TDC-GP2 T&R 2K - AMS AG

Target Speciﬁcation

Time-to-Digital Converter

TDC-GP2

26th May 2010

Document-No.: DB_GP2_e V2.0

2-channel Universal

Time-to-Digital Converter

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

TDC

TDC-GP2

Disclaimer / Notes

The information provided by this data sheet is believed to be accurate and reliable. However, no

responsibility is assumed by acam for its use, nor for any infringements of patents or other rights

of third parties that may result from its use. The information is subject to change without notice

and is provided „as is“ without warranty of any kind (expressed or implied). Picostrain is a registered

trademark of acam. All other brand and product names in this document are trademarks or service

marks of their respective owners.

Support

For a complete listing of direct sales, distributors and sales representatives visit the acam website

at:

http://www.acam.de/company/distributors

For technical support you can contact the acam support team in the headquarter in Germany or the

Distributor in your country. The contact details of acam in Germany are:

sales@acam.de or by phone +49-7244-74190.

Published by acam-messelectronic gmbh

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 2

Table of Contents

TDC-GP2

1 Overview 1-4

2.4 Package Drawings

2 Characterisitcs &

2-7

2.3 Pin Description 2-6

2.2 Timings 2-3

2.1 Electrical Characteristics 2-1

2.5 Power Supply 2-8

2.6 Register settings 2-10

3 Measurement Mode 1 3.2 Measurement Flow 3-2

3.1 General Description 3-1

4 Measurement Mode 2

4.3 Stop Masking 4-7

4.2 Measurement Flow 4-2

4.1 General Descriptionweis 4-1

5.4 SPI-interface

5 Details and Special Functions 5-10

5.3 Temperature Measurement 5-7

5.2 Fire-pulse Generator 5-4

5.1 Oscillator 5-1

5.5 Fast Initialization 5-11

5.6 Noise Unit 5-11

6 Applications 6-1

7 Miscellaneous 7-1

Specifications

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

TDC-GP2

Time-to-Digital-Converter

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 1-4

TDC-GP2

1 Overview

1.1 Introduction

TDC-GP2 is the next generation of acam generalpurpose TDCs. Higher resolution and smaller pa-

ckage size make it ideal for cost sensitive industrial applications. With special functional blocks like a

ﬁre-pulse generator, stop-enable, temperature measurement, and clock control it is perfectly suited

for ultrasonic ﬂow-meter and heat-meter applications.

1.2 Features

Measurement Mode 1

2 channels with typ. 50 ps resolution rms

Measurement range 3.5 ns to 1.8 µs (0 to

1.8µs between stop channels)

15 ns pulse-pair resolution with 4-fold multihit

capability

4 events can be measured arbitrarily against

each other

Trigger to rising or/and falling edge

Windowing for precise stop enable

Measurement Mode 2

1 channel with typ. 50 ps resolution rms

Measurement range 500 ns to 4 ms

2 x CLKHS pulse-pair resolution with 3-fold

multihit capability

Trigger to rising or/and falling edge

Each of the 3 events can be assigned to an

adjustable measuring window with 10ns reso-

lution

Temperature Measurement

2 or 4 sensors

PT500/PT1000 or higher

Very high resolution: 16 Bit eff.

(0.004 °C resolution for platinum sensors)

Ultra low current (0.08 µA when measuring

every 30 seconds)

General

QFN 32 package

I/O voltage 1.8 V to 5.5 V

Core voltage 1.8 V to 3.6 V

1 MHz continuous data rate max.

Temperature range – 40 °C to 125 °C

4 wire SPI interface

Fire pulse generator

Clock calibration unit

Precise stop enable by windowing

Trigger to rising and/or falling edge

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

1-5

TDC-GP2

Time-to-Digital-Converter

1.3 Blockdiagram

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 2-1

TDC-GP2

2 Characterisitcs & Specifications

2.1 Electrical Characteristics

Absolute Maximum Ratings

Supply voltage

Vcc vs. GND - 0.3 to 4.0 V

Vio vs. GND - 0.3 to 7.0 V

Continous current into Output-Pin (Iout) 30 mA

Storage temperature (Tstg) - 65 to 150 °C

Junction temperature (Tj) max.125 °C

Recommended Operating Conditions

Symbol Parameter Conditions Min Typ Max Unit

Vcc Core supply voltage* Vio > Vcc 1.8 3.6 V

Vio I/O supply voltage 1.8 5.5 V

tri Normal Input Rising Time 50 ns

tfa Normal Input Falling Time 50 ns

tri Schmitt Trigger Rising Time 5 ms

tfa Schmitt Trigger Falling Time 5 ms

Ta Ambient Temperature Tj must not exceed

125°C

-40 120 °C

*including the oscillator pins XIN, XOUT, Clk32In, Clk32Out

DC Characteristics (Vio = Vcc = 3.3 V ± 0.3 V, Tj = -40 to +85°C)

Symbol Parameter Conditions Min Typ Max Unit

I32 Current 32 kHz Icc + Iio, only 32kHz os-

cillator running,

Vcc = 3.6 V

4.5 µA

Ihs Current 4 Mhz Icc + Iio, only ClkHS run-

ning cont. at 4MHz,

Vcc = 3.6 V

260 µA

Itmu Current time measuring

unit

only during active time

measurement

15 mA

Iddq Quiescent current all clocks off, Vio = Vcc =

3.6 V @ 85 °C

<150 nA

Il Input Leakage Current -1 +1 µA

Voh High Level Output Voltage Ioh= tbd mA Vio=Min. Vio-

0.4

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

2-2

TDC-GP2

Time-to-Digital-Converter

Symbol Parameter Conditions Min Typ Max Unit

Vol Low Level Output Voltage Iol = tbd mA, Vio=Min 0.4 V

Vih High Level Input Voltage LVTTL Level, Vio = Max. 2.0 V

Vil Low Level Input Voltage LVTTL Level, Vio = Min. 0.8 V

Vth High Level Schmitt Trigger

Voltage

1.1 2.4 V

Vtl Low Level Schmitt Trigger

Voltage

0.6 1.8 V

Vh Schmitt Trigger Hysteresis 0.1 V

Terminal Capacitance

Symbol Terminal Condition Rated Value Unit

Min. Typ. Max.

Ci Input measured @ Vcc = Vio,

f = 1 MHz,

Ta = 25°C

10 pF

Co Output 10

Cio Bidirectional 10

Time Measuring Unit

Symbol Terminal Condition Rated Value Unit

Min. Typ. Max.

LSB Vio = Vcc = 3.3 V -40 °C

3.6 V

25 °C

3.3 V

85 °C

3.0 V

35 63 111

Vio = Vcc = 2.5 V -40 °C

2.75 V

25 °C

2.5 V

85 °C

2.25 V

38 76 156

Standard Deviation Vio=3.3 V, Vcc =3.3 V 2.75 V 2.5 V

Ta = 25°C - 50 -

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 2-3

TDC-GP2

Temperature Measuring Unit

Symbol Terminal Condition Rated Value Unit

Min. Typ. Max.

Resolution RMS Vio = Vcc =3.3 V

PT1000

150 nF Capacitance

16.0 Bit

SNR 96 dB

Absolute Gain-Error 0,1 %

Gain-Drift vs. Vio 0,08 %/V

Gain-Drift vs. Temp 0,0008 %/K

Uncalibrated Offset <0.01 %

Offset Drift vs.

Temp

<0,2 ppm/K

PSRR >100 dB

2.2 Timings

At Vcc = 3.3 V ± 0.3 V, ambient temperature -40 °C to +85 °C unless otherwise speciﬁed

Oscillator

Symbol Parameter Min. Typ. Max. Unit

Clk32 32 kHz reference oscillator 32,768 kHz

toszst Oscillator start-up time with ceramic resonator 200 µs

toszst Oscillator start-up time with crystal oscillator 5 ms

ClkHS High-speed reference oscillator 2 8 MHz

Serial Interface

Symbol Parameter Max. @ Vio = Unit

2.0 V 2.5 V 3.3 V

fclk Serial clock frequency 10 20 25 MHz

Symbol Parameter Min. @ Vio = Unit

2.0 V 2.5 V 3.3 V

tpwh Serial clock, pulse width high 50 25 20 ns

tpwl Serial clock, pulse width low ns

tsussn SSN enable to valid latch clock 20 40 10 ns

tpwssn SSN pulse width between write cycles 50 30 20 ns

thssn SSN hold time after SCLK falling 70 40 25 ns

tsud Data set-up time prior to SCLK falling 10 5 5 ns

thd Data hold time before SCLK falling 10 5 5 ns

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

2-4

TDC-GP2

Time-to-Digital-Converter

Symbol Parameter Max. @ Vio = Unit

1.8 V 2.5 V 3.3 V

tvd Data valid after SCLK rising 30 20 16 ns

Serial Interface (SPI compatible, Clock Phase Bit =1, Clock Polarity Bit =0):

Figure 2-1: SPI Write

Figure 2-2: SPI Read

8-Bit Opcodes:

