ISD · 2727 North First St reet, San Jose, CA 95134 · TEL: 408/943-6666 · FAX: 408/544-1787 · http://www.isd.com
May 2000
—Advan c e In formation —
ISD-T360SB
VoiceDSP™ Digital Speech Processor with
Full-Duplex Speakerphone, Master/Slave CODEC Interface and
Multiple Flash Support
The Vo iceDS P™ prod uct family combine s multiple
digital signal processing functions on a single chip
for cost -effect ive soluti ons in tel ephony, a utomo-
tive and consumer applications.
The Voic eDS P pro cesso r of fers nece ssary f eature s
to modern telephony products, such as: high-
quality speech record and playback, electrical
and acoustic echo cancellation for full-duplex
hands-free speakerphone operation.
The T360SB V oiceDSP can be used in va rious ap-
plications:
•Digital telephony with add-on speech
proc ess ing: Digital Teleph one Answer ing
Machines (DTADs), hands-free speakerpho ne
operation for ISDN, DECT, Digital Spread
Spectrum, and analog cordless applications;
CT0/1 Base stations.
•An add-on chi p for corded tel eph ones
featuring DTAD functions and/or full-duplex,
hands-free speakerphone operation.
•Stand-alone digital answering machines with
full-duplex, hands-free speakerphone
operation.
•Voice memo recording
•Automotive applications employing full-
duplex speakerphone operations for hands-
free, in-car communications, and for car
status and information announcements.
Based on ISD’s unique concept which combines
16-bit DSP (Digital Speech Processor) and 16-bit
RISC core technology, the ISD-T360SB is a high-
performing chip solution for various applications.
To facilitate incorporating the VoiceDSP proces-
sor, it features system support functions such as an
interrupt control unit, codec interface (master,
slave), Microwire interface to the system micro-
controller, as well as a mem ory handler fo r Flash
memory dev ices. Design of high-end, p rice opti -
mized systems are possible with ISD’s VoiceDSP
flexible system interfaces (codec, microcontroller,
and memory management suppo rt ) .
The ISD-T360SB processor operates as a peripheral
controlled by the system microcontroller via an
enhanced, serial Microwire interface . The system
microcontroller typically controls the analog cir-
cuits, buttons and display, as well as activates
functions through commands. The Voice DSP e xe-
cutes these commands and returns status infor-
mation to the Microcontroller.
The VoiceDSP software resides in the on-chip
ROM. It includes DSP-based algorithms, system
support functions, and a software interface to
hardware perip her als .
ISD-T360SB
ii Voice Solutions in Silicon™
FEATURES AT A GLANCE
DTAD MANAGEMENT
•Highest quality speech recording in PCM
forma t for music on ho l d or other OGM (Out
Going Message) recording and IVS
•Selectable high-quality speech compression
rate of 4.7 Kbit/s, 6.7 Kbit/s or 8.7 Kbit/s
•Up to approximately 15 minutes recording on
a 4-Mbit Flash
•Up to 4 hour speech r ecording
•High-quality music compression for music on
hold (64 Kbit/s PCM)
•Programmable message tag for message
categorization, e.g., Mailboxes, Incoming
Messages (ICM), OutGoing Messages (OGM)
•Message management
•Skip forw ar d or backwar d du r ing message
playback
•Variable speed playback
•Real-time clock: Day of Week, Hours, Minutes
•Multi-lingual speech synthesis using
Inter national Voca bulary Supp or t (IVS)
•Vocabularies available in: English, Japanese,
Mandarin , G er man, F r ench a n d S p anish
•Conversation Recor ding
•Software Automatic Gain Con trol (A G C)
SPEAKERPHONE
•Digital fu ll-dupl ex speakerphon e
•Acoustic- and line-echo cancellation
•Continuous on-the-fly monitoring of external
and inte rnal condit io ns (ac ou st ic and line)
provides high-quality, hands-free,
conversation in a changing environment
•Minimu m microcontr ol ler control interven tion
(Launch-and-forget)
•Supports: On, Off, Mute, and Hold functions
•Additi on al Softw ar e A u tom atic Gain Con trol
for imp roved speakerphone per formanc e
CALL AND DEVICE MANAGEMENT
•Digita l v ol ume control
•Least cost routing support (LCR)
•Power-down mode
•3.3V or 5V select abl e power supply
•32.7 MHz internal oper ati on using 4.096 MHz
external crystal oscillator
•Availa ble in the PQFP 100-pin pac k age
•DTMF generation and detection
•Telephone line functions including busy, dial
tone detection an d V O X detectio n
•Single tone genera ti on
•DTMF detec ti on du rin g mes s ag e re cor d an d
playback
ISD-T360SB
iii
ISD
PERIPHERAL CONTROL
Codec
•µ-Law, A-Law, and 16-bit linear codec input
support
•Selectable master /sl av e codec interf ace. In
addition, the slave mode is implemented with
respe ct to IOM-2TM/GCI specifications
•Supports two selectable in-c om ing lines in
slave mode without speakerphone for DTAD
recording
•Suppor ts one in- coming line in s la ve mode
and one mast er speakerphone in terface
•Supports up to 32 user-selectable speech
channe ls in sl ave mode
•Softw ar e A u tom ation Ga in Con trol ( A G C)
•Supports long-frame and short-frame codecs
•Single/double bit clock rate for slave mode
•On-chip code c cl ock ge n eration
Memory
•Supports up to four 4-Mbit, four 8-Mbit, or four
16-Mbit Flash devices from Toshiba or
Samsung
•The number of messages that can be stored is
limited only by me mor y size
•Direct access to message memory
•Message storage contains all data in a
concate n ated chain of mem or y bl ocks.
•Memory mapping and pr odu ct floor test
included
•Support s exter n al vo cabularies, u s i ng Fl ash
memory or expansion ROM
Microwire
•MICROWIRE slave interface to an exte rn al
microcontroller
•Sophisticated command language to
optimize syste m cod e size
ISD-T360SB
iv Voice Solutions in Silicon™
INTERNATIONAL VOCABULARY SUPPORT (IVS)
For manufacturing recorded voice prompt and
speech synthesis, the ISD International Vocabulary
Support delivers pre-recorded voice prompts in
the same high-quality of the user-recorded
speech. For complete control over quality and
memory management, the IVS features adjust-
able speech compressions. In addition, several
pre-recorded voice prompt sets are available in
various languages for further convenience.
Availabl e L angua g es :
•English
•Japanese
•Mandarin
•German
•French
•Spanish
You can al so develop a ne w vocabulary using the
ISD-T360 IVSTOOL application. This PC-Windows95/
98/2000™-based application synthesizes record-
ed .wav files into the ISD-T360SB’s various com-
pression rates (including PCM). The IVSTool
applicatio n supports various languages, includi ng
their unique grammar str ucture s.
ISD’s VoiceDSP pro du ct s st or e IV S v ocabular i es on
either Flash memory or expansion ROM memo-
ries, thus DTAD manufac tur ers can d esign a pro d-
uct for multiple countries, featuring various
languages.
For more details about IVS, refer to the
IVS User’s
Guide
.
Figure 1-1: ISD-T360SB Block Diagr am—Ba sic Configurati on wit h Four 4/8/16Mbit
NAND Flash Devices (Samsung/Toshiba)
ISD-T360SB
v
ISD
Figure 1-2: ISD -T3 60SB Bloc k D iagram-Basic Configuration with IV S EPROM
Speakerphone
Codec
VFRO
VFXI
DR
DX
CLK
FS
Line Codec
VFRO
VFXI CLK
FS
CDOUT
CDIN
CCLK
CFS0
CFS1
A(0:15)
IVS
EPROM
A0
CE OE
D2
D3
D4
D5
D6
D7
D1
D0
A1
A4
A3
A5
A6
A7
A10
A8
A9
A11
A12
A13
A14
A15
A16
A17
A2
A18
D(0:7)
EMCS
PB0
PB1
PB2
ISD-T360SB
MWCLK
MWRQST
MWCS
MWRDY
MWDOUT
MWDIN
Microcontroller
DX
DR
ISD-T360SB
vi Voice Solutions in Silicon™
Table of Contents
ISD vii
Chapter 1—HARDWARE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1
1.1 PIN ASSIGNMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1
1.1.1 Pin-Signal Assignment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1- 1
1.2 DESCRIPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3
1.2.1 Resetti ng . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3
1.2.2 Clocking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3
1.2.3 Power-D ow n Mo de . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1- 4
1.2.4 Power an d Grounding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-5
1.2.5 Memory Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-6
1.2.6 The Codec Inte r fa ce . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-13
1.3 SPECIFICATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-19
1.3.1 Absolute Maximum Rat in gs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-19
1.3.2 Electrica l Characteristi cs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-19
1.3.3 Switch ing Character is ti cs — P rel i mi nar y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-21
1.3.4 Synch ronous Timing Ta bl es . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-24
1.3.5 Timing Dia grams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-2 6
Chapter 2—SOFTWARE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1
2.1 SYSTEM OPERATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1
2.1.1 The State Mac hin e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1
2.1.2 Comman d Ex ecution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-2
2.1.3 Event Han dl in g . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-2
2.1.4 Message Ha ndling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3
2.1.5 Tone Generat ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-4
2.1.6 Initialization and Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-4
2.1.7 Power-D ow n Mo de . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2- 5
2.2 PERIPHERALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-5
2.2.1 Microcontr oller interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-5
2.2.2 Memory Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-8
2.2.3 codec inter f ace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-9
2.3 ALGORITHM FEATURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-9
2.3.1 VCD (Voice Compression and Decompre ss ion ) . . . . . . . . . . . . . . . . . . . . . . . . . . 2-10
2.3.2 DTMF Det ect ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-12
2.3.3 Tone and Ener gy Detection (Cal l Progr ess) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-13
ISD-T360SB
viii Voice Solutions in Silicon™
2.3.4 Full-Duplex Speakerphone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-15
2.3.5 Speec h Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-17
2.3.6 Calle r ID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 -22
2.4 VOICEDSP PROC ESSOR COMMANDS—QUICK REFERENCE TABLE . . . . . . . . . . . . . .2-22
2.5 COMMAND DESCRIPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-25
Chapter 3—SCHEMATIC DIAGRAMS . . . . . . . . . . . . . . . . . . . 3-1
3.1 APPLICATION INFORMATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-1
Chapter 4—PHYSICAL DIMENSIONS . . . . . . . . . . . . . . . . . . . 4-1
1—HARDWARE ISD-T360SB
1-1
ISD
Chapter 1—HARDWARE
1.1 PIN ASSIGNMENT
The following sections detail the pins of the ISD-
T360SB process or . S la s h es separa te the nam es of
signals that shar e the same pi n.
1.1.1 PIN-SIGNAL ASSIGNMENT
Table 1-1 shows all the pins and the signals that
use them in different configurations. It also shows
the type and direction of each signal.
Figure 1-1: 100- P QFP Pa c kag e Connection Diagr am
80
79
78
77
76
75
74
73
72
71
70
69
68
67
66
65
64
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
84858687888990919293949596979899100
4746454443424140393837363534333231
NC
TST / PA0 / W R 0
PA1 / WR1
PA2 / CTTL
PA4 / MWRD Y
PA5 / MWDOUT
V
SS
MWRQST / PA7 / B ST1
V
CC
PD0 / MWC LK
PD1 / MWD IN
A10
NC
NC
NC
D0
D1
D2 / RA11
V
SS
D3
VCCHI
Vcc
D4
D5
D6
NC
Vss
NC
VccA
V
CC
A8
A7
A6
A5
A4
A3
A2
V
CC
A1
V
SS
PB7 / D15
PB6 / D14
PB5 / D13
PB4 / D12
PB3 / D11
V
SS
PB2 / D10
V
CC
PB1 / D9
PB0 / D8
PC7 / BMCS / E N V1
PC6 / EMCS / ENV0
PC5 / IOCS / ENV3
PC4 / A15 / BE1 / ENV2
PC3 /A14 / BE0
PC2 / A13
NC
18
19
20
21
22
23
24
25
26
27
28
29
30
D7
PC0 / A11
PC1/ A12
504948
INT3 / MWCS
63
62
61
60
59
58
57
56
55
54
53
52
51
CCLK
CDIN
CFS0
CDOUT
RESET
818283
A0 / A16 / DDIN
NC
NC
NC
VssA
X2 / CLKIN
X1 / PLI
NC
NC
NC
NC
NC
NC
NC
NC
NC
A9
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
100-PQFP
Top View
ISD-T360SB
PA3 / PFS
PA6 / BST0
CFS1 / PWM
1—HARDWARE
ISD-T360SB
1-2 Voice Solutions in Silicon™
1. Input during reset, otherwise output.
2. Virtual address lines for IVS ROM, bits [0:2]
3. Chip select lines for Flash devices, bits [3:6]
4. Schmitt trigger input.
Table 1-1: VoiceDSP Pin Signal Assignment
Pin Name Signal Name Type Description
A(0:15) A(0:15) Output Address bits 0 through 15
CCLK CCLK I/O Codec Master/slave Clock
CDIN CDIN Input Data Input from Codec
CDOUT CDOUT Output Data Output to Codec
CFS0 CFS0 I/O Codec 0 Frame Synchronization
CFS1 CFS1 Output Codec 1 Frame Synchronization
D(0:7) D(0:7) I/O Data bits 0 through 7
EMCS/ENV0 EMCS Output Expansion Memory Chip Select
EMCS/ENV0 ENV0 Input1Environment Select
MWCLK MWCLK Input4MICROWIRE Clock
MWCS MWCS Input4MICROWIRE Chip Select
MWDIN MWDIN Input4MICROWIRE Data Input
MWDOUT MWDOUT Output MICROWIRE DATA Output
MWRDY MWRDY Outp ut MICROWIRE Ready
MWRQST MWRQST O utp ut MICROWIRE Reque st Sign al
PB(0:7)2,3 PB(0:7) I/O Port B, bits 0 through 7
PC(0:7) PC(0:7) Output Port C, bits 0 through 7
RESET RESET Input4Reset
TST TST Input Test pin
VCC VCC Power 3.3 V power supply pin
VCCAV
CCA Power 3.3 V analog circuitry power supply pin
VCCHI VCCHI Power 5 V power supply pin. Connect to VCC if 3.3 V power supply is used.
VSS VSS Power Ground for on-chip logic and output drivers
VSSAV
SSA Power Ground for on-chip analog circuitry
X1 X1 Oscillator Crystal Oscillator Interface
X2/CLKIN X2 Oscillator Crystal Oscillator Interface
1—HARDWARE ISD-T360SB
1-3
ISD
1.2 DESCRIPTION
This section provides details of the functional char-
acteri stics of the V oiceDS P processor . It is div ided
into th e following sec ti o n s :
• Resetting
• Clocking
• Power-Down Mode
• Power and Grounding
• Memory In terfac e
• Codec Interface
1.2.1 RESETTING
The RESET pin is used to reset the VoiceDSP proces-
sor.
On applic ati on of power, RESET must be held low
for at least tpwr after VCC is stable. This ensures that
all on-chip voltages are completely stable before
oper at io n. W he nev er RESE T is applied, it must also
remai n active fo r not less than tRST, see Table 1-9
and Table 1-10. During this period, and for 100 ms
after, the TST signal, which is an internal signal,
must be hig h. This can be don e with a pul l-u p re-
sistor on the TST pin.
The va lu e of MWRDY is undefined during the reset
period, and for 100 ms after. The micr ocontroller
should either wait before polling the signal for the
first t ime, or the sig nal should be p ulled hig h dur-
ing this period.
Upon reset, the ENV0 signal is sampled to deter-
mine the operating environment. During reset, the
EMCS/ENV0 pin is used for the ENV0 input signals.
An internal pull-up resistor sets ENV0 to 1.
After rese t, th e sam e pin is used for EMCS.
System Load on ENV0
For any loa d on the ENV0 pin, the vo ltage shou ld
not drop below VENVh (see Table 1-8).
If the l oad on the ENV 0 pin causes t he current t o
exceed 10 µA, use an external pull-up resistor to
keep the pin at 1.
Figu re 1-2 s how s a re comme nded cir cuit for ge n-
erating a reset signal when the power is turned on.
Figure 1-2: Recomme nde d Po wer-On Reset
Circuit
1.2.2 CLOCKING
The VoiceDSP processor provides an internal oscil-
lator that interacts with an external clock source
through the X1 and X2/CLKIN pins. Either an exter-
nal single-phase clock signal, or a crystal oscilla-
tor, may be used as the clock source.
External Single-Phase Clock Signal
If an ext er n al single-phase clock sou r ce i s us ed, it
should be connected to the CLKIN signal as
shown in Figure 1-3, and should conform to the
voltage-level requirements for CLKIN stated in
“ELECTR ICA L CH A RAC TERISTICS” on pag e 1-16.
NOTE the CLKIN signal is not 5V tolerant.
VCC
RESET
ISD-T360
VCC
VSS
1—HARDWARE
ISD-T360SB
1-4 Voice Solutions in Silicon™
Figure 1-3: External Clock Source
Crystal Oscillator
A cry sta l osc ill ato r i s co nne ct ed t o th e o n- ch ip os -
cillat or circu it via the X1 and X2 sign als, as shown
in Figu r e 1-4.
Figure 1-4: C onne ct ions for an External
Crysta l Osc illator
Keep stray capacitance and inductance, in the
oscillator circuit, as low as possible. The crystal res-
onator, an d the exter nal compone nts, should b e
as close to the X1 an d X 2/CLKIN pi ns as possi ble ,
to keep the trace lengths in the printed circuit to
an absolute minimum.
You can use crystal oscillators with maximum load
capacitance of 20 pF, although the oscillation
frequency may differ from the crystal’s specified
value.
Table 1-2 lists th e c ompone nts in the crysta l osc il-
lato r circu it
1.2.3 POWER-DOWN MODE
Power-down mode is useful during a power failure
or in a power-saving model, when the power
source for the processor is a backup ba ttery or in
battery-powered devices, while the processor is in
idle mode.
In power-down mode, the clock frequency of the
VoiceDSP p rocessor is reduced and some of the
processor modules are deactivated. As a result,
the ISD-T360SB consumes considerably less power
than in normal-power mode. Although the
VoiceDSP processor does not pe rform al l its usual
functions in power-down mode, it does retain
stored messages and maintain the date and
time.
NOTE In power-down mode all the chip select
signals, CS0 to CS3, a re set to 1 . To guar-
antee that there is no current flow from
these signals to the Flash devices, the
power supply to these devices must not
be disconnected.
The ISD-T360SB stores messages and all memory
management information in Flash memory.
When Flash memory is used for memory manage-
ment, power does not need to be maintained to
the processor to preserve stored messages.
To keep power consumption low during power-
down mode, the RE SET, MWCS, MWCLK and MW-
DIN signals should be held above VCC –0.5 V or
below VSS +0.5 V.
X2/CLKIN
Clock Generator
X1
VoiceDSP
Single-phase Clock Signal
X1X2
C1
R1
C2
ISD-T360
crystal
Table 1-2: Compon ents of Crys tal Osc illat or
Circuit
Component Values Tolerance
Crystal Resonator 4.096 MHz 50 PPM
Resistor R1 10 MW5%
Capacitors C1, C2 33 pF 20%
1—HARDWARE ISD-T360SB
1-5
ISD
1.2.4 POWER AND GROUNDING
Power Pin Connections
The ISD-T36 0SB can operat e over t wo suppl y volt-
age ranges 3.3 V ±10% and 5 V ±10%. The pow-
er supply and ground pins (VCC, VSS, VCCA, VSSA
and VCCHI) must be connected as shown in
Figure 1-5 when operating in a 3.3 V environment,
and as show n in Fig ur e 1-6 when opera ti n g in a 5
V environment. Failure to correctly connect the
pins may result in damage to the device.
The Capacitor and Resistor values are given in
Table 1-3.
Figure 1-5: 3.3 V Power Connection Diagram
Table 1-3: Components of Supply Circuit
Component Values Tolerance
Resistor R1 10 W5%
Capacitors C1, C2,
C3, C4, C5, C6, C7 0.1 µF Ceramic 20%
Capacitors C8, C9,
C10, C11, C12, C13,
C14
2.2 µF Tantal um 20%
97 95 87 86 84
76
66
64
17
19
20
40 42
V
SSA
C
1
V
SS
V
CC
V
CCHI
V
SS
V
CC
V
CC
V
SS
ISD-T360
C
5
V
SS
V
CC
V
CCA
V
CC
V
SS
C
4
C
3
C
7
R
1
C
6
C
9
C
10
C
8
C
11
C
12
C
13
+
+
+
+
+
+
+
3.3V Supply
1—HARDWARE
ISD-T360SB
1-6 Voice Solutions in Silicon™
Figure 1-6: 5 V Power Connection Diagram
For optimal noise immunity, the power and
ground pins should be connected to VCC and the
ground planes, respectively, on the printed circuit
board. If VCC and the ground planes are not used,
single conductors should be run directly from
each VCC pin to a power point, and from each
GND pin to a ground point. Avoid daisy-chained
connections. The VoiceDSP does not perform its
usual functions in power-down mode but it still pre-
serves stored messages, maintains the time of
day.
When you build a prototype, using wire-wrap or
other methods, solder the capacitors directly to
the power pi ns of the Voice DSP pro cesso r socket ,
or as close as possible, with very short leads.
1.2.5 MEMORY INTERFACE
Flash Support
The ISD-T360SB VoiceDSP supports Flash devices
for storing recorde d data, thus, power can be dis-
connected to the ISD-T360SB witho ut losing dat a.
The ISD-T360 SB supports NAND Flash device inter-
faces, such as TC58V16BFT, TC5816BFT,
TC58A040F, KM29N040T, KM29W8000T/IT, and
KM29W16000AT/AIT. The ISD-T360SB may be con-
nected to up to four Flash devices, resulting with
maximum recording storage of 16-Mbits x 4 =
64 Mbits (up to 4 hours of recording time).
The following flash devices are supported:
97 95 87 86 84
76
66
64
17
19
20
40 42
V
SSA
C
1
V
SS
V
CC
V
CCHI
V
SS
V
CC
V
CC
V
SS
5 V Supply
ISD-T360
C
5
V
SS
V
CC
V
CCA
V
CC
V
SS
C
4
C
3
C
7
R
1
C
6
C
14
C
9
C
10
C
8
C
11
C
12
C
13
+
+
+
+
+
+
+
1-7
1—HARDWARE ISD-T360SB
ISD
Internal Memory Organization
The Flash devices detailed in Table 1-4 divide internally into basic 4-Kbyte block units. The ISD-T360SB uses
one block on each device for memory management, leaving the rest of the blocks available for record-
ing. Using at least one block for a single recorded message yields a maximum of
NUM_OF_BLOCKS_IN_MEM – 1 (see Table 2-10 for parameter definition) messages per device.
Figure 1-7: Memory Interface with Four 4/8/16 Mbit, NAND Flash Devices
(Samsung, Toshiba)
Table 1-4: Su pported Fla sh Devices
Manufacturer Memory Device Name Characteristics Operating Voltage Memory Size
Toshi ba TC58V16 BFT 2Mx8 3.3 V 16-Mbit LV
Toshiba TC5816BFT 2Mx8 5.0 V 16-Mbit
Samsung KM29W040T 512Kx8 2.7V - 5.0 V 4-Mbit
Samsung KM29W8000T/IT 1Mx8 2.7V - 5.0 V 8-Mbit
Samsung KM29W16000AT/AIT 2Mx8 2.7V - 5.0 V 16-Mbit
1—HARDWARE
ISD-T360SB
1-8 Voice Solutions in Silicon™
NAND Flash (Samsung, Toshiba)
The VoiceDSP processor supports up to four NAND
interface Flash memory devices for storing mes-
sage s. A flash device with t he NAND interface uses
a single 8bit I/O port to set the address and ac-
cess the data. The ISD-T360SB supports three types
of Flash volumes (4Mbit, 8Mbit and 16Mbit as list-
ed inTa ble 1- 4) while al l the conn ected Fl ash de-
vices must be of the same type. Ports B and C are
used to connect ISD-T360SB to the Flash devices
using port B for address and data transfer and port
C for communication control and chip select.
