action request command was completed with
an ’action request failed’ condition:
ARF=’0’ No action request was performed or no ’action request
failed’ condition occured completing an action request.
ARF=’1’ The last action request command was completed with an
’action request failed’ condition.
ARACK Action Request Acknowledge Interrupt
This bit indicates that an action request command was completed
successfully:
ARACK=’0’ No action request was performed or completed
successfully.
ARACK=’1’ The last action request command was completed
successfully.
PEB 20534
Host Memory Organization
Semiconductor Group 391 Data Sheet 09.98
11.2.2.2 DMA Controller Interrupt Vector
DMA controller interrupt vectors are transferred to the corresponding channel and
direction specific interrupt queues IQSCCiRX and IQSCCiTX respectively.
Table 103 DMA Controller Interrupt Vectori
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
0Source ID 0 0 0000000HIFIERR
1514131211109876543210
0000000000000000
Source ID 0002Receive Channel 0 Interrupt Vector (IQSCC0RX)
0012Receive Channel 1 Interrupt Vector (IQSCC1RX)
0102Receive Channel 2 Interrupt Vector (IQSCC2RX)
0112Receive Channel 3 Interrupt Vector (IQSCC3RX)
1002Transmit Channel 0 Interrupt Vector (IQSCC0TX)
1012Transmit Channel 1 Interrupt Vector (IQSCC1TX)
1102Transmit Channel 2 Interrupt Vector (IQSCC2TX)
1112Transmit Channel 3 Interrupt Vector (IQSCC3TX)
HI Host Initiate d interrupt (Rx/Tx Channel)
This bit indicates that an Host Initiated (HI) interrupt occured, i.e. the
corresponding DMA controller channel detects the ’HI’ bit set to ’1’ in the
receive or transmit descriptor before branching to the next descriptor.
HI=’0’ No Host Initiated (HI) interrupt is indicated by this vector.
HI=’1’ An Host Initiated (HI) interrupt is indicated by this vector.
PEB 20534
Host Memory Organization
Semiconductor Group 392 Data Sheet 09.98
FI Frame Indication interrupt (Rx/Tx Channel)
This bit indicates that an Frame Indication (FI) interrupt occured.
Receive direction:
FI=’1’ indicates, that a frame has been received completely or was
stopped by a DMAC receiver reset command or a hold condition set in a
receive descriptor. It is set when the DSCC4 branches from the last
descriptor belonging to the current frame (or block) (FE=’1’) to the first
descriptor of a new frame. It is also set when the descriptor in which the
frame/block is finished contained a hold condition.
Transmit direction:
Issued if the ’FE’ bit is detected in the transmit descriptor. It is set when
the DSCC4 branches to the next transmit descriptor, belonging to a new
frame or when ’HOLD’ bit is set in conjunction with ’FE’ bit. Only ’ERR’
indication without ’FI’ is set, if a transmit descriptor contains a ’HOLD’
(hold condition) but no ’FE’ bit.
FI=’0’ No Frame Indication (FI) interrupt is indicated by this
vector.
FI=’1’ An Frame Indication (FI) interrupt is indicated by this
vector.
ERR ERROR Indication interrupt (Rx/Tx Channel)
This bit indicates that an Error interrupt occured.
Receive direction:
Issued if the current frame/block could not be transferred to the shared
memory completely, because of a hold condition in a receive descriptor
not providing enough bytes for the frame/block or the frame/block was
aborted by a DMAC receiver reset command.
Transmit direction:
Issued if a transmit descriptor contains a hold condition but FE=’0’ or if
the last descriptor had NO=0 and FE=’0’.
ERR=’0’ No Error (ERR) interrupt is indicated by this vector.
ERR=’1’ An Error (ERR) interrupt is indicated by this vector.
PEB 20534
Host Memory Organization
Semiconductor Group 393 Data Sheet 09.98
11.2.2.3 SCC Interrupt Vector
Serial Channel (SCC) related interrupt vectors are transferred to the corresponding
channel and direction specific interrupt queues IQSCCiRX and IQSCCiTX respectively.
Note: Interrupt vectors generated by the SCCs might contain interrupt indications for
both, receive AND transmit direction. But in receive interrupt queues only the
receive interrupt indications need to be served and in transmit interrupt queues
only transmit interrupt indications need to be served by the software.
Table 104 SCC Interrupt Vectori
Mode
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
H0Source ID 0 0 1000000
ALLS
1
XDU
A0Source ID 0 0 1000000
ALLS
1
XOFF
B0Source ID 0 0 1000000
ALLS
1
XDU
Mode
1514131211109876543210
H
TIN
CSC
XMR
XPR
1111
RDO
RFS
RSC
PCE
PLLA
CDSC
RFO
FLEX
A
TIN
CSC
XON
XPR
11
BRK
BRKT
TCD
TIME
PERR
FERR
PLLA
CDSC
RFO
1
B
TIN
CSC
XMR
XPR
1111
TCD
1
PERR
SCD
PLLA
CDSC
RFO
1
PEB 20534
Host Memory Organization
Semiconductor Group 394 Data Sheet 09.98
Source ID 0002Receive Channel 0 Interrupt Vector (IQSCC0RX)
0012Receive Channel 1 Interrupt Vector (IQSCC1RX)
0102Receive Channel 2 Interrupt Vector (IQSCC2RX)
0112Receive Channel 3 Interrupt Vector (IQSCC3RX)
1002Transmit Channel 0 Interrupt Vector (IQSCC0TX)
1012Transmit Channel 1 Interrupt Vector (IQSCC1TX)
1102Transmit Channel 2 Interrupt Vector (IQSCC2TX)
1112Transmit Channel 3 Interrupt Vector (IQSCC3TX)
Bit field 18..0 of the SCC interrupt vector is a copy of the SCC Interrupt Status Register
ISR. The meaning of the bit field depends on the selected protocol mode (HDLC (H),
ASYNC (A), BISYNC (B)).
(For detailed information on Interrupt Status Register ISR refer to Table 87, “ISR:
Interrupt Status Register,” on page 342.)
PEB 20534
Host Memory Organization
Semiconductor Group 395 Data Sheet 09.98
11.2.2.4 SSC Interrupt Vector
SSC interrupt vectors are transferred to the peripheral interrupt queue IQP.
Table 105 SSC Interrupt Vectori
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
1100000R/TD/E0000ERRRXTX
1514131211109876543210
ISW(15:0)
(Source ID)
bit field
31..28
11002SSC Interrupt Vector (IQP)
R/T Receive/ Tra n smi t Indicator
This bit indicates whether the SSC vector is generated for transmit or
receive interrupt events:
R/T=’0’ SSC transmit interrupt vector.
R/T=’1’ SSC receive interrupt vector.
D/E Data/Error Indicator
This bit indicates whether the SSC vector is generated for data
(transmit/receive) or error interrupt events:
D/E=’0’ SSC error interrupt vector.
D/E=’1’ SSC data interrupt vector.