MSB LSB Description

1 0 0 0 0 ADR2 ADR1 ADR0 Write into address ADR

1 0 1 1 0 ADR2 ADR1 ADR0 Read from address ADR

01110000Init

0 1 0 1 0 0 0 0 Power On Reset

0 0 0 0 0 0 0 1 Start_Cycle

0 0 0 0 0 0 1 0 Start_Temp

0 0 0 0 0 0 1 1 Start_Cal_Resonator

0 0 0 0 0 1 0 0 Start_Cal_TDC

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 2-5

TDC-GP2

Disable Timings

Figure 2-3:

Spec Description Min (ns) Max (ns)

tS-EN Enable Setup Time 5 ns -

tSH-EN Enable Hold Time 5 ns -

Reset Timings

Figure 2-4:

Spec Description Min (ns) Max (ns)

tph Reset pulse width 50 ns -

trfs Time after rising edge of reset

pulse before hits are accepted

200 ns -

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

2-6

TDC-GP2

Time-to-Digital-Converter

2.3 Pin Description

No. Name Description Buffer type Value If not

used

1 Xin Oscillator driver in GND

2 Xout Oscillator driver out

3 Vio I/O – supply voltage

4 GND Ground

5 Fire1 Fire pulse generator output 1 48 mA

6 Fire2 Fire pulse generator output 2 48 mA

7 Fire_In Signal input for quasi “Sing Around” GND

8 INTN Interrupt ﬂag 12 mA LOW active

9 SSN Slave select Schmitt trigger LOW active

10 SCK Clock serial interface Schmitt trigger

11 SI Data input serial interface Schmitt trigger

12 SO Data output serial interface 12 mA tristate

13 RSTN Reset input Schmitt trigger LOW active

14 Vcc Core supply voltage

15 Clk32Out Output 32 kHz clock generator n. c.

16 Clk32In Input 32 kHz clock generator GND

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 2-7

TDC-GP2

No. Name Description Buffer type Value If not

used

17 SenseT Sense input temperature measure-

ment

Schmitt trigger GND

18 LoadT Load output temperature measure-

ment

24 mA n.c.

19 PT4 Port 4 temperature measurement 48 mA

20 PT3 Port 3 temperature measurement 48 mA

21 GND Ground

22 Vio I/O – supply voltage

23 PT2 Port 2 temperature measurement 48 mA

24 PT1 Port 1 temperature measurement 48 mA

25 En_Stop2 Enable pin stop input 2 Schmitt trigger HIGH active Vio

26 En_Stop1 Enable pin stop input 1 Schmitt trigger HIGH active Vio

27 Stop2 Stop input 2 GND

28 GND Ground

29 Vcc Core supply voltage

30 Stop1 Stop input 1 GND

31 Start Start input

32 En_Start Enable pin start input Schmitt trigger HIGH active Vio

2.4 Package Drawings

Suitable socket: Plastronics 32QN50S15050D

Symbol Dimension in Millimeters

Center pad shall not be connected to GND.

Min. Nom. Max.

D - 5 -

E - 5 -

A - - 1

A1 0 - -

b 0.17 - 0.3

e - 0.5 -

L 0.3 - 0.5

G 3.24

Thermal resistance

Roughly 40 K/W at 0 m/s air ﬂow, 37 K/W at 1 m/s air ﬂow, 35 K/W at 2m/s air ﬂow (values

just for reference).

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

2-8

TDC-GP2

Time-to-Digital-Converter

Soldering Temperature Profile

The temperature proﬁle for infrared reﬂow furnace (in which the temperature is the resin’s surface

temperature) should be maintained within the range described below.

Maximum temperature

The maximum temperature requirement for the resin surface, given 260ºC as the peak tempera-

ture of the package body’s surface, is that the resin surface temperature must not exceed 250ºC

for more than 10 seconds. This temperature should be kept as low as possible to reduce the load

caused by thermal stress on the package, which is why soldering for short periods only is recommen-

ded. In addition to using a suitable temperature proﬁle, we also recommend that you check carefully

to conﬁrm good soldering results.

2.5 Power Supply

Supply voltage

Although the TDC-GP2 is a fully digital circuit, some analog measures affect the circuit. The reason

is that the TDC is based on the internal analog measure ‚propagation delay time‘ which is inﬂuenced

by temperature and supply voltage. A good layout of the supply voltage is essential for good measure-

ment results. It should be high capacitive and of low in-ductance.

The TDC-GP2 provides two pairs of power supply terminals:

Vio - I/O supply voltage

Vcc - Core supply voltage

All ground pins should be connected to a ground plane on the printed circuit board. Vio and Vcc

should be provided by a battery or ﬁxed linear voltage regulator. Do not use switched regulators to

avoid disturbances caused by the I/O supply.

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 2-9

TDC-GP2

The measurement quality of a time-to-digital converter depends on a good power supply. The chip

sees mainly pulsed current and therefore a sufﬁcient bypassing is mandatory:

Vcc 100 µF (minmum 47 µF)

Vio 10 µF (minimum 1 µF)

The supply voltage should be provided through analog regulators. We strongly recommend not to use

switch mode power supplies.

Current consumption

The current consumption is the sum from different parties (all data for Vio = Vcc = 3.6V):

Iddq < 150 nA Quiescent current,

I32 typ. 4.5 µA Current into the 32 kHz oscillator, turned on only if the 32 kHz

oscillator is connected

Ihs typ. 260 µA/s Current into the high speed oscillator,

* (active runtime) Example: In ultrasonic ﬂow-meters the high-speed oscillator is

on for about 2ms only. The average current consumption is

260 µA/s * 2 ms = 0.52 µA

Itmu typ. 15 mA/s Current into the time measuring unit, In measurement range 1

* (active measuring time) the time measuring unit is active for the start-stop time inter

val plus the calibration time interval of 2 periods of the refe

rence clock per measurement.

In measurement range 2 the time measuring unit is on for

average 4 periods of the reference clock per measurement,

two for the time measurement and two for calibration.

Example: With 10 measurements per second in measurement

range 2 and a 4 MHz reference clock the time measuring unit

is active for only about 10 µs.

The average current is 15 mA/s * 10 µs = 0.150 µA.

Ialu typ. 7 nA per calculation Current into ALU during data proccesing including

a calibration calculation.

Example: At 1000 measurements per second with 3 stops per

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

2-10

TDC-GP2

Time-to-Digital-Converter

start the ALU average current is 7 nA * 3000 = 21 µA.

IT typ. 2.5 µAs * measure rate The current for a full temperature measurement is typ.2.5µAs.

In heat-meters the temperature is measured typically once

every 30 seconds. The average current is about 0.085 µA

2.6 Register settings

Service bits are for acam testing and security purposes only, Please use the recommended values.

2.6.1 Write registers

Default values in second rows

Bit Reg0 (*) Reg1 (*) Reg2 (*) Reg3 (*) Reg4 (*) Reg5 (*)

23 FIRE# 0 HIT2 0 EN_INT 0 s.c. 0 s.c. 0 CONF_FIRE 0

22 0 1 0 s.c. 0 s.c. 0 0

21 0 0 1 EN_ERR_VAL 0 s.c. 1 0

20 0 1 RFEDGE2 0 SEL_TIMO_

MR2

1 s.c. 0 EN_STARTNOISE 0

19 DIV_FIRE 0 HIT1 0 RFEDGE1 0 1 s.c. 0 DIS_PhaseNoise 0

18 0 1 DELVAL1 0 DELVAL2 0 DELVAL3 0 REPEAT_FIRE 0

17 0 0 0 0 0 0

16 0 1 0 0 0 0

15 CALRES# 0 EN_FAST_INIT 0 0 0 0 PHASE_FIRE 0

14 0 s.c. 1 0 0 0 0

13 ClkHSDiv 0 HITIN2 0 0 0 0 0

12 0 0 0 0 0 0

11 START_ClkHS 0 0 0 0 0 0

10 1 HITIN1 0 0 0 0 0

9 PORT# 1 0 0 0 0 0

8 TCycle 0 0 0 0 0 0

7 No_FAKE 0 n.c 0 0 0 0 0

6 SelClkT 1 n.c 0 0 0 0 0

5 Calibrate 1 n.c 0 0 0 0 0

4 DisAutoCal 0 n.c 0 0 0 0 0

3 MRange2 1 n.c 0 0 0 0 0

2 NEG_STOP2 0 n.c 0 0 0 0 0

1 NEG_STOP1 0 n.c 0 0 0 0 0

0 NEG_START 0 n.c 0 0 0 0 0

s.c. = Special acam conﬁguration bits, n.c. = not in use

(*) = Default value after Power On Reset

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 2-11

TDC-GP2

Short description of the bits:

Bits Name Descritpion Value

Reg 0

0 NEG_START Negation start input 0 = non-inverted input signal – rising edge

1 = inverted input signal – falling edge

1 NEG_STOP1 Negation stop 1 input 0 = non-inverted input signal – rising edge

1 = inverted input signal – falling edge

2 NEG_STOP2 Negation stop 2 input 0 = non-inverted input signal – rising edge

1 = inverted input signal – falling edge

3 MRange2 Switch to measurement range 2 0 = measurement range 1

1 = measurement range 2

4 DisAutoCal Enables/disables auto-calibration

run in the TDC

0 = auto-calibration after measurement

1 = auto-calibration disabled

5 Calibrate Enables/disables calibration

calculation in the ALU

0 = calibration off (only MR 1)