Connecting less than four Flash devices requires
connecting the Flash devices sequentially, start-
ing from PC4 up to PC7 (see Figure 1-7). Upon ini-
tializ ation, the ISD-T360SB scans the flash devices
to detect defected blocks. The defected blocks
are marked in a special map, which is located in
the last block of each device.
Flash Endurance
A Flash memory may be erased a limited number
of times. To maximize the Flash use, the memory
manager utilizes the Flash’s blocks evenly (i.e.,
each block is erased more or less the same num-
ber of times), to ensure that all blocks have the
same lifetime. Refer to the respective Flash mem-
ory device data sheets for specific endurance
specifications.
A VoiceDS P processo r mes sage uses at le ast one
block. Th e maximum recording ti me depends on
four factor s :
1. The basic compr ession rate (4.7 Kbit/ s, 6.7
Kbit /s or 8. 7 Kbit/s).
2. The amount of silence in the recorded
speech
3. The n u m ber of bad bloc ks
4. The number of recorded messages. (The
basic memory allocation unit for a mes-
sage is a 4-Kbyt e block, which mean s that
half a block, on av erage, is wast ed per re -
corded messag e.)
Rom interface
IVS vocabularies can be stored in either Flash
memory and/or ROM. The VoiceDSP processor
supports IVS ROM devices through an Expansion
Memory mechanism. Up to 64 Kbytes (64K x 8) of
Expansi on Memory are directl y support ed. Never-
theless, the processor uses bits of the on-chip port
(PB) to further extend the 64 Kbytes address space
up to 0.5 Mbytes ad dr ess space.
ROM is conne cted to the VoiceDS P processor us-
ing the data bus, D(0:7), the address bus, A(0:15),
the extend ed address signals, EA(16:18), and Ex-
pansion Memory Chip Select, EMCS, controls. The
number of extended address pins to use may
vary, depen ding on the size and con figurat ion of
the ROM. ISD-T360SB configured with NAND Flash
memory cannot support extension ROM.
Table 1-5: Recording Time with No Silence
Compression
Memory Size Compression
Rate
Total Recording
Time
4 Mbit 4.7 Kbit/s 14.5 Minutes
4 Mbit 6.7 Kbit/s 10.2 Minutes
4 Mbit 8.7 Kbit/s 7.8 Minutes
8 Mbit 4.7 Kbit/s 29.1Minutes
8 Mbit 6.7 Kbit/s 20.4 Minutes
8 Mbit 8.7 Kbit/s 15.7 Minutes
16 Mbit 4.7 Kbit/s 58.1Minutes
16 Mbit 6.7 Kbit/s 40.8 Minutes
16 Mbit 8.7 Kbit/s 31.4 Minutes
32 Mbit 4.7 Kbit/s 116.2 Minutes
32 Mbit 6.7 Kbit/s 91.5 Minutes
32 Mbit 8.7 Kbit/s 62.8Minutes
1—HARDWARE ISD-T360SB
1-9
ISD
1.2.6 THE CODEC INTERFACE
The ISD-T360SB provides an on chip interface for analog and digital telephony, supporting master and
slave codec interface modes. In master mode, the ISD-T360SB controls the operation of the codec for use
in analog telephony. In the slave mode, the ISD-T360SB codec interface is controlled by an external source.
This mode is used in digital telephony (i.e., ISDN or DECT lines). The slave mode is implemented with respect
to IOM-2™/GCI specifications.
See Tabl e 1-6 f or codec o ptions f or the IS D-T360SB (I SD suppo rts compati ble code cs in add ition to those
list ed bel ow ).
The codec interface su pports the fo ll owing featu r es:
• Ma ster Mode or Slave M ode.
• 8- or 16-bit Channel Width.
• Long (Variable) or Short (Fixed) Frame Protocol.
• Single or Double Bit (Slave Mode Only) Clock Rate.
• Single or Dual Channel Codecs
• On e or Two Code cs
• Multipl e Clock A n d Sample R ates.
• On e or Two Frame S ync Signal s
Tabl e 1-6: Su pporte d C odec Options
Codec Configuration Options
The codec interface may be configured so that the ISD-T360SB is connected to either:
• one single codec (the default option) (using channel 0)
• two single codecs (using channels 0, 2)
• one dual codec (using channel 0, 1)
Manufacturer Codec Device Name Characteristics Operating Voltage Conversion Type
National Semiconductor TP3054 Single codec 5 V µ-Law
Natio nal Semico nduc tor T P 3057 Single codec 5 V A -Law
Oki MSM7533V Dual codec 5 V µ-Law, A-Law
Oki MSM 7704 Dual codec 3.3 V µ-Law, A-Law, LV
Lucent T7502 Dual codec 5 V A-Law
Lucent T7503 Dual codec 5 V µ-Law
Motorola MC14 54 8 2 Single codec 5 V 16 - bit linear
Motor ola MC14 54 84 Sing le code c 5 V µ-Law, A-Law
Winbond W6612 Dual codec 5 V µ-Law, A-Law
1—HARDWARE
ISD-T360SB
1-10 Voice Solutions in Silicon™
The Codec Interface Signals
The codec interface consists of five signals: CDIN,
CDOUT, CCLK, CFS0 and CFS1. The first codec is
connected to CDIN, CDOUT, CCLK and CFS0 (see
Figure 1-9). The second codec is connected to
CDIN, CDOUT, CCLK an d CFS 1 (see Figure 1-10).
Data is read from the codec through the CDIN in-
put pin, while data is simultaneously transferred to
the codec through the CDOUT output pin (see
Figure 1-8).
The Codec Frame Sync 0 signal (CFS0) is generat-
ed by the codec i nterface in master mode, and
is ge nerated extern a l ly in sla v e m ode.
The Codec Frame Sync 1 signal (CFS1) is generat-
ed by the codec interface in master and slave
modes. The signal is generated with a delay rela-
tive to the CFS0 signal. CFS1 signal may be de-
layed between 0 and 255 CCLK cycles (see
Figure 1-8).
The Codec Clock (CCLK) signal is used to synchro-
nize the data in and out of the codec interface.
The CCLK s ignal i s used to indica te a delay (rela -
tive to CF S0) of data and CF S1 signal s. The CCLK
signal is generated according to the pro-
gram med cl ock rate in maste r mode, and is gen -
erated externally in slave mode.
In Slave Mo de o perat io n, the C CLK sign al is inpu t
to the ISD-T360SB and controls the frequency of
the codec interface operation. The CCLK may
take on any frequency between 500 KHz and
2.048 MHz.
In master mode, the codec interface is config-
ured to single bit rate clock. In slave mode, a dou-
ble bit rate clock is available as well. When using
double bit rate clock, the codec interface divides
the frequency in half and the resulting clock is
used as CCLK. Note that the signals are sampled
on the fa lling edge of CCLK.
Short-frame / Long-frame Formats
The co dec may be configured in o ne of the fol-
lowin g formats:
• A sh ort-f r ame form at (see Figu re 1-24)
• A long-frame format (see Figure 1-25)
In short- frame o peratio n, CFS0 a nd CFS1 ar e ac-
tive fo r one C CLK cy cle. In long-f ram e opera tion ,
CFS0 and CFS1 are active for the width of the
transaction.
Short-frame and long-frame formats are available
in both master and slave modes (for more infor-
mation r efer to the software descript ion on pa ge
2-9).
1—HARDWARE ISD-T360SB
1-11
ISD
Channel Width
The Codec interface supports both 8-bit and 16-bit channel width in both master and slave modes. The
channel width determine s how many data bits are transferred in a single codec transaction per channel.
Refer to Table 1-7 for typical codec applications.
Table 1-7: Typical Codec Applications
Figure 1-8: Codec Protocol-Short Frame—8-Bit Channel Width
Application Codec
Type
No. of
Channels
Master/
Slave
Channel
Width
(No. Bits)
Long/
Short
Frame
Protocol
Bit Rate
CCLK
Freq.
(MHz)
Sample
Rate (Hz)
No. of
Frame
Syncs
Analog µ-Law single 1 Master 8 short or
long 1 2.048 8000 1
ISDN—8 bit
digital—A-Law dual 2 Slave 8 short 1 or 2 2.048 8000 1
Linear single 1 Master 16 short 1 2.048 8000 1
IOM-2/GCI single or
dual 1–2 Sl ave 8 short 1 or 2 1.536 8000 1
266
Compatibility single or
dual 1 or 2 Master 8 long or
short 1 2.048 8000 1 or 2
1—HARDWARE
ISD-T360SB
1-12 Voice Solutions in Silicon™
Figure 1-9: Codec Interface with One Single Codec, NSC TP3054, for Single Line Operation
Figure 1-10: Codec Interfa ce Diagr am w ith Tw o, Sing le Codec s, NSC TP3054, and NS C TP3057,
for Speakerphone Operation
1—HARDWARE ISD-T360SB
1-13
ISD
Figure 1-11: Codec Interface for Dual Line or Single Line and Speakerphone Operation
with OKI Dual Codec
Figure 1-12: Codec Interface for Dual Line or Single Line and Speakerphone
with Lucent Dual Codec
1—HARDWARE
ISD-T360SB
1-14 Voice Solutions in Silicon™
Figure 1-13: Codec Interface with one single Mot or ola codec, MC145482, for single line
operation
Figure 1-1 4: C odec Interface with two s ingle Motorol a co decs, MC145482, for speakerphone
operation
1—HARDWARE ISD-T360SB
1-15
ISD
Figure 1-15: Codec Interface for Dual Line or Single Line and Speakerphone Operation with
Winb ond Dual Codec
NOTE: To use this codec, you must define the codec interface to operate with two single codecs.
1—HARDWARE
ISD-T360SB
1-16 Voice Solutions in Silicon™
1.3 SPECIFICATIONS
1.3.1 ABSOLUTE MAXIMUM RATINGS NOTE Absolute maximum ratings indicate limits
beyond which permanent damage may
occur. Continuous operat ion at these li mits
is not intended; operation should be limited
to the conditions specified below.
1.3.2 ELECTRICAL CHARACTERISTICS
TA = 0ºC to +70ºC, V CC = 5 V ±10% Or 3.3 V ±10%, GND = 0 V
Storage temperature –65°C to +150°C
Temperature under bias 0°C to +70°C
All input and output voltages
with respect to GND –0.5 V to +6.5 V
Table 1-8: Electrical Characteristics
(All Para me ters with Refere nc e t o VCC = 3.3 V)
Symbol Parameter Conditions Min Typ Max Units
CXX1 and X2 Capacitance117.0 pF
ICC1 Active Supply Current Normal Operation Mode,
Running Speech
Applications2
40.0 80.0 mA
ICC2 Standby supply current Normal Operation Mode,
DSPM Idle230.0 mA
ICC3 Power-down Mode Supply
Current Power -down Mod e2,3 12.0 mA
ILInput Load Current 0 V £ VIN £ VCC –5.075.0 mA
IO (Off) Output Leakage Current (I/O
pins in Input Mode) 0 V £ VOUT £ VCC –5.075.0 mA
tWRa WR0 Active After R.E. CTTL, T1 tCTp/
2+2
tWRCSh WR0 Hold after EMCS4R.E. EMCS R.E. to R.E. WR0 7.5
tWRh WR0 Hold After R.E. CTTL tCTp/2 – 6
tWRia WR0 Inactive After R.E. CTTL, T3 tCTp/
2+2
VENVh ENV0 Input, High Voltage 2.0 V
VHh CMOS Input with Hysteresis,
Logical 1 Input Voltage 2.1 V
VHl CMOS Input with Hysteresis,
Logical 0 Input Voltage 0.8 V
VHys Hysteresis Loop Width10.5 V
1—HARDWARE ISD-T360SB
1-17
ISD
1. Guaranteed by design.
2. IOUT =0, TA 25°C, VCC =3.3 V for V
CC pins and 3.3 V or 5 V on VCCHI pins, operating from a 4.096 MHz crystal
and running from internal memory with Expansion Memory disabled.
3. All input signals are tied to 0 (above VCC – 0.5 V or below VSS + 0.5 V), except ENV0, which is tied to VCC.
4. Measured in power-down mode. The total current driven, or sourced, by all the VoiceDSP processor’s output signals
is less than 50 µA.
5. Guaranteed by design, but not fully tested.
6. CLKIN signal is not 5V tolerant.
7. Negative hold times are allowed since they are relative to the internal signal CTTL
VIH TTL Input, Logical 1 Input
Voltage 2.0 VCC +
0.5 V
VIL TTL Input, Logical 0 Input
Voltage –0.570.8 V
VOH Logical 1 TTL, Output Voltage IOH = –0.4 mA 2.4 V
VOHWC MMCLK, MMDOUT and EMCS
Logical 1, Output Voltage IOH = –0.4 mA 2.4 V
IOH = –50 µA5VCC – 0.2 V
VOL Logical 0, TTL Output Voltage IOL = 4 mA 0.45 V
IOL = 50 µA50.2 V
VOLWC MMCLK, MMDOUT and EMCS
Logical 0, Output Voltage IOL = 4 mA 0.45 V
IOL = 50 µA50.2 V
VXH CLKIN Input, High Voltage External Clock62.0 V
VXL CLKIN Input, Low Voltage External Clock60.8 V
Table 1-8: Electrical Characteristics
(All Para me ters with Refere nc e t o VCC = 3.3 V)
Symbol Parameter Conditions Min Typ Max Units
1-18
1—HARDWARE
ISD-T360SB
Voice Solutions in Silicon™
1.3.3 SWITCHING CHARACTERISTICS
Definitions
All timing specifications in this section refer to 0.8 V or 2.0 V on the rising or falling edges of the signals, as
illustrated in Figure 1-16 through Figure 1-22, unless specifically stated otherwise.
Maximum times assume capacitive loading of 50pF. CLKIN crystal frequency is 4.096 MHz.
NOTE CTTL is an internal signal and is used as a reference to explain the timing of other signals. See
Figure 1-31.
Figure 1-16: Synchronous Output Signals (Valid, Active and Inactive)
NOTE: Signal valid, active or inactive time, after a rising edge of CTTL or MWCLK.
Figure 1-17: Synchronous Output Signals (Valid)
NOTE: Signal valid time, after a falling edge of MWCLK.
2.0V
0.8V
CTTL or
MWCLK
Signal
tSignal
2.0V
2.0V
0.8V
MWCLK
Signal
tSignal
0.8V
1—HARDWARE ISD-T360SB
1-19
ISD
Figure 1-18: Synchronous Output Signals (Hold), after Rising Edge of CTTL
NOTE: Signal hold time, after a rising edge of CTTL.
Figure 1-19: Synchronous Output Signals (Hold), after Falling Edge of MWCLK
NOTE: Signal hold time, after a falling edge of MWCLK.
Figure 1-20: Synchronous Input Signals
NOTE: Signal setup time, before a rising edge of CTTL or MWCK, and signal hold time after a rising edge of CTTL or
MWCK
2.0V
0.8V
CTTL
Signal
2.0V
tSignal
2.0V
0.8V
MWCLK
Signal
tSignal
2.0V
2.0 V
CTTL or
MWCLK
2.0 V
0.8 V
2.0 V
0.8 V
tSignal Setup tSignal Hold
Signal
1—HARDWARE
ISD-T360SB
1-20 Voice Solutions in Silicon™
Figu re 1-21: Asyn chronous Si gnals
NOTE: Signal B starts after rising or falling edge of signal A.
The RESET has a Schmitt trigger input buffer. Figure 1-22 shows the input buffer characteristics.
Figure 1-22 : Hyste r esis Input Charac teristics
2.0 V
0.8 V
2.0 V
0.8 V
Signal A
Signal B
tSignal
Vout
VHh
VHl
VHys
1—HARDWARE ISD-T360SB
1-21
ISD
1.3.4 SYNCHRONOUS TIMING TABLES
In this section, R.E. means Rising Edge and F.E. means Falling Edge.
1. In normal operation mode, tCTp must be 30.5 ns; in power-down mode, tCTp must be 50,000 ns.
2. Guaranteed by design, but not fully tested.
3. Negative hold times are allowed since they are relative to the internal signal CTTL.
Table 1-9: O u tput Sig n als
Symbol Figure Description Reference Conditions Min (ns) Max (ns)
tAlh [10:0] Address Hold After R.E. CTTL -1.53
tAhh [15:11] Address Hold After R.E. CTTL 0.0
tAv Address Valid After R.E. CTTL, T1 9.0
tCCLKa CCLK Active After R.E. CTTL 12.0
tCCLKh CCLK Hold After R.E. CTTL 0.0
tCCLKia CCLK Inactive After R.E. CTTL 12.0
tCDOh CDOUT Hold After R.E. CTTL -2.03
tCDOv CDOUT Valid After R.E. CTTL 12.0
tCTp CTTL Clock Period1R.E. CTTL to next R.E. CTTL 30.5 250,000
tEMCSa EMCS Active After R.E. CTTL, T2W1 12.0
tEMCSh EMCS Hold After R.E. CTTL 0.0
tEMCSia EMCS Inactive After R.E. CTTL T3 12.0
tFSa CFS0 Active After R.E. CTTL 25.0
tFSh CFS0 Hold After R.E. CTTL -2.53
tFSia CFS0 Inactive After R.E. CTTL 25.0
tMWDOf MICROWIRE Data Float1After R.E. MWCS 70.0
tMWDOh MICROWIRE Data Out Hold2After F.E. MWCLK 0.0
tMWDOnf MICROWIRE Data No Float2After F.E. MWCS 0.0 70.0
tMWDOv MICROWIRE Data Out Valid2After F.E. MWCLK 70.0
tMWITOp MWDIN to MWDOUT Propagation Tim e 70.0
tMWRDYa MWRDY Active After R.E. of CTTL 0.0 40.0
tMWRDYia MWRDY Inactive After F.E. MWCLK 0.0 70.0
tPCh PC hold After R.E. CTTL -0.53
tPABh PA, PB and MWRQST hold After R.E. CTTL, T2W1 -3.03
tPABCv PB and MWRQST After R.E. CTTL, T2W1 12.0
1—HARDWARE
ISD-T360SB
1-22 Voice Solutions in Silicon™
1. Guaranteed by design, but not fully tested in power-down mode.
2. Guaranteed by design, but not fully tested.
Table 1-10 : Input Signals
Symbol Figure Description Reference Conditions Min (ns) Max (ns)
tCCLKSp Codec Clock Period (slave) R.E. CCLK to next R.E. CCLK 244
tCCLKSh Codec Clock High (slave) At 2.0 V (both edges) 120
tCCLKSl Codec Clock Low (slave) At 0.8 V (both edges) 120
tCDIh CDIN Hold After R.E. CTTL 0.0
tCDIs CDIN Setup Before R .E. CTTL 25.0
tCFS0Ss CFS0 Setup Before R.E. CCLK TBD
tCFS0Sh CFS0 Hold After R.E. CCLK TBD
tDIh Data in Hold (D0:7) After R.E. CTTL T1, T3 or TI 0.0
tDIs Dat a in Setup (D 0:7 ) Before R.E. CTTL T1 , T3 or TI 19.0
tMWCKh MICROWIRE Clock High (slave) At 2.0 V (both edges) 100.0
tMWCKl MICROWIRE Clock Low (slave) At 0.8 V (both edges) 100.0
tMWCKp MICROWIRE Clock Period
(slave)1R.E. MWCLK to next R.E. MWCLK 2.5 µs
tMWCLKh MWCLK Hold After MWCS becomes inactive 50.0
tMWCLKs MWCLK Setup Before MWCS becomes active 100.0
tMWCSh MWCS Hold After F .E . MW C L K 75.0
tMWCSs MWCS Setup Before R.E. MWCLK 100.0
tMWDIh MWDIN Hold After R.E. MWCLK 50.0
tMWDIs MWDIN Setup Before R.E. MWCLK 100.0
tPWR Power Stable to RESET R.E.2After VCC reaches 4.5 V 30.0 ms
tRSTw RESET Pulse Width At 0.8 V (both edges) 10.0 ms
tXh CLKIN High At 2.0 V (both edges) tX1p/2 – 5
tXl CLKIN Low At 0.8 V (both edges) tX1p/2 – 5
tXp CLKIN Clock Period R.E. CLKIN to next R.E. CLKIN 244.4
1—HARDWARE ISD-T360SB
1-23
ISD
1.3.5 TIMING DIAGRAMS
Figure 1-23: ROM Read Cycle Timing
1. This cycle may be either TI (Idle), T3 or T3H.
2. Data can be driven by an external device at T2W1, T2W, T2 and T3.
3. This cycle may be either TI (Idle) or T1.
Figu re 1- 24: Cod e c Shor t Frame Timin g
NOTE: The CCLK and CFS0 timing is shown for Master Mode only. For Slave Mode, see Figure 1-26.
CTTL
CCLK
CFS0/
CDOUT
CDIN
tCTp tCTp
tCTp tCTp tCTp
tCTp 4xtCTp tCTp tCTp
tCTp tCTp 4xtCTp
tCCLKa
tCCLKia
tFSa
tCTp tCTp
CFS1 tFSia
tCDIs tCDIh
BIT 7
BIT 7
tCDOv
tCDOh
tCCLKh tCCLKh
tFSh tFSh
tCTp
tCTp
4xtCTp tCTp tCTp tCTp
tCTp
4xtCTp tCTp tCTp
tCDOv
1—HARDWARE
ISD-T360SB
1-24 Voice Solutions in Silicon™
Figure 1-25: Codec Long Frame Timing
NOTE: The CCLK and CFS0 timing is shown for Master Mode only. For Slave Mode, see Figure 1-26.
Figure 1-26: Slave Codec CCLK and CFS0 Timing
NOTE: For CFS1, CDIN, CDOUT timing, see Figure 1-24 and Figure 1-25.
CTTL
CCLK
CFS0/
CDOUT
CDIN
tCTp tCTp
tCTp tCTp tCTp
tCTp 4xtCTp tCTp tCTp
tCTp tCTp 4xtCTp
tCCLKa
tCCLKia
tFSa
tCTp
CFS1 tFSia
BIT 0
tCDOv
tCDOh
tCCLKh tCCLKh
tFSh tFSh
tCTp
tCTp
4xtCTp tCTp tCTp
tCTp
BIT 7
BIT 7
CCLK
CFS0 tCFS0Ss tCFS0Sh
tCCLKSh tCCLKSl
tCCLKSp
1—HARDWARE ISD-T360SB
1-25
ISD
Figur e 1- 27: MICROWIR E Tran s ac tion Timing — D ata Transmitted to Ou tput
1—HARDWARE
ISD-T360SB
1-26 Voice Solutions in Silicon™
Figur e 1- 28: MICROWIRE Transaction Tim in g — D ata Echoed to Ou tput
Figure 1-29: Output Signal Timing for Port PB and MWRQST
NOTE: This cycle may be either TI (Idle), T2, T3 or T3H.
1—HARDWARE ISD-T360SB
1-27
ISD
Figure 1-30: CLKIN Timing
Figure 1-31: CTTL Timing
Figure 1-32: Reset Timing When Reset Is Not at Power-Up
1—HARDWARE
ISD-T360SB
1-28 Voice Solutions in Silicon™
Figure 1-33: Reset Timing When Reset Is at Power-Up
2—SOFTWARE ISD-T360SB
2-1
ISD
Chapter 2—SOFTWARE
The VoiceDSP software resides in the on-chip
ROM. It includes DSP-based algorithms, system
support functions and a software interface to
hardware peripherals.
2.1 SYSTEM OPERATION
This sect ion prov ides details of the syst em su pport
functions and their principle operation. It is divided
into the followin g subj ects:
• The State M ach i n e
• Command Executio n
• Event Handling
• Message Handling
• Tone Generation
• Initial izatio n and Co nfi gura ti o n
• Power Down Mode (PDM)
2.1.1 THE STATE MACHINE
The ISD-T360SB operates in two modes, normal
mode (DTAD) and speakerphone mode. To
change the mode use the Set Speakerphone
Mode (SSM) command. The VoiceDSP processor
functi ons as a state m ach ine u nder ea ch m ode.
It changes state either in response to a command
sent by the microcontroller, after execution of the
command is completed, or as a result of an inter-
nal event (e.g. memory full or power failure). For
more information see “VoiceDSP PROCESSOR
COMMANDS—QUICK REFERENCE TABLE” on pa ge
2-22. The VoiceDSP state s in DTAD mo de are de-
scribed below.