ERR Error interrupt
This bit indicates that an error interrupt occured:
ERR=’0’ No error interrupt is indicated by this vector.
ERR=’1’ An error interrupt is indicated by this vector.
Bit field 15..0 contains the corresponding bit field 15..0 of
register SSCCON (operation mode).
PEB 20534
Host Memory Organization
Semiconductor Group 396 Data Sheet 09.98
RX Receive interrupt
This bit indicates that a receive interrupt occured:
RX=’0’ No receive interrupt is indicated by this vector.
RX=’1’ A receive interrupt is indicated by this vector.
Bit field 15..0 contains the corresponding bit field 15..0 of
register SSCRB (receive buffer containing the received
data bits).
TX Transmit interrupt
This bit indicates that a transmit interrupt occured:
TX=’0’ No transmit interrupt is indicated by this vector.
TX=’1’ A transmit interrupt is indicated by this vector.
Bit field 15..0 contains a constant zero value. This
interrupt means, that the transmit buffer can be reloaded
with new transmit data (register SSCTB).
ISW(15:0) Interrupt Status Word
The contents of this bit field depends on the SSC interrupt type:
Type: Meaning:
error int.
(D/E=’0’) ISW(15:0) = SSCCON(15:0)
(Register SSCCON in operational mode.)
receive int.
(D/E=’1’,
R/T=’1’)
ISW(15:0) = SSCRB(15:0)
(Register SSCRB.)
transm. int.
(D/E=’1’,
R/T=’0’)
ISW(15:0) = 0000H
PEB 20534
Host Memory Organization
Semiconductor Group 397 Data Sheet 09.98
11.2.2.5 LBI Interrupt Vector
LBI interrupt vectors are transferred to the peripheral interrupt queue IQP.
Table 106 SSC Interrupt Vectori
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
1101000000000000
1514131211109876543210
0000000000000000
(Source ID)
bit field
31..28
11012LBI Interrupt Vector (IQP)
This interrupt vector (with no additional bit information) is generated, if a state transition
inactive to active is detected at LBI interrupt input signal LINTI1.
The polarity (high or low active) of input signal LINTI1 is determined by bit ’LINTIC’ in
register LCONF.
PEB 20534
Host Memory Organization
Semiconductor Group 398 Data Sheet 09.98
11.2.2.6 GPP Interrupt Vector
GPP interrupt vectors are transferred to the peripheral interrupt queue IQP.
Table 107 GPP Interrupt Vectori
31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16
1111000000000000
1514131211109876543210
GPDATA(15:0)
(Source ID)
bit field
31..28
11112GPP Interrupt Vector (IQP)
GPDATA
(15:0) General Purpose Port Status Information
This bit field 15..0 contains the corresponding bit field 15..0 of general
purpose data register GPD ATA.
Whenever a transition on at least one general purpose pin is detected, a
GPP interrupt vector is generated and the current status of all GPP pins
can be evaluated in bit field ’GPDATA’ of the interrupt vector.
PEB 20534
Test Configuration
Semiconductor Group 399 Data Sheet 09.98
12 Test Configuration
12.1 JTAG Boundary Scan Interface
In the DSCC4 a Test Access Port (TAP) controller is implemented. The essential part of
the TAP is a finite s tate mac hine (16 states) c ontrolli ng the di fferent op erational modes
of the boundary s can. Both, TAP controll er and boundary sc an, meet the requireme nts
given by the JTAG standard: IEEE 1149.1. Figure 82 gives an overview about the TAP
controller.
Figure 82 Block Diagram of Test Access Port and Boundary Scan Unit
If no boundary scan operation is planned TRST has to be connected with VSS. TMS and
TDI do not need to be connected since pull-up transistors ensure high input levels in this
case. Nevertheless it would be a good practice to put the unused inputs to defined levels.
In this case, if the JTAG is not used:
TMS = TCK = ’1’ is recomme nde d.
Test handling (boundary scan operation) is performed via the pins TCK (Test Clock),
TMS (Test Mode Select), TD I (Test Data Input) and TDO (T est Data Output) when the
TAP controller is not in its reset state, i.e. TRST is connected to VDD or it remains
unconnecte d due to its interna l pull-up. Test data at TDI are loaded w ith a 4-MHz clo ck
Clock Generation
Test Access Port (TAP)
TAP Controller
- Finite State Machine
- Instruction Register (3 bit)
- Test Signal Generator
CLOCK
TCK
TRST
TMS
Reset
Data in
TDI
Test
Control
TDO
Enable
Data out
CLOCK
BS Data IN
Identific ation Scan (32 bit)
Boundary Sc an (n bit)
6
Control
Bus
ID Data out
SS Data
out n
.
.
.
.
.
.
1
2
Pins
PEB 20534
Test Configuration
Semiconductor Group 400 Data Sheet 09.98
signal c onne cte d to TCK. ‘1’ or ‘0’ on TM S cau ses a tra nsi tion from one controller state
to another; constant ’1’ on TMS leads to normal operation of the chip.
Table 108 Boundary Scan Sequence of the DSCC4
TDI ->
Seq.
No. Pin I/O Number of
Boundary Scan Cells Constant Value
In, Out, Enable
1INTA I/O 3 001
2 AD0 I/O 3 000
3 AD1 I/O 3 000
4 AD2 I/O 3 000
5 AD3 I/O 3 001
6 AD4 I/O 3 101
7 AD5 I/O 3 100
8 AD6 I/O 3 000
9 C/BE0 I/O 3 100
10 AD7 I/O 3 000
11 AD8 I/O 3 110
12 AD9 I/O 3 000
13 AD10 I/O 3 000
14 AD11 I/O 3 000
15 AD12 I/O 3 000
16 AD13 I/O 3 000
17 AD14 I/O 3 000
18 AD15 I/O 3 000
19 C/BE1 I/O 3 000
20 PAR I/O 3 000
21 SERR I/O 3 000
22 PERR I/O 3 000
23 STOP I/O 3 000
24 DEVSEL I/O 3 000
25 TRDY I/O 3 000
26 IRDY I/O 3 000
PEB 20534
Test Configuration
Semiconductor Group 401 Data Sheet 09.98
27 FRAME I/O 3 000
28 C/BE2 I/O 3 000
29 AD16 I/O 3 000
30 AD17 I/O 3 000
31 AD18 I/O 3 000
32 AD19 I/O 3 000
33 AD20 I/O 3 000
34 AD21 I/O 3 000
35 AD22 I/O 3 000
36 AD23 I/O 3 000
37 IDSEL I/O 3 000
38 C/BE3 I/O 3 000
39 AD24 I/O 3 000
40 AD25 I/O 3 000
41 AD26 I/O 3 000
42 AD27 I/O 3 000
43 AD28 I/O 3 000
44 AD29 I/O 3 000
45 AD30 I/O 3 000
46 AD31 I/O 3 000
47 REQ I/O 3 000
48 GNT I/O 3 000
49 CLK I 1 0
50 RST I/O 3 000
51 W/R I/O 3 000
52 DEMUX I/O 3 000
53 RTS3 I/O 3 000
54 CD3 I/O 3 000
55 CTS3 I/O 3 000
56 TXD3 I/O 3 000
Seq.