1 = calibration on

6 SelClkT Select reference signal for internal

cycle clock for tem-perature

measurement

0 = use 32,768 kHz as cycle clock

1 = use 128 * CLKHS as period for cycle clock

(32µs with 4 MHZ high speed clock signal )

7 FAKE# Number of dummy cycles at the

beginning of a tempera-ture

measurement

0 = 2 Fake measurements

1 = 7 Fake measurements

8 TCycle Sets cycle time for tempera-ture

measurement

0 = 128 µs cycle time @ 4 MHz

1 = 512 µs cycle time @ 4 MHz (recommended)

9 PORT# Sets number of ports used for

temperature measure-ment

0 = 2 temperature ports (PT1 and PT2)

1 = 4 temperature ports

10-11 START_CLKHS Switch on high-speed oscilla-tor 0 = Oscillator off

1 = Oscillator on

2 = settling time = 640 µs

3 = settling time = 1280 µs (see Bug Report)

12-13 ClkHSDiv Sets predivider for CLKHS 0 = divided by 1

1 = divided by 2

2 = divided by 4

3 = divided by 4

14-15 CALRES# Sets number of periods used for

calibrating the ceramic resonator

0 = 2 periods = 61.035 µs

1 = 4 periods = 122.07 µs

2 = 8 periods = 244.14 µs

3 = 16 periods = 488.281 µs

16-19 DIV_FIRE Sets predivider for internal clock

signal of ﬁre pulse generator

0 = not permitted1 = divided by 2

2 = divided by 3

3.= divided by 4

…

15 = divided by 16

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

2-12

TDC-GP2

Time-to-Digital-Converter

20-23 FIRE# Sets number of pulses gen-erated

by ﬁre pulse generator

0 = off

1 = 1 pulse

2 = 2 pulses

3 = 3 pulses

…

15 = 15 pulses

Reg 1

8-10 HITIN1 Number of expected hits on

channel 1

0 = stop channel 1 disabled

1 = 1 hit

2 = 2 hits

3 = 3 hits

4 = 4 hits

5 to 7 = not permitted

11-13 HITIN2 Number of expected hits on

channel 2

0 = stop channel 2 disabled

1 = 1 hit

2 = 2 hits

3 = 3 hits

4 = 4 hits

5 to 7 = not permitted

15 EN_FAST_INIT Enables fast init operation 0 = Fast init mode disabled

1 = Fast init mode enabled

16-19 HIT1 Deﬁnes operator for ALU data

post-processing

MRange1: HIT1-HIT2

MRange2: HIT2-Start

MRange1:

0 = Start

1 = 1. Stop Ch1

2 = 2. Stop Ch1

3 = 3. Stop Ch1

4 = 4. Stop Ch1

5 = no action

6 = Cal1 Ch1

7 = Cal2 Ch1

9 = 1. Stop Ch2

A = 2. Stop Ch2

B = 3. Stop Ch2

C = 4. Stop Ch2

MRange2:

1 = Start

20-23 HIT2 Deﬁnes operator for ALU data

post-processing

MRange1: HIT1-HIT2

MRange2: HIT2-Start

MRange1:

0 = Start

1 = 1. Stop Ch1

2 = 2. Stop Ch1

3 = 3. Stop Ch1

4 = 4. Stop Ch1

5 = no action

6 = Cal1 Ch1

7 = Cal2 Ch1

9 = 1. Stop Ch2

A = 2. Stop Ch2

B = 3. Stop Ch2

C = 4. Stop Ch2

MRange2:

2 = 1. Stop Ch1

3 = 2. Stop Ch1

4 = 3. Stop Ch1

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 2-13

TDC-GP2

Reg 2

0-18 DELVAL1 Delay value for internal stop ena-

ble unit, hit 1 channel 1. Fixed

point number with 14 integer

and 5 fractional digits in multip-

les of Tref

DELVAL1 = 0 to 16383.96875

19 RFEDGE1 Edge sensitivity channel 1 0 = rising or falling edge

1 = rising and falling edge

20 RFEDGE2 Edge sensitivity channel 1 0 = rising or falling edge

1 = rising and falling edge

21-23 EN_INT Activates interrupt sources wired

by OR

Bit 23 = Timeout interrupt enable

Bit 22 = End Hits interrupt enable

Bit 21 = ALU interrupt enable

Reg 3

0-18 DELVAL2 Delay value for internal stop

enable unit, hit 2 channel 1.

Fixed point number with 14

integer and 5 fractional digits

in multiples of Tref

DELVAL1 = 0 to 16383.96875

19-20 SEL_TIMO_MR2 Select predivider for timeout in

measurement range 2

0 = 64 µs

1 = 256 µs

2 = 1024 µs

3 = 4096 µs

@ 4 MHz ClkHS

21 EN_ERR_VAL Timeout forces ALU to write

0xFFFFFFFF into the output

0 = disabled

1 = enabled

Reg 4

0-18 DELVAL1 Delay value for internal stop ena-

ble unit, hit 3 channel 1. Fixed

point number with 14 integer

and 5 fractional digits in multip-

les of Tref

DELVAL1 = 0 to 16383.96875

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

2-14

TDC-GP2

Time-to-Digital-Converter

Reg 5

0-15 PHASE_FIRE Enables phase reversing for each

pulse of a sequence of up to 15

possible pulses

0 = no inversion

1 = inversion

16-18 REPEAT_FIRE Number of pulse sequnce repeti-

tion for "quasi-sing-arround"

0 = no signal repetition

1 = 1 signal repetition

2 = 2 signal repetition

...

7 = 7 signal repetition

19 DIS_PHASENOISE Phase noise unit, has to be disa-

bled, See bug report section 7.2

1 = disable phase shift

Setting 1 is mandatory

20 EN_STARTNOISE Enables additional noise for start

channel

1 = switch on noise unit

21-23 CONF_FIRE Output conﬁguration for pulse

generator

Bit 23 = 1: negate output Fire2

Bit 22 = 1: disable output Fire2

Bit 21 = 1: disable output Fire1

2.6.2 Read registers / Output data format

ADR Symbol Bits Description

0 RES_0 32 Measurement result 1, ﬁxed-point number with 16 integer and 16 fractional digits

215 20,2-1 2-16

1 RES_1 32 Measurement result 2, ﬁxed-point number with 16 integer and 16 fractional digits

2 RES_2 32 Measurement result 3, ﬁxed-point number with 16 integer and 16 fractional digits

3 RES_3 32 Measurement result 4, ﬁxed-point number with 16 integer and 16 fractional digits

4 STAT 16 15 - 13 12 11 10 9 8 - 6 5 - 3 2 - 0

n.c. Error

short

Error

open

Timeout

Precounter

Timeout

TDC

# of hits

Ch2

# of hits

Ch1

Pointer

result

5 REG_1 8 Content of highest 8 Bits of write register 1, to be used for testing the communication

The data structure and the occupancy of the result registers depends on the operation mode and

whether calibrated or non-calibrated data are stored. Several cases must be distinguished:

Only in measurement range 1 negative results are possible.

In measurement range 2 only positive results are possible, given as unsigned numbers.

A non-calibrated measurement is only possible in measurement range 1.

In measurement range 1 with calibrated data (ALU) the time intervals that have to be measured

can not exceed twice the period of the calibration clock. When measuring bigger time intervals an

ALU - overﬂow will occur and 0xFFFFFFFF is written in the appropriate result register.

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 2-15

TDC-GP2

a. Measurement range 1 with calibrated data (Calibrate = 1)

The results are given in multiples of the internal reference clock (= external reference

clock divided by 1, 2 or 4 (DIV_CLKHS)). Calibrated data are 32 bit ﬁxed point numbers

with 16 integer bits and 16 fractional bits. Any calibrated result covers therefore 1 re-

sult register. The serial output begins with the highest bit (215) and ends with the lowest

one (2-16). The numbers are available in complements of 2.

Time = RES_X * Tref * 2ClkHSDiv = RES_X * Tref * N , with N = 1, 2 or 4

Time < 2 * Tref * 2ClkHSDiv

b. Measurement range 1 without calibration (Calibrate = 0)

Non-calibrated data are of the type ‚Signed Integer‘ and are stored as a 16 bit value in

the high word of the result registers. The bits of the low word are set to zero. The result

is represented as number of LSB and is available in complements of 2.

Time = RES_X * LSB ~ RES_X * 65 ps

c. Measurement range 2

In measurement range 2 the TDC-GP2 only supports calibrated measurement. The

results are given in multiples of the internal reference clock (= external reference clock

divided by 1, 2 or 4 (DIV_CLKHS)). Calibrated data are 32 bit ﬁxed point numbers with

16 integer bits and 16 fractional bits. Any calibrated result covers therefore 1 result

(2-16). The numbers are available in complements of 2.