RESET
The VoiceDSP processor is initialized to the RESET
state a fter a ful l hardware re set by the RESET signal
(See “RESETTING” on page 1-3). In this state the
processor detectors (VOX, constant energy, call
progress tones and DTMF) ar e not acti ve. In al l oth-
er states, these detectors are active. (See the SDET
and RDET commands for further detail s ) .
IDLE
This is the state from which most commands are
executed. A s soon as a command a n d al l its pa-
rameters are received, the processor starts exe-
cuting the command.
RECORD
In this state a message is compressed (unless
stored in PCM format) and recorded into the mes-
sage memory.
PLAY
In this state a message is decompressed (unless
stored in PC M form at), and played back.
SYNTHESIS
An individual word or a sentence is synthesized
from an external v ocabular y in th i s state.
TONE_GENERATE
In the TONE_GENERATE state, the VoiceDSP pro-
cessor generates single or DTMF tones.
MSG_OPEN
The VoiceDSP processor either reads or writes 32
bytes to the message memory, or sets the mes-
sage read/write pointer on a 32 byte boundary.
POWER DOWN MODE
In this state, the power consumption is reduced for
the Voice DSP proce ssor. For more info rmati on, re-
fer to the Power Down Mode section on page 2-5.
2—SOFTWARE
ISD-T360SB
2-2 Voice Solutions in Silicon™
2.1.2 COMMAND EXECUTION
A VoiceDSP processor command is represented
by an 8-bit opcode. Some commands require
parameters. Some commands have return val-
ues. Commands are either synchronous or asyn-
chronous.
SYNCHRONOUS COMMANDS
A synchronous command must complete execu-
tion before the microcontroller can send a new
command (e.g. GMS, GEW). A command se-
quence begins when the microcontroller sends
an 8-bit opcode to the processor, followed by the
command’s parameters (if any). The VoiceDSP
pro cesso r then execu tes the comma nd an d, if r e-
quired, transmits a return value to the microcon-
troller. Upon completion, the processor notifies the
microcontroller that it is ready to accept a new
command.
ASYNCHRONOUS COMMANDS
An asynchronous command starts execution in
the background and notifies the microcontroller,
which ca n send more commands while the cur-
rent command is still running (e.g. R, P). After re-
ceiving an asynchronous command, such as P
(Playback), R (R ecord), SW (Say Word s) or GT (Gen-
erate Tone), the VoiceDSP processor switches to
the appropriate state and executes the com-
mand until finished or a S (Stop) or PA (Pause)
command is received from the microcontroller.
When compl eted, the E V_NORMAL _END event is
set and the processor switches to the IDLE state.
“VoiceDSP PROCESSOR COMMANDS—QUICK REF-
ERENCE TABLE” on page 2–22 displays all the pro-
cessor commands, the valid source states in
which these com mand s are vali d, and the st ates
resul ting from the command.
2.1.3 EVENT HANDLING
STAT U S WORD
The 16-bi t sta tu s wor d indicat es ev ents that occur
during nor m al operati on. Th e V oi ceDS P pr ocessor
activates the MWRQST signal, to indicate a
change in the status word. This signal remains ac-
tive until the processor receives a GSW (Get Status
Word) command.
For detailed description of the Status Word and
the meaning of each bit, see “GSW Get Status
Word” on page 2-34 .
ERROR WORD
The 16-b it error wor d indicates er rors that occurr ed
during execution of th e last comman d. If an er r or
is det ected, the command is n ot proce ssed; th e
EV_E RROR bit in th e sta tus wor d is set to 1 , and the
MWRQST signal is activated.
ERROR HANDLING
When the microcontroller detects the active
MWRQST signal, it issu es the GSW co mmand, de-
activating the MWRQST signal. Then, the micro-
controller tests the EV_ERROR bit in the status word,
and, i f set, sends the GEW (Get Erro r Word) c om-
mand to read the er ror word for detai ls .
For detailed description of the Error Word and the
meaning of each bit, se e “GEW Get Er ror Word” on
page 2-31.
2—SOFTWARE ISD-T360SB
2-3
ISD
2.1.4 MESSAGE HANDLING
A message is the b asi c uni t on whi ch mo st of t he
VoiceDSP commands operate. A VoiceDSP pro-
cessor message, stored on a flash memory de-
vice, c an be reg arded as a c omputer file stor ed
on a flash mass-storage device.
The ISD-T360SB manages messages for a wide
range of app lica tion s, which requ ire di ffe ren t lev -
els of DTAD functionality. The Voic eDSP processor
features advanced memory-organization fea-
tures such as multiple OutGoing Messages
(OGMs), mailboxes, and the ability to distinguish
between InComing Messages (ICMs) and OGMs.
A message is created with either the R (Record) or
the CMSG (Create Message) command. Once
created, the message is assigned a time-and-
day stamp a nd a message tag which is rea d by
the micr ocontrol ler. Th e R comm and takes vo ice
samples from the codec, compresses them, and
stores them in the message memory.
When a message is created with the CMSG com-
mand the data to be recorded is provided by the
microcontroller, via the WMSG (Write Message)
command and not throug h the c ode c. Here , the
data is transf erred di rectly to th e message mem-
ory, and not compressed by the ISD-T360SB voice
compression algorithm.
WMSG, RMSG (Read Message) and SMSG (Set
Message Pointer) are message-data access
commands used to store and read data to or
from any location in the message memory (see
“VoiceDSP PROCESSOR COMMANDS—QUICK REF-
ERENCE TABLE” on page 2-22 for more details). Us-
ing these commands, the microcontroller can
utiliz e message s f or feature s s u ch as a Te le phone
Directory and Caller Numbers List (Caller IDs of
those who called but did not leave a message.)
A message can be played back (P command)
and deleted (DM command). Redundant data
(e.g. trailing tones or silence) can be removed
from the message tail with the CMT (Cut Message
Tail) command.
The PA (Pause) and RES (Resume ) c ommand s, re-
spectively, suspend the P (Playback) and R
(Record) commands, and resume them from the
point at which they were suspended.
CURRENT M ESSAGE
The GTM (Get Tagged Message) command se-
lects the current message. Most message han-
dling commands (P, DM, RMSG) operate on the
current message.
Deleting the current message does not cause a
different message to become current; the current
message is undefi ned. If you issue the GTM com-
mand to skip to the next message, the first mes-
sage, newer than the just deleted message,
becomes the current message.
MESSAGE TAG
Each mes sage ha s a 2-byte message t ag which
is used to categorize messages, and implement
feature s such as OutGoing Messages, mail boxes
and different handling of old and new messages.
The tag is created during the R (Record) com-
mand. Use the GMT (Get Message Tag) and SMT
(Set Message Tag) commands to handle mes-
sage tags.
NOTE Message tag bits can only be cleared and
are set only when a message is first cre-
ated. This limitation, inherent in Flash
memories, allows bits to be changed only
from 1 to 0 (changing bits from 0 to 1
requires a special erasure procedure).
However, the usual reason for updating an
existing tag is to mark a message as old.
This can be done when a message is first
created by using one of the bits as a new/
old indicator, setting the bit to 1 and later
clearing it when necessary.
2—SOFTWARE
ISD-T360SB
2-4 Voice Solutions in Silicon™
2.1.5 TONE GENERATION
The VoiceDSP processor generates DTMF tones
and single-frequency tones from 300Hz to 3000Hz
in increments of 100Hz. The ISD-T360SB tone gen-
eration conforms to the EIA-470-RS standard.
Note, however, that value of some tunable pa-
rameters may need adjusting to meet the stan-
dard specifications since the energy level of
generated tones depends on the analog circuits
used.
1. Tune the DTMF_GEN_TWIST_LEVEL parame-
ter to control the twist level of the generat-
ed DTMF tones.
2. Use the VC (Volume Control) command,
and tune the TONE_GEN_LEVEL parameter,
to control the energy level at which these
tones are generated.
3. Use the GT (Generate Tone) command to
specify the DTMF tones, and the frequency
at which single tones are generated.
Refer to Table 2-5, VC command and GT com-
mand of the Command Description section for
further details of the relevant tunable parameters
and command s.
NOTE The DTMF detector performance is
degraded during tone generation, espe-
cially if the frequency of the generated
tone is close to the frequency of one of
the DTMF tones.
2.1.6 INITIALIZATION AND CONFIGURATION
Use the following procedures to initialize the
VoiceDSP processor:
NORMAL INITIALIZATION
Reset the VoiceDSP processor by activating the
RESET signal. (See “RESETTING” on page 1-3.)
1. Issue a CFG (Configure VoiceDSP proces-
sor) command to change the configura-
tion acc ording to you r env i ronmen t.
2. Issue an INIT (Initialize System) command to
initialize the VoiceDSP firmware.
3. Issue a ser ies of TUNE com mands to ad just
the VoiceDSP processor to the require-
ments of your system.
NOTE The tuning of NUM_OF_BLOCKS_IN_MEM
(tunable parameter number 62) should be
done before t he CFG command. This tune
operati on should be execut ed when using
flash memory other than 4Mb. The tuning
of NUM_OF_BLOCKS_FOR_TEST (tunable
parameter number 63) should be done
before the INIT command.
TUNABLE PARAMETERS
The VoiceDSP processor can be adjusted to the
specific system’s requirements using a set of tun-
able parameters. These parameters are set to
their default values after reset and can be later
modified with the TUNE command. By tuning
these parameters, you can control various as-
pects of the VoiceDSP processor’s operation, such
as silence com pression, tone dete ction, and no-
energy de tection.
Table 2-4 to Table 2-12 of the Command Descrip-
tion section describes all the tunable parameters
in detail.
2—SOFTWARE ISD-T360SB
2-5
ISD
2.1.7 POWER-DOWN MODE
The PDM (Go To Power-Down Mode) command
switches the ISD-T360SB to power-down mode.
The purpose of the PDM command is to save
power during battery operation, or for any other
power saving cause. During power-down mode
only basic functions, such as time and date up-
date, are ac tive (for mo re deta ils re fer to POWER -
DOWN MODE description on page 1-4) .
This PDM command may only be issued when the
processor is in IDLE mode (for an explanation of
the ISD-T360SB states, see “The State Machine” on
page 2-1). If it is necessary to switch to power-
down mode from any other state, the controller
must first issue a S (Stop) command to switch the
processor to the IDLE state, and then issue the
PDM command. Sending any command while in
power-down mode resets the VoiceDSP processor
detectors, and returns it to normal operation
mode.
NOTE Entering or exiting power-down mode can
distort the real-time clock by up to 500 µs.
Thus, to maintain the accuracy of the real-
time clock, enter or exit the power-down
mode as infrequently as possible.
2.2 PERIPHERALS
This section provides detai ls of the peri pherals in-
terface support functions and their principle oper-
ation. I t i s divided into th e followi n g sub je cts:
• Mi crocontroller Interface (Slav e MI CROWIRE)
• Memory In terfac e
• Codec Interface
2.2.1 MICROCONTROLLER INTERFACE
MICROWIRE/PLUS™ is a synchronous serial com-
munication protocol minimizes the number of
connecti ons, and thus t he cost, of com municat-
ing with peri ph erals .
The VoiceDSP MICROWIRE interface implements
the MICROWIRE/PLUS interface in slave mode, with
an additional ready signal. It enables a microcon-
troller to interface efficiently with the VoiceDSP pro-
cessor application.
The microcontroller is the protocol master and
provides the clock f or t he p rotoc ol. The Voi ceDSP
processor supports clock rates of up to 400 KHz.
This transfer rate refers to the bit transfer; the actual
throughput is slower due to byte processing by the
VoiceDSP processor and the microcontroller.
Communication is handled in bursts of eight bits
(one byt e). In each bur st the Voice DSP processor
is able to receive and tr ansmit eig ht bits of data.
After eight bits have been transferred, an internal
interrupt is issued for the VoiceDSP processor to
process the byte, or to prepare another byte for
sending. In parallel, the VoiceDSP processor sets
MWRDY to 1, to s ignal the mi croco ntrol ler t hat i t is
busy with the byte processing. Another byte can
be transferred only when the MWRDY signal is
cleared to 0 by the VoiceDSP processor. When the
VoiceDSP processor transmits data, it expects to
receive the value 0xAA before each transmitted
byte. The VoiceDSP processor reports any status
change by clearing the MWRQST si gna l to 0.
If processor command’s parameter is larger than
one byte, the microcontroller transmits the Most
Significant Byte (MSB) first. If a return value is larger
than one byte, the VoiceDSP processor transmits
the MSB first.
The following signals are used for the interface
protocol. Input and output are relative to the
VoiceDSP processor.
INPUT SIGNALS
MWDIN
MICROWIRE Data In. Used for input only, for trans-
ferring data from the microcontroller to the
VoiceDSP processor.
2—SOFTWARE
ISD-T360SB
2-6 Voice Solutions in Silicon™
MWCLK
MICROWIRE Clock. Serves as the synchronization
clock during communication. One bit of data is
transferred on every clock cycle. The input data is
available on MWDIN and is latched on the clock
rising edge. The transmitted data is output on
MWDOUT on the clock falling edge. The signal
should re ma in low when switchin g MWCS .
MWCS
MICROWIRE Chip Select. The MWCS signal is
cleared to 0, to indicate that the VoiceDSP pro-
cessor is being accessed. Setting MWCS to 1
causes the VoiceDSP processor to start driving
MWDOUT with bit 7 of the transmitted value. Set-
ting the MWCS signal resets the transfer-bit counter
of the pr otocol, so t h e signal can be u sed to sy n -
chronize between the VoiceDSP processor and
the micro con troller.
To prevent false detection of access to the
VoiceDSP p rocessor due to spikes on the MWCL K
signal, use this chip select signal, and toggle the
MWCLK input signal, only when the VoiceDSP pro-
cessor is accessed.
OUTPUT SIGNAL S
MWDOUT
MICROWIRE Data Out. Used for output only, for
transferring data from the VoiceDSP processor to
the microcontroller. When the VoiceDSP processor
receiv es data it is echoed back to th e m i crocon-
troller on this signal, unless the received data is
0xAA. In this case, the VoiceDSP processor echoes
a command’s retur n val u e.
MWRDY
MICROWIRE Ready. When active (0), this signal in-
dicates that the VoiceDSP processor is ready to
transfer (receive or transmit) another byte of data.
This signal is set to 1 by the VoiceDSP processor af-
ter ea ch byte tr ansf er has b een co mple ted. It re -
mains 1, while the VoiceDSP processor is busy
reading the by te, writ ing the next byt e, or execu t-
ing t he re ce ived c om mand ( afte r the las t para m-
eter has been received). MWRDY is cleared to 0
after reset. For proper operation after a hardware
reset, this signal should be pulled up.
MWRQST
MICROWIRE Request. When active (0), this signal
indic ates t hat new s tat us i nfo rmati on i s av aila bl e.
MWRQST is deactivated (set to 1), after the
VoiceDSP processor receives a GSW (Get Status
Word) command from the microcontroller. After
reset, this signal is active (0) to indicate that a reset
occurred. MWRQST, unlike all the signals of the
communication protocol, is an asynchronous line
that is controlled by the VoiceDSP firmware.
SIGNAL USE IN TH E INTE RFACE PROT OC OL
Afte r reset, both MWRQ ST and MWRDY are cleared
to 0.
The MWRQ ST sign al is acti vate d to in dicat e that a
reset occ urred. The EV_R ESET bit in the stat us reg -
ister is u s ed to indicate a reset condit ion.
The GSW command should be issued after reset
to verify that the EV_RESET event occurred, and to
deactivate the MWRQST signal.
While th e MWCS signal is active (0), the VoiceDSP
processor read s data from MWDIN on ev ery r ising
edge of MWCLK. VoiceDSP processor also writes
every bit back to MWDOUT. This bit is either the
same bit which was read from MWDIN (in this case
it is written back as a synchronization echo after
some propa gation del ay), or it is a bit of a value
the VoiceDSP processor transmits to the micro-
controller (in this case it is written on every falling
edge of the clock).
When a co mmand ha s more than on e para me-
ter/r eturn-val ue, the pa rameters/ return-values are
transmitted in the order of appearance. If a pa-
rameter/return-value is more than one byte long,
the bytes a re tran smitt ed fro m the mo st sig nifican t
to the least significant.
2—SOFTWARE ISD-T360SB
2-7
ISD
The MWRDY sign al is use d as follows :
1. Active (0) MWRDY signals the mi croco ntrol -
ler tha t the las t eight bi ts of da ta transf erred
to/fr om the vo ice module were ac cepted
and processed (see below).
2. The MWRDY signal is deactivated (set to 1
by the VoiceDSP processor) after 8-bits of
data were transferred to/from the VoiceDSP
processor . The bit is set following the falling
edge of the eighth MWCLK clock-cycle.
3. The MWRDY signal is activated (cleared to
0) by the VoiceDSP processor when it is
ready to r eceive the fi rst parameter byte (if
there are any parameters) and so on till the
last byte of parameters is transferred. An
active MWRDY signal after the last byte of
parameters indicates that the command
was parsed and (if possible) executed. If
that co mman d has a ret urn val ue, t he mi -
crocontroller must read the value before is-
suing a new command.
4. When a return value is transmitted, the
MWRDY signal is deactivated after every
byte, and activated again when the
VoiceDSP processor is ready to send anoth-
er byte, or to receive a new command.
5. The MWRDY signal is activated (cleared to
0) after reset, and after a protocol time-out.
(See “INTERFACE PROTOCOL TIME-OUTS”)
The MWRQ ST signal is used as follows:
1. Th e MWRQST signal is activated (cleared to
0), w hen the stat u s w ord is c h an ged.
2. The MWRQST signal remains active (0), until
the VoiceDSP processor receives a GSW
command.
Figure 1-27 and Figure 1-28 illustrate the se-
quence of activities during a MICROWIRE data
transfer between VoiceDSP and the microcon-
troller.
INTERFACE PROTOCOL TIME-OUTS
Depending on the VoiceDSP processor’s state, if
more than 100 milliseco nds elapse betwe en the
assertion of the MWRDY signal and the transmis-
sion 8th bit of the next byte pertaining to the same
command transact ion, a time-out event occur s,
and the VoiceDSP processor responds as follows:
1. Sets the error bit in the status word to 1.
2. S ets the EV _TIME OUT bit in the err or w ord t o
1.
3. Activa tes the MW RQST signal (clears it to 0).
4. A cti v ates the MWRDY signal (clear s it to 0).
5. Waits for a new command. (After a time-
out occurs, i.e., the microcontroller re-
ceived MWRQST during the command
transfer, or result reception, t he microcon-
troller must wait at least four milliseconds
before issu in g the n ext command.)
ECHO MECHANISM
The VoiceDSP processor echoes back to the mi-
crocontroller all the bits received by the VoiceDSP
processor. Upon detection of an error in the echo,
the microcontroller should stop the protocol
clock, which eventually causes a time-out error
(i.e., ERR_TIMEOUT bit is set in the error word).
For commands that have a return value, the mi-
crocontroller must transmit a byte value of 0xAA
for each byte returned. In response, the VoiceDSP
processor transmits the return values instead of
echoing th e 0xA A byte.
Upon det ection of an error the Voice DSP proces-
sor activates the MWRQST signal, and sets the
ERR_COMM bi t in the error word.
2—SOFTWARE
ISD-T360SB
2-8 Voice Solutions in Silicon™
2.2.2 MEMORY INTERFACE
DEVICE NUMBER AND TYPE
The Voice DSP processor su pports va rious types of
Flash memory devices. Up to four devices may be
connected to the VoiceDSP, where all the con-
nected d evices must be of the same t ype. Ea ch
memory device may be of 4Mbit, 8Mbit or
16Mbit; thus a total of 64Mbit non-volatile memo-
ry may be connected for message storage (up to
4 hours of voice recording).
See “MEMORY INTERFACE” on page 1-6, for de-
tailed de scription of the supported Flash and the
hardware connectivity.
Use the CFG command to define the type and
number of installed memory devices (see “CFG
Configure VoiceDSP config_value” on page 2-26).
MEMO RY D E VICE SIZ E
The memory manager handles the memory de-
vices in basic units of 4Kbyts blocks. This approach
is defined due to the nature of Flash devices
where the basic unit that can be erased is a
4Kbytes blo ck.
Memory blocks cannot be shared by different
voice messages. Therefore, the maximum num-
ber of messages per memory device, equals to
the number of memory blocks minus one (one
block per device is used for memory manage-
ment).
The size of the connected memory devices, is de-
fined by the number of memory blocks in each
device. Refer to tunable parameter index 62, in
Table 2-10, for detailed description of the avail-
able number of blocks for Flash.
PRODUCTION LINE TESTING
In many cases it is desired to test the ISD-T360SB in
the pr oduction li ne as part of the whole applica -
tion. Usua lly in these case s, the testing time is an
important factor and should be minimized as pos-
sible. The initialization time of the memory devices
is significant and should be avoided during pro-
duction (Refer to Table 1-4). Therefore, a dedicat-
ed parameter is defined in order to allow a
producti on line testi ng while usi ng a small part of
the real co nnect ed mem ory siz e.
It should be noted that in case of power failure
during the production line testing, the connected
memory devices should be replaced, and the
process shou ld be repeated. R efer to parameter
index 63, in Tabl e 2-10, for further explanatio n of
the prod u ction line testing.
2—SOFTWARE ISD-T360SB
2-9
ISD
2.2.3 CODEC INTERFACE
SUP PORTED FUN CTIONALI TY
The VoiceDSP processor supports analog and dig-
ital telephony in various configurations. For analog
telephony the VoiceDSP operates in master
mode, where it provides the clock and the syn-
chronization signals. It supports a list of single
channel and dual channel codecs, as listed in
Table 1-6. For digital telephony the VoiceDSP op-
erates in slave mode, where the control signals
are provided by an extern al source.
The codec interface is designed to exchange
data in short frame format as well as in long frame
format. The channel width may be either 8 bits (u-
Law format or A-Law format), or 16 bits (linear for-
mat). In slave mode the clock may be divided by
two, if required (two bit rate clock mod e).
The VoiceDSP processor supports up to 2 voice
channels, where the line should be connected as
channel 0 (in master mode or in slave mode - de-
pends on the configuration), and the speaker-
phone (speaker and microphone) should be
connected as channel 1 or as channel 2, de-
pends on the configuration (channel 2 is always
connected as master).
See “The Codec Interface” on page 1-9, for de-
tailed d escri ption of the supp orted cod ec devic -
es and the har dware conn ect ivity.
Use the CFG command to define the codec
mode (master or slave), the data frame format
(short or long) , the channel width (8 bit s or 16 bits),
the cloc k bit rate ( sing le or dual) and the n umber
and type of codecs (one or two, single channel or
dual channel). See “CFG Configure VoiceDSP
config_ v al u e” on pag e 2-26.
DATA CHANNEL TIMING
To provide additional flexibility (e.g., when using
speakerphone), the codec interface provides a
programmable delay of data and synchroniza-
tion sign als, rela tive to the CFS0 signal.
Refer to tunable parameters index 65 to index 69,
in Table 2-11, for detailed description of the delay
registers and their significance.
2.3 ALGORITHM FEATURES
This section provides details of the VoiceDSP algo-
rithms and their principle operation. It is divided
into the followin g subj ects:
• VCD (Voice Compression and
Decompression)
• DTMF Detection
• Tone and Energy Detection (Call Progress)
• Speakerphone
• Speech Synthe sis
• SW Automatic Gain Control
2—SOFTWARE
ISD-T360SB
2-10 Voice Solutions in Silicon™
2.3.1 VCD (VOICE COMPRESSION AND
DECOMPRESSION)
The VoiceDSP processor implements a state of the
art VCD algorithm of the CELP family. The algo-
rithm provides 3 compression rates that can be
selected dynamically. PCM recording (no com-
pression) is al so pr ov ided.
The lowest co mpr ession r ate of 4.7 Kbi t/s enab le s
about 30 minutes of recording on an 8-Mbit de-
vice. The mid compr ession rate of 6.7 Kbit/s p ro-
vides about 20 minutes of voice recording time.