No. Pin I/O Number of
Boundary Scan Cells Constant Value
In, Out, Enable
PEB 20534
Test Configuration
Semiconductor Group 402 Data Sheet 09.98
57 RXD3 I/O 3 000
58 TXCLK3 I/O 3 000
59 RXCLK3 I/O 3 000
60 RTS2 I/O 3 000
61 CD2 I/O 3 000
62 CTS2 I/O 3 000
63 TXD2 I/O 3 000
64 RXD2 I/O 3 000
65 TXCLK2 I/O 3 000
66 RXCLK2 I/O 3 000
67 LALE I/O 3 000
68 LINTO I/O 3 000
69 LINTI2 I/O 3 000
70 LINTI1 I/O 3 000
71 LRDY I/O 3 000
72 LBHE I/O 3 000
73 LWR I/O 3 000
74 LRD I/O 3 000
75 LCSI I/O 3 000
76 LCSO I/O 3 000
77 LBREQ I/O 3 000
78 LHLDA I/O 3 000
79 LHOLD I/O 3 000
80 LD0 I/O 3 000
81 LD1 I/O 3 000
82 LD2 I/O 3 000
83 LD3 I/O 3 000
84 LD4 I/O 3 000
85 LD5 I/O 3 000
86 LD6 I/O 3 000
Seq.
No. Pin I/O Number of
Boundary Scan Cells Constant Value
In, Out, Enable
PEB 20534
Test Configuration
Semiconductor Group 403 Data Sheet 09.98
87 LD7 I/O 3 000
88 LD8 I/O 3 000
89 LD9 I/O 3 000
90 LD10 I/O 3 000
91 LD11 I/O 3 000
92 LD12 I/O 3 000
93 LD13 I/O 3 000
94 LD14 I/O 3 000
95 LD15 I/O 3 000
96 LA0 I/O 3 000
97 LA1 I/O 3 000
98 LA2 I/O 3 000
99 LA3 I/O 3 000
100 LA4 I/O 3 000
101 LA5 I/O 3 000
102 LA6 I/O 3 000
103 LA7 I/O 3 000
104 LA8 I/O 3 000
105 LA9 I/O 3 000
106 LA10 I/O 3 000
107 LA11 I/O 3 000
108 LA12 I/O 3 000
109 LA13 I/O 3 000
110 LA14 I/O 3 000
111 LA15 I/O 3 000
112 RTS1 I/O 3 000
113 CD1 I/O 3 000
114 CTS1 I/O 3 000
115 TXD1 I/O 3 000
116 RXD1 I/O 3 000
Seq.
No. Pin I/O Number of
Boundary Scan Cells Constant Value
In, Out, Enable
PEB 20534
Test Configuration
Semiconductor Group 404 Data Sheet 09.98
-> TDO
An input p in (I) uses one bounda ry scan cell (data in), a n output pi n (O) uses two cells
(data out, ena ble ) an d a n I/ O-pi n (I/ O) use s t hree ce lls (dat a i n, d ata ou t, e nabl e). No te
that some functional output and input pins of the DSCC4 are tested as I/O pins in
boundary scan, hence using three cells. The boundary scan unit of the DSCC4 contains
a total of n = 374 scan cells.
The right column of table 108 gives the initialization values of the cells.
The desired test mode is selected by serially loading a 3-bit instruction code into the
instruction register via TDI (LSB first); see table 109.
EXTEST is used to examine the interconnection of the devices on the board. In this test
mode at first all input pins capture the current level on the corresponding external
interconnection line, whereas all output pins are held at constant values (‘0’ or ‘1’,
117 TXCLK1 I/O 3 000
118 RXCLK1 I/O 3 000
119 TEST1 I1 0
120 RTS0 I/O 3 000
121 CD0 I/O 3 000
122 CTS0 I/O 3 000
123 TXD0 I/O 3 000
124 RXD0 I/O 3 000
125 TXCLK0 I/O 3 000
126 RXCLK0 I/O 3 000
Table 109 Boundary Scan Test Modes
Instruction (Bit 2 … 0) Test Mode
000
001
010
011
111
others
EXTEST (external testing)
INTEST (internal testing)
SAMPLE/PRELOAD (snap-shot testing)
IDCODE (reading ID cod e)
BYPASS (bypass operation)
handled like BYPASS
Seq.
No. Pin I/O Number of
Boundary Scan Cells Constant Value
In, Out, Enable
PEB 20534
Test Configuration
Semiconductor Group 405 Data Sheet 09.98
accordi ng to table 108). The n the conte nts of the bound ary scan i s shifted to TDO. At
the same time the next scan vector is loaded from TDI. Subsequently all output pins are
updated acco rding to the new bo und ary sc an c ontents and al l inpu t pins aga in capture
the current external level afterwards, and so on.
INTEST supports internal testing of the chip, i.e. the output pins capture the current level
on the corresponding internal line whereas all input pins are held on constant values (‘0’
or ‘1’, according to table 108). The resulting boundary scan vector is shifted to TDO.
The next test vector is serially loaded via TDI. Then all input pins are updated for the
following test cycle.
Note: In capture IR-state the code ‘001’ is automatically loaded into the instruction
register, i.e. if INTEST is wanted the shift IR-state does not need to be passed.
SAMPLE/PRELOAD is a test mode which provides a snap-shot of pin levels during
normal operation.
IDCODE: A 32-bit identification register is serially read out via TDO. It contains the
version num ber (4 b its), the device code (16 bits) and the manufacturer c ode (11 b its).
The LSB is fixed to ‘1’.
Note: Since in test logic reset state the code ‘011’ is automatically loaded into the
instructi on register, the ID code can easily be read out in shift DR state which is
reached by TMS = 0, 1, 0, 0.
BYPASS: A bit entering TDI is shifted to TDO after one TCK clock cycle.
TDI -> 0010 0000 0000 0011 0110 0000 1000 001 1 -> TDO
PEB 20534
Application Hints
Semiconductor Group 406 Data Sheet 09.98
13 Application Hints
13.1 DSCC4 PCI Bus Utilization
This chapter will be added in next version of the Data Sheet.
PEB 20534
Application Hints
Semiconductor Group 407 Data Sheet 09.98
13.2 Application Example: De-multiplexed Bus Interface
This chapter will be added in next version of the Data Sheet.
PEB 20534
Application Hints
Semiconductor Group 408 Data Sheet 09.98
13.3 SSC Example: Connecting A Serial EEPROM
This chapter will be added in next version of the Data Sheet.
PEB 20534
Electrical Characteristics
Semiconductor Group 409 Data Sheet 09.98
14 Electrical Characteristics
14.1 Im por tan t Electrical Requireme nts
VDD3 = 3.3 V ± 0.3 V VDD3 max = 3.6 V
VDD5 = 5.0 V ± 0.25 V VDD5 max = 5.25 V
During all DSCC4 power-up and power-down situations the difference |VDD5 –VDD3|
may not exceed 3.6V. The absolute maximums of VDD5 and VDD3 should never be
exceeded.