Time = RES_X * Tref * 2ClkHSDiv = RES_X * Tref * N , with N = 1, 2 or 4

d. Temperature measurement

Discharge tiem in the same format as in c. measurement mode 2.

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

2-16

TDC-GP2

Time-to-Digital-Converter

2.6.3 Status register

Bits Name Description Values

2 – 0 Pointer result register Pointer to the next free result register

5 – 3 # of hits Ch 1 Number of hits registered on channel 1

8 – 6 # of hits Ch 2 Number of hits registered on channel 2

9 Timeout TDC Indicates an overﬂow of the TDC unit 1 = overﬂow

10 Timeout Precounter Indicates an overﬂow of the 14 bit precounter in MR 2 1 = overﬂow

11 Error open Indicates an open sensor at temperature measurement 1 = open

12 Error short Indicates a shorted sensor at temperature measurement 1 = short

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 3-1

TDC-GP2

3 Measurement Mode 1

3.1 General Description

2 stop channels referring to one start channel

Each of typ. 50 ps RMS resolution

LSB width typ. 65 ps

15 ns pulse pair resolution

4-fold multihit capability for each stop channel

Measurement range 3.5 to 1.8 µs (0 to 1.8µs between stop channels)

Selectable rising/falling edge sensitivity for each channel

Enable pins for powerful windowing functionality

The possibility to arbitrarily measure all events against each other

Digital TDCs use internal propagation delays of signals through gates to measure time intervals with

very high precision. Figure 5 clariﬁes the principal structure of such an absolute-time TDC. Intelligent

circuit structures, redundant circuitry and special methods of layout on the chip make it possible to

reconstruct the exact number of gates passed by the signal. The maximum possible resolution stron-

gly depends on the maximum possible gate propagation delay on the chip.

Figure 3-1

The measuring unit is actuated by a START signal and stopped by a STOP signal. Based on the positi-

on of the ring oscillator and the coarse counter the time interval between START and STOP is calcula-

ted with a 20 Bit measurement range.

The BIN size (LSB) is typically 65 ps at 3.3 V and 25 °C ambient temperature. The RMS noise is

about 50 ps (0.7 LSB). The gate propagation delay times strongly depend on temperature and vol-

tage. Usually this is solved doing a calibration. During such a calibration the TDC measures 1 and 2

periods of the reference clock.

The measurement range is limited by size of the counter:

tyy = BIN x 26224 ~ 1.8 µs

Figure 3-2

3.2.1 Configuration

At the beginning the TDC-GP2 has to be

conﬁgured. The main settings for measu-

rement range 1 are:

a. Select measurement range1

setting register 0, Bit3, MRange2 = 0.

b. Select the reference clock

(see also section 5.1)

deﬁnes the switch-on behavior of the

high-speed clock. If only the 32kHz is used

this is be set to „0“. If only the high-speed

clock is used this is be set to „1“(conti-

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

3-2

TDC-GP2

Time-to-Digital-Converter

Input circuitry

Each input separately can be set to be sensitive to rising or falling edge or both edges. This is done

in register 0, Bits 0 to 2. (NEG_START, NEG_STOP1, NEG_STOP2) and register 2, Bit 19&20,

RFEDGEx.

Furthermore all Start/Stop-inputs support a high active enable pin.

3.2 Measurement Flow

Figure 3-3

Time (Condition) Description

tph 2,5 ns (min.) Minimum pulse width

tpl 2,5 ns (min.) Minimum pulse width

tss 3.5 ns ns (min)

1.8 µs (max.)

Start to Stop

trr 15 ns (typ.) Rising edge to rising edge

tff 15 ns (typ.)

tva 560 ns uncalibrated

4.6 µs calibrated

Last hit to data valid

txx No timing limits

tyy 1,8 µs (max) Max. measuring range

nuously on). In case both oscillators are used for current saving reasons this should be set to „2“ for

ceramic oscillators and to „3“ for quartz oscillators“.

4). This is important for calibrated measurements in measurement range 1 because the ALU works

correctly only if 2*Tref(intern) is bigger than the maximum time interval to be measured. Otherwise

the ALU output is 0xFFFFFFFF.

Make also sure that 2*Tref(intern) < 1.8 µs to avoid a timeout during calibration.

c. Set the number of expected hits

In register 1, Bits 8 to 10 and 11 to 13, HITIN1 and HITIN2 the user has to deﬁne the number of

hits the TDC-GP2 has to wait for. A maximum of 4 on each channel is possible. The TDC-GP2 measu-

res until the set number of hits is registered or a timeout occurs.

d. Select calibration

As the BIN size varies with temperature and voltage the TDC-GP2 ALU can internally calibrate the

results. This option is switched on by setting register 0, Bit5, Calibrate = „1“. It is recommended to

do this.

For the calibration the TDC measures 1 and 2 cycles of the reference clock. The two data are stored

as Cal1 and Cal2.

There are two ways to update the calibration data Cal1 and Cal2:

- Separate calibration by sending opcode Start_Cal_TDC via the SPI interface

- Automatic update by setting register 0, Bit 4, DisAutoCal = „0“. In most applications this will be the

preferred setting.

e. Define ALU data processing

While the TDC unit can measure up to 4 hits on each channel the user is free in his deﬁnition what

the ALU shall calculate. The settings are done in register 1, Bits 16 to 19 and 20 to 23, HIT1 and

HIT2. Both parameters can be set to:

0 = Start

1 = 1. Stop Ch1 9 = 1. Stop Ch2

2 = 2. Stop Ch1 A = 2. Stop Ch2

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 3-3

TDC-GP2

3 = 3. Stop Ch1 B = 3. Stop Ch2

4 = 4. Stop Ch1 C = 4. Stop Ch2

6 = Cal1 Ch1

7 = Cal2 Ch1

The ALU calculates Hit1 - Hit2.

In case calibration is active the ALU does the full calibration calculation (except when reading the cali-

bration values. In this case the ALU writes the Cal1/Cal2 raw data to the output register).

N = 1, 2 or 4.

Figure 3-4

Examples:

Reg1 = 0x01xxxx - 1st Stop Ch1-Start

Reg1 = 0x2Bxxxx - 3rd Stop Ch2-2nd

Stop Ch1

Reg1 = 0x06xxxx - Cal1

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

3-4

TDC-GP2

Time-to-Digital-Converter

f. Select input sensitivity

In register 2, Bits 19 & 20, RFEDGE1 and RFEDGE2, the user can select whether the stop in-

puts are sensitive to either rising or falling edges (RFEDGE = „0“) or to both rising and falling edges

(RFEDGE = „1“).

In register 0, Bits 0 to 2 the user can add an internal inverter to each input, Start, Stop1 and

Stop2. With RFEDGE = „0“ this is the same as rising edge (NEG_X = „0“) or falling edge

(NEG_X = „1“).

g. Interrupt behavior

The interrupt pin 8, INT can have different sources. They are selected in register 2, Bits 21 to 23,

EN_INT.

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 3-5

TDC-GP2

EN_INT = 0 no Interrupt source

1 ALU ready

2 The set number of hits is there

4 Timeout of the TDC unit

The different options are wired by OR to enable more than one source.

There are further conﬁguration options that will be described later in this chapter.

After the conﬁguration the user has to initialize the TDC-GP2 by sending opcode „Init“ so that the TDC

accepts Start and Stop hits.

3.2.2 Measurement

After an initialization the TDC unit will start with the ﬁrst pulse on the Start input. It will run until:

the set number of hits has been seen (maximum 4 on both stop channels in MR1)

or until a timeout occurs at the end of the measurement range (at about1.8µs in ,MR1).

The time measurement raw data are internally stored. The number of hits can be seen from the

status register, bits 3 to 8. In case calibration is active the TDC now measures one and two periods

of the internal reference clock (Tref * 1,2 or 4). The calibration raw data Cal1 and Cal2 are also

internally stored.

Figure 3-5

3.2.3 Data Processing

At the end of the measurement the ALU starts to process the data according to the HIT1, HIT2 set-

tings and transfers the result to the output register. In case calibration is off the ALU transfers the

16 Bit raw data to the output register. With calibration the ALU calculates according to 3.1.1.d and

transfers the 32Bit ﬁxed point number to the output register.

The ALU can be switched off conﬁguring HIT1=HIT2=5.

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

3-6

TDC-GP2

Time-to-Digital-Converter

The time it takes the ALU depends on whether calibration is on or not and the supply voltage.

Table 3-1: ALU timings

un-calibrated calibrated

3.3 V 220 ns 2.3 µs

2.5 V 310 ns 2.5 µs

2.0 V 580 ns 2.8 µs

As soon as the data is available from the output

ALU interrupt is enabled, see reg 2, EN_INT). Further

the load pointer of the output register is increased

by 1 and points to the next free memory. The actual

position of the load pointer can be seen in the status

3.2.4 Reading Data

Now the user can read the data sending the opcode 10110ADR. With the next 16 cycles (un-calib-

rated data) or 32 cycles (calibrated data) the TDC-GP2 will send the result, beginning with the most

signiﬁcant Bit (MSB).

a. Un-calibrated data format:

16 Bit Signed integer in complements of 2. 1BIN = uncalibrated gate delay is about 65ps at 5V and

25°C.