The highest compress ion rate of 8.7 Kbit/s, which
provides the best compressed voice quality,
stores approximately 10 minutes on an 8-Mbit de-
vice. For detailed information about recording
times refer to Table 1-5.
Before recording each message, the microcon-
troller select s one of th e th ree compre ssio n rate s,
or PCM re cordi ng, wit h the co mpre ssi on_rat e pa -
rameter of the R (Record) command. During mes-
sage playback the V oi ceDSP processor reads this
one byte pa ramete r and selec ts the ap propriat e
speech decompressi on algorithm.
IVS vocabularies can be pr epared in ei th er of the
three compression rates, or in PCM format, using
the IVS tool. All messages in a single vocabulary
must be recorded using the same algorithm. (See
the IVS User’s Guide for more details). During
speech synthesis, the VoiceDSP processor auto-
matically selects the appropriate speech decom-
pression al g or it hm.
VARIABLE SPEED PLAYBACK
This feat ure increases o r decreases the speed of
messages and synthesized messages during
playback. Use the SPS (Set Playback Speed) to set
the speed of m e ssag e playback. Th e new speed
applies to all recorded messages and synthesized
messages (only if synthesized using IVS), until
changed by another SPS command. If this com-
mand is issued while the VoiceDSP processor is in
the PLAY state, the speed also changes for the
message currently being played.
The speedup / slowdown algorithm is designed to
maintain the pitch of the original speech. This ap-
proach provides the same speech tone while
playback speed varies.
CONVERS AT ION RE C ORD ING / PLAY BACK
The VoiceDSP supports the conversation recording
feature, whic h en ables th e user to simultaneou sly
record the far end caller and near end caller. The
VoiceDSP simultaneously records data from chan-
nels 0 and 1 by summing the two channels and
then compressing the summation. The data is
then saved in the flash memory. When Playing the
compressed message, the data is simultaneously
broadca st to both channels. Th is feature is avail-
able for all compression rates. This feature can be
operated using the SLC command. For more de-
tails, refer to the SLC command in the Command
Description section.
PCM RECORDING
The VoiceDSP is capable of recording data in
PCM fo rmat (th at is t he or iginal sam ples f ormat ei-
ther in 8 bit u- L aw format, 8 bit A - law format or 16
bit linear format). The PCM data uses more stor-
age space , but it provides the highest quality for
OGM, music-on-hold or IVS data. The PCM record-
ing may be selected as one of the available
compression rates during the R command
(compression_rate = 0).
Variabl e Speed Pla yback is not acc essi ble dur ing
PCM recording and playback.
2-11
2—SOFTWARE ISD-T360SB
ISD
2.3.2 DTMF DETECTION
The VoiceDSP processor detects DTMF tones, thus enabling remote control operations. Detection is active
throughout the ope ration of the Voi ceDSP processor. Dete ction can be conf igured using the SDET (Set
Detectors Mask) command, which controls the reporting of the occurrence of tones, and the RDET (Reset
Detector s) command wh ich reset s the detect ors. The accu racy of the tone lengt h, as reporte d by the
tone detectors, is ±10 ms.
DTMF detection may be reported at the starting point, ending point, or both. The report is made through
the status word (for further details, see GSW command).
For further details about tunable parameters refer to Table 2-6 of the Command description section. The
DTMF dete ctor perfor mance, as mea sure d on t he li ne inp ut us ing an ISD-DS360 -DAA board , is summa-
rized below in Table 2-1.
* Performance d epends on the DAA design. For reliable DTMF detection:
• A hardware echo canceller, that attenuates the echo by at least 6 dBm, is required during playback.
• The HW AGC, if present, must be disabled during playback.
1. Tune parameters 60 and 61 affect DTMF detection sensitivity. The detection sensitivity is -34 dBm when these
tunable parameters get their maximum value of 5.
2. Performance with echo canceler is 10 dB better than without echo canceler. For a silent message, detection
sensitivity is -34 dBm, with echo canceler.
3. The accuracy of reported DTMF tones is ±10 ms.
4. If the interval between two consecutive identical DTMF tones is less than, or equal to, 20 ms, the two are detected
as one long DT MF tone. If the interva l betwee n two consec utive identi cal DTMF tones is b etween 20 ms an d 45
ms, separate detection is unpredictable.
5. Determined by the DTMF_REV_TWIST tunable parameter value.
DTMF SW AGC
In order to remove the linkage between the HW AGC and the detection level of the DTMF detector, two
new t u nabl e pa ram ete rs ar e ad ded . T he se tun abl e par ame t ers de fin e th e gai n o f the SW AG C fo r DTM F
signals.
DTMF_DET_AGC_IDLE - SW AGC for DTMF detection in Idle and Record states. When incrementing this tun-
able by 1, the dynamic range is increased by 3 dB.
DTMF_DET_AGC_PLAY - SW AGC for DTMF detection in Play and Tone_Generate states. When increment-
ing thi s tu n abl e by 1, th e dynamic ran g e is increase d by 3 dB.
For more information on the SW AGC commands, refer to index values 60 and 61 in Table 2-6 of the
Command Description section.
Table 2-1: DTMF Detector Performance*
Play/IVS Synthesis Record/Idle
Detection Sensitivity1Performance depends on the message being played.2-34 dBm
Accepted DTMF Length3>50 ms >40 ms
Frequency Tolerance ±1.5% ±1.5%
S/N Ratio 12 dB 12 dB
Minimum Spacing4>50 ms >45 ms
Normal Twist 8 dB 8 dB
Reverse Twist54 dB or 8 dB 4 dB or 8 dB
2—SOFTWARE
ISD-T360SB
2-12 Voice Solutions in Silicon™
ECHO CANC ELLATION
Echo cancellation is a technique used to improve
the performance of DTMF detection during
speech synthesis, tone generation, and OGM
playback. For echo cancellation to work properly,
HW AGC must not be active in parallel. Thus, to
take ad van tag e o f ec ho can cel la ti on, th e mi cr o-
controll er must contro l the HW AGC, if exists, (i.e .,
disable the HW AGC during PLAY, SYNTHESIS and
TONE_GENER A TE states and e n able it again after-
wards). If HW AGC can not be disabled, do not
use echo ca n cel lation.
The microcontroller should use the CFG com-
mand to activate/deactivate echo cancellation.
NOTE Normally, a HW AGC is not required with
The ISD-T360SB, since SW AGC is optional
for the VCD al gorithm, DTMF d etection and
the speakerphone module.
2.3.3 TONE AND ENERGY DETECTION (CALL
PROGRESS)
The VoiceDSP processor detects busy and dial
tones, constant energy level, and no-energy
(VOX). This enab les call progress tracking. Detec-
tion is active throughout the operation of the
VoiceDSP processor. Detection can be config-
ured using the SDET (Set Detectors Mask) com-
mand, which controls the reporting of the
occurrenc e of tones, a nd th e R DET (Reset De tec -
tors) command which resets the detectors. The
accuracy of the tone length, as reported by the
tone detectors, is ±10 ms.
TUNABLE PARAMETERS
Tunable parameters control the detection of busy
and dial tones, constant energy level (in the fre-
quency range 200–3400Hz), and no-energy.
These parameters should be tuned to fit the sys-
tem hardware. In addition, changes may be re-
quired to the tunable parameters according to
the setting (On or Off) of the HW Automatic Gain
Control (HW AGC), if exists. For more information
refer to Table 2-7 and Table 2-8 of the Command
Description section.
Figure 2- 1: Bus y an d Dial-Tone Band-Pass Filt er Fre qu ency Respons e
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Frequency (Hz)
–40
–30
–20
–10
0
Magnitude dB
2—SOFTWARE ISD-T360SB
2-13
ISD
BUSY AND DIAL TONES
Busy and dial-tone detectors work with a band-
pass filter that limits the frequency range in which
tones can be detected to 0–1100Hz. Figure 2-2
shows the frequency response of this band-pass
filter.
The design of the busy-tone detector allows very
high flexibility in detecting busy tones with varying
cadences.
The tunable parameters are divided into four sets:
1. Busy Tone On-time and Off-time Range
Specification:
BUSY_DET_MIN_ON_TIME
BUSY_DET_MIN_OFF_TIME
BUSY_DET_MAX_ON_TIME
BUSY_DET_MAX_OFF_TIME
2. Busy T one Ca dence Contro l Specific ation
BUSY_DET_VERIFY_COUNT
BUSY_DET_DIFF_THRESHOLD
BUSY_DET_TONE_TYPE
BUSY_DET_VERIFY_COUNT determines the
number of On/Of f cadence s that det ector
should detect before reporting busy tone
presence.
BUSY_DET_DIFF_THRESHOLD describes the
maximum allowed difference between
two compared On or Off periods, as deter-
mined by the BUSY_DET_TONE_TYPE tunable
parameter.
BUSY_DET_TONE_TYPE specifies the type of
cadence s th at ar e su pported.
Legal value s are:
Two cadence s on ly
Three ca den ces only
Both two and three cadences.
The acceptance criteria for two cadences:
[E1–E3] < BUSY_ DET _DIF F _THRESHOLD
and
[S1–S3] < BUSY _DET _ DIFF_THRESHOLD
The acceptance criteria for three cadenc-
es:
[E1–E4] < BUSY_ DET _DIF F _THRESHOLD
and
[S1–S4] < BUSY _DET _ DIFF_THRESHOLD
3. Busy and D i al Tone Ener g y Thre s h olds
TONE_DET_ON_ENERGY THRESHOLD
TONE_ D ET _O FF_ENERG Y THRESHOL D
4. Busy Detection Time
BUSY_DET_MIN_TIME
Figure 2-2: Busy-Tone Detector—Default Cadence Specification
E1E2E3
S1S2S3
[E1 – E3] < 100 ms [S1 – S3] < 100 ms 100 < Ei < 1680 ms 70 < Si < 1220 ms
2—SOFTWARE
ISD-T360SB
2-14 Voice Solutions in Silicon™
CONSTANT ENER GY
The constant-energy detector reports the pres-
ence of constant energy in the range of 200Hz to
3400Hz. It is intended to detect both white and
pink noise and can be used to detect line d iscon-
nection during recording.
It is recommended to use the constant energy
mechanism in conjunction with the no-energy
(VOX) mechani sm .
The fo llowing tuna ble par ameters co ntrol the op -
eration of the constant-energy detector:
CONST_NRG_DET_TIME_COUNT
CONST_NRG_DET_TOLERANCE_TIME
CONST_NRG_DET_LOW_THRESHOLD
CONST_NRG_DET_HIGH_THRESHOLD
NO ENERGY (VOX )
The no-energy detector reports when the energy
in the frequency range of 200Hz to 3400Hz re-
mains below a pre-programmed threshold for a
pre-programmed time-out. A programmable tol-
erance is allowed.
It is recommended to use the no-energy (VOX)
mechanism in conjunction with the constant-en-
ergy mechan i s m .
The fo llowing tuna ble par ameters co ntrol the op -
eration of the no-energy (VOX) mechanism:
VOX_DET_ENERGY_THRESHOLD
VOX_DET_TIME_COUNT
VOX_DET_TOLERANCE_TIME
2.3.4 FULL-DUPLEX SPEAKERPHONE
The speakerphone f eature lets the use r communi-
cate through a telephone line, using the unit’s
speaker and microphone instead of its handset.
The speakerphone processes signals sent from the
line to the speaker, and from the m icrophone to
the line. It also performs the nece ssary switching,
attenuation and echo cancellation on the signals
pre s en t on the line/sp eaker.
The ISD-T360SB speakerphone is simple to use; it
requires no special hardware or training for the
echo cancelers. The gain control is fully digital,
which eliminates the need for analog gain control
hardware.
The speakerphone features two types of echoes,
the electrical echo (line or circuit) and the
acoustic echo. The electrical echo is a result of an
imperfect impedance match between the 4- to
2-wire in terface (h ybrid) an d the l i n e i m pedance.
The electrical echo, relatively short term, has a
transfer function that varies slowly. The second
echo, the acoustic echo, is a line impedance re-
turning from t he speaker to the microphone . This
echo is relatively long term, and its transfer func-
tion may vary quite quickly if anyone, or anything,
moves in the room. Both echoes must be can-
celed to achieve a high-quality hands-free
system.
For more details of the speakerphone tunable pa-
rameters refer to Table 2-9 of the Command De-
scription secti on.
SPEAKERPHONE TERMIN OLOGY
Send Path
The signal path from the microphone (near-end
speaker) to the line (far-end listener). The micro-
phone i s the input p ort, a nd line -out is the outp ut
port of this signal path.
Receive Path
The sign al path from the line (far-en d speaker) to
the loudspeaker (near-end listener). The line-in is
the inpu t port, and t he speaker is th e output po rt
for this signal p ath.
AEC
Acoustic Echo Controller. The part in the speaker-
phone algorithm that controls the echo in the
send path.
2—SOFTWARE ISD-T360SB
2-15
ISD
EEC
Electric Echo Controller. The part in the speaker-
phone algorithm that controls the echo in the re-
ceive pat h .
SPEAKERPHONE MODES OF OPERATION
Full-D up le x (ON)
The speakerphone works in full-duplex mode,
meaning both pa rties can speak and hear each
other simultaneously. In this mode both the
acou s tic and electric ec h o controlle rs are a ctive.
The VoiceDSP processor tone detectors are not
active in this mo d e.
Mute
In this mode of operation, the speakerphone gen-
erates silence to the line. The near-end listener
can hear the far-end speaker but not vice versa.
Tone dete ctors are not ac ti ve .
Hold
During Ho ld mode sile nce i s genera ted to the line
and the speaker, and neither side can hear the
other.
Restart
In Restart mode the speakerphone re-initializes it-
self to the last speakerphone mode (full-duplex,
transparent or mute). This mode should be used to
resume the speakerphone operation when there
is a significant change in the environmental con-
ditions (e.g., parallel pickup) that may affect the
speakerphone quality.
Transparent
While in Transparent mode, the speakerphone
works in full-duplex mode but without echo can-
cellation.
Samples from the microphone are transferred to
the line, and samples from the line are transferred
to the speaker, with no processing. This mode
should be used only for tuning and testing the sys-
tem.
Silence
In silence mode, only the electric echo cancellor
is active.
Listen
In Listen mod e the line is audi ble on the spea ker,
and the processor tone detectors are active.
During Listen mode, dialing with the GT command
and call progress detection are possible, since
the busy and dial tone detectors are active.
The follo wing pseud o-code dem onstrates how to
make a call from speakerphone mode:
2—SOFTWARE
ISD-T360SB
2-16 Voice Solutions in Silicon™
Figu re 2-3 : Sp eaker ph one Ps eu do Code Representation
2.3.5 SPEECH SYNTHESIS
Speech synthesis is the technology use d to create
messages out of predefined words and phrases
stored in a vo cabula r y.
There are two kinds of predefined messages: fixed
messages (voice menus in a voice-mail system)
and programmable messages (time-and-day
stamp, or the You have n messages announce-
ment in a DTAD).
A vocabulary includes a set of p redefined word s
and phrases, needed to synthesize messages in
any language. Applications which support more
than one language require a separate vocabu-
lary for each lang uage.
INTERNATIONAL VOCABULARY SUPPORT
(IVS)
IVS is a mechanism by which the VoiceDSP pro-
cessor utilizes several vocabularies stored on an
external storage device. IVS enables the ISD-
T360SB to synthesize messages with the same
meaning, but in different languages, from sepa-
rate vocab u la ries.
IVS Features
• Multiple vocabularies stored on a single
storage device .
• Plug-and-play. The same microcontroller
code is used for all languages.
• Synth esized and reco rded message s use the
same voice compression algorithm to
achieve equal qua lity.
• Argumen ted sentences. (For example: You
have <n> messag es.)
while () {
EV = wait_event()
case EV of:
spkr_button_pressed:
if (speakerphone_on) {
SSM 0 // Put VoiceDSP in idle mode
first_digit = TRUE
deactivate_digit_timeout_event()
}
else
SSM 1 // Put VoiceDSP in full-duplex speakerphone mode
digit_pressed:
if (first_digit) {
SSM 4 // Enter LISTEN mode
first_digit = FALSE
}
GT <dtmf_of_digit> // Dial the digit
S // Stop. Note that after the S command
// the VoiceDSP is still in speakerphone mode
enable_digit_timeout_event() // To "guess" when dialing is completed.
digit_timeout_event:
SSM 1 // Dialing is completed, Go back to full-duplex mode
deactivate_digit_timeout_event()
}
2—SOFTWARE ISD-T360SB
2-17
ISD
• Auto-synthesized t ime-and-day stamp (drive n
by the VoiceDSP processor’s clock).
• Support for various language and sentence
structures:
– One versus many. (For example: You
have one message versus You have
two messages.)
– None versus many. (For example: You
have no messages versus You have
two messages.)
– Number synthesis (English—Eighty ver-
sus French—Quatre-vingt).
– W or d or der (Engli sh — Twenty-one ver-
sus German—Einundzwanzig).
– D ay s of th e week ( Mon day throug h
Sunday versus Sunday through
Saturday).
VOCABUL ARY DESIGN
There are several issues, sometimes conflicting,
which must be addressed when designing a vo-
cabulary.
Vocabul ary Con te n t
If memory space is not an issue, the vocabulary
could contain all the required sentences, each
recorde d separately.
If memory space is a concern, the vocabulary
must be com pact; it should cont ain the minimum
set of words and phrases required to synthesize all
the sentences. The least memory is used when
phrases and words that are common to more
than on e s ente nce a re reco rd ed o nly on ce, a nd
the IVS too l is used to syn thesize sent ences out of
them.
A good combination of sentence quality and
memory space is achieved if you take the “com-
pact” approach, and extend it to solve pronunci-
ation problems. For example, the word twenty is
pronounced differently when used in the sentenc-
es You have twe nty messages and You have
twenty-two messages. To solve this problem,
words that are pronounced differently should be
recorded more than once, each in the correct
pronunciation.
Vocab u lary Recordin g
When recording vocabulary words, there is a
compromise between space and quality. The
words should be re corded and saved in a com-
pressed form , and you should use the best vo ice
compression for that purpose. However, lower
compression rates do affect the voice qu ality.
Another issue to consider is the difference in voice
quality between synthesized and recorded mes-
sages (e.g., between time-and-day stamp and
ICMs in a DTAD environment). It is more pleasant
to the human ear to hear both messages have
the same sound quality.
Vocabulary Access
Sometimes compactness and high quality are
not enough. There should be a simple and flexible
interface to access the vocabulary elements. Not
just the vocabulary but the code to access the vo-
cabulary should be compact.
When designing for a multi-lingual environment,
there are ev en more iss ues to conside r. Each vo-
cabulary should be able to handle language-
specific structures and designed in a cooperative
way with the ot her vocabul aries so th at the code
to access each vocabulary is the same. When
you use the command to synthesize the sentence
Monday. 12:30 P.M., you should not care in what
langu age it is going to be play ed back .
2—SOFTWARE
ISD-T360SB
2-18 Voice Solutions in Silicon™
IVS VOCABULARY COMPON EN TS
This section describes the basic concept of an IVS
vocabulary, its components, and the relationships
between th em.
Basic C o n c ep ts
An IVS vocabulary consists of words, sentences,
and special codes that control the behavior of
the algorithm which VoiceDSP processor uses to
synthesize sentences .
Word Table
The words are the basic units in the vocabulary.
Create synthesized sentences by combining
words in the vocabulary. Each word in the vocab-
ulary is given an index which identifies it in the
word table.
Note that, depending on the language structures
and sentences synthesized, you may need to
record some words more than once in the vocab-
ulary. For example, if you synthesize the sentenc-
es: you have twenty messages an d you have
twenty-five messages, the word twenty is pro-
nounced differently. In this example,
twenty
should be def ined as two differ en t words.
Number Tables
The number tables allow you to treat numbers dif-
ferently depending on the context.
Example 1: The number 1 c an be announce d as
one as in message number one or as first as in
first message .
Example 2: The number 0 c an be announce d as
no as in you have no messages or as oh as in
mond ay , ei ght oh five A.M.
A separate number table is required for each par-
ticular type of use. The number table contains the
indices of the words in the vocabulary that are
used to synthesize the number. Up to nine number
tables can be included in a vocabulary.
Sentence Table
The sentence table describes the predefined sen-
tences in the vocabulary. The purpose of this table
is to make the microcontroller that drives the
VoiceDSP processor independent of the lan-
guage being synthesized. For example, if the
Flash and/or ROM memory contains vocabularies
in various languages, and the first sentence in
each vocabulary means you have n messages,
the microcontroller switches languages by issuing
the followi ng com m and to V oiceDSP proce ssor :
SV <storage_media>, <vocabulary_id> -Select a
new vocabulary
The microcontroller software is thus independent
of the gr ammar of t he langua ge in use. Th e sen-
tences consist of words, which are represented by
their indices in the vo ca bulary.
Sentence 0
All sentences but one are user defined. The
VoiceDSP processor treats the first sentence in the
sentence table (se ntenc e 0) speci ally, to support
time-and-day stamp. The processor assumes that
the sentence is designed for both system time,
and message time-and-day stamp announce-
ment, and ha s one argume nt which is int erpreted
as follows:
0 Sys tem time is announce d
1 The time-and-day stamp of the cur-
rent message is announced.
Example 1: When the microcontroller sends the
command: SAS 0, 0
The system ti m e an d day is an n ou n ced.
Example 2: When the microcontroller sends the
command: SAS 0, 1
The curren t messag e ti me-and -day sta mp
is announced.
The fo llowing Figure 2-4 s hows the interrel ationship
betw een the three types of table s .
2—SOFTWARE ISD-T360SB
2-19
ISD
Figure 2-4: The Interrelationship between the Word, the Number, and the Sentence Tables
Control and Option Codes
The list of word indices alone cannot provide the
entire r ange of senten ces that the V oiceDSP pro -
cessor is able to synthesize. IVS control and option
codes send special instructions to control the
speech synthesis algorithm’s behavior in the pro-
cessor.
For example, if the sentence should announce
the time of day, the VoiceDSP processor should
be able t o substitut e the curr ent day and time in
the sentence. These control words do not repre-
sent recorded words, r ather they instruct th e pro-
cessor to take special actions.
THE IVS TOOL
The IVS tool in cl u des two utilities :
1. The DOS - based IVS Compil er
2. IVSTOOL for Windows. A Windows 95/98/
2000 utility
The tools help create vocabularies for the VoiceD-
SP processor. They take you from designing the vo-
cabulary structure, through defining the
vocabul ary sentences, to rec ording the vocabu-
lary words.
IVS Comp il er
The IVS compiler runs on MS-DOS (version 5.0 or
later) and enables you to insert your own vocabu-
lary, (i.e., basic words and data used to create
numbers and sentences, as directories and files in
MS-DOS). The IVS compiler then outputs a binary
file containing that vocabulary. In turn, this infor-
mation can be burned into an EPROM or Flash
memory to be used by the VoiceDSP software.
IVS Voic e Com pression
Each IVS vocabulary can be compiled with either
the 4.7 Kbit /s , the 6 .7 Kbi t/s or t he 8 .7 K bit/s vo ice
compre ssion algorit hm, or in PC M format. Def ine
the bit rate before compilation. The VoiceDSP pro-
cessor automatically selects the required voice
decompression algorithm when the SV com-
mand choo ses th e act ive vocabul ar y.
Senten ce Tabl e
Number Table
Word Table
You have OP T_NUMBER CONTROL_SING MESSAGES five
twenty
You have
messages
message
2-20
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
Graphical User Inter face (GUI)
The IVS package includes a Windows utility to assist the vocabulary designer to synthesize sentences. With
this util it y, you can recor d words, compo se sente nces an d liste n to words and sen ten ces in the spec ifi c
compre ssi on rat e qu al ity selec ted.
HOW TO USE THE IVS TO OL WITH THE
VOICEDSP PROCESSOR
The IVS tool creates IVS vocabularies, and stores them as a binary file. This file is burnt into a ROM device
or programmed into a Flash memory device using the INJ (Inject IVS) command. The VoiceDSP processor
SO (Say One Word) command is used to select the required vocabulary. The SW (Say Words), SO, SS (Say
Sentence) and SAS (Say Argumented Sentenc e) command s are used to synthesize the required word o r
sentence. The typical vocabulary-creation process using the IVS Tool software is as follows:
1. Design th e v ocabu lary.
2. Creat e the voc abulary files ( as descri bed in d etail be low). Us e IVS TOO L for W indows to simplif y this
process.