Figure 83 shows that both VDD3 and VDD5 can take on any time sequence, not exceeding
the maximum slew rate for VDD3 mentioned later and not exceeding a voltage difference
of 3.6V, for up to 50 milliseconds at power-up and power-down.Within 50 milliseconds of
power-up the voltages must be within their respective absolute voltage limits. At power-
down, within 50 milliseconds of either voltage going outside its operational range, the
voltage difference should not exceed 3.6V and both voltages must be returned below
0.1V:
Figure 83 Power-up and Power-down scenarios
3.3V
+/- 0.3V
5V
+/- 0.25V
U/V
50
power up
V
DD3
limit
V
DD5
limit
t/ms
N
power down
V
DD3
limit
V
DD5
limit
N+50
0
0.1V
Within the grey box es any shape of VDD3 and VDD5 signal is allowed with the requirement s that the absolute
limits of each signal are not exceeded, the slew rate recommendation for VDD3 is met to guarantee proper
boundary scan reset and the volt age difference do es not ex ce ed 3. 6V.
Outside the grey box es t he v olt ages provided to VDD3 and VDD5 should be inside the normal operation range.
In this p ower-up example V DD5 is enabled after VDD3 reached it s minimum operation value which is a typical
implementation.
For power-down VDD5 is switche d of f be fo re VDD3.
PEB 20534
Electrical Characteristics
Semiconductor Group 410 Data Sheet 09.98
Similar criteria also apply to power down in case of power failure situations:
Figure 84 Power-Failure scenarios
Additional recommendations:
The pin TEST1 has to be tied to VSS (refer to pin description table).
3.3V
+/- 0.3V
5V
+/- 0.25V
U/V
50 t/ms
N
power failure:
V
DD3
break down
V
DD3
limit
V
DD5
limit
N+15
0
3.3V
+/- 0.3V
5V
+/- 0.25V
U/V
t/ms
N
power failure:
V
DD5
break down
V
DD3
limit
V
DD5
limit
0
0.1V 0.1V
Within t he grey boxes a ny sha pe of VDD3 and VDD5 s ignal is allowed with the requirem ents tha t the absol ute
limits of eac h si gnal are not exceed ed and the specified voltage differences are not exc eeded.
a. In case of VDD5 break-dow n the 3.6V difference is not exce eded anyway . T he voltages m ust return below
0.1V within 50 milliseconds.
b. In case of V DD3 break -dow n the max imum vo ltage di ffere nce mus t not excee d 4.5 V fo r a maxim um of 15
millisecon ds . Th e v olt ages must return below 0.1V within 50 millisecond s .
This scenario is allowed for 2000 power failure cycles.
PEB 20534
Electrical Characteristics
Semiconductor Group 411 Data Sheet 09.98
14.2 Absolute Maximum Ratings
Note: Stresses above those listed here may cause permanent damage to the device.
Exposure to absolute maximum rating conditions for extended periods may affect
devi ce reliability.
14.3 Thermal Package Characteristics
Table 110 Absolute Maximum Ratings
Parameter Symbol Limit Values Unit
min. max.
Ambient temperature under bias TA070°C
Junction temperature under bias TJ125 °C
Storage temperature Tstg – 65 125 °C
Voltage at any pin with respect to ground VS– 0.4 VDD5 + 0.4 V
Table 111 Thermal Package Characteristics
Parameter Symbol Value Unit
Thermal Package Resistance Junction to Ambient
Airflow: Ambient Temperature:
without airflow TA=-40°C θJA(0,-40) 42.3 °C/K
without airflow TA=+25°C θJA(0,25) 37.2 °C/K
airflow 1 m/s TA=+25°C θJA(1,25) 34.9 °C/K
airflow 2 m/s TA=+25°C θJA(2,25) 33.3 °C/K
airflow 3 m/s TA=+25°C θJA(3,25) 32.2 °C/K
PEB 20534
Electrical Characteristics
Semiconductor Group 412 Data Sheet 09.98
14.4 DC Characteristics
a) Non-PCI Interface Pins and Power Supply Pins
Note: 1. The listed characteristics are ensured over the operating range of the
integrated circuit. Typical characteristics specify mean values expected over
the product ion spread. If not otherwise specified, typical characteristics apply
at
TA
=25
°
C and the given supply voltage.
Note: 2. The electrical characteristics described in section 14.2 also apply here!
Table 112 DC Characteristics (Non-PCI Interface Pins and Power Supply Pins)
TA = 0 to + 70 °C; VDD5 = 5 V ± 5 %, VDD3 = 3.3 V ± 0.3 V, VSS = 0 V
Parameter Symbol Limit Values Unit Test Condition
min. max.
L-input voltage VIL –0.4 0.8 V
H-input voltage VIH 2.0 VDD5 + 0.4 V
L-output voltage VQL 0.45 V IQL = 7 mA
(pin TXD)
IQL = 2 mA
(all others / non-PCI)
H-output voltage VQH 2.4 VDD3 VIQH
=
– 400 µA
Power
supply
current
VDD3
operational ICC3 < 300 mA inputs at VSS/VDD,
no output loads
power down
(no clocks) ICC3 < 2 mA
Power supply current
VDD5 ICC5 < 2 mA
Power dissipation P< 1090 mW
typical
values operational
current ICC3 typ. < 250 mA VDD3 = 3.3V,
inputs at VSS/VDD,
no output loads
power
dissipation
P
typ.
< 830 mW
Input leakage current
Output leakage current ILI
ILQ
1µA0 V < VIN < VDD to 0 V
0 V < VOUT < VDD to 0 V
(pins with internal pull-
ups excluded)
PEB 20534
Electrical Characteristics
Semiconductor Group 413 Data Sheet 09.98
b) PCI Pins
According to the PCI specification V2.1 from June 1, 1995
(Chapter 4.2.1: Electrical DC Specifications for 5 V signaling)
.
14.5 Capacitances
a) Non-PCI Interface Pins
b) PCI Pins
According to the PCI specification V2.1 from June 1, 1995 (Chapter 4: Electrical
Specification for 5 V signalling)
Table 113 DC Characteristics PCI Interface Pins
TA = 0 to + 70 °C; VDD5 = 5 V ± 5 %, VDD3 = 3.3 V ± 0.3 V, VSS = 0 V
Parameter Symbol Limit Values Unit Test Condition
min. max.
L-input voltage VIL –0.5 0.8 V
H-input voltage VIH 2.0 VDD5 + 0.5 V
L-output voltage VQL 0.45 V IQL = 3mA
H- output voltage VQH 2.4 V IQH
=
– 2 mA
Table 114 Capacitances (Non-PCI Interface Pins)
TA = 25 °C; VDD5 = 5 V ± 5%, VDD3 = 3.3 V ± 0.3 V, VSS = 0 V
Parameter Symbol Limit Values Unit Test Condition
min. max.