Time = RES_X x 65ps

b. Calibrated data format:

32 Bit ﬁxed-point number in complements of 2.

Given in multiples of the reference clock.

Time = RES_X * Tref * N, N = 1, 2 or 4

The measured time interval may not exceed ,

otherwise the ALU will go into overﬂow and will

write the data 0xFFFFFFFF to the output regis-

ter.

The conﬁguration of the ALU allows only one

hit calculation at the time. In case more than

one hit has been measured it is necessary to

write new commands to HIT1/HIT2 to instruct

the ALU for calculating the other hits. After

writing to HIT1/HIT2 it is necessary to wait for

minimum 4.6µs (calibrated data) or 580ns (un-

calibrated data) before reading or writing again

to HIT1/HIT2.

Example:

conﬁguration

...

write reg1=0x104400 4 hits on channel 1,

calculate Hit1-Start

...

Initialize

...

while(Check interrupt ﬂag)

write reg1=0x204400 calculate Hit2-Start

wait(4.6µs)

write reg1=0x304400 calculate Hit3-Start

wait(4.6µs)

write reg1=0x404400 calculate Hit4-Start

wait(4.6µs)

Now all Hit data are available from registers 0 to 3.

The load pointer value is 4.

At the end the TDC-GP2 has to be initialized again to be ready for the next measurement. This is

done by sending the opcode „Init“ so that the TDC accepts new Start and Stop hits.

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 4-1

TDC-GP2

4 Measurement Mode 2

4.1 General Description

1 stop channels referring to one start channel

Typ. 50 ps RMS resolution

LSB width typ. 65 ps

2 x Tref pulse pair resolution

3-fold multihit capability

Measurement range 2 x Tref to 4 ms @ 4MHz

Selectable rising/falling edge sensitivity

Integrated programmable windowing for each single stop with 10 ns precision

Digital TDCs use internal propagation delays of signals through gates to measure time intervals with

very high precision (see also measurement range 1, section 4).

In measurement range 2 the maximum time interval is extended using a pre-divider. The resolution

in LSB remains unchanged by that. In this mode the high-speed unit of the TDC does not measure

the whole time interval but only time intervals from START and STOP to the next rising edge of the

reference clock (ﬁne-counts). In between the ﬁne-counts the TDC counts the number of periods of the

reference clock (coarse-count).

Figure 4-1

The BIN size (LSB) is typically 65 ps at 3.3 V and 25 °C ambient temperature. The RMS Noise is

approx. 50 ps (0.7 LSB). The gate propagation delay times strongly depend on temperature and vol-

tage. In measuring range 2 the result is the sum of different ﬁne and coarse-count results. Therefore

it is necessary in measuring range 2 to make a calibration. During a calibration the TDC measures 1

and 2 periods of the reference clock.

The measurement range is limited by size of the coarse counter:

tyy = Tref x 214 = 4.1ms @ 4MHz

The time interval between START and STOP is calculated with a 26 Bit measurement range.

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

4-2

TDC-GP2

Time-to-Digital-Converter

Figure 4-2 Time (Condition) Description

tph 2,5 ns (min.) Minimum pulse width

tpl 2,5 ns (min.) Minimum pulse width

tss 2*Tref Start to Stop @

Dis_Phasenoise=1

trr 2*Tref Rising edge to rising edge

tff 2*Tref Falling edge to falling edge

tva 4.6 µs (max.) ALU start to data valid

tyy 4 ms (max) Max. measuring range

Input circuitry

Each input separately can be set to be sensitive to rising or falling edge. This is done in register 0,

Bits 0 to 2. (NEG_START, NEG_STOP1).

Further all Start/Stop-inputs support a high active enable pin.

Note:

In case the Start-Stop interval is less than the lower limit tzz the TDC will ignore more and more

events the smaller it is. In no case there will be wrong results.

4.2 Measurement Flow

Figure 4-3

4.2.1 Configuration

At the beginning the TDC-GP2 has to be conﬁgu-

red. The main settings for measurement range 2

are:

a. Select measurement range2

setting register 0, Bit3, MRange2 = 1.

b. Select the reference clock

(see also section 5.1)

In measurement range 2 the TDC-GP2 needs

the high-speed clock for the time measurement.

In case of low-power applications this clock can

be switched of in between measurements. The

a 32.768 kHz clock is necessary for the timing

control during the oscillator power-on.

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 4-3

TDC-GP2

only the high-speed clock is used this is be set to „1“(continuously on). In case both oscillators are

used for current saving reasons this should be set to „2“ for ceramic oscillators and to „3“ for quartz

oscillators“.

4). The choice has an inﬂuence on the minimum time interval

tmin = 2 * Tref * 2ClkHDiv

and the maximum time interval

tmax = 214 * Tref * 2ClkHDiv

Further, it is necessary that

2 * Tref * 2ClkHDiv < 1.8 µs.

Otherwise the ALU will go into an overﬂow during calibration and write 0xFFFFFFFF as output data.

c. Set the number of expected hits

In register 1, Bits 8 to 10, HITIN1 the user has to deﬁne the number of hits the TDC-GP2 has to

wait for. A maximum of 3 on channel 1 is possible. The number HITIN1 always has to be higher by 1

than the number of expected hits. The reason is that the Start is also counted as a hit. The TDC-GP2

measures until the set number of hits is registered or a timeout occurs. register 0, Bits 11 to 13,

HITIN2 have to be set to „0“.

Example:

2 stop pulses are expected:

HITIN1 = 3, HITIN2 = 0

d. Select calibration

The calibration is switched on by setting register 0, Bit5, Calibrate = „1“. It is mandatory to do this.

For the calibration the TDC measures 1 and 2 cycles of the reference clock. The two data are stored

as Cal1 and Cal2.

There are two ways to update the calibration data Cal1 and Cal2:

Separate calibration by sending opcode Start_Cal_TDC via the SPI interface

Automatic update by setting register 0, Bit 4, DisAutoCal = „0“. In most applications this will be

the preferred setting.

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

4-4

TDC-GP2

Time-to-Digital-Converter

e. Define ALU data processing

While the TDC unit can measure up to 3 hits the ALU can calculate only one hit at once. The settings

are done in register 1, Bits 16 to 19 and 20 to 23, HIT1 and HIT2. The Start pulse is internally

handled like a Stop pulse because of the special measuring method in measurement range 2.

Reg1 = 0x21xxxx = Calculate 1st Stop Ch1-Start

Reg1 = 0x31xxxx = Calculate 2nd Stop Ch1-Start

Reg1 = 0x41xxxx = Calculate 3rd Stop Ch1-Start

The ALU calculates the time interval as:

Time = RES_X * Tref * 2ClkHSDiv

f. Select input sensitivity

In register 2, Bits 19 & 20, RFEDGE1 and RFEDGE2, the user can select whether the stop in-

puts are sensitive to either rising or falling edges (RFEDGE = „0“) or to both rising and falling edges

(RFEDGE = „1“). In register 0, Bits 0 to 2 the user can add an internal inverter to each input, Start,

Stop1 and Stop2. With RFEDGE = „0“ this is the same as rising edge (NEG_X = „0“) or falling edge

(NEG_X = „1“).

g. Interrupt behavior

The INT pin can have various sources, to be selected in register 2, Bits 21 to 23, EN_INT.

EN_INT = 0 no Interrupt source

1 ALU ready

2 The set number of hits is there

4 Timeout of the TDC unit

The different options are wired by OR.

After the conﬁguration the user has to initialize the TDC-GP2 by sending opcode „Init“ so that the TDC

accepts Start and Stop hits.

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 4-5

TDC-GP2

4.2.2 Measurement

After an initialization the TDC unit will start with the ﬁrst pulse on the Start input. It will run until:

the set number of hits has been seen (maximum 3 on channel 1 in measurement range 2)

or until a timeout occurs. The timeout can be programmed in multiples of the reference clock

setting Reg 3, Bits 19&20, SEL_TIMO_MR2. At 4 MHz the values are:

SEL_TIMO_MR2 (@ 4 MHz, ClkHSDiv = 0)

= 0 = 64 µs

= 1 = 256 µs

= 2 = 1024 µs

= 3 = 4096 µs

At the end of the time measurement the TDC measures 2 periods of the reference clock for calibra-

tion.

4.2.3 Data processing

At the end of the measurement the ALU starts to process the data according to the HIT1, HIT2

settings and transfers the result to the output register. The ALU calculates according to 4.1.2.e and

transfers the 32Bit ﬁxed point number to the output register.

The time it takes the ALU depends on the supply voltage:

Table 4-1: ALU timings

calibrated

3.3 V 2.3 µs

2.5 V 2.8 µs

2.0 V 3.1 µs

As soon as the data is available from the output register the interrupt ﬂag is set (assumed that the

ALU interrupt is enabled, see reg 2, EN_INT). Further the load pointer of the output register is incre-

ased by 1 and points to the next free memory. The actual position of the load pointer can be seen in

the status register, Bits 0 to 2.