3. Reco rd the words using any standard PC so und card and sound editing software, t hat can cre-
ate.wav files.
4. Run the IVS compiler to compress the.wav files, and compile them and the vocabulary tables into
an IVS vocabulary file.
5. Repeat steps 1 to 4 to create a separate IVS vocabulary for each language that you want to use.
Note that each language file must have a different ID number. To specify an ID Number, go to
the “Vocabu lary ID” option in th e Vo cabu l ary/Buil d Opti on s m en u .
6. Exit IVS Tool and burn the IVS vocabulary files into a ROM (or Flash memory) d evice. Use the INJ
(Injec t I V S) com m a n d to program th e data into a Fla s h dev ice.
7. For inje cting voca bul aries with multiple languages into the flash memory:
• From a DOS prompt, issue the following DOS command: copy voc1.bin + voc2.bin
voc_bot h .bi n (wh er e v oc1.bin and voc2 .bin are the two vocabu lary files an d vo c_both.bi n is
the c om bined fi le.)
• The two files are now combined into a s ingle file.
• Inject the united vocabulary file using the INJ command. (See page 2-40 for details.)
• To switch between the two languages in real time operation, use the SV command. (See page
2-52 for details.) The SV
type id
comman d sw it ch es between speci fi ed v ocabulary IDs.
Once the v ocabulary is in place , the speech synthesis com mands of the Voic eDSP processor can be
used to synthesize sentences.
Figure 2-5 shows the vocabul ar y-c reation pr ocess for a sing l e table on a ROM or Flash mem or y de vi ce.
2-21
2—SOFTWARE ISD-T360SB
ISD
Figure 2-5: Crea tion of an IVS Vo cabulary
2.3.6 SW AGC
The Software Automatic Gain Control (SWAGC) algorithm can be activated with the compression algorithm or
with the Speakerphone algorithm. When used with the compression algorithm, it regulates the input signal to a
dynamic range that can provide higher compression quality. When used with the speakerphone algorithm, it
amplifies the input signal from the line to the speaker and this improves the quality for attenuated far end
speakers.
The SWAGC algorithm senses the energy level and updates the signal gain in order to amplify low energy sig-
nals and to avo id signal sat uration. The SWAGC feat ure eliminates the need for an external har dware AGC, thus
reducing hardware costs and complexity. Hardware AGC may be used with the compression algorithm to
avoid signal saturation prior to sampling the signal. Note that the hardware AGC cannot be used with the
speakerphone algorithm.
A set of tunable parameters is available for controlling the SWAGC algorithm. Different values can be used for
the compre ssion and s peakerphone al gorithms. The SWA GC may even be tur ned completely off for ei ther one
of the al gori thms. The way t o use dif fer ent tunable val u es is by updati ng th e tunabl e par amet er valu es prior to
activating each one of the algorithms. For more details on the SWAGC tunable parameters, see Table 2-4 in
the Command Description secti o n.
Figure 2-6 : SWAGC Char acteristi cs with Def ault Tunab l e Paramete r Values
.wav Files
Number
Tables
Sentence
Table
PC
+
Sound
Card
IVS C o mpiler
Compressed
Files (.vcd)
IVS Vocabulary
Files
ROM
Programmer
ROM
.wav File
Editor
Flash
INJ
Command
.ini File
IVS TOOL
for
Windows
JQM JPR JPM JOR JOM JNR JNM JR M
JQM
JPR
JPM
JOR
JOM
JNR
JNM
Outp u t Pow e r (dBm @ 600 ohm)
Input Power (dBm @ 600 ohm)
2-22
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
2.4 VOICEDSP PROCESSOR COMMANDS—QUICK REFERENCE TABLE
Table 2-2: Speech Commands
Command
Description Opcode
Hex Source State Result
State
Command Parameters Return Value
Name S/A* Description Bytes Description Bytes
CCIO S Configure
Codec I/O 34 RESET, IDLE No change Config_value 1 None -
CFG S Configure
VoiceDSP 01 RESET No change Config_value 3 None -
CMSG S Create
Message 33 IDLE MSG_OPEN Tag, Num_of_
blocks,
comp_rate
2+2+1 None -
CMT S Cut Message
Tail 26 IDLE No change Time_length 2 None -
CVOC S Check
Vocabulary 2B IDLE No change None - Test result 1
DM S Delete
Message 0A IDLE No change None - None -
DMS S Delete
Messages 0B IDLE No change Tag_ref,
Tag_mask 2 + 2 None -
GCFG S Get
Configuration
Value
02 RESET, IDLE No change None - Version 1
GEW S Get Error Word 1B All states No change None - Error word 2
GI S Get Informa tion
item 25 PLAY, RECORD,
SYNTHESIS, TONE_
GENERATE, IDLE
No change Item 1 Item value 2
GL S Get Length 19 IDLE No change None - Message
length 2
GMS S Get Memory
Status 12 IDLE No change Type 1 Remaining
memory
blocks
2
GMT S Get Message
Tag 04 IDLE No change None - Message tag 2
GNM S Get Number of
Messages 11 IDLE No change Tag_ref,
Tag_mask 2 + 2 Numbe r of
messages 2
GSW S Get Status Word 14 All states No change None - Status wor d 2
GT A Generate Tone 0D IDLE TONE_
GENERATE Tone
(single Tone or
DTMF)
1None -
GTD S Get Time and
Day 0E IDLE No change Time_day_
option 1 Time and day 2
GTM S Get Tagged
Message 09 IDLE No change Tag_ref,
Tag_mask,Dir 2+2+1 Message
found 1
GTUNE S Get Tunable
Parameter 06 IDLE, RESET No change Index 1 Parameter_
value 2
INIT S Initialize System 13 RESET, IDLE IDLE No ne - None -
INJ S Inject IVS data 29 IDLE No change N,
byte1...byten
4 + n None -
2-23
2—SOFTWARE ISD-T360SB
ISD
MR S Memory Reset 2A RESET, IDLE No change None - None -
P A Playback 03 IDLE PLAY None - None -
PA S Pause 1C PLAY, RE CORD,
SYNTHESIS,
TONE_G EN ERA T E,
IDLE*
No change None - None -
PDM S Go To Power-
Down Mode 1A IDLE No change None - None -
R A Record
Message 0C IDLE RECORD Tag (message
Tag),
Compression_
rate
2 + 1 None -
RDET S Reset Detectors 2C IDLE No change Detectors_
reset_mask 1None -
RES S Resume 1D P LA Y , RE C O R D,
SYNTHESIS,
TONE_GENERATE
IDLE*
No change None - None -
RMSG S Read Message 32 IDLE, MSG_OPEN MSG_OPEN None - Data 32
S S Stop 00 All states but
RESET IDLE None - None -
SAS A Sa y
Argumented
Sentence
1E IDLE SYNTHESIS Sentence_n,
Arg 1 + 1 None -
SB S Skip Backward 23 PLAY, IDLE* No change Time_length 2 None -
SDET S Set Detectors
Mask 10 IDLE No change Detectors_
mask 1None -
SE S Skip to End of
Message 24 PLAY, IDLE* No change None - None -
SETD S Set Time and
Day 0F IDLE No change Time_and_
day 2None -
SF S Skip Forward 22 PLAY, IDLE* No change Time_length 2 None -
SLC S Set Line
Channels 38 IDLE No change Active_chann
els 1None -
SMSG S Set Message
Pointer 30 IDLE, MSG_OPEN MSG_OPEN Num_of_
pages 2None -
SMT S Set Message
Tag 05 IDLE No change Message_tag 2 None -
SO A Say One Word 07 IDLE SYNTHESIS Word_number 1 N one -
SPS S Set Playback
Speed 16 PLAY, SYNTHESIS,
IDLE No change Speed 1 None -
SS A Say Sentence 1F IDLE SYNTHESIS Sentence_n 1 None -
Table 2-2: Speech Commands (Continued)
Command
Description Opcode
Hex Source State Result
State
Command Parameters Return Value
Name S/A* Description Bytes Description Bytes
2-24
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
NOTE: * Command is valid in IDLE state, but has no effect.
S = Synchronous command
A= Asynchronous command
NOTE: * Command is valid in IDLE state, but has no effect.
S = Synchronous command
A = Asynchronous command
SV S Set Vocabulary
Type 20 IDLE No change Type, Id 1 + 1 None -
SW A Say Words 21 IDLE SYNTHESIS N,
word1...wordn
1 + N None -
TUNE S Tune
Parameters 15 IDLE, RESET No change Index,
Parameter_
value
1 + 2 None -
VC S Volume Control 28 PLAY, SYNTHESIS,
IDLE,
TONE_GENERATE
No change
vol_level
(increment/
decrement)
1None -
WMSG S Write Message 31 IDLE, MSG_OPEN MSG_OPEN Data 32 None -
Table 2-3: Speakerphone Commands
Command
Description Opcode
Hex Source State Result
State
Command Parameters Return Value
Name S/A* Description Bytes Description Bytes
GEW S Get Error Word 1B TONE_GENERATE,
IDLE No change None - Error word 2
GSW S Get Status Word 14 TONE_GENERATE,
IDLE No c hange None - Status word 2
GT A Generate Tone 0D IDLE TONE_GEN
ERATE Tone
(Single Tone or
DTMF)
1None -
RDET S Reset Detectors 2C IDLE No change Detectors_reset
_mask 1None -
S S Stop 00 TONE_GENERATE,
IDLE IDLE None - None -
SDET S Set Detectors
Mask 10 IDLE No change Detectors_
mask 1None -
SSM S Set
Speakerphone
Mode
2F IDLE No change Mode 1 None -
TUNE S Tune
Parameters 15 IDLE, RESET No change Index,
Parameter_
value
1 + 2 None -
VC S Volume Control 28 TONE_GENERATE,
IDLE No chan ge
vol_level
(increment/
decrement)
1None -
Table 2-2: Speech Commands (Continued)
Command
Description Opcode
Hex Source State Result
State
Command Parameters Return Value
Name S/A* Description Bytes Description Bytes
2-25
2—SOFTWARE ISD-T360SB
ISD
2.5 COMMAND DESCRIPTION
The commands are listed in alphabetical order.
The execution time for all commands, when specified, includes the time required for the microcontroller
to receive the return value, where appropriate.
The executi on time does not incl ude the protoco l timing overhead , i.e., the execution times are mea-
sured from th e m om en t that the comma n d is detect ed as va lid until the command is fu l ly ex ecu ted.
NOTE: Each command description includes an example application of the command. The examples
show the opcode i ssued by the microco ntro ller and the r esponse ret ur ned by the Voic eDSP pr o-
cessor. For commands which require a return value from the processor, the start of the return
value is indicated by a thick vertical line.
CCIO Configure Codec I/O
config_value
Configur es the voic e samp le pat hs i n v ario us state s. It s hou ld be used to c hang e th e defa ult Voi ceDSP
processor configuration. It is relevant only when two codecs are used and speakerphone is connected.
The
config_value
parameter is encoded as follows:
Bit 0
Loopback control.
0 Loopback disabled (default).
1 Loopback enabled. In the RECORD state, the input samples are echoed back, unchanged
(i.e., no volume control), to the same codec.
Bit 1
Codec inp u t con trol.
0 Input is rec eived via th e li ne codec (default) .
1 Input is received via the speakerphone codec.
Bits 2–3
Codec out pu t con trol.
00 In PL AY, IDLE, SYNTH ESIS and TONE _GENERATE st ates, outpu t is to bo th codec s. In RECORD
state, output is to the non-input codec (no volume control). If the loopback control bit is set,
output in RE COR D state is to both codecs as well (def au l t) .
01 Output in all states is to the line codec.
10 Output in all states is to the speakerphone codec.
11 Output in all states is to both codecs.
Bits 4–7
Reserved.
2-26
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
Example
CFG Configure VoiceDSP
config_value
Configur es the Voice DSP proce ssor in vari ous hardwa re enviro nments. It shou ld be used to change the
default pr ocessor configu r ation.
The
config_value
parameter is encoded as follows:
Bits 0–3
Memory type.
0000 No Memory (d ef aul t) .
0001 Reserved.
0010 Reserved.
0011 Reserved.
0100 Samsung NAND Flash.
0101 Toshiba NAND Flash.
0110 Reserved.
0111 Reserved.
1000 - 1111 Reserv ed.
Bits 4–5
Number of installed memory devices.
00 1 (Default)
01 2
10 3
11 4
Bit 6 -10
Reserved
Bit 11 - 14
Reserved
CCIO 01
Byte sequence: Microcontroller 34 01
VoiceDSP 34 01
Description: Enable loopback
2-27
2—SOFTWARE ISD-T360SB
ISD
Bit 1 5
Echo Cancella ti on Cont ro l (fo r DTM F Detect io n).
0 Echo can cellation off.
1 Echo ca ncella tion is on during playb ack. (recommended)
Echo cancellation improves the performance of DTMF detection during playback. Echo cancellation
can be turned on only with a system that can disable HW AGC (if present) during playback. A system fea-
turing HW AGC, that cannot be controlled by the microcontroller (i.e., disabled or enabled), must not turn
on this bit.
Bit 1 6
Clock bit rate (in Slave Mode only).
0 One bit rate clock (default).
1 Two bit rate clock.
Bit 1 7
Codec conf i guration .
0 Short-frame format (def aul t) .
1 Long-fra me for m at (guaranteed by design bu t n ot tested).
Bits 18–19
Codec type.
00 16-bit linear (default).
01 µ-Law.
10 A-Law.
Bit 2 0
Codec int er f ace mode.
0 Master codec interface (default).
1 Slave codec interface.
Bits 21–22
Number and type of codec s
00 One single codec (default).
01 Two sin gle codecs.
10 One dual codec.
11 Reserved.
2—SOFTWARE
ISD-T360SB
2-28 Voice Solutions in Silicon™
The codecs should be conne cted as follows:
Teleph on e li n e or equ iv alent - always con nected as channel 0.
Speaker and microphone - connected as channel 1 in case of one dual codec, connected as channel
2 in case of two single codecs.
Bit 23
Reserved.
Example
CMSG Create Message
tag
num_of_blo cks compress ion_ rat e
Creates a new message with a message tag
tag
, and compression rate
compression_rate
allocates
num_of_blocks
4-Kbytes blocks for the new message, and sets the message pointer to the beginning of
the message data. CMSG switches the VoiceDSP processor to the MSG_OPEN state.
The memory space available for the message data is compu ted as follows:
(127 x
num_of_blocks
– 2) x 32 bytes.
Once a messag e is o pen (the proce ssor is in th e M SG_OPEN s tate) , the message p ointer can be set t o
any po sition on a pag e (32 by tes) boun dary within the me ssage with the SM SG comm and. Modi fy th e
message contents with the WMSG command, and read with the RMSG command.
The microcontroller must issue an S command to close the message and switch the VoiceDSP processor
to the IDLE state.
If the mem ory is fu ll, EV_M EMFULL is set in th e status wo rd and no m essage i s created . If the memory is
not full but there is insufficient memory and the processor cannot allocate more memory space,
EV_MEMLOW is set in the status word and no message is created.
The compre ssion ra te may be define d as 0 for PCM recor ding or eith er 1,2,3 for compr ession rate s 4.7
Kbit/s, 6.7 Kbit /s, 8.7 Kbit/s respectivel y. See sectio n “VCD” on page 2- 10 for a description of the compres-
sion algorithm.
Example
CFG 188013
Byte sequence: Microcontroller 01 18 80 13
VoiceDSP 01 18 80 13
Description: Configure the VoiceDSP to work with:
Single codec in Slave Mode and A-Law compressed samples.
Data in Short Frame form at and Single Bit Rate interfa ce.
Two Serial Toshiba Flash devices.
Echo Cancellation for DTMF detection is On.
CMSG 0101 0 001
Byte sequence: Microcontroller 33 01 01 00 01
VoiceDSP 33 01 01 00 01
Description: Create a new message with a tag=0101, and allocate 1 block (4 Kbytes) for its data.
2-29
2—SOFTWARE ISD-T360SB
ISD
CMT Cut Message Tail time_length
Cut
time_length
units, in 10 ms segments, off the end of the current message. The maximum value of
time_length
is 6550. In case of silence, cut-time accurac y is 0.1 to 0.2 seconds (dep ends on compres-
sion rate).
If time_length is greater than 6550 the ERR-PARAM is set in the ERR_WORD.
NOTE If time_length is longer than the total duration of the message, the EV_NORMAL_END event is set
in the status word and the message becomes empty but not deleted.
A compressed f rame repr esents 21 ms of speec h, th us the m inimum meaningf ul paramet er is 3,
(i.e., a 30 ms cut).
CMT 1 or CMT 2 have no effect.
The CMT command can not be used on data messages.
In PCM recording mode the accuracy of µ-,
A-Law
configuration is 0.50 sec while the accuracy
of linear configuration is 0.25 sec.
Use the DM (Delete Message), or DMS (Delete Messages) command to delete the message.
Example
CVOC Check Vocabulary
Checks (checksum) if the IVS data of the currently selected vocabulary was correctly programmed to the
ROM or Flas h device.
If the voca bu l ary data is corr ect the return value is 1. Otherwise the return val ue is 0.
If the curren t v ocabulary is u n def ined, ERR_INV A LI D is reported.
Example
CMT 02BC
Byte sequence: Microcontroller 26 02 BC
VoiceDSP 26 02 BC
Description: Cut the last seven seconds of the current message.
CVOC
Byte sequence: Microcontroller 2B AA
VoiceDSP 2B 01
Description: Check the current vocabulary.
The VoiceDSP processor responds that the vocabulary is OK.
2-30
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
DM Delete Message
Deletes the current message. Deleting a message clears its message tag.
Dele ting the curr ent me ssag e do es not caus e a differ ent messa ge to be come curre nt. T he c urren t mes -
sage is undefined. If, for example, you issue the GTM command to skip to the next message, the first mes-
sage that is new er than th e just de le ted message is select e d as th e current message.
The memory space releas ed by the delete d m essag e is immediat el y availabl e for recordin g n ew me s-
sages.
Example
DMS Delete Messages
tag_ref tag_mask
Deletes all messages whose message tags match the tag_ref parameter. Only bits set in tag_mask are
compared i.e., a match is considered successful if:
message tag
and
tag_mask
=
tag_ref
and
tag_mask
where and is a bitwise AND operation.
After the command completes execution, the current message is undefined. Use the GTM command to
select a messag e to be th e current messa ge .
The memory space released by the deleted messages is immediately available for recording new mes-
sages.
Example
DM
Byte sequence: Microcontroller 0A
VoiceDSP 0A
Description: Delete current message.
DMS FFC2 003F
Byte sequence: Microcontroller 0B FF C2 00 3F
VoiceDSP 0B FF C2 00 3F
Description: Delete all old incoming messages from mailbox Number 2 in a system where the message tag
is encoded as follows:
Bits 0–2: mailbox ID 8 mailboxes indexed: 0 to 7
Bit 3: new/old message indicator
0—Message is old
1—Message is new
Bits 4–5: message type 00—ICM/memo
01—OGM
10—Call transfer message
Bits 6–15: not used
Note: the description of the tag is an example only. All bits of the tag are user-
definable.
2-31
2—SOFTWARE ISD-T360SB
ISD
GCFG Get Configuration Value
Returns one byte with the following information:
Bits 0–7
Magic number, which specifie s the VoiceDSP firmware version.
Example
GEW Get Error Word
Return s th e 2-byte error word.
ERROR WORD
The 16-bit error word indicates errors that occurred during execution of the last command. If an error is
detected, the command is not processed; the EV_ERROR bit in the status word is set to 1, and the
MWRQST signal is activated (set to 0).
The GEW command reads the error word. The error word is cleared during reset and after execution of
the GEW command.
If errors ERR_C OMMAND or ERR_PARA M occur during the execution o f a command that has a return val-
ue, t he ret urn val ue is und efined . The mi croco ntrol ler must still r ead the r eturn va lue, t o ensur e prop er syn-
chronization.
The bits of the error wo rd are used as follows :
ERR_OPCODE
Illegal opcode. The VoiceDSP processor does not recognize the opcode.
ERR_COMMAND
Illegal command sequence. The command is not legal in the current state.
ERR_PARAM
Illega l parameter . Th e va lu e of the parameter is out of ran g e, or is no t appropr iate for the com m an d.
GCFG
Byte sequence: Microcontroller 02 AA
VoiceDSP 02 01
Description: Get the VoiceDSP processor magic number.
The VoiceDSP processor responds that it is Version 1.
159876543 2 10
Res ERR_CID ERR_
INVALID ERR_
TIMEOUT ERR_
COMM Res ERR_
PARAM ERR_
COMMAND ERR_
OPCODE Res
2-32
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
ERR_COMM
Microcontroller MICROWIRE communication error.
ERR_TIMEOUT
Time-out erro r. Depending on the V oiceDSP proc essor’s state, more t han 100 milli seconds elap sed be-
tween the arrival of two consecutive bytes (for commands that have parameters).
ERR_INVALID
Command can not be perform ed in current cont ext.
ERR_CID
Error during CID d ete ct io n.
Example
GI Get Information item
Returns th e 16-bit value specified by
item
from one of the in ternal re gi sters of the VoiceDSP processor.
item
may be one of th e following :
0 The duration of the last detected DTMF tone, in 10 ms units. The return value is meaningful
only if DTMF detection is enabled, and the status word shows that a DTMF tone was de-
tected.
1 The duration of the last detected busy tone in 10 ms units.
2 The energy lev el of the sam ples in the last 10 m s.
3 The energy level of the samples, in the last 10 ms, that are in the frequency range de-
scribed in Figure 2-1 on page 2-12. The return value is meaningful only if one of the tone
detectors is enabled (bits 0,1 of the detectors mask; see the description of SDET com-
mand).
The return va lue is unpre di ctable for any othe r valu e of
item
.
GEW
Byte sequence: Microcontroller 1B AA AA
VoiceDSP 1B 00 02
Description: Get the VoiceDSP processor error word (typically sent after GSW when EV_ERROR is reported in
the status word).
The VoiceDSP processor responds: ERR_OPCODE:
2-33
2—SOFTWARE ISD-T360SB
ISD
Example
GL Get Length
Return s th e length of the current messag e i n multiple s of 4 Kby tes ( blocks).
The retu rne d value incl ude s the messa ge di recto ry in for mat ion (64 byt es for t he fir st bl ock an d 32 byte s
for every other block), the message data, and the entire last block of the message, even if the message
occupies only a portion of the last block. Since a memory block includes 4096 bytes, the returned length
may be bigger than the actual message length by up to 4095 bytes. The minimum length of a message
is one block.
Example
GMS Get Memory Status
type
Returns the total remaining memory blocks as a 16bit unsigned integer. The parameter type must be 0.
The estimate d rema ining recording time may be calc ulated as follows:
Time = (Num_of_blocks x 4096 x 8) / (Compression_rate x 1000)
This estimate assumes no silence compression: a real recording may be longer, according to the
amount of silence detected and compressed.
Example
GI 0
Byte sequence: Microcontroller 25 00 AA AA
VoiceDSP 25 00 00 06
Description: Get the duration of the last detected DTMF tone.
The VoiceDSP processor responds: 60 ms.
GL
Byte sequence: Microcontroller 19 AA AA
VoiceDSP 19 00 04
Description: Get the length of the current message.
The VoiceDSP processor responds:
4
i.e., the message occupies 16384 (4 * 4096) bytes.
GMS 00
Byte sequence: Microcontroller 12 00 AA AA
VoiceDSP 12 00 00 28
Description: Return the remaining memory blocks.
The VoiceDSP responds:
40 blocks.
2-34
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
GMT Get Message Tag
Returns the 16-bit tag associated with the current message. If the current message is undefined,
ERR_INVALID is reported.
Example
GNM Get Number of Messages tag_ref tag_mask
Returns the number of messages whose message tags match the
tag_ref
parameter. Only bits set in
tag_mask
are compa red, a match is considered su ccessful if:
message tag
and
tag_mask
=
tag_ref
and
tag_mask
where and is a bitwise AND operation.