Input capacitance CIN 15pF
Output cap aci tanc e COUT 510pF
I/O-capacitance CIO 615pF
PEB 20534
Electrical Characteristics
Semiconductor Group 414 Data Sheet 09.98
14.6 AC Characteristics
a) Non-PCI Interface Pins
TA = 0 to + 70 °C; VDD5 = 5 V ± 5%; VDD3 = 3.3 V ± 0.3 V
Inputs are driven to 2.4 V for a logical “1” and to 0.4 V for a logical “0”. Timing
measurements are made at 2.0 V for a logical “1” and at 0.8 V for a logical “0”.
The A C testing in put/ou tput waveforms are shown be low.
Figure 85 Input/Output Waveform for AC Tests
b) PCI Pins
According to the PCI specification V2.1 from June 1, 1995 (Chapter 4: Electrical
Specification for 5 V signalling)
14.6.1 PCI Bus Interface Timing
The A C testing in put/ou tput waveforms are shown in figures 86 and 87 below.
Figure 86 PCI Output Timing Measurement Waveforms
ITS09800
= 50 pF
Load
C
Test
Under
Device
0.45
2.4 2.0
0.80.8
2.0 Test Points
ITS09801
Clock
V
test
t
val
V
test
Output Delay Device
Under
Test
V
th
V
tl
TRI-STATE
Output
V
test
V
test
t
on
t
off
C
Load
= 50 pF
PEB 20534
Electrical Characteristics
Semiconductor Group 415 Data Sheet 09.98
Figure 87 PCI Input Timing Measurement Waveforms
The timings below show the basic read and write transaction between an initiator
(Master) and a target (Slave) device. The DSCC4 is able to work both as master and
slave.
As a master the DSCC4 reads/writes data from/to host memory using DMA and burst.
The slave mode is used by an CPU to access the DSCC4 PCI Configuration Space, the
on-chip registers and to access peripherals connected to the DSCC4 Local Bus Interface
(LBI).
14.6.1.1 PCI Read Transaction
The transaction starts with an address phase which occurs during the first cycle when
FRAME is activated (clock 2 in figure 88). During this phase the bus master (initiator)
outputs a val id address on AD(31:0) and a valid bus command o n C/BE(3:0). The first
clock of the first da ta phase is cloc k 3. Duri ng the dat a phase C/BE i ndic ate whic h b yte
lanes on AD(31:0) are involved in the current data phase.
The first data phase on a read transaction requires a turn-around cycle. In figure 88 the
addre ss is valid on c lock 2 a nd then the m ast er stop s dr iving AD. The ta rget drives the
AD lines following the turnaround when DEVSEL is asserted. (TRDY cannot be driven
until DEVSEL is asserted.) The earliest the target can provide valid data is clock 4. Once
enabled, the AD output buffers of the target stay enabled through the end of the
transaction.
Table 115 PCI Input and Output Measurement Conditions
Symbol Value Unit
Vth 2.4 V
Vtl 0.4 V
Vtest 1.5 V
Vmax 2.0 V
PEB 20534
Electrical Characteristics
Semiconductor Group 416 Data Sheet 09.98
A data phase may consist of a data transfer and wait cycles. A data phase completes
when data is transferred, which occurs when both IRDY and TRDY are asserted. When
either is dea sserted a wait cyc le is inserted. In the exam ple below, data is su ccessfull y
transferred on clocks 4, 6 and 8, a nd w ait cyc les are in serte d on cloc ks 3, 5 a nd 7. Th e
first data phase completes in the minimum time for a read transaction. The second data
phase is extended on clock 5 because TRDY is deasserted. The last data phase is
extended because IRDY is deasserted on clock 7.
The Master knows at clock 7 that the next data phase is the last. However, the master is
not ready to complete the last transfer, so IRDY is d eassert ed on cl ock 7, and FR AME
stays assert ed. On ly when IR DY is assert ed can FRAM E be dea sserted, whic h occurs
on clock 8.
Figure 88 PCI Read Transaction
ITD07575
123456789
BE’s
CLK
FRAME
AD
C/BE
IRDY
TRDY
DEVSEL
Bus Transaction
Data 1 Data 2 Data 3
Data Transfer
Wait
Wait
Data Transfer
Phase
Data
Phase
Address
Bus CMD
Address
Wait
Data Transfer
Data
Phase Data
Phase
PEB 20534
Electrical Characteristics
Semiconductor Group 417 Data Sheet 09.98
14.6.1.2 PCI Write Transaction
The transaction starts when FRAME is activated (clock 2 in figure 89). A write
transaction is similar to a read transaction except no turnaround cycle is required
following the address phase. In the example, the first and second data phases complete
with z ero wait cycles . The thi rd data phase has three wa it c yc les inserted by the target.
Both initiator and target insert a wait cycle on clock 5. In the case where the initiator
inserts a wait cycle (clock 5), the data is held on the bus, but the byte enables are
withdrawn. The last data phase is characterized by IRDY being asserted while the
FRAME signal is deasserted. This data phase is completed when TRDY goes active
(clock 8).
Figure 89 PCI Write Transaction
ITD07576
Address
123456789
Bus CMD
Wait
CLK
FRAME
AD
C/BE
IRDY
TRDY
DEVSEL
Address
Phase Data
Phase Bus Transaction
BE’s-1
Data Transfer
Data Transfer
Data Transfer
Wait
Wait
Data 1 Data 2 Data 3
BE’s-2 BE’s-3
Phase
Data Data
Phase
PEB 20534
Electrical Characteristics
Semiconductor Group 418 Data Sheet 09.98
14.6.1.3PCI Timing Characteristics
When the DSCC4 operates as a PCI Master (initiator) and it either reads or writes a burst
– as controlled by the on-chip DMA controller – it does not deactivate IRDY between
consecutive data. In other words, no wait states are inserted by the DSCC4 as a
transactio n ini tiator. The num bers of wait stat es, in serted by th e DSCC4 as ini tiator are
listed in table 116.
When the DSCC4 operates as a PCI Slave (target), it inserts wait cycles by deactivating
TRDY. The numbers of wait states, typically inserted by the DSCC4 are listed in
table 116:
The number of wait states inserted by the DSCC4 as target is not critical because
accesses to/via the DSCC4 are usually kept to a minimum in a system.
Table 116 Number of Wait States Inserted by the DSCC4 as Initiator
Transaction Number of Wait States
1st Data Cycle 2nd and Subsequent Data Cycles
Memory read burst 0 0
Memory write burst 0 0
Fast Back-to-back burst;
1st transaction 00
Fast Back-to-back burst;
2nd and subsequent
transactions
10
Table 117 Number of Wait States Inserted by the DSCC4 as Slave
Transaction Number of Wait States
Configuration read 2
Configuration write 0
Register read 3
Register write 0
LBI read 3
LBI write 0
PEB 20534
Electrical Characteristics
Semiconductor Group 419 Data Sheet 09.98
Figure 90 PCI Clock Specification
Note: Rise and fall times are specified in terms of the edge rate measured in V/ns. This
slew rate must be met across the minimum peak-to-peak portion of the clock
waveform as shown in figure 90.