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

4-6

TDC-GP2

Time-to-Digital-Converter

4.2.4 Reading Data

Now the user can read the data sending the opcode 10110ADR. With the next 32 cycles (calibrated

data) the TDC-GP2 will send the result, beginning with the main signiﬁcant Bit (MSB).

The 32 Bit ﬁxed-point number in complements of 2 represent the time interval in multiples of the

reference clock.

Time = RES_X * Tref * 2ClkHSDiv

The conﬁguration of the ALU allows only one hit calculation at the time. In case more than one hit

has been measured it is necessary to write new commands to HIT1/HIT2 to instruct the ALU for

calculating the other hits. After writing to HIT1/HIT2 it is necessary to wait for minimum 4.6µs (cali-

brated data) or 580ns (un-calibrated data) before reading or writing again to HIT1/HIT2.

Example:

conﬁguration

...

write reg1=0x214400 3 hits on channel 1, calculate Hit1-Start

...

Initialize

...

while(Check interrupt ﬂag)

write reg1=0x314400 calculate Hit2-Start

wait(4.6µs)

write reg1=0x414400 calculate Hit3-Start

wait(4.6µs)

Now all hit data are available from registers 0 to 2. The load pointer value is 3.

At the end the TDC-GP2 has to be initialized again to be ready for the next measurement. This is

done by sending the opcode „Init“ so that the TDC accepts new Start and Stop hits.

4.3 Stop Masking

The TDC-GP2 can set time-based masking windows for each of the 3 hits on Stop1 input when no

hits are accepted. The masking refers to the start event and has an accuracy of less than 10 ns.

The internal enable unit is connected to the external enable pin by a logical AND. The external enable

pin must be set to “1” to use the internal masking unit. The conﬁguration settings are made in regis-

ters 2 to 4, DELVAL1, DELVAL2 and DELVAL3:

DELVAL1 … DELVAL3 are ﬁxed point numbers with 14 bit integer and 5 bit fractional digits, in

multiples of the internal reference clock

Delaymin = DELVALX / 25 * Tref * 2ClkHSDiv

The minimum mask size is 3 clock cycles

The mask values must have an ascending order. Each mask value must be 3 clock cycles bigger

than the previous value

It is mandatory that if not all registers are used the mask values that are not required are set to “0”.

When all DELVAL registers are set to 0, the complete unit is disabled.

Example:

4 Mhz reference, ClkHSDiv = 1

DELVAL1 = 0x3200 1st Stop not accepted before 200 µs after Start

(128000/32 * 250ns * 21 = 200 µs)

DELVAL2 = 0x3300 2nd Stop not accepted before 204 µs after Start

(13056/32 * 250ns * 21 = 204 µs)

DELVAL1 = 0x3400 3rd Stop not accepted before 208 µs after Start

(13312/32 * 250ns * 21 = 208 µs)

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 4-7

TDC-GP2

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

4-8

TDC-GP2

Time-to-Digital-Converter

Figure 5-1

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 5-1

TDC-GP2

5 Details and Special Functions

5.1 Oscillator

The TDC-GP2 uses up to 2 clock signals depending on the operating mode:

High-speed clock, typically 4 MHz, for calibration and as a predivider for the TDC measuring unit

in measurement range 2

32 kHz clock signal used for internal timer functions.

5.1.1 High-Speed Oscillator

Generally the TDC-GP2 needs a 2 to 8 MHz high-speed clock for calibration. Operating in measure-

ment range 2 the TDC-GP2 needs the high-speed clock signal also as a part of the time measuring

unit. The oscillator takes an average current of 260 µA when running all the time.

Because it is needed only during the time measu-

rement the TDC-GP2 has the capability to control

the on-time by itself. The high-speed clock can be

switched on automatically with a INIT opcode only

for the period of time measurement. An additional

delay cares for the settling time of the oscillator.

The settings are done in register 0, Bits 10 & 11,

START_CLKHS:

START_CLKHS = 0 Oscillator off

= 1 Oscillator on

= 2 The measurement is started with 640 µs delay.

= 3 same as ‘2’, but with 1280 µs delay

The programmable delay guarantees that the oscillator has settled before the measurement starts.

For ceramic resonators 640 µs will be sufﬁcient.

Note:

Notice the Bug report in section 7 when using Start_CLKHS with quartz oscillators.

By this measure the average current consumption can be drastically reduced.

Example:

At one ToF measurement in an ultrasonic ﬂowmeter (forth/back) per second the high-speed oscilla-

tor is active only for about 2 ms. The average current consumption is 260 µAs * 2 ms = 0.52 µA.

Figure 5-2

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

5-2

TDC-GP2

Time-to-Digital-Converter

5.1.2 32.768 kHz Oscillator

The TDC-GP2 needs a 32.768 kHz reference for the start-up control of the high-speed clock and the

clock calibration. It therefore offers a an integrated driver:

There is no need for the 32.768 kHz clock if the high speed clock is permanently running (START_

CLKSHS = 1) and the high speed clock calibration is not used (e.g. in case of a quartz oscillator).

The 32.768 kHz oscillator is permanently running and has a current consumption of about 4.5 µA at

3.6 V. It is also possible to provide an external low-

frequency rectangular clock at the CLK32Out pin

(3.6 V max.). This signal could be generated by an

external microprocessor. It will reduce the current

consumption down to 1.2 µA. For this reason the

32 kHz oscillator should only be used if no external

low-current 32.768 kHz clock is available. The

settling time of this oscillator is about 2 s.

The external circuit is necessary only if the

32.768 kHz oscillator is used. Otherwise CLK32In has to be connected to GND.

5.1.3 Calibrating a Ceramic High-speed Oscillator

Using a ceramic oscillator for the 2 to 8 MHz clock will be attractive because it is of low cost and

has a fast settling time. Unfortunately it has a poor tolerance of 0.3 to 0.5 % and shows a tempe-

rature drift. For this reason the TDC-GP2 allows to execute a calibration measurement that allows

to compensate this behavior. This measurement is based on the very precise 32.768 kHz clock. The

TDC-GP2 generates start/stop pulses from the 32.768 kHz and measures this time interval with its

TDC unit. The result is stored in the result register and the interrupt ﬂag is set. The frequency error

of the ceramic resonator can be calculated by the microprocessor. The calibration is conﬁgured by

setting register 0, CALRES# and is started with „START_Cal_Resonator“ - instruction by the micro-

processor.

The time interval to be measured is set by CALRES# which deﬁnes the number of periods of the

32.768kHz clock:

CALRES# = 0 2 periods = 61.035 µs

= 1 4 periods = 122.07 µs

= 2 8 periods = 244.14 µs

= 3 16 periods = 488.281µs

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 5-3

TDC-GP2

The results is given in multiples of the high-speed clock and (divided by 1, 2 or 4 (DIV_CLKHS)) as 32

bit ﬁxed point numbers with 16 integer bits and 16 fractional bits.

The microcontroller can compare this measured value to the theoretical value and calculate the cor-

rection factor RES_X/REStheor.

Example:

The system shall work with a 4 MHz resonator. With CLKHSDIV=0 and CALRES#=1 the theore-

tical result is 122.0703125µs/250ns = 488.28125 (RES_0 = 0x01E84800). If the ceramic

resonator in use is not exactly at 4MHz but only 3.98MHz the calibration measurement will show

485,83984375 (RES_0 = 1E5D700). The correction factor for the microcontroller is 1.005.

Note:

During clock calibration the start input has to be enabled.

5.1.4 How to use Clock Calibration

a. Application

This option is dedicated especially to ultrasonic ﬂow and heat meters. In those applications the use

of ceramic oscillators shows two main advantages: lower cost and less current consumption. Mainly

because of the short oszillation start up time of the ceramic oscillator the operating current can be

reduced by several µA. Referring to 10 years of operation this saves several 100 mAh in battery

capacitance. There is no negative effect on the resolution when using this option the correct way.

b. Jitter of the 32 kHz clock and consequences

The 32 kHz clock is very precise in frequency with only a few ppm error. However the phase jitter is

about 3 to 5 ns peak-peak. For this reason also a calibration measurement (Start_Cal_Resonator)

has this error. When multiplying a measurement result with the calibration result, the jitter of the

calibration is transferred to the result by the ratio calibration measurement time (see CALRES#) to

measurement time. Using a permanently updated calibration value will add a considerable jitter to the

measurement result.

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

5-4

TDC-GP2

Time-to-Digital-Converter

c. Application of this option in ultrasonic flow meters

A measurement result is always made of two single time-of-ﬂight measurements in ultrasonic ﬂow

meters, with and against the ﬂow direction. The difference between those measurements is a mea-

sure for the ﬂow. To avoid an inﬂuence of the calibration jitter on this measurement result it is ne-

cessary only to use the same calibration for both ToF measurements. Following this the differenc

between the two ToF measurements will be free of the jitter of the clock calibration measurement.