The
tag_ref
and
tag_mask
parame ters ar e each two bytes; the re tu rn val u e is also two bytes l ong.
See “Messag e Ta g ” on pag e 2-3 for a descr iption of messag e-tag encodi n g . If
tag_mask
= 0, the total
number of all existing messages is ret urne d, re ga rd le s s of the
tag_ref
value.
Example
GSW Get Status Word
Returns th e 2-byte status word.
STAT U S WORD
The VoiceDS P processor has a 16-bit sta tus wor d to indicate ev en ts that occur during n or m al oper ation.
The VoiceDSP processor asserts the MWRQST signal (clears to 0), to indicate a change in the status word.
This signa l re mains active until the VoiceDSP pro c ess o r receives a G SW co mma nd.
GMT
Byte sequence: Microcontroller 04 AA AA
VoiceDSP 04 00 0E
Description: Get the current message tag.
In a system where the message tag is encoded as described in the DMS command, the
VoiceDSP processor return value indicates that the message is a new ICM in mailbox Number
6.
GNM FFFE 0003
Byte sequence: Microcontroller 11 FF FE 00 03 AA AA
VoiceDSP 11 FF FE 00 03 00 05
Description: Get the number of messages which have bit 0 cleared, and bit 1 set, in their message tags.
VoiceDSP processor responds that there are five messages which satisfy the request.
2-35
2—SOFTWARE ISD-T360SB
ISD
The status word is cleared during reset, and upon a successful GSW command.
The bits in the status word are used as follows:
EV_DTMF_DIGIT
DTMF digit. A value indicating a detected DTMF digit. (See the description of DTMF code in the GT com-
mand.)
EV_DTMF_END
1 = Ended detection of a DTMF tone. The detected digit is held in EV_DTMF_DIGIT.
EV_NORMAL_END
1 = Normal completion of operation, e.g., end of message playback.
EV_MEMFULL
1 = Mem or y is full.
EV_ERROR
1 = Error detected in the last command. You must issue the GEW command to return the error code and
clear th e er ror con dition.
EV_BUSY
1 = Busy tone detected. Use this indicator for call progress and line disconnection.
EV_DIALTONE
1 = Dial tone detected. Use this indicator for call progress and line disconnection.
EV_MEMLOW
1 = Not enough memory. (See CMSG command for further details.)
EV_CONST_NRG
1 = A period of constant energy was detected. Use this indicator for line disconnection. (See
CONST_NRG_DET_TIME_COUNT in Table 2-8.)
EV_VOX
1 = A perio d of silen ce (n o en ergy) w as d etect ed on the telep hone l ine . Use t his i ndica tor for l ine d iscon-
nectio n . (See V O X_ DET_TIME_C OUNT in Table 2-8.)
EV_RESET
When the VoiceDSP processor completes its power-up sequence and enters the RESET state, this bit is set
to 1, and the MWRQST signal is activated (cleared to 0). Normally, this bit changes to 0 after performing
15 14 13 12 11 10 9 8 7 6 5 4 3 0
EV_
DTMF EV_
RESET EV_
VOX EV_
CONST_
NRG
Res EV_
MEMLOW EV_
DIALTONE EV_
BUSY EV_
ERROR EV_
MEMFULL EV_
NORMAL_
END
EV_
DTMF_
END
EV_
DTMF_
DIGIT
2-36
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
the INIT comm an d. If this bit i s set during no rm al operation of the VoiceDSP processor, it indicates an i n -
terna l Vo iceDS P p rocesso r er ror . T he mi croc ontr oll er ca n r ecov er f rom su ch an erro r by re-i ni tial izin g t he
system.
EV_DTMF
1 = Started detection of a DTMF tone.
Example
GT Generate Tone tone
Generates the tone specified by the 1-byte tone parameter. The VoiceDSP state changes to
TONE_GENERATE. The tone generation continues until an S command is received.
A DTMF or a single frequency to ne may be genera t ed as shown:
To gen erate a D TMF enco de th e bits as follows :
Bit 0
1
Bits 1–4
DTMF code.
Where the DTMF code is encoded as fo llows:
Bits 5–7
0
GSW
Byte sequence: Microcontroller 14 AA AA
VoiceDSP 14 00 40
Description: Get the VoiceDSP processor Status Word (typically sent after the MMRQST signal is asserted by
the VoiceDSP processor which indicates a change in the status word).
The VoiceDSP processor responds that the memory is full.
Value (Hex) DTMF Digit
0 to 9 0 to 9
AA
B*
C#
DB
EC
FD
2-37
2—SOFTWARE ISD-T360SB
ISD
To generate a single freq ue ncy tone enco de the bi ts as follows:
Bit 0
0
Bits 1-5
3–30
The value in bits 1–5 is multiplied by 100 to generate the required frequency (300Hz–3000Hz).
Bits 6-7
0
The VoiceDSP processor does not check for the validity of the tone specification. Invalid specification
yields un pr edi ctable results.
Example
GTD Get Time and Day
time_day_option
Returns the time and day as a 2-byte value.
time_day_option
may be one of th e fol l owin g :
0 Get the system time and day.
1 Get the current message time-and-day stamp.
Any other
time_day_option
returns the time-and-day stamp of the current message.
Time and day are encoded as follows:
Bits 0–2
Day of the week (1 through 7).
Bits 3–7
Hour of the day (0 through 23).
Bits 8–13
Minute of the hour (0 through 59).
Bits 14–15
00 The time was not set before the current message was recorded.
11 The ti m e was set, i.e., the SETD (Set Ti me of Day) com m and was executed.
NOT E If th e cur re n t me ss ag e is und e fi n e d, and tim e _d a y _optio n is 1, a n E RR _I N VA L ID er ro r is rep o r te d.
GT 20
Byte sequence: Microcontroller 0D 20
VoiceDSP 0D 20
Description: Generate a single-frequency 1600Hz tone.
2-38
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
Example
GTM Get Tagged Message tag_ref tag_mask dir
Select s the curr en t messag e, accord in g to instructi o ns in
dir
, to be the fir st, n th next or nth pr eviou s mes-
sage which com pl ies with the equa ti on :
message tag
and
tag_mask
=
tag_ref
and
tag_mask
.
where and is a bitwise AND operation.
dir
is one of the following:
0 Selects the first (oldest) message.
-128 Selects the last (newest) message.
n Selects the nth next message starting from the current message.
–n Selects the nth previo u s message start in g fr om th e cu rrent messag e.
NOTE To select the nth, or -nth, message with a given tag to be the current message you must first
select the first message (dir=0), or the last message (dir=-128), that complies with the above
equation, and then issue another GTM command with n-1 (for next message), or -n+1 (for previous
message), as a parameter, to skip to the nth, or -nth, message respectively.
If a message is found, it becomes the current message and 1 (TRUE) is returned. If no message
is foun d , the cu rr e nt m es sa g e rem a i ns u n ch a ng e d an d 0 (F AL SE ) is retu rn e d.
If dir is not 0, and the current message is undefined the return value is unpredictable. After the
command execution the current message may either remain undefined or changed to any existing
message. The only exception is when the GTM command is executed just after the DM command.
(See the DM command for further details.)
To access the nth message, when n > 127, a sequence of GTM commands is required.
GTD 1
Byte sequence: Microcontroller 0E 01 AA AA
VoiceDSP 0E 01 E8 29
Description: Get the current message time-and-day stamp.
The VoiceDSP processor responds that the message was created on the first day of the week at
5:40 A.M. The return value also indicates that the SETD command was used to set the system
time and day before the message was recorded.
Note: If the SAS command is used to announce the time-and-day stamp, “Monday” is
announced as the first day of the week. For an external vocabulary, the
announcement depends on the vocabulary definition (See the IVS User’s Manual for
more de t a ils ).
2-39
2—SOFTWARE ISD-T360SB
ISD
Example
GTUNE Get Tune index
Gets the value of the tunable parameter identified by
index
(one byte) as the 2-byte value,
parameter_value
. This command may be used to read and identify the parameter value that was set to
tune the VoiceDSP.
If
index
does not point to a valid tunable parameter, ERR_PARAM is set in the error word.
This index parameter should be in a he xidecimal value.
The GTUNE c ommand may be u sed in IDL E state o r RESE T state. If TUNE c ommand was not us ed to set
the tunable parameters, then the GTUNE command will read the default parameter value.
Example
Table 2-4 through Table 2-12 describe the tunable parameters, their index numbers in decimal, and their
default v alues.
GTM FFCE 003F 00
Byte sequence: Microcontroller 09 FF CE 00 3F 00 AA
VoiceDSP 09 FF CE 00 3F 00 01
Description: Select the ol dest of the new ICM s, in mailbox number 6, to be the cur rent message, for a s ystem
where the mess age tag is encoded as descri bed in the ex ample fo r the DMS command.
The VoiceDSP processor returns a value indicates that there is such a message.
The following pseudo-code demonstrates how to play all new ICMs in mailbox number 6. The
messages are marked as old after being played:
Return_val = GTM(FFCE, 003F, 00) /*Get the oldest message with the defined tag*/
While (Return_Val == TRUE)
Begin
P() /* Play */
Message_tag = GMT() /* Get message tag */
SMT(FFF7) /* Mark the message as ‘old’ */
GTM(FFCE, 003F, 01) /* Get next message with the same tag */
End
GTUNE 17
Byte sequence: Microcontroller 06 17 AA AA
VoiceDSP 06 17 02 BC
Description: Get the minimum period for busy detection
ComactSPEECH responds:
700 (7 seconds).
2-40
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
INIT Initialize System
Execute this command after the VoiceDSP processor has been configured (see CFG command). INIT
performs a soft reset of the VoiceDSP as follows:
• Initializes the message directory information.
• Messages are not deleted. To delete the messages, use the DM and DMS commands.
• Sets the detectors mask to 0.
• Set s the vo lume level tha t is controlle d by the VC command, to 0.
• Sets the playback speed to normal (0).
• Switches to the IDLE state.
• Initializes the tone de tector s .
The current message is undefined after INIT execution.
The tunable parameters are not affected by this command. They are set to their default values only dur-
ing RESET .
Example
INJ Inject IVS Data n byte1... byten
Injects vocabulary dat a of size
n
bytes to good Flash blocks.
This command programs Flash devices, on a production line, with IVS vocabulary data. It is optimized for
speed; all VoiceDSP processor detectors are suspended during execution of the command. Use the
CVOC comma nd to check wh et h er progr amm i n g was succe ssful.
If ther e is n ot enoug h memo ry space for the voca bulary d ata, ER R_PARAM is set in th e erro r word, a nd
execution stops.
Flash blocks that include IVS data can not be used for recording, even if only one byte of the block con-
tains IV S data (e.g., if th e v ocabular y size is 4K + 100 byt es, two blocks of th e Flash are not a va il able for
message recording).
Example
INIT
Byte sequence: Microcontroller 13
VoiceDSP 13
Description : Initialize the VoiceDSP proc essor .
INJ 00000080
Data
Byte sequence: Microcontroller 29 00 00 00 80 Vocabulary Data
VoiceDSP 29 00 00 00 80 Echo of Data
Description: Inject 128 bytes of vocabulary data.
2-41
2—SOFTWARE ISD-T360SB
ISD
MR Memory Rese t
Erases all memory blocks and initializes the memory and message management. Bad blocks, and
blocks which are used for IVS vocabularies, are not erased. This command can be issued in either RESET
or IDLE states.
NOTE When Memory Reset is used in RESET state, it must be issued after the CFG command is issued,
or the memory type and number of devices are not defined. In this case the result is
unpredictable.
NOTE The command erases all messages and should be used with care.
Example
P Playback
Begins playback of the current message. The VoiceDSP processor state changes to PLAY. When playback
is complete, the VoiceDSP processor sets the EV_NORMAL_END bit in the status word, and activates
(clears to 0) the MWRQST si gnal. The s tate then changes to IDLE. Playback can be p aused with the PA
command, and can be resumed later with the RES command. Playback can be stopped with the S
command.
If the current messag e is undefi ne d, ERR _INVALID is re port ed.
Example
PA Pause
Suspend s the execution of the cu rrent GT, P, R, S AS, SO, SW, or SS co mmands. The PA command doe s
not change the state of the VoiceDSP processor; execution can be resumed with the RES command.
NOTE DTMF and tone detectors remain active during Pause, Play and Record commands.
MR
Byte sequence: Microcontroller 2A
VoiceDSP 2A
Description: Erase all memory blocks.
P
Byte sequence: Microcontroller 03
VoiceDSP 03
Description: Play the current message.
2-42
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
Example
PDM Go To Power-Down Mode
Switches the VoiceDSP processor to power-down mode (see “POWER-DOWN MODE” on page 1-4 for de-
tails). Sending a ny command whi le in power -down mode r esets the processor d etectors, and returns it
to normal operation mode.
NOTE If an event report is pending (MWRQST is active) and not processed by the microcontroller prior
to issuing the PDM command, the event is lost.
Example
R Record
tag compression_rate
Records a new message with message tag
tag
and compression rate
compression_rate
. The VoiceDSP
processor state changes to RECORD. The R command continues execution until stopped by the S com-
mand. Recording can be paused with the PA command, and can be resumed later with the RES com-
mand.
If the memory becomes full, recording stops and EV_MEMFULL is set in the status word.
See “Mess ag e Tag ” on pag e 2-3 for a description of messag e-tag encod in g .
The compre ssion ra te may be define d as 0 for PCM recor ding or eith er 1,2,3 for compr ession rate s 4.7
Kbits/s, 6.7 Kbits/s, 8.7 Kbits/s respectively. See “VCD” on page 2-10 for a description of the compression
algorithm.
NOTE A time-and-day st amp is automaticall y attached to each message. Before using the R command
for th e first time, use the SETD co mmand. Failure to do so resul ts in undefi ned values f or the time-
and-day stamp.
The CID_RECORD tunable parameter (index 51) may be used to attach the contents of the CID buffer to the
message memory durin g recording.
To read CID information attached to the message, use the GCID command. See page 2-31 for
details.
PA
Byte sequence: Microcontroller 1C
VoiceDSP 1C
Description: Suspend playback of current message.
PDM
Byte sequence: Microcontroller 1A
VoiceDSP 1A
Description: Put the VoiceDSP processor in power-down mode.
2-43
2—SOFTWARE ISD-T360SB
ISD
Example
RDET Reset Dete ctors detectors_reset_mask
Resets the VoiceDSP processor tone and energy detectors according to the value of the
detectors_reset_mask parameter. A bit set to 1 in the mask, resets the corresponding detector. A bit
cleared to 0 is ignored.
The 1-byte
detectors_reset_mask
is encoded as fo llows:
Bit 0
Reset the busy and dial tone detectors.
Bits 1–3
Reserved. Must be cleared to 0.
Bit 4
Reset the constant energy detector.
Bit 5
Reset the no energy (VOX) detector.
Bit 6
Reset the DTMF detector.
Bit 7
Reserved. Must be cleared to 0.
Example
R 000E 03
Byte sequence: Microcontroller 0C 00 0E 03
VoiceDSP 0C 00 0E 03
Description: Reco rd a new ICM in mailbox Number 6 in a system wher e the message tag is enco ded as
described in the example of the DMS command. The compression rate is defined as 8.7 Kbit/s
RDET 20
Byte sequence: Microcontroller 2C 20
VoiceDSP 2C 20
Description: Reset the VOX detector.
2-44
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
RES Resume
Resumes the activity that was susp en ded by the PA, SB, SE or S F com m an ds.
Example
RMSG Read Message
Returns 32 bytes of data from the current position of the message pointer, and advances the message
pointer by 32 bytes.
If the VoiceDSP processor was in the IDLE state, the command opens the current message, switches the
VoiceDSP processor to the MSG_OPEN state, sets the message pointer to the beginning of the message
data, and retu r n s the 32 bytes of data.
The microcontroller must issue an S command to close the message, and switch the VoiceDSP processor
to the IDLE state.
If the current messag e is undefi ne d, ERR _INVALID is re port ed.
Trying to read beyond the end of the message sets the EV_NORMAL_END event, and the VoiceDSP pro-
cessor switc h es to th e IDLE state. In this case, the return va lue is undefin ed and should be ignored.
Example
SStop
Stops exec u ti on of the cu r rent comman d an d switches the Vo iceDSP processor to the IDL E st ate. S may
be used to stop the execution of CMSG, SMSG, WMSG, RMSG and all asynchronous commands. See Ta-
ble 2-2 on page 2-22 for details.
This comman d may also b e used to swi tch the Voic eDSP proce ssor state t o IDLE in order to issue c om-
mands that are allowed to execute only while in the IDLE state.
Example
RES
Byte sequence: Microcontroller 1D
VoiceDSP 1D
Description: Resume playback which was suspended by either the PA, SF or SB command.
RMSG
Byte sequence: Microcontroller 32 AA AA ...
VoiceDSP 32 32 bytes of data
Description: Read 32 bytes from the current message memory.
S
Byte sequence: Microcontroller 00
VoiceDSP 00
Description: Stop current activity (e.g., playback, recording) and change the VoiceDSP processor to IDLE
state.
2-45
2—SOFTWARE ISD-T360SB
ISD
SAS Say Argumented Sentence
sentence_n arg
Announces sentence number
sentence_n
of the currently selected vocabulary, and passes
arg
.
sentence_n
and
arg
are each 1-byte long. The Voic eDS P proc essor state chan ge s to SYNT HE S IS .
When pla yi n g is com pl ete, the VoiceDSP processor sets the EV_ NO R M AL_END bit in the status wor d, an d
activates the MWRQST signal. The state then changes to IDLE.
If the curren t v ocabulary is u n def ined, ERR_INVA L I D is reported.
Example
SB Ski p Back ward time_length
Skips backward in the current message
time_length
units, in 0.2 second segments, and pauses message
playbac k. The RES command must be issue d to cont inu e playba ck.
time_length
is a 2-byte parameter
that can have any value up to 320 (64 seconds). The skip accuracy is five percent. SB is meaningful only
in the PLA Y state.
If the beginning of the message is detected during the execution of the SB command, execution termi-
nates, the EV_NORMAL_END bit in the status register sets, the MWRQST signal activates, and the processor
switches to the IDLE state.
If
time_length
is greater than 320, ERR_PARAM is set in the error word.
Playback speed does not affect the behavior of this command.
Example
SDET Set Detect ors Mask detectors_mask
Controls the reporting of detection of tones and energy detectors according to the value of the
detectors_mask
parameter. A bit set to 1 in the mask, enables the reporting of the corresponding detec-
tor. A bit cl eared to 0 disa bles the repor ting.
Disabling reporting of a detector does not sto p or rese t th e detector.
The 1-byte
detectors_mask
is encod ed as follows:
SAS 00 03
Byte sequence: Microcontroller 1E 00 03
VoiceDSP 1E 00 03
Description: Announce the first sentence in the sentence table of the currently selected vocabulary with ‘3’
as the actual parameter.
SB 0019
Byte sequence: Microcontroller 23 00 19
VoiceDSP 23 00 19
Description: Skip backwards five seconds from the current position in the message being played.
2-46
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
Bit 0
Report detection of a busy tone.
Bit 1
Report detection of a dial tone.
Bit 2 -3
Reserved. Must be cleared to 0.
Bit 4
Report detection of a constant energy.
Bit 5
Report detection of no ener g y ( VOX ) on the line.
Bit 6
Report the ending of a detected DTMF.
Bit 7
Report the start of a detected DTMF (up to 40 ms after detection start).
Example
SE Skip to End of Message
This co mma n d i s valid onl y i n the PLAY state. When inv oked, playback is suspended (a s f or th e P A com-
mand), and a jump to the end of the message is performed. Playback remains suspended after the
jump.
Example
SDET BB
Byte sequence: Microcontroller 10 BB
VoiceDSP 10 BB
Description: Set reporting of all VoiceDSP processor detectors, except for end-of-DTMF.
SE
Byte sequence: Microcontroller 24
VoiceDSP 24
Description: Skip to end of current message.
2-47
2—SOFTWARE ISD-T360SB
ISD
SETD Set Time and Da y time_and_day
Sets the system time and day as specified by the 2-bytes
time_and_day
parameter. The
time_and_day
parameter is en coded as follows:
Bits 0–2
Day of the week (1 through 7).
Bits 3–7
Hour of the day (0 through 23).
Bits 8–13
Minute of the hour (0 through 59).
Bits 14–15
Must be set to 1.
If
time_and_day
value is not valid , ERR_PARAM is set in the error word.
Example
SF Skip Forward
time_length
Skips for ward in the curr ent messag e
time_length
units, in 0.2 second segments, and causes message
playback to pause. The RES command must be issued to continue playback.
time_length
is a 2-byte pa-
rameter that can have any value up to 320 (64 seconds). The skip accuracy is five percent. This com-
mand is meaningful only in the PLAY sta te.
If the end of the mess age is detected d uring exec ution of SF, exe cution of t he command ter minates,
the EV_NORMAL_END bit in the status word sets, the MWRQST signal activates, and the processor switches
to the IDLE state.
If
time_length
is greater than 320, ERR_PARAM is set in the error word.
Playback speed does not affect the behavior of this command.
Example
SETD DE09
Byte sequence: Microcontroller 0F DE 09
VoiceDSP 0F DE 09
Description: Set time and day to Monday 1.30 A.M. (where Monday is the first day of the week)
SF 0019
Byte sequence: Microcontroller 22 00 19
VoiceDSP 22 00 19
Description: Skip forward five seconds from the current position in the message being played.
2-48
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
SLC Set Line Channels
active_channels
Configures the VoiceDSP to activate/deactivate channel 1. The activation of channel 1 is used to enable
the conversation recording feature. When activating this feature, the VoiceDSP sums channel 0 and
channel 1 during the record command and then compresses the sum. When playing the message, it is
decompressed and b ro adcasted to th e active chan n els as set by th is command. Note that ch ann el 0
is always active and is used as the default line channel.
Use the following parameter values to activate the channels:
Example
SMSG Set Message Pointer num_of_pages
Sets the message pointer to
num_of_pages
x 32 bytes from the beginning of the current message data.
If the VoiceDSP processor was in the IDLE state, the command opens the current message and switches
the VoiceDSP processor to the MSG_OPEN state. The microcontroller must issue an S command to close
the message, and switch the VoiceDSP processor to the IDLE state.
If
num_of_pages
x 32 is greater than the message length, EV_NORMAL_END is set in the status word, the
message pointer is set to the end of the message, and the VoiceDSP processor switches to the IDLE state.
If the current messag e is undefi ne d, ERR _INVALID is re port ed.
Example
SMT Set Message Tag message_tag
Sets the tag of the current message. The 2-byte
message_tag
can be used to implement mailbox func-
tions by including the mailbox number in the tag, or to handle old and new messages differently by using
one bit in the tag to mark the m essag e as old or new. See “Mes sag e Tag ” on pag e 2-3.
To change the message tag, you should first get the tag using the GMT command, read the tag, modify
it, and write it bac k.
Active Channel value Channel 1
1 Non-active
3Active
SLC 03
Byte sequence: Microcontroller 0x38 03
VoiceDSP 0x38 03
Description: Enables both channel 0 and 1
SMSG 00 0 A
Byte sequence: Microcontroller 30 00 0A
VoiceDSP 30 00 0A
Description: Set the message pointer to 10 pages (320 bytes) from the beginning of the current message
data.
2-49
2—SOFTWARE ISD-T360SB
ISD
NOTE Message tag bits can only be cleared. Message tag bits are set only when a message is first
created.
If the current message is undefined, ERR_INVALID is reported.
Example
SO Say One Word word_number
Plays the word number
word_number
in the curr ent vocabular y. The 1-byte
word_number
may be any
value from 0 through the index of the last word in the vocabulary. The VoiceDSP processor state changes
to SYNTHESIS.
When playback of the selected word has been completed, the VoiceDSP processor sets the
EV_NO RMAL_END bit in the status wo rd, and acti vates the MWRQST signal. The state then changes to IDLE.
If
word_number
is not defined in the current vocabulary, or if it is an IVS control or option code,
ERR_PARAM is set in the error word.
If the curren t v ocabulary is u n def ined, ERR_INV A LI D is reported.
Example
SPS Set Playback Speed speed
Sets the speed of message playback as specified by speed. The new speed applies to all recorded mes-
sages and synthesized messages (only if synt hesized using IVS), until changed by another SPS co mmand.