Table 118 PCI Clock Characteristics
Parameter Symbol Limit Values Unit
min. typ. max.
CLK cycle time T30 ns
CLK high time tH11 ns
CLK low time tL11 ns
CLK slew rate (see note) 1 4 V/ns
ITD07577
0.4 V
1.5 V
2.4 V
2.0 V
0.8 V
t
HL
t
T
Voltage (V) 2 Vpp min
PEB 20534
Electrical Characteristics
Semiconductor Group 420 Data Sheet 09.98
Note 1 Minimum times are measured with 0 pF equivalent load; maximum times are
measured with 50 pF equivalent load.
Note 2 REQ and GNT are point-to-point signals. All other signals are bussed
GNT setup (min) time: 10ns
Table 119 PCI Interface Signal Characteristics
Parameter Limit Values Unit Remarks
min. typ. max.
CLK to signal valid delay
bussed signals (2) 11 ns Notes 1, 2
CLK to signal valid delay
point-to-point (2) 12 ns Notes 1, 2
Float to active delay 2 (3) ns
Active to float delay 20 ns
Input setup time to CLK
bussed signals 7nsNote 2
Input setup time to CLK
point-to-point 10 ns Note 2
Input hold time from CLK 0 ns
PEB 20534
Electrical Characteristics
Semiconductor Group 421 Data Sheet 09.98
14.6.2 De-multiplexed Bus Interface
Figure 91 Master Single READ Transaction followed by a Master Single WRITE
Transaction in De-multiplexed Bus Configuration
Figure 92 Master Burst WRITE/READ Access in De-multiplexed Bus Configura-
tion
The timing provided in Table 37 and Table 38 can also be applied to the de-multiplexed
bus interface.
TRDY
W/R READ Access
A (31 : 2)
BE (3 : 0)
D (31 : 0)
FRAME
CLK
Address
don´t care
Address
BE (3 : 0)
Data
WRITE Access
ITT10451
Address
Address
BE (3 : 0)
Data
don´t care
BE (3 : 0)
Data 2
WRITE/READ Access
Address
TRDY
W/R
A (31 : 2)
BE (3 : 0)
D (31 : 0)
FRAME
CLK
don´t care
Address
BE (3 : 0)
Data 1
ITT10452
BE (3 : 0)
Data 4Data 3
BE (3 : 0)
PEB 20534
Electrical Characteristics
Semiconductor Group 422 Data Sheet 09.98
Note: The PCI parity signal PAR is not generated in de-multiplexed mode. It is driven
active low by the DSCC4.
Table 120 Additional De-multiplexed Interface Signal Characteristics
Parameter Limit Values Unit Remarks
min. typ. max.
CLK to address bus signal
valid delay 12 ns
CLK to W/R signal valid
delay 12 ns
Address bus Input setup
time to CLK 8ns
Address bus Input hold time
to CLK 0ns
W/R signal Input setup time
to CLK 8ns
W/R signal Input hold time
to CLK 0ns
PEB 20534
Electrical Characteristics
Semiconductor Group 423 Data Sheet 09.98
14.6.3 Local Bus Interface Timing
Figure 93 Synchronous LBI Read Cycle Timing Multiplexed Bus
Figure 94 Synchronous LBI Write Cycle Timing Multiplexed Bus
LALE
LCLKO
LA[15..0],
LD[15..0]
LRD
LCSO
3130
32 33
34
35 36
37
3938
40
35
61
LALE
LCLKO
LA[15..0],
LD[15..0]
LWR
LCSO
3130
34
35 41
37
42
40
35
62
32 33
PEB 20534
Electrical Characteristics
Semiconductor Group 424 Data Sheet 09.98
Note: T
LBICLK
is the LBI clock time period which depends on the LBI clock division factor.
MCTC is t he number of mas ter clo ck w ait st ates (i n LB I clock cy cles) sele cted i n
registe r LCON F.
T
LRDY
is the number of additional wait states (in LBI clock cycles) introduced by
LRDY control signal if enabled via bit ’RDEN’ in register LCONF.
T
PCICLK
is the PCI clock time period.
1) If extended LALE timing is selec t ed v ia bit ’EALE’ in register LC ON F
Table 121 LBI Timing (synchronous, multiplexed bus)
No. Parameter Limit Values Unit
min. max.
30 LALE to LCLKO delay 5 20 ns
31 LALE pulse width 1 TLBICLK
(1.25 TLBICLK)1) (ns)
32 Address phase width 2 TLBICLK (ns)
33 LA, LBHE hold to LCLKO delay 5 20 ns
34 LCSO active to LCLKO delay 5 20 ns
35 LRD, LWR active/inactive to LCLKO delay 5 20 ns
36 LRD pulse width 2TLBICLK +MCTC+TLRDY (ns)
37 LCSO inactive to LCLKO delay 5 20 ns
38 LD to LCLKO setup time 0 ns
39 LD to LCLKO ho ld time 25 ns
40 LRD, LWR active to cycle start delay 2 TLBICLK (ns)
61 cycle end to next cycle start delay 2 TLBICLK (ns)
41 LWR pul se wid th 2TLBICLK +MCTC+TLRDY (ns)
42 LD to LCLKO ho ld de lay 1 TPCICLK
+ 5 1 TPCICLK
+ 20 ns
LD to LWR inactive delay 1 TPCICLK (ns)
LD tristate delay to LCLKO 1 TPCICLK
+ 20 ns
62 cycle end to next cycle start delay 5 TLBICLK (ns)
PEB 20534
Electrical Characteristics
Semiconductor Group 425 Data Sheet 09.98
Figure 95 Synchronous LBI Read Cycle Timing De-multiplexed Bus
Figure 96 Synchronous LBI Write Cycle Timing De-multiplexed Bus
LCLKO
LD[15..0]
LRD
LCSO
LA[15..0]
44 45
43
47
46
48 4950
51 63
LCLKO
LD[15..0]
LWR
LCSO
LA[15..0]
52 53
60
56
55
54
57
59 64
58
PEB 20534
Electrical Characteristics
Semiconductor Group 426 Data Sheet 09.98
Note: T
LBICLK
is the LBI clock time period which depends on the LBI clock division factor.
MCTC is t he number of mas ter clo ck w ait st ates (i n LB I clock cy cles) sele cted i n
registe r LCON F.
T
LRDY
is the number of additional wait states (in LBI clock cycles) introduced by
LRDY control signal if enabled via bit ’RDEN’ in register LCONF.
T
PCICLK
is the PCI clock time period.
Table 122 LBI Timing (synchronous, de-multiplexed bus)
No. Parameter Limit Values Unit
min. max.