The clock can be calibrated only between measurements that are not directly substracted from each

other.

5.2 Fire-pulse Generator

5.2.1 General Description

The ﬁre-pulse generator generates a sequence of pulses which is highly programmable in frequency,

phase and number of pulses. The high-speed oscillator frequency divided by the factor selected for

ClkHSDiv is used as the basic frequency. This frequency is internally doubled and can freely be divided

by a factor of 2 to 15. It is possible to generate 1 to 15 pulses. For each pulse the phase can be ad-

justed per register conﬁguration. The ﬁre-pulse generator is activated by sending opcode Start_Cycle.

The ﬁre-pulse generator provides 2 outputs, Fire1 and Fire2. The driver strength of each output is

48mA @5V. These 2 outputs can be paralleled to increase the driver strength up to 96 mA. Fur-

thermore Fire2 output signal can be inverted to double the signal amplitude. The outputs can be set

individually high-Z.

The ﬁre-pulse generator allows to generate and send pulse sequences multiple times for use in a qua-

si “sing-around” method. Using this feature the received pulse sequence is fed into TDC-GP2 Fire_In

input. It is digitally ampliﬁed and directly forwarded to the output buffer for an immediate re-emittance

without any clock delay.

Note: When the fire pulse generator is use it is mandatory to send a start pulse.

5.2.2 Configuration

Number of pulses:

FIRE# = 0 Switch off ﬁre-pulse generator

1 1 pulse

2 2 pulses

… …

15 15 pulses

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 5-5

TDC-GP2

Phase:

The phase of each pulse can be deﬁned in register 5, Bits 0 to 15, PHASE_FIRE. „0“ stands for

HIGH-LOW and „1“ for LOW-HIGH. The pulse sequence begins with the LSB and ends with the MSB

Example:

Fire# = 7, PHASE_FIRE = 0x0055

Fire-pulse frequency:

The input signal ﬁreclk1 for the ﬁre pulse generator is derived from the high speed clock CLKHS and

the selected value for the high speed clock divider CLKHS_DIV.

Figure 5-3

This Signal is internally doubled and divided by DIV_FIRE.

DIV_FIRE = 0 not permitted

1 divided by 2

2 divided by 3

... ...

15 divided by 16

improve statistical behaviour when averaging.

DIS_PHASESHIFT = 0 Phase shift on

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

5-6

TDC-GP2

Time-to-Digital-Converter

DIS_PHASESHIFT = 1 Phase shift off

ﬁreclk2 is used as reference signal for the Fire1 / Fire2 - signal which is emitted by the output buf-

fers Fire1 / Fire 2 of the ﬁre pulse generator.

Figure 5-4

As shown in Figure 16 at least 2 clock periods Tﬁreclk2 are required to send one Fire_Pulse. One for

the high phase and one for the low phase of the Fire1/Fire2 output signal.

Example:

CLKHS = 4 MHz, CLKHS_DIV = 1, DIV_FIRE = 1

Max. frequency of the Fire1 / Fire2 output signal:

Driver outputs:

The output drivers are conﬁgured in register 5, Bits 21 to 23, CONF_FIRE:

Bit 23 = 1 Inverted output on FIRE2

Bit 22 = 1 FIRE2 disabled (High-Z)

Bit 21 = 2 FIRE1 disabled (High-Z)

Pulse-burst repetition (quasi sing-around):

In register 5, Bits 16 to 18, REPEAT_FIRE the number of repetitions of the pulse sequence is deﬁ-

ned.

REPEAT_FIRE = 0 no repetition

= 1 1 repetition

... ...

= 7 7 repetitions

Only the number of pulses set under FIRE# will be repeated. With a period of 5 µs without a pulse

the TDC-GP2 detects the end of a pulse sequence.

Note:

It is strongly recommend that the summarized time of ﬂight of the up to 7 repetitions does not ex-

ceed the measurement range of the GP2!!!

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 5-7

TDC-GP2

5.3 Temperature Measurement

Especially for heat meter applications the TDC-GP2 has a PICOSTRAIN based temperature measuring

unit that offers high resolution and very low current consumption.

The measurement is based on measuring discharge times. Therefore a capacitor is discharged alter-

nately through the sense resistors and the reference resistors.

Figure 5-5 Figure 5-6

The unit has 4 resistor ports with the following function:

PT1 reference resistor lower temperature

PT2 sense resistor lower temperature

PT3 sense resistor higher temperature

PT4 reference resistor higher temperature

The temperature sensor should have a minimum resistance of 500 Ohm. The TDC-GP2 measures

the discharge times of the RC-networks made of each resistor and the capacitor.

The precision of the temperature measurement is about 0.004°C several times better than needed

for heat meters.

The temperature measurement is fully automated. It is triggered by the µC sending the opcode

„Start_Temp“. The TDC-GP2 controls the 4 measurements by itself. After the 4 measurements have

ﬁnished the interrupt ﬂag is set. The four data are found in registers 0 to 3.

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

5-8

TDC-GP2

Time-to-Digital-Converter

From Res_2/RES_1 and RES_3/RES_4 the microcontroller can calculate the ratio Rtemp/Rref. By

means of a look-up table it can calculate the temperature for the special type of sensor in use.

It is not possible with TDC-GP2 to use 4-wire temperature sensors.

Configuration

Tcycle = 0 128 µs cycle time @ 4MHz

Tcycle = 1 512 µs cycle time @ 4MHz

Port# = 0 2 ports = 1 sensor

Port# = 1 4 ports = 2 sensors

ture measurement. This is necessary to overcome mechanical effects of the load capacitor.

Fake# = 0 2 dummy measurements

Fake# = 1 8 dummy measurements

A full temperature measurement last for

2 x (# of ports) + (# of fakes) cycles.

Recommended Capacitor Values

The discharge time should be about 150 µs. Therefore the capacitor should have the following value:

PT500: 220 nF

PT1000: 100 nF

Please set Tcycle = 1 to avoid Timeout Error.

Recommended Capacitor Type

To get accurate results we recommend capacitor types with very low dC/dU. We recommend:

CfCap Series from Tayo Yuden

For heatmeter application please do note use X7R or similar capacitors.

Current consumption

By means of the TDC technology the temperature measurement needs an extremely low current,

much less than an A/D converter does.

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 5-9

TDC-GP2

A full temperature measurement with 2 sensors, 2 references am PT1000 sensor type, including

all calculations takes less than 2.5µAs. With one temperature measurement in 30 seconds (typical

for heatmeters) the average current consumption is 0.08 µA only. This is about 50 times less than

other solutions. A PT500 sensor doubles the current.

Note:During temperature measurement the start input has to be enabled.

Error detection

Additionally the temperature unit checks the plausibility of the results. It is able to detect a short cir-

cuit of the sensor or an open sensor. The TDC-GP2 provides in the relevant output register an error

code instead of a measurement value.

1. Short circuit: equivalent to a very short time interval (< 8 x Tref = 2 µs @ 4 MHz). The TDC-GP2

writes a 0x0 to the output register of the shorted sensor.

2. Broken sensor: equivalent to no stop signal or timeout. The TDC-GP2 writes a 0xFFFFFFFF into

the output register of the open sensor.

Table 5-1: Analog speciﬁkation

Parameter Value (typ.) Unit

Resolution RMS 16.0 Bit

SNR 96 dB

Absolut Gain-Error 0,1 %

Gain-Drift vs. Vio 0,08 %/V

Gain-Drift vs. Temp 0,0008 %/C

Uncalibrated Offset <0.01 %

Offset Drift vs. Temp <0,2 ppm/C

PSRR >100 dB

Condition: Vio = Vcc = 3.3 V, PT1000, 150nF charging capacitor

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

5-10

TDC-GP2

Time-to-Digital-Converter

5.4 SPI-interface

The serial interface is compatible with the 4-wire SPI standard. It needs the SerialSelectNot (SSN)

and can not operated as 3-wire interface.

SSN - Slave Select

SCK - SPI Clock

SI - SPI Data In

SO - SPI Data Out

The TDC-GP2 does only support the following SPI mode:

Clock Phase Bit =1

Clock Polarity Bit =0

The timings are shown in ﬁures setion 2.2. It is mandatory to set the SSN – line to High-state for at

least 50ns between each Read-/Write sequence.

SSN as Reset

The SerialSelectNot (SSN) line is the HIGH-active reset for the serial interface. After SSN is set to

LOW different operations can be addressed, not depending on the status of the interface before the

reset.

OPCodes

MSB LSB Description

1 0 0 0 0 ADR2 ADR1 ADR0 Write into address

ADR

followed by

24 bit data

1 0 1 1 0 ADR2 ADR1 ADR0 Read from address

ADR

01110000Init

0 1 0 1 0 0 0 0 Power On Reset

0 0 0 0 0 0 0 1 Start_Cycle

0 0 0 0 0 0 1 0 Start_Temp

0 0 0 0 0 0 1 1 Start_Cal_Resonator

0 0 0 0 0 1 0 0 Start_Cal_TDC

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 5-11

TDC-GP2

The transfer starts with the MSB and is ﬁnished sending the LSB. After sending the last Bit TDC-GP2

transfers the data into the target register or executes the command. It is not possible to do incre-

mental writing. Each register must be addressed separately.