If this command is issued while the VoiceDSP processor is in the PLAY state, the speed also changes for
the message currently being played.
Speed
may be one of 13 values, from –6 to +6. A value of 0 represents normal speed. If
speed
is not in
the –6 to +6 range, ERR_PARAM is set in the error word.
SMT FFF7
Byte sequence: Microcontroller 05 FF F7
VoiceDSP 05 FF F7
Description: Mark the current message as old in a system where the message tag is encoded as described
in the example of the DMS command.
Note that the VoiceDSP processor ignores bits in the tag which are set to 1; only bit 3
is modi fi e d in the me ss a ge ta g .
SO 00
Byte sequence: Microcontroller 07 00
VoiceDSP 07 00
Description: Announce the first word in the word table of the currently selected vocabulary.
2-50
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
NOTE A negative speed value r epresents an increase i n speed, a positive value r epresents a decr ease
in speed.
NOTE The playback speed control is not applicable when the stored messages or the IVS data are not
compressed (Stored in PCM format).
The change in speed is approximate, and depends on the recorded data. In any case, if i < j, playback
speed with parameter i is the same or faster than with parameter j.
Example
SS Say Sentence sentence_n
Say senten ce num ber
sentence_n
of the cur rentl y selec ted voc abul ary .
sentence_n
is 1-byte long. The
VoiceDSP processor state changes to SYNTHESIS.
If the sentence has an argume n t, 0 is passed as the value for this ar g u m en t.
When play ing has be en comple ted, the VoiceDSP p rocessor sets the EV_NORMAL _END bit in the s tatus
word, and activates the MWRQS T signal. The state then changes to IDLE.
If
sentence_n
is not defi n ed in th e cu r r e n t v ocabulary, ERR_PARAM is set in the error word.
If the curren t v ocabulary is u n def ined, ERR_INV A LI D is reported.
Example
SPS FB
Byte sequence: Microcontroller 16 FB
VoiceDSP 16 FB
Description: Set playback speed to –5.
SS 00
Byte sequence: Microcontroller 1F 00
VoiceDSP 1F 00
Description: Announce the first sentence in the sentence table of the currently selected vocabulary.
2-51
2—SOFTWARE ISD-T360SB
ISD
SSM Set Speakerphone Mode mode
Sets the speakerphone to the
mode
mod e of o p er atio n. Th e co mm an d is val id w hen th e Vo ic eDS P pr o-
cessor is in IDLE state.
mode
can be one of:
0 OFF Deac tivate the sp eakerphone, and return the VoiceDSP processor to nor -
mal operation mode.
1 ON Put the VoiceDSP processor in speakerphone mode and activate speaker-
phone in full-duplex mode i.e., with full cancellation of both the acous tic
and the el ectric al ech oes. To ne det ectors are not acti ve. G ains in the Sen d
and Receive paths are set by the relevant tunable parameters.
2 TRANSPARENT Activate the speakerphone with no echo cancellation (this mode is used
for system tuning).
3 MUTE Activat e the sp eakerp hon e, while g enera ti ng sile nce to the line. T he ne ar-
end-listener can hear the far-end-speaker, but not vice versa. Tone detec-
tors are n ot active .
4 LISTEN The line is audibl e o n the sp eak er. Tone det ector s are a ctive. T his mod e is
used for call generation.
5 Reserved.
6 RESTART Restart the current speakerphone mode. This mode differs from ON; it does
not require full initialization of the speakerphone. It should be used to re-
sume the speakerphone operation to adjust to an environment change
(e.g., parallel pickup ).
7 HOLD Stop the codec interrupts. Neither side can hear each other.
8 SILENCE The sp eakerphone i s disabled. El ectric ech o cancellati on is activ e for the
CAS detection (CIDCW).
See “Full-duplex Sp eakerphon e” on page 2- 14 for more detail s.
NOTE Only commands that are specif ied in Table 2-3, are active during all speakerpho ne modes (other
than 0).
Example
SSM 01
Byte sequence: Microcontroller 2F 01
VoiceDSP 2F 01
Description: Put the VoiceDSP processor into Speakerphone mode, and set the speakerphone to full-duplex
mode.
2-52
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
SV Set Vocabulary Type type id
Selects the vocabulary table to be used for voice synthesis. The vocabulary type is set according to the
1-byte
type
paramete r:
0 For compati bility on ly.
1 External vocab ula ry in RO M.
2 External vocab ulary in Flash.
3-7 Reserved.
The host is responsible for selecting the current vocabulary, with SV command, before using an SAS, SO,
SS or S W command. Each external v ocabular y table h as a un ique id which is part of the vocabu lary in -
ternal header (See the IVS User’s Guide for more details). If type is 1 or 2, the VoiceDSP processor search-
es for the on e by te
id
parameter in each vocabul ary tab le header until a match is found.
If the
id
parameter doe s no t p oint to a valid IVS voc a bulary, ERR_PARA M is se t in th e error wo rd .
Example
SW Say Words n word1...word
n
Plays n words, indexed by
word
1
to
word
n
(
<9
). The VoiceDSP processor state changes to SYNTHESIS.
On completion, t he EV_NORMAL _E ND bit in t h e status wor d is set, an d the MWRQ S T signal goes low. The
state then changes to IDLE.
If one of the wor ds is not defined in the current vocabulary, or if it is an IVS contro l or option code, or if
n > 8, ERR_ PARAM is re port ed .
If the curren t v ocabulary is u n def ined, ERR_INV A LI D is reported.
Example
SV 02 03
Byte sequence: Microcontroller 20 02 03
VoiceDSP 20 02 03
Description: Select the vocabulary with vocabulary-id 3, which resides on a Flash, as the current
vocabulary.
SW 02 00 00
Byte sequence: Microcontroller 21 02 00 00
VoiceDSP 21 02 00 00
Description: Announce the first word, in the word table of the currently selected vocabulary, twice.
2-53
2—SOFTWARE ISD-T360SB
ISD
TUNE Tune index parameter_value
Sets the value of the tunable parameter identified by
index
(one byte) to the 2-byte value,
parameter_value
. This command may be used to tune the DSP algorithms to a specific Data Access Ar-
rangem ent (DAA) in terface, o r to change oth er paramet ers. If you do not use TUNE, th e VoiceDSP p ro-
cessor uses default values.
If
index
does not point to a valid tunable parameter, ERR_PARAM is set in the error word.
NOTE The tunable parameters are assigned with their default values on application of power. The INIT
command does not affect these parameters.
The index param e te r of th is c om m an d s ho u ld be in h e xid e ci m a l n um b ers w hi le t he in d ex o f th e
tunable parameter should be a decimal number.
Example
The tables 2-4 through 2-13 describe the tunable parameters, their index numbers and their default val-
ues, group ed by their functional ity.
TUNE 17 02BC
Byte sequence: Microcontroller 15 17 02 BC
VoiceDSP 15 17 02 BC
Description: Set the minimum period for busy detection to 700 (7 seconds).
Table 2-4: TUNABLE PARAMETERS: Software Automatic Gain Control
Index Parameter Name Description Default
70 SWAGC enabling:
AGC_enable_flag Enables the SWAGC. Value 1 to enable, value 0 to disable.
Different values of the tunable parameter can be used for the
compression and Speakerphone algorithms. In order to change
the value of the tunable parameter use the TUNE command prior
to entering each of the two algorithms.
Required value for speakerphone mode is 0 - disabling the
SWAGC.
Legal values: 0,1
1
86 Input threshold for gain
modification:
AGC_VAD_THR
Power threshold for Voice activity detection. Only signals stronger
than the threshold are considered voice and are increased by the
SWAGC algorithm. The default value is equivalent to –38dBm input.
Each multiplication of the value by 2 adds approximately 3dB to
the input level.
Different values of the tunable parameter can be used for the
compression and Speakerphone algorithms. In order to change
the value of the tunable parameter use the TUNE command prior
to entering each of the two algorithms.
Legal values: 0 - 32767
4
2-54
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
87 SWAGC time scale
operation:
AGC_DECAY_COEFF
Determines the desired time scale for SWAGC algorithm operation.
Increa sing this parameter decr eases the time scale, and the
SWAGC reacts quicker to changes in input levels and can become
more sensitive to fluctuations in signal power. Default value is
equivalent to 5 sec per 12dB change.
Different values of the tunable parameter can be used for the
compression and Speakerphone algorithms. In order to change
the value of the tunable parameter use the TUNE command prior
to entering each of the two algorithms.
Legal values: 0 - 32767
2048
85 SWAGC low power
bound:
AGC_POWER_LOW
Low bound of desired signal power. The SWAGC algorithm
amplifies the gain whenever the output signal power decreases
below AGC_POWER_LOW.
The default value is equivalent to –17dBm output power.
Each multiplication of the value by 2 adds approximately 3dB to
the output power.
Different values of the tunable parameter can be used for the
compression and Speakerphone algorithms. In order to change
the value of the tunable parameter use the TUNE command prior
to entering each of the two algorithms.
Legal values: 0 - 32767
512
11 SWAGC High power
bound:
AGC_POWER_HIGH
High bound of desired signal power. The SWAGC algorithm
atten uate s the gain whene ver the output signal powe r increases
above AGC_POWER_HIGH.
The default value is equivalent to
–11dBm output power.
Each multiplication of the value by 2 adds approximately 3dB to
the output power.
Different values of the tunable parameter can be used for the
compression and Speakerphone algorithms. In order to change
the value of the tunable parameter use the TUNE command prior
to entering each of the two algorithms.
Required value for speakerphone mode is 5000. This value should
be given only if the SWAGC is enabled in this mode.
Legal values: 0 - 16384
2048
88 SWAGC maximum gain
step:
AGC_MAX_GAIN
This tunable parameter controls the maximum gain of the SWAGC
algorithm. Adding 1 to the tunable parameter increases the gain
by 6dB.
Default is equivalent to 12dB gain.
Different values of the tunable parameter can be used for the
compression and Speakerphone algorithms. In order to change
the value of the tunable parameter use the TUNE command prior
to entering each of the two algorithms.
Legal values: 0 - 3
2
Table 2-4: TUNABLE PARAMETERS: Software Automatic Gain Control
Index Parameter Name Description Default
2-55
2—SOFTWARE ISD-T360SB
ISD
Table 2-5: TUNABLE PARAMETERS: Tone Generation and Message Playback
Index Parameter Name Description Default
27 DTMF Generation:
DTMF_GEN_TWIST_LEVEL A one-byte value that controls the twist level of a DTMF tone,
generated by the GT command, by controlling the energy level of
each of the two tones (low frequency and high frequency)
composing the DTMF tone. The Least Sig nificant Nibble (LSN)
controls the low tone and the Most Significant Nibble (MSN) controls
the high tone. The energy level of each tone, as measured at the
output of a TP3054 codec (before the DAA) connected to the
VoiceDSP processor is summarized in the following table:
Nibble Value Tone Energy (dB-Volts)
0 0
1 –17.8
2–14.3
3–12.9
4–12.4
5–12.0
6–11.9
7–11.85
8 - 15 –11.85
The volume of the generated DTMF tone during measurements
was 6. (TONE_GEN_LEVEL+
vol_level
= 6).
Note: the vol_level parameter is controlled by the VC command.
For the default level, the high tone is –14.3 dBV and the low tone is
–12.4 dBV, which gives a DTMF twist level of 1.9 dB.
The energy level of a single generated tone is the level of the low
tone.
66
16 Tone Generation:
TONE_GEN_LEVEL Controls the energy level at which DTMF and other tones are
generated. Each unit represents 3 dB. The default level is the
reference lev el.
For example, if you set this parameter to 4, the energy level is 6 dB
less than the default level. The actual output level is the sum of
TONE_GEN_L EV EL a nd the
vol_level
parameter, controlled by the
VC command. The tones are distorted when the level is set too
high.
Legal values: 0 £ TONE_GEN_LEVEL + vol_le vel £ 12.
6
21 VCD Playback and Voice
Synthesis:
VCD_PLAY_LEVEL
Controls the energy during playback and external voice synthesis.
Each unit represents 3 dB. The default level is the reference level.
For example, if you set this parameter to 4, the energy level is 6 dB
less than the default level. The actual output level is the sum of
VCD _PLAY _LE VEL a nd th e
vol_level
parameter (controlled by the VC
command). Speech is distorted when the level is set too high.
Legal values: 0 £ VCD_PLAY_LEVEL + vol_level £ 12.
6
2-56
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
Table 2-6: TUNABLE PARAMETERS: DTMF Detection
Index Parameter Name Description Default
17 Energy Level:
DTMF_DET_MIN_ENERGY Minimum energy level at which DTMF tones are detected. If you
divide (multiply) the value by 2, the detection sensitivity decreases
(increases) by 3 dB.
Legal values: 8 to 4096
32
24 Echo Canceler:
DTMF_DET_ECHO_DELAY The near-echo delay in samples. The sampling rate is 8000Hz (i.e.,
125 ms per sample).
Legal values: 0 to 8.
4
26 Twist Level:
DTMF_DET_REV_TWIST Controls the reverse twist level at which the VoiceDSP process or
detects DTMF tones. While the normal twist is set at 8 dB, the reverse
twist can be either 4 dB (default) or 8 dB (if this parameter is set to
1).
0
60 SW AGC:
DTMF_DET_AGC_IDLE SW AGC for DTMF in Idle/Record modes. When incrementing the
tunable by 1, the dynamic range is increased by 3 dB.
Legal values: 0 to 5.
0
61 SW AGC:
DTMF_DET_AGC_PLAY SW AGC for
PLAY
and
TONE_GENERATE
mod es . W hen i n creme nti n g
the tunable by 1, the dynamic range is increased by 3 dB.
Legal values: 0 to 5.
3
Table 2-7: TUNAB LE PARA METE RS : Tone Detection
Index Parameter Name Description Default
18 Dial Tone:
TONE_DET_TIME_COUNT Controls the duration of a tone before it is reported as a dial tone,
in 10 msec units. The accuracy of the constant is ±10 ms.
Legal values: 0 to 65535.
700
19 Busy and Dial Tone:
TONE_DET_ON_ENERGY_
THRESHOLD
Minimum energy level at which busy and dial tones are detected
as ON (after 700Hz filtering). If you divide (multiply) the value by 2
you get about 3 dB decrease (increase) in the threshold.
The mapping between energy level and the parameter value is as
follows (measured on the codec output when a 400Hz tone was
injected to the codec input):
Tunable value Energy threshold (dB-Volts)
10 –31.8
20 –28.6
100 –21.7
500 –14.7
8000 –2.5
Legal values: 0 to 65535.
160
2-57
2—SOFTWARE ISD-T360SB
ISD
20 Busy and Dial Tone:
TONE_DET_OFF_ENERGY_
THRESHOLD
Maximum energy level at which busy and dial tones are detected
as OFF (after 700Hz filtering). If you divide (multiply) the value by 2
you get about 3 dB decrease (increase) in the threshold.
The mapping between energy level and the parameter value is the
same as for TONE_DET_ON_ENERGY_THRESHOLD
Legal values: 0 to 65535.
110
23 B u sy Ton e :
BUSY_DET_MIN_TIME Minimum time period for busy detection, in 10 ms units. The
accuracy of the constant is ±10 ms.
Legal values: 0 to 65535.
600
53 B u sy Ton e :
BUSY_DET_MIN_ON_TIME Minimum period considered as On period for busy tone detection.
Note that for weak signals:
(–30 dB and below) the maximum value is 12 (i.e., 120 ms
minimum detection time).
Unit: 10 ms. Accuracy is ±20 ms.
Legal values: 10 to 1000.
10
54 B u sy Ton e :
BUSY_DET_MAX_ON_TIME Maximum period considered as On for busy-tone detection. Unit:
10 ms. Accuracy is ±20 ms .
Legal values: 10 to 1000.
170
55 B u sy Ton e :
BUSY_DET_MIN_OFF_TIME Minimum period considered as Off for busy-tone detection. Unit:
10 ms. Accuracy is ±20 ms .
Legal values: 5 to 1000.
7
56 B u sy Ton e :
BUSY_DET_MAX_OFF_TIME Maximum period considered as On for busy-tone detection. Unit:
10 ms. Accuracy is ±20 ms .
Legal values: 5 to 1000.
130
57 B u sy Ton e :
BUSY_DET_VERIFY_COUNT Number of On/Off cadences that must be detected prior to
reporting busy-tone presence.
Legal values: 9 to 127.
9
58 B u sy Ton e :
BUSY_DET_TONE_TYPE Specifies the type of busy tone to detect:
1 —Two cadences
2 —Three cadences
3 —Both two and three cadences
1
59 B u sy Ton e :
BUSY_DET_DIFF_THRESHOLD The maximum allowed difference between two compared On or
Off periods. Unit: 10 ms.
Legal values: 0 to 255.
9
Table 2-7: TUNAB LE PARA METE RS : Tone Detection
Index Parameter Name Description Default
2-58
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
Table 2-8: TUNABLE PARAMETERS: Energy Detection
Index Parameter Name Description Default
10 Silence (VOX):
VOX_DET_ENERGY_THRES
HOLD
This parameter determines the minimum energy level at which
voice is detected. Below this level, it is interpreted as silence.
If you divide (multiply) the value by 2, you get approximately a 3db
decrease (increase) in the threshold. The mapping between
energy level and the parameter value is as follows (measured on
the codec output when a 400Hz tone was injected into the codec
input): Tunable value Energy threshold (dB-Volts)
10 –45
20 –42
40 –39
80 –36
160 –33
320 –30
640 –27
1280 –24
2560 –21
Legal values: 1 to 32767.
12
12 Silence (VOX):
VOX_DET_TIME_COUNT This parameter, in units of 10 ms, determines the period of silence
before the VoiceDSP processor reports silence. The accuracy of the
constant is ±10 ms.
Legal values: 0 to 65535.
700
22 Silence (VOX):
VOX_DET_TOLERANCE_TIM
E
Control s the maxim um ene rgy -perio d, in 10 ms units, that does
NOT reset the vo x de te ctor.
Legal values: 0 to 255.
3
47 Constant Energy:
CONST_NRG_DET_TIME_
COUNT
Minimum elapsed time until the VoiceDSP processor reports
constant energy level. Units: 10 ms. Accuracy: ±10 ms
Legal values: 1 to 65534
700
48 Constant Energy:
CONST_NRG_DET_
TOLERANCE_TIME
Variations in constant energy, up to this time, do not reset the
constant energy detector. Units: 10 ms.
Legal values: 0 to 255
5
49 Constant Energy:
CONST_NRG_DET_LOW_
THRESHOLD
Determines the minimum energy level that is treated as constant
energy. The minimum energy is calculated as follows:
(1—1/2CONST_NRG_DET_LOW_THRESHOLD) * average_energy
Legal values: 1 to 16
1
50 Constant Energy:
CONST_NRG_DET_HIGH_
THRESHOLD
Determines the maximum energy level that is treated as constant
energy. The maximum energy is calculated as follows:
(1 + 1/2CONST_NRG_DET_HIGH_THRESHOLD) * average_energy
Legal values: 0 to 16
1
2-59
2—SOFTWARE ISD-T360SB
ISD
Table 2-9: TUNABLE PARAMETERS: Speakerphone
Index Parameter Name Description Default
31 Acoustic Echo Canceler
(AEC):
SP_AEC_PRIORITY_BIAS
Controls the bias in priority between the Send and Receive paths.
If send priority-bias is preferred, the value should be greater than
zero.
For no priority bias, the value should be zero.
For priority bias for the Receive path, the value should be negative.
Steps are 3 dB each (e.g., +3 is 9 dB bias for the Send path, –2 is 6
dB bias for the Receive path).
Legal values: –4 to 4.
Note: this parameter is represented as a signed byte in the system.
When reading its value with the GTUNE command, ignore the
upper byte of the returned va lue.
0
32 Acoustic Echo Canceler
(AEC):
SP_AEC_COUPLING_
LOSS_THRESHOLD
This parameter limits the acoustic return loss. Its value
(SP_AEC_COUPLING_LOSS_THRESHOLD / 32767) is compared with
the RMS value of Sout divided by the RMS value of Rin, during a
single-talk event. The loop gain is decreased, if necessary, to
control the TCL level.
For SP_AEC_COUPLING_LOSS_THRESHOLD = 32767 this loop is
disabled.
Legal values: 0 to 32767.
2047
34 Acoustic Echo Canceler
(AEC):
SP_AEC_LR_LEVEL
Controls the speakerphone gain from the microphone to the line-
out. The total attenuation, or gain, depends on both the analog
gain and this value. The gain is:
K * signal
where:
K = SP_AEC_LR_LEVEL/4096.
Legal values: 0 to 16000.
14000
35 Electric Echo Canceler
(EEC):
SP_EEC_LR_LEVEL
Controls the speakerphone gain from the line-in to the speaker. The
total attenuation, or gain, depends on both the analog gains and
this value. The gain is:
K * signa l
where:
K = (SP_EEC_LR_LEVEL/4096) * (2(6 + vol_level)/2)
(vol_level is controlled by the VC command)
Legal values: 0 to 400.
281
2-60
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
36 Acoustic Echo Canceler
(AEC):
SP_AEC_CLIP_POS
Specifies the positive peak-value at which the analog circuit of the
speaker saturates. Codec analog full scale corresponds to µLAW
full scale values after expansion. Assume that positive saturation
occurs at amplitudes higher than those of a sine wave at X [dBm0].
The SP_AEC_CLIP_POS value is set as:
SP_AE C _C LIP _P O S = 3263 6 * 10((X – 3.17)/20)
Note: a sine wave with amplitude 4 * 8159 = 32636
corresponds to 3.17 dBm0.
Example:
For X = –6.2761 dBm0, the value is:
SP_AEC_CLIP_POS = 32636 * 10((–6.2761– 3.17)/20) = 0.33 71 * 32 636
=11000;
Legal values: 0 to 32767.
16000
37 Acoustic Echo Canceler
(AEC):
SP_AEC_CLIP_NEG
Specifies the negative peak value at which the analog circuit of
the speaker saturates. Codec analog full scale corresponds to
µLAW full scale values after expansion.The value of
SP_AEC_CLIP_NEG is set as shown for SP_AEC_CLIP_POS, above.
Legal values: –32768 to 0.
–16000
38 Electric Echo Canceler
(EEC):
SP_EEC_CLIP_POS
Specifies the positive peak value at which the analog circuit of the
line-out saturates. Codec analog full scale corresponds to µLAW full
scale values after expansion. The value of SP_EEC_CLIP_POS is set
as shown for SP_AEC_CLIP_POS, above.
Legal values: 0 to 32767.
16000
39 Electric Echo Canceler
(EEC):
SP_EEC_CLIP_NEG
Specifies the negative peak value at which the analog circuit of
the line-out saturates. Codec analog full scale corresponds to
µLAW full scale values after expansion. The value of
SP_EEC _CL IP_N EG is set as shown for SP_ AEC_C LIP _ PO S, above.
Legal values: –32768 to 0.
–16000
43 Acoustic Echo Canceler
(AEC):
SP_AEC_VOX_HYST
Controls the hysteresis in near-talker detection. (The speakerphone
state machine has a built-in hysteresis mechanism to prevent
fluctuations in the talker identification process i.e., identifying the
active side.) The value of this parameter is a dimensionless
number, which should be evaluated during the tuning process for
specific hardware.
Larger values for the parameter correspond to a wider hysteresis
loop. Negativ e values increase the pro bability that the state
machine remains in the last state.
Legal values: –127 to 127.
Note: this parameter is represented as a signed byte in the system.
When reading its value with the GTUNE command, ignore the
upper byte of the returned va lue.
10
Table 2-9: TUNABLE PARAMETERS: Speakerphone
Index Parameter Name Description Default
2-61
2—SOFTWARE ISD-T360SB
ISD
NOTE: Refer to the Software Automatic Gain Control tunable parameters, listed in Table 2-4, for more details on
speakerphone-related tunable parameters.
44 Electric Echo Canceler
(EEC):
SP_EEC_VOX_HYST
Controls the hysteresis in far-talker detection. (The speakerphone
state machine has a built-in hysteresis mechanism to prevent
fluctuations in the talker identification process i.e., identifying the
active side.) The value of this parameter is a dimensionless
number, which should be evaluated during the tuning process for
specific hardware.