43 LA to LCLKO delay 5 20 ns
44 LCSO active to LCLKO delay 5 20 ns
45 LCSO inactive to LCLKO delay 5 2 0 ns
46 LRD pulse width 2TLBICLK +MCTC+TLRDY (ns)
47 LRD, LWR inactive to LCLKO delay 5 20 ns
48 LD to LCLKO setup time 0 ns
49 LD to LCLKO ho ld time 25 ns
50 LRD, LWR active to LCLKO delay 5 20 ns
51 LRD active to cycle start delay 1 TLBICLK (ns)
63 cycle end to next cycle start delay 2 TLBICLK ns
52 LCSO active to LCLKO delay 5 20 ns
53 LCSO inactive to LCLKO delay 5 2 0 ns
54 LWR active to LCLKO delay 5 20 ns
55 LWR pul se widt h 2TLBICLK +MCTC+TLRDY (ns)
56 LRD, LWR inactive to LCLKO delay 5 20 ns
57 LD vali d after LCLKO delay 5 20 ns
58 LD hold after LCLKO delay 1 TPCICLK
+ 5 1 TPCICLK
+ 20 ns
LD to LWR inactive delay 1 TPCICLK (ns)
LD tr istate delay after LCLKO 1 TPCICLK
+ 20 ns
59 LWR active to cycle start delay 1 TLBICLK (ns)
60 LA to LCLKO delay 5 20 ns
64 cycle end to next cycle start delay 5 TLBICLK (ns)
PEB 20534
Electrical Characteristics
Semiconductor Group 427 Data Sheet 09.98
Figure 97 LRDY Timing
Table 123 LBI LRDY Timing
No. Parameter Limit Values Unit
min. max.
65 LRDY to LCLKO setup time 0 ns
66 LRDY to LCLK O hold ti me 25 ns
LCLKO
LD[15..0]
(read cycle)
LRD, LWR
LRDY
LD[15..0]
(write cycle)
2 cycles fix MCTC
first LRDY
evaluation last LRDY
evaluation
(LRDY active)
transaction
termination
Notes:
- MCTC wait state configuration is assumed to one cycle in this figure
- LRDY is evaluated the first time with the clock cycle following the MCTC related wait states
- Transaction is terminated one clock cycle after detecting LRDY active
- LRDY is evaluated only if LRDY-control is enabled
65 66
PEB 20534
Electrical Characteristics
Semiconductor Group 428 Data Sheet 09.98
14.6.4 Local Bus Interface Arbitration Timing
Figure 98 LBI Arbitration Timing
Table 124 LBI Arbitration Timing
No. Paramete r Limit Values Unit
min. max.
70A LHOLD inactive to LBREQ inactive delay 30 ?? ns
70B First master cycle start to LHOLDA inactive delay 120 ns
71A LHOLD inactive to LHOLDA inactive delay 30 ns
71B Last master cycle terminated to LHLDA active
again 90 ns
71C LBREQ active again to LHLDA inactive 60 ns
71A
70A
71B
DSCC4
requests
the local bus
DSCC4 releases
the local bus on request
of another device
LHOLD
(input)
LHLDA
(output)
LBREQ
(output)
LCSO
LRD/LWR
70B
71C
PEB 20534
Electrical Characteristics
Semiconductor Group 429 Data Sheet 09.98
14.6.5 PCM Serial Interface Timing
14.6.5.1 Clock Input Timing
Figure 99 Clock Input Timing
Table 125 Clock Input Timing (non high speed modes)
No. Parameter Limit Values Unit
min. max.
81 RxCLK clock period 50 ns
82 RxCLK high time 23 ns
83 RxCLK low time 23 ns
84 TxCLK clock perio d 50 ns
85 TxCLK high time 23 ns
86 TxCLK low time 23 ns
87 XTAL1 clock period (internal oscillator used) 50 ns
XTAL1 clock period (TTL clock signal supplied) 30 ns
88 XTAL1 high time (internal oscillator used) 23 ns
XTAL1 high time (TTL clock signal supplied) 13 ns
89 XTAL1 low time (internal oscillator used) 23 ns
XTAL1 low time (TTL clock signal supplied) 13 ns
RxCLK
TxCLK
XTAL1
81,84,87
82,85,88 83,86,89
PEB 20534
Electrical Characteristics
Semiconductor Group 430 Data Sheet 09.98
Table 126 Clock Input Timing (high speed mode)
No. Parameter Limit Values Unit
min. max.
81 RxCLK clock period 19.2 ns
82 RxCLK high time 8.6 ns
83 RxCLK low time 8.6 ns
84 TxCLK clock period 19.2 ns
85 TxCLK high time 8.6 ns
86 TxCLK low time 8.6 ns
PEB 20534
Electrical Characteristics
Semiconductor Group 431 Data Sheet 09.98
14.6.5.2 Receive Cycle Timing
Figure 100 Receive Cycle Timing
Note:
1. Whichever supplies the receive clock depending on the selected clock mode:
externally clocked via RxCLK or XTAL1 or
internally clocked via DPLL, BRG or BCR.
(No edge relation can be measured if the internal receive clock is derived from the
external clock source by devision stages (BGR, BCR) or DPLL)
2. NRZ, NRZI and Manchester data encoding
3. FM0 and FM1 data encoding
4. If Carrier Detect auto start feature enabled (not for clock modes 1 and 5)
90
91 92
91 92
93 94
Receive Clock
(Note 1)
RxD
(Note 2)
RxD
(Note 3)
CD
(Note 4)
91 92
PEB 20534
Electrical Characteristics
Semiconductor Group 432 Data Sheet 09.98
Table 127 Receive Cycle Timing
No. Parameter Limit Values Unit
min. max.
Receive
data rates ext ernal ly clo cked
(HDLC, high speed mode) 52 MBit/s
external ly clo cke d
(non high speed ) 10 MBit/s
internally clocked
(DPLL modes) 2 MBit/s
internally clocked
(non DPLL modes) 2 MBit/s
90 Clock
period external ly clocked 19.2 ns
internally clocked
(DPLL modes) 480 ns
internally clocked
(non DPLL modes) 480 ns
91 RxD to RxCLK setup time 5 ns
92 RxD to RxCLK hold time 15 ns
93 CD to RxCLK rising edge setup time 10 ns
94 CD to RxCLK falling edge hold time 10 ns
PEB 20534
Electrical Characteristics
Semiconductor Group 433 Data Sheet 09.98
14.6.5.3 Transmit Cycle Timing
Figure 101 Transmit Cycle Timing
Note:
1. Whichever supplies the transmit clock depending on the selected clock mode:
externally clocked via TxCLK, RxCLK or XTAL1 or
internally clocked via DPLL, BRG or BCR.
(No edge relation can be measured if the internal transmit clock is derived from the
external clock source by devision stages (BGR, BCR) or DPLL)
2. NRZ, NRZI and Manchester data encoding
3. FM0 and FM1 data encoding
4. If TxCLK output feature is enabled (only in some clock modes)
5. The timing is valid for non bus configuration modes and bus configuration mode 1. In
bus configuration mode 2, TxD and RTS are right shifted for 0.5 TxCLK periods i.e.
driven by t he falling TxCLK ed ge.