Larger values for the parameter correspond to a wider hysteresis
loop. Negativ e values increase the pro bability that the state
machine remains in the last state.
Legal values: –127 to 127.
Note: this parameter is represented as a signed byte in the system.
When reading its value with the GTUNE command, ignore the
upper byte of the returned va lue.
10
45 Acoustic Echo Canceler
(AEC):
SP_AEC_DTD_TH
Controls the sensitivity of the system. Low values correspond to high
sensitivity, with a greater false alarm probability (i.e., an echo is
considered a real talker).
High values correspond to low sensitivity, with slower switching. This
parameter is affected by the loop gain and the specific hardware
characteristics.
Legal values: 0 to 127.
73
46 Electric Echo Canceler
(EEC):
SP_EEC_DTD_TH
Controls the sensitivity of the system. Low values correspond to high
sensitivity, with a greater false alarm probability (i.e., an echo is
considered a real talker).
High values correspond to low sensitivity, with slower switching. This
parameter is affected by the loop gain and the specific hardware
characteristics.
Legal values: 0 to 127.
82
33 Attenuation:
SP_BLOCK_LEVEL Controls the maximum attenuation level of the speakerphone
suppressors. It affects the speakerphone stability and its subjective
quality. The maximum attenuation is calculated according to:
SP_BLOCK_LEVEL/228
Legal values: 550 to 32000.
10922
42 Tone Generation:
SP_TONE_GEN_
LEVEL
Controls the energy level at which DTMF, and other tones, are
generated to the line (codec 0) while the speakerphone is active.
Each unit represents 3 dB.
Legal values: 0 to 10.
Note: the energy level at whi ch the tones ar e generated to
th e sp e a ke r (c o de c 1) while th e spe a ke r p ho ne is ac ti v e , is
controlled by the TONE_GEN_LEVEL tunable parameter and
th e vo l_lev e l.
Note: vol_level is controlled by the VC command.
6
Table 2-9: TUNABLE PARAMETERS: Speakerphone
Index Parameter Name Description Default
2-62
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
Table 2-10: TUNABLE PARAMETERS: Memory Support
Index Parameter Name Description Default
62 Memory Device Size:
NUM_OF_BLOCKS_IN_ME
M
Defines the number of blocks (each block is of 4096 bytes) in every
Flash memory device. The number and type of connected
devices are defined by the CFG command.
Flash Device Size (Mbits) Number of Blocks Value
4128
8256
16 512
Note: this parameter must be tuned before the CFG command.
128
63 Memory Size for Testing
NUM_OF_BLOCKS_FOR_TE
ST
Defines the number of blocks (each block is of 4096 bytes) in every
Flash memory device for production line testing purposes.
The number should be small to minimize testing time during the
production sequence. However, the number of blocks should be
larger than the number of expected bad blocks in the memory
device.
In case of v alue=0, no production test is performe d.
In any case other than value= 0, the number of blocks is defined
by the parameter value, and a production testing cycle is
performed after RESET.
Legal values: 0 to the value of tune 62.
Note: If power fails during production testing cycle, the
memory status is unpredicted. The memory device should be
replaced and the production test should be repeated.
Note: this parameter must be tuned before the INIT command.
0
Table 2-11: TUNABLE PARAMETERS: Codec Support (Samples)
Index Parameter Name Description Default
65 Channel 0 Delay:
CFRD0 The delay of codec channel 0 from Frame Synch 0 (CFS0) to
start of valid data.
Legal values: 0 to 255
1
66 Channel 1 Delay:
CFRD1 The delay of codec channel 1 from Frame Synch 0 (CFS0) to
start of valid data.
Legal values: 0 to 255
9
67 Channel 2 Delay:
CFRD2 The delay of codec channel 2 from Frame Synch 0 (CFS0) to
start of valid data.
Legal values: 0 to 255
17
2-63
2—SOFTWARE ISD-T360SB
ISD
Table 2-12: TUNABLE PARAMETERS: Clocking
68 Frame Synch Delay:
CFSD The delay of Frame Synch 1 (CFS1) from Frame Synch 0
(CFS0).
Legal values: 0 to 255
16
69 D ata Valid Delay:
CFET The delay between Frame Synch 0 (CFS0) to end of valid
data of all channels.
Legal values: 0 to 255
25
Index Parameter Name Description Default
79 PLL_Configuration:
PLL_MODE Controls the internal clock PLL multiplication factor, which
generates the master system clock. The default value (0)
allows using external crystal (or oscillator) of 4.096Mhz.
Setting this value to 1 allows using external crystal (or
oscillator) of 65.536Mhz.
Legal values: 0, 1
0
Table 2-11: TUNABLE PARAMETERS: Codec Support (Samples)
Index Parameter Name Description Default
2-64
2—SOFTWARE
ISD-T360SB
Voice Solutions in Silicon™
VC Volume Cont rol
vol_level
Controls the energy level of all the output generators (playback, tone generation, voice synthesis and
speakerphone), with one command. The resolution is ±3 dB.
The actual outpu t level is composed of the tunable level variable, plu s th e
vol_level
. The valid range for
the actual output level of each output generator is defined in Table 2-5.
For example, if the tunable variable VCD_PLAY_LEVEL (parameter number 21) is 6, and
vol_level
is –2,
then the output level equals VCD_PLAY_LEVEL +
vol_level
= 4.
Example
WMSG Write Message
data
Writes 32 bytes of data to the current position of the message pointer, and advances the message point-
er by 32 byte s .
If the VoiceDSP processor is in IDLE state, the command opens the current message, switches the
VoiceDSP processor to MSG_OPEN state, sets the message pointer to the beginning of the message da-
ta, and writes the 32 bytes of data.
To add data at the end of an existing message, issue the SMSG command to the last page of the mes-
sag e. Iss ue the WM SG co mma nd wit h a b uff er con sis tin g of 32 FF by tes (t his ha s no e ff ect o n t he cur ren t
data in the page). A subsequent WMSG command adds a new block to the message, and writing con-
tinues at the beginning of the new block.
The microcontroller must issue an S command to close the message and switch the VoiceDSP processor
to IDLE state.
If the current messag e is undefi ne d, ERR _INVALID is re port ed.
NOTE When updating an existing message, bits can only be cleared, but not set.
Example
VC 04
Byte sequence: Microcontroller 28 04
VoiceDSP 28 04
Description: Set the volume level to VCD_PLAY_LEVEL + 4.
WMSG 32 bytes
Byte sequence: Microcontroller 31 32 bytes of data to write
VoiceDSP 31 echo 32 bytes of data
Description: Write 32 bytes in the message memory.
2-65
2—SOFTWARE ISD-T360SB
ISD
2—SOFTWARE
ISD-T360SB
2-66 Voice Solutions in Silicon™
3-1
3—SCHEMATIC DIAGRAMS ISD-T360SB
ISD
Chapter 3—SCHEMATIC DIAGRAMS
3.1 APPLICATION INFORMATION
The follo wing schem atic diagrams are extrac ted from a Voi ceDSP demo un it, based on t he ISD-ES360
board.
The demo includes three basic clusters:
• COP888EEG Microcontroller
• VoiceDSP processor cluster working with 3.3v, including two TP3054 codecs and an ISD-T360SB
controlling Samsung 16Mb NAND flash.
• User interface that incl u des a 12-key (4x3) keypad.
For more details about the demo, please refer to the ISD-ES360 Demo Operating Instructions.
3—SCHEMATIC DIAGRAMS
ISD-T360SB
3-2 Voice Solutions in Silicon™
Figure 3-1: VoiceDSP ISD-T360SB
CFS0
PB7
C51
2.2UF
R73 10K
PB5
PB3
CDIN
C34
2.2UF
C46
33PF
C41
2.2UF
RESET-
CCLK
PB2
PC1
C39
2.2UF
Vcc3.3
MWRDY
R74 10K
MWRQST
PB[0..7]
C47
.1UF
R64
10M
C43
.1UF
Vcc3.3
Vcc3.3
PC[0..4]
TST
MWCLK
PB1
C45
33PF
C40
.1UF
PB4
Y1
4.096MHz
Vcc3.3
PB6
C35
2.2UF
C33
.1UF
MWDIN
C44 .1UF
R63 10K
PC0
CDOUT
ISD-T360SA
U17
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
60
59
58
57
56
55
54
53
52
51
50
49
48
47
46
45
44
43
42
41
80
79
78
77
76
75
74
73
72
71
70
69
68
67
66
65
64
63
62
61
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
NC
NC
NC
NC
NC
NC
NC
NC
A9
A10
NC
NC
nISE
D0
D1
D2/RA11
Vss
D3
VCCHI
Vcc
D4
D5
D6
D7
PC0/A11
PC1/A12
NC
NC
NC
NC
NC
PC2/A13
PC3/A14/BE0
PC4/A15/BE1
PC5/nIOCS/ENV3
PC6/nEMCS
PC7/nBMCS/ENV1
PB0/D8
PB1/D9
Vcc
CDIN
CFS0
CDOUT
CFS1/PWM
nRESET
NC
NC
NC
NC
NC
NC
NC
nINT3/nMWCS
PB7/D15
PB6/D14
PB5/D13
PB4/D12
PB3/D11
Vss
PB2/D10
NC
NC
NC
NC
VssA
X2/CLKIN
X1/PLI
TST/PA0/NwrO
PA1/nWR1
PA2/CTTL
PA3/PFS
PA4/nMWRDY
PA5/MWDOUT
PA6/BST0
Vss
nMWRQST/PA7/BST1
Vcc
PD0/MWCLK
PD1/MWDIN
CCLK
NC
NC
NC
Vss
NC
VccA
Vcc
A8
A7
A6
A5
A4
A3
A2
Vcc
A1
Vss
A0/A16/nDDIN
NC
NC
CFS1
MWRQST
PC4
R59 10 ohm
Vcc3.3
MWCS
MWRDY
PB0
MWDOUT
PC2
C48 2.2UF
VoiceDSP ISD-T360SA
PC3
C36 .1UF
CFS0
PB7
C51
2.2UF
R73 10K
PB5
PB3
CDIN
C34
2.2UF
C46
33PF
C41
2.2UF
RESET-
CCLK
PB2
PC1
C39
2.2UF
Vcc3.3
MWRDY
R74 10K
MWRQST
PB[0..7]
C47
.1UF
R64
10M
C43
.1UF
Vcc3.3
Vcc3.3
PC[0..4]
TST
MWCLK
PB1
C45
33PF
C40
.1UF
PB4
Y1
4.096MHz
Vcc3.3
PB6
C35
2.2UF
C33
.1UF
MWDIN
C44 .1UF
R63 10K
PC0
CDOUT
ISD-T360SA
U17
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
60
59
58
57
56
55
54
53
52
51
50
49
48
47
46
45
44
43
42
41
80
79
78
77
76
75
74
73
72
71
70
69
68
67
66
65
64
63
62
61
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
NC
NC
NC
NC
NC
NC
NC
NC
A9
A10
NC
NC
nISE
D0
D1
D2/RA11
Vss
D3
VCCHI
Vcc
D4
D5
D6
D7
PC0/A11
PC1/A12
NC
NC
NC
NC
NC
PC2/A13
PC3/A14/BE0
PC4/A15/BE1
PC5/nIOCS/ENV3
PC6/nEMCS
PC7/nBMCS/ENV1
PB0/D8
PB1/D9
Vcc
CDIN
CFS0
CDOUT
CFS1/PWM
nRESET
NC
NC
NC
NC
NC
NC
NC
nINT3/nMWCS
PB7/D15
PB6/D14
PB5/D13
PB4/D12
PB3/D11
Vss
PB2/D10
NC
NC
NC
NC
VssA
X2/CLKIN
X1/PLI
TST/PA0/NwrO
PA1/nWR1
PA2/CTTL
PA3/PFS
PA4/nMWRDY
PA5/MWDOUT
PA6/BST0
Vss
nMWRQST/PA7/BST1
Vcc
PD0/MWCLK
PD1/MWDIN
CCLK
NC
NC
NC
Vss
NC
VccA
Vcc
A8
A7
A6
A5
A4
A3
A2
Vcc
A1
Vss
A0/A16/nDDIN
NC
NC
CFS1
MWRQST
PC4
R59 10 ohm
Vcc3.3
MWCS
MWRDY
PB0
MWDOUT
PC2
C48 2.2UF
VoiceDSP ISD-T360SA
PC3
C36 .1UF
ISD-360SB
3-3
3—SCHEMATIC DIAGRAMS ISD-T360SB
ISD
Figure 3-2: Nand Flash Device
PC0
PB0
PC3
PB6
VCC
PB5
PC1
PB[0..7]
PC2
PB4
C37
.1UF
PC4
PB2
PB7
PB1
C38
.1UF
NAND FLASH
U9
KM29W16000AT
1
2
3
4
5
6
7
8
9
10
13
14
15
16
17
18
19
20
21
22
32
31
30
29
28
27
26
25
24
23
44
43
42
41
40
39
38
37
36
35
VSS
CLE
ALE
WE
WP
N.C
N.C
N.C
N.C
N.C
N.C
N.C
N.C
N.C
N.C
I/O0
I/O1
I/O2
I/O3
VSS
N.C
N.C
N.C
N.C
N.C
I/O7
I/O6
I/O5
I/O4
VCC
VCC
CE
RE
R/B
GND
N.C
N.C
N.C
N.C
N.C
R46
4.7K
PC[0..4]
VCC
PB3
R47
100K
3-4
3—SCHEMATIC DIAGRAMS
ISD-T360SB
Voice Solutions in Silicon™
Figure 3-3: Line and Speakerphone Codec
R54
100K
CDOUT
GSXA
CINA
CINB
LINE CODEC
VFXPA
CINA
VCCA J3
CON12
1
2
3
4
5
6
7
8
9
10
11
12
R52
10K
R39 10
C31
.1UF
CDIN
C26
.1UF
C29
.1UF
SPEAKERPHONE CODEC
R51 0
VFXPB
COUTA
VCCA
CFS0
C23
22UF
C28
.1UF
NVCC
VFXNA
GSXB
VFXNB
R55
0
C25
.1UF
U12
TP3054
1
2
3
4
5
6
7
8 9
10
11
12
13
14
15
16
VBB
GNDA
COUT
VCC
FSR
DR
BCLKR
MCLKR MCLKX
BCLKX
DX
FSX
TSX
GSX
VFX-
VFX+
R49
0
CINB
C30
.1UF
CFS1
COUTB
R48
10K
CCLK
CDOUT
NVCC
U10
TP3054
1
2
3
4
5
6
7
8 9
10
11
12
13
14
15
16
VBB
GNDA
COUT
VCC
FSR
DR
BCLKR
MCLKR MCLKX
BCLKX
DX
FSX
TSX
GSX
VFX-
VFX+
R50
100K
R41
10
COUTA
CCLK
CDIN
COUTB
CFS1
R56 0
C24
22UF
CFS0
3-5
3—SCHEMATIC DIAGRAMS ISD-T360SB
ISD
Figure 3-4: Microcontroller and User Interface
S5
1 3
OFFHK\
6
7
S6
1 3
MWDIN
#
RING-
RES
E
VCC
COL2
S12
1 3
S8
1 3
MWCLK
S7
1 3
COL1
COL1
U19
MAX701
46
5
7
8
3
2
1
GND RESET
RESET
N.C
Vcc
N.C
N.C
MR
R60
10K
8
C50
39PF
ROW1
MWDOUT
COL3
ROW4
COP888EG1
U20
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
39
38
37
36
35
34
33
32
31
30
29
44
43
42
41
40
6
5
4
3
2
1
CKI
VCC
I0
I1
I2
I3
I4
I5
I6
I7
L0
L1
L2
L3
C4
C5
C6
C7
L4
L5
L6
L7
G0
RESET
GND
D7
D6
D5
D4
D3
D2
D1
D0
C1
C0
G3
G2
G1
G7
G6
G5
G4
C3
C2
D11
SD103
2 1
VCC
2
MICROCONTROLLER AND USER INTERFACE
R69
1M
1 3
C52
39PF
S9
1 3
R58
10K
Y2
3.5795MHz
MWRQST
WD
COL2
ROW1
S10
1 3
*
9
ROW2
D14
GREEN LED
12
S13
1 3
VCC
MWRDY
ROW4
VCC
R66
10K
RESET
SWITCH
R70
560
1
S11
1 3
D13
SD103
2 1
VCC
R67
10K
5
R65
10K
SPSDWN
4
ROW2
MWCS
R72
10K
ROW3
S4
1 3
3
COL3
ROW3
S3
1 3
0
D12
SD103
2 1
S14
1 3
S5
1 3
OFFHK\
6
7
S6
1 3
MWDIN
#
RING-
RES
E
VCC
COL2
S12
1 3
S8
1 3
MWCLK
S7
1 3
COL1
COL1
U19
MAX701
46
5
7
8
3
2
1
GND RESET
RESET
N.C
Vcc
N.C
N.C
MR
R60
10K
8
C50
39PF
ROW1
MWDOUT
COL3
ROW4
COP888EG1
U20
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
39
38
37
36
35
34
33
32
31
30
29
44
43
42
41
40
6
5
4
3
2
1
CKI
VCC
I0
I1
I2
I3
I4
I5
I6
I7
L0
L1
L2
L3
C4
C5
C6
C7
L4
L5
L6
L7
G0
RESET
GND
D7
D6
D5
D4
D3
D2
D1
D0
C1
C0
G3
G2
G1
G7
G6
G5
G4
C3
C2
D11
SD103
2 1
VCC
2
MICROCONTROLLER AND USER INTERFACE
R69
1M
1 3
C52
39PF
S9
1 3
R58
10K
Y2
3.5795MHz
MWRQST
WD
COL2
ROW1
S10
1 3
*
9
ROW2
D14
GREEN LED
12
S13
1 3
VCC
MWRDY
ROW4
VCC
R66
10K
RESET
SWITCH
R70
560
1
S11
1 3
D13
SD103
2 1
VCC
R67
10K
5
R65
10K
SPSDWN
4
ROW2
MWCS
R72
10K
ROW3
S4
1 3
3
COL3
ROW3
S3
1 3
0
D12
SD103
2 1
S14
1 3
S5
1 3
OFFHK\
6
7
S6
1 3
MWDIN
#
RING-
RES
E
VCC
COL2
S12
1 3
S8
1 3
MWCLK
S7
1 3
COL1
COL1
U19
MAX701
46
5
7
8
3
2
1
GND RESET
RESET
N.C
Vcc
N.C
N.C
MR
R60
10K
8
C50
39PF
ROW1
MWDOUT
COL3
ROW4
COP888EG1
U20
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
39
38
37
36
35
34
33
32
31
30
29
44
43
42
41
40
6
5
4
3
2
1
CKI
VCC
I0
I1
I2
I3
I4
I5
I6
I7
L0
L1
L2
L3
C4
C5
C6
C7
L4
L5
L6
L7
G0
RESET
GND
D7
D6
D5
D4
D3
D2
D1
D0
C1
C0
G3
G2
G1
G7
G6
G5
G4
C3
C2
D11
SD103
2 1
VCC
2
MICROCONTROLLER AND USER INTERFACE
R69
1M
1 3
C52
39PF
S9
1 3
R58
10K
Y2
3.5795MHz
MWRQST
WD
COL2
ROW1
S10
1 3
*
9
ROW2
D14
GREEN LED
12
S13
1 3
VCC
MWRDY
ROW4
VCC
R66
10K
RESET
SWITCH
R70
560
1
S11
1 3
D13
SD103
2 1
VCC
R67
10K
5
R65
10K
SPSDWN
4
ROW2
MWCS
R72
10K
ROW3
S4
1 3
3
COL3
ROW3
S3
1 3
0
D12
SD103
2 1
S14
1 3
3-6
3—SCHEMATIC DIAGRAMS
ISD-T360SB
Voice Solutions in Silicon™
4-1
4—PHYSICAL DIMENSIONS ISD-T360SB
ISD
Chapter 4—PHYSICAL DIMENSIONS
Figure 4-1: 100L PQF P (14x20x2.75mm foot pr int 3.2 mm) IQC
EH
AA2
Y
Seating Plane L
L1
See Detail
G
Controlling dimension : Millimeters
A1
E
D
H
D
eb
c
0.08
070
0.003
1.60
0.95
17.40
0.80
17.20
0.65
17.00
0.063
0.037
0.921
0.685
0.031
0.913
0.677
0.025
0.905
0.669
0.65
20.10
14.10
0.20
0.40
2.87
20.00
14.00
2.72
19.90
13.90
0.10
0.20
2.57
0.791
0.555
0.008
0.016
0.113
0.787
0.551
0.107
0.026
0.783
0.547
0.004
0.008
0.101
Symbol
Min Nom Max Max
Nom
Min
Dimension in inch Dimension in mm
A
b
c
D
e
HD
HE
L
Y
A1
A2
L1
E
0.012
0.006 0.15
0.30
23.00 23.20 23.40
7
0.020 0.032 0.50 0.80
0.350.25
0.010 0.014 0.018 0.45
G
GG
G
3.30
0.130
4-2
4—PHYSICAL DIMENSIONS
ISD-T360SB
Voice Solutions in Silicon™
4—PHYSICAL DIMENSIONS ISD-T360SB
4-3
ISD
The warranty for each product of ISD (Information Storage
Devices, Inc.), is contained in a written warranty which governs
sale and use of such product. Such warranty is contained in the
printed terms and conditions under which such product is sold, or
in a separate written warranty supplied with the product. Please
refer to such written warranty with respect to its applicability to
certain applications of such product.
These products may be subject to restrictions on use. Please
contact ISD, for a list of the current additional restrictions on
these products. By purchasing these products, the purchaser of
these products agrees to comply with such use restrictions. Please
contact ISD for clarification of any restrictions described herein.
ISD, reserves the right, without further notice, to change the ISD
VoiceDSP product specifications and/or information in this
document and to improve reliability, functions and design.
ISD assumes no responsibility or liability for any use of the ISD
VoiceDSP products. ISD conveys no license or title, either
expressed or implied, under any patent, copyright, or mask work
right to the ISD VoiceDSP products, and ISD makes no warranties
or representations that the ISD VoiceDSP products are free from
patent, copyright, or mask work right infringement, unless
otherwise specified.
Application examples and alternative uses of any integrated
circuit contained in this publication are for illustration purposes
only and ISD makes no representation or warranty that such
applications shall be suitable for the use specified.
The ISD products described herein may not be used in
recordable greeting cards intended for resale in Japan without
ISD’s prior written consent.
Recordable greeting cards are articles, usually constructed of
paper or cardboard (but not limited to these materials), that
include artwork(including pre-printed or customer-affixed
photographs or artwork) and/or editorials that have the look and
feel of greeting cards and are able to be inserted into an envelope
(but do not necessarily need to be inserted) and primarily
intended to be presented or delivered to an individual as a
greeting card, that allow the sender to record and the recipient to
play back, and audio signal message. The foregoing restriction
applies only to recordable greeting cards intended, directly or
indirectly, for retail distribution. Recordable greeting cards
incorporating ISD products may not be sold, directly or indirectly,
through retail channels without ISD’s prior written consent.
The 100-year retention and 100K record cycle projections are
based upon accelerated reliability tests, as published in the ISD
Reliability Report, and are neither warranted nor guaranteed by
ISD.
Information contained in this ISD VoiceDSP data sheet
supersedes all data for the ISD VoiceDSP products published by
ISD prior to August, 1999.
This data sheet and any future addendum to this data sheet is
(are) the complete and controlling ISD VoiceDSP product
specifications. In the event any inconsistencies exist between the
information in this and other product documentation, or in the
event that other product documentation contains information in
addition to the information in this, the information contained
herein supersedes and governs such other information in its entirety.
Copyright© 1999, ISD (Information Storage Devices, Inc.) All rights
reserved. ISD is a registered trademark of ISD. ChipCorder is a
trademark of ISD. VoiceDSP is a trademark of ISD. All other
trademarks are properties of their respective owners.
IMPORTANT NOTICES
2727 North First Street
San Jose, California 95134
Tel: 408/943-6666
Fax: 408/544-1787 Part No. 052000DS360SB