100
101
Transmit Clock
(Note1)
TxD
(Note2,5)
TxD
(Note3)
TxCLK
(Note4)
106
102
103
104 105
106
102
103
CxD
CTS
RTS
(Note5)
PEB 20534
Electrical Characteristics
Semiconductor Group 434 Data Sheet 09.98
Table 128 Transmit Cycle Timing
No. Parameter Limit Values Unit
min. max.
Transmit
data rates externally clocked
(HDLC, high speed mode) 52 MBit/s
externally clocked
(non high speed) 10 MBit/s
internally clocked
(DPLL modes) 2 MBit/s
internally clocked
(non DPLL modes) 2 MBit/s
100 Clock
period externally clocked 19.2 ns
internally clocked
(DPLL modes) 480 ns
internally clocked
(non DPLL modes) 480 ns
101 TxD to TxCLK delay (NRZ, NRZI encoding) 35 ns
102 TxD to TxCLK delay (FM0, FM1, Manchester
encoding) 35 ns
103 TxD to TxCLK(out) delay (if TxCLKO is enabled) 0 5 ns
104 CxD to TxCLK setup time,
CTS to TxCLK setup time 10 ns
105 CxD to TxCLK hold time,
CTS to TxCLK hold time 10 ns
106 RTS to TxCLK delay (not bus configuration mode) 35 ns
RTS to TxCLK delay (bus configuration mode) 35 ns
PEB 20534
Electrical Characteristics
Semiconductor Group 435 Data Sheet 09.98
14.6.5.4 Strobe Timing (clock mode 1)
Figure 102 Strobe Timing
Note:
1. No bus configuration mode and bus configuration mode 1
2. Bus configuration mode 2
3. TxD Idle is either
active high
or
high impedance
if ’open drain’ output type is selected.
Table 129 Strobe Timing (clock mode 1)
No. Parameter Limit Values Unit
min. max.
110 Receive strobe to RxCLK setup 5 ns
111 Receive strobe to RxCLK hold 15 ns
112 Transmit strobe to RxCLK setup 5 ns
113 Transmit strobe to RxCLK hold 15 ns
114 TxD to RxCLK delay 35 ns
115 TxD to RxCLK high impedance delay 35 ns
110 111
valid
112 113
114
114
115
115
RxCLK
CD
(RxStrobe)
RxD
(Note1)
TxCLK
(TxStrobe)
TxD
(Note1,3)
TxD
(Note2,3)
PEB 20534
Electrical Characteristics
Semiconductor Group 436 Data Sheet 09.98
14.6.5.5 F rame Synchronisation Timing (clock mode 5)
Figure 103 Frame Synchronisation Timing
Table 130 Frame Synchronisation Timing (clock mode 5)
No. Parameter Limit Values Unit
min. max.
130 Sync pulse to Rx CLK setup time 5 ns
131 Sync pulse to RxCLK hold time 5 ns
132 TxCLKout to RxCLK delay (time slot monitor) 40 55 ns
132
132
132
132
130 131
RxCLK
CD
(FSC)
TxCLK
Note1
TxCLK
Note2
PEB 20534
Electrical Characteristics
Semiconductor Group 437 Data Sheet 09.98
14.6.5.6 High Speed Receive Cycle Timing
Figure 104 High Speed Receive Timing
Table 131 High Speed Receive Timing
No. Parameter Limit Values Unit
min. max.
140 RCG setup time 5 ns
141 RCG hold time 5 ns
142 RxD setup time 5 ns
143 RxD hold time 5 ns
RxCLK
RCG
RxD
140
141
142
143
PEB 20534
Electrical Characteristics
Semiconductor Group 438 Data Sheet 09.98
14.6.5.7 High Speed Transmit Cycle Timing
Figure 105 High Speed Transmit Timing
Note: T
TxCLK
is the TxCLK signal time period.
Timing 147 results from a constant functional one clock offset + signal delay.
Table 132 High Speed Transmit Timing
No. Parameter Limit Values Unit
min. max.
145 TCG setup time 5 ns
146 TCG hold time 5 ns
147 TxD to TxCLK delay 16.5 ns
148 TxD to TxCLKout delay 0 5 ns
149 TxD to TCG active delay 1 TTxCLK
+16.5 (ns)
TxCLK
TCG
TxD
TxCLKout
145
146
149
148
147
PEB 20534
Electrical Characteristics
Semiconductor Group 439 Data Sheet 09.98
14.6.6 Reset Timing
Figure 106 Reset Timing
Note: RST may be asynchronous to CLK when asserted or deasserted. RST may be
asserted during power-up or asserted after power-up. Nevertheless deassertion
must be clean, bounce-free edge as recommended by PCI Spec. Revision 2.1.
Note: RST signal timing is independent of whether PCI or De-multiplexed mode is
selected via pin DEMUX.
Table 133 Reset Timing
No. Parameter Limit Values Unit
min. max.
150 RESET pulse width 120 ns
150 Number of CLK cycles during RST
active 2CLK
cycles
power-on
VDD3
CLK
RST
150
151
PEB 20534
Electrical Characteristics
Semiconductor Group 440 Data Sheet 09.98
14.6.7 J TA G-Boundary Scan Timing
Figure 107 JTAG-Boundary Scan Timing
Table 134 JTAG-Boundary Scan Timing
No. Parameter Limit Values Unit
min. max.
160 TCK period 166 ∞ns
161 TCK high time 80 ns
162 TCK low time 80 ns
163 TMS setup time 30 ns
164 TMS hold time 10 ns
165 TDI setup time 30 ns
166 TDI hold time 20 ns
167 TDO valid delay 60 ns
160
161 162
163 164
165 166
167
TCK
TMS
TDI
TRST
TDO
PEB 20534
Electrical Characteristics
Semiconductor Group 441 Data Sheet 09.98
14.6.8 SSC Serial Interface Timing
Figure 108 SSC Interface Timing (Master)
Note: T
CLK
is the CLK signal time period.
1) In SSC ’Slav e M ode’, signal MRST is out put and MTSR is inpu t.
Table 135 SSC Interface Timing
No. Parameter Limit Values Unit
min. max.
170 MCLK high to MCSi active delay 2TCLK + 40 ns
171 MCLK high to MTSR delay (Master) 2TCLK + 40 ns
MCLK high to MRST delay (Slave)1) 4TCLK + 40 ns
172 MRST setup time TCLK+20 ns
173 MRST hold time TCLK+20 ns
171
172
173
170
MCLK
MTSR
MRST
MCSi
PEB 20534
Package Outlines
Semiconductor Group 442 Data Sheet 09.98
15 Package Outlines
Figure 109 Package Outline
P-FQFP-208-7
(Plastic Metric Quad Flat Package)
Sorts of Packing
Package outlines for tubes, trays etc. are contained in our
Data Book “Package Information”. Dimensions in mm
SMD = Surface Mounted Device