When reading from the chip it is necessary to send the opcode ﬁrst, too. With the ﬁrst positive edge

of the clock following the opcode the TDC-GP2 sends the MSB of the addressed register to SO out-

put. Each positive edge transfers the next lower Bit to the output.

5.5 Fast Initialization

In measurement range 1 the TDC-GP2 offers the possibility of a fast initialization. Activated by setting

is already prepared for the next measurement while the data can be read out. This mode is for high-

est speed applications only. It is most reasonable for un-calibrated measurements with only one stop.

5.6 Noise Unit

In case the user wants to improve the measuring results by averaging it is necessary that the values

do not always display exactly the same time difference. Instead the user should provide some ‚noise‘

so that different quantization steps of the characteristic curve of the TDC are involved. This can not

happen with very constant time differences. One would constantly hit the same LSB.

The noise unit enables the use of weighted averaging even for constant time differences. The noise

unit adds a random offset to the start. It is dedicated to applications where the TDC gets a dummy

start and measures the time difference between STOP1 and STOP2 (e.g. laser range ﬁnders).

The noise unit is switched on by setting register 5, Bit 20, EN_STARTNOISE = „1“

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

5-12

TDC-GP2

Time-to-Digital-Converter

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 6-1

TDC-GP2

6 Applications

6.1 Ultrasonic Heatmeter

6.1.1 General Description

The TDC-GP2 is perfectly suited for low-cost ultrasonic heatmeter designs. Thanks to the implemen-

ted functionality, including precision temperature measurement, ﬁre pulse generation, windowing and

clock calibration it is sufﬁcient to add a simple microprocessor (without A/D converter) and a trans-

ducer dependant driver and receiver. The extremely low current consumption guarantees the neces-

sary long battery lifetime in such applications.

The measurement is fairly automated by the TDC-GP2. The microprocessor just sends a start com-

mand. The TDC then ﬁres the transducers and measures the time of ﬂight. It calibrates the results

and provides them to the microprocessor.

Figure 6-1

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

6-2

TDC-GP2

Time-to-Digital-Converter

6.1.2 Configuration

A typical conﬁguration could look like:

Port# = 4, TCycle = 0, Fake# = 2, SelClkT = 1, Calibrate = 1,

DisAutoCal = 0, MRange2 = 1, NEG_STOP2 = 1, NEG_STOP1 = 1,

NEG_START = 1

DELVAL1 = 400

REPEAT_FIRE = 0, PHASE_FIRE = 0

All inputs are set to rising edges

Measurement range 2 is used with Auto-calibration.

The temperature measurement uses the high-speed clock with 128µs cycle time and 2 fake mea-

surements on 4 ports (2 sensors, for cold and hot water)

The high-speed clock is switched on only for the time measurement with 640µs delay

The 4 MHz high-speed clock is used without a divider

The 4 MHz clock calibration is based on 8 periods of the 32.768 kHz clock (244.14 µs)

For the ﬁre-pulse generation the 4 MHz is internally doubled and the divided by 4 making 1 MHz.

The generator sends 3 pulses

The ﬁrst stop is accepted after 100 µs, the second one after 102 µs and the third one after

104 µs.

The TDC is set to measure 3 hits on STOP1.

The ALU is set to calculate ﬁrst Hit1 – Start

All interrupt options are activated, timeout will be given after 1024 µs

Phase-shifting is disabled

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 6-3

TDC-GP2

6.1.3 Measurement Flow

Power-on reset:

Send SO = 0x50

Conﬁguration:

Send SO = 0x80338AE8

Send SO = 0x82114000

Send SO = 0x82E03200

Send SO = 0x83083300

Send SO = 0x84203400

Send SO = 0x85080000

Calibrate Clock:

Send SO = 0x03 Start_Cal_Resonator

Check-loop INTN = 0?

Send SO = 0xB0, Read SI = RES_0

Correction factor = 488.28125/RES_0

Measurement loop:

Temperature measurement every 30 seconds:

Send SO = 0x02 Start_Temp

Check-loop INTN = 0?

Send SO = 0xB4, Read SI = STAT

STAT&0x1E00 > 0: -> Error routine

Send SO = 0xB0, Read SI = RES_0

Send SO = 0xB1, Read SI = RES_1

Send SO = 0xB2, Read SI = RES_2

Send SO = 0xB3, Read SI = RES_3

Rhot/Rref = RES_0/RES_1

Rcold/Rref = RES_3/RES_2

Go to look-up table to get the temperatures

Time-of-ﬂight measurement every half second:

Send SO = 0x70 Initialize TDC

Send SO = 0x01 Start_Cycle triggers ﬁre-pulse

generator.

Check-loop INTN = 0?

Send SO = 0xB4, Read SI = STAT

STAT&0x0600 > 0: -> Error routine

Send SO = 0x81314000 calculate HIT2-Start

Wait for 4.6µs (ALU time)

Send SO = 0x81414000 calculate HIT3-Start

Wait for 4.6µs (ALU time)

Send SO = 0xB0, Read SI = RES_0

Send SO = 0xB1, Read SI = RES_1

Send SO = 0xB2, Read SI = RES_2

µP can now start the data post-processing and

calculate the ﬂow and the heat.

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

6-4

TDC-GP2

Time-to-Digital-Converter

6.1.4 Current consumption

a. Time measurement

32.768kHz: are provided by the µP: < 1.3 µA

4 MHz: 0 µA in power down, 270 µA when active

TDC: < 150 nA if not active, 15 mA during active time of the high-speed unit

With 2 measurements per second (forth and back ﬂow) the total consumption is < 3 µA.

b. Temperature measurement

A full measurement over all four ports takes less than 2.5 µAs current. Typically the temperature is

measured once in 30 seconds. The average current then is about 0.085 µA. This is about 50 times

less than with solutions without a TDC.

c. Quiescent current

Thanks to the current optimized 0.35µ technology the quiescent current is less than 150 nA typ.

d. Total system current

The complete current for the measuring unit (TDC, analog part, Transducer) will depend on the ana-

log circuit part. It will be in the range of 4 to 6 µA.

With a low-power µP (e.g. MSP430 series from TI) the average current consumption of the total

device might be in the range of 11 to 16 µA.

It is possible to operate the system from a lithium-thionylchloride AA cell for 10 years without chan-

ging the battery. At 6 years runtime it maybe even possible to work with a low-cost 3V CR2450 coin

cell battery.

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 7-1

TDC-GP2

7 Miscellaneous

7.1 Bug Report

7.1.1 Quartz Oscillator Start Up time

Bug: When using a quartz resonator for the oscillator the oscillation start up time is about

3 ms even with an optimized circuit. The TDC-GP2‘s automatic on-time control for the

high speed clock is based on 640 µs or 1280 µs delay without any activity. In case of a

quartz resonator this delay is too short and will cause malfunction.

Corrective: There are several options to deal with.

1.Oscillator permanently on

The possibility of switching off the high speed oscillator is added only for current sa-

ving. This oscillator needs about 270 µA when running continuously. In all application

without attention to current consumption we recommend to run the high speed oscilla-

tor continuously (START_CLKHS=1).

2.Ceramic Resonator

With a ceramic resonator the oscillation start up time is about 200 µs. There is no

problem with TDC-GP2‘s internal delays. We recommend the use of ceramic resona-

tors in all applications targeting ratio measurements (like ultrasonic ﬂow metering).

In all applications that need a quartz resonator and the switch-off functionality we re-

commend the following measures:

3.Lower Clock Frequency at CLK32In

A clock with lower frequency increases the internal timer delays. In this case we recom-

mend about 4 kHz. The timer delay is increased to 10 ms then. This is enough for a

save oscillation start up time of the quartz.

4.Start/Stop Oscillator by Software

Writing a „1“ or „0“ into Start_CLKHS starts or stops the oscillator by software. This

way a µP can start the oscillator by command and then after 10 ms can send a mea-

surement opcode like START_CYCLE. At the end of the measurement the µP has to

switch off the oscillator by a further command.

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de

7-2

TDC-GP2

Time-to-Digital-Converter

7.1.2 DIS_PHASENOISE

Bug: The TDC-GP2 offers a Phase-Noise function that decouples the calibration clock from

the ﬁre pulse generator. This option was implemented to provide the necessary statis-

tics for averaging in case the user is looking for a resolution much better than 1 LSB

(about 65 ps) and has a a very strong coupling between the start and the reference

clock.

Because of a design bug the phase noise unit might fail in case of distortions on the

power supply. This is relevant only in measurement range 2.

Corrective: The phase noise unit has to be switched off by setting DIS_PHASE_NOISE = 1. In case

the start is already asynchronous to the referenceclock this has no effect on the ca-

pability of averaging.

7.2 Last Changes

05.10.2006 3.2.1b&c, 5.1.3, DIS_PHASENOISE=1 required

27.11.2006 5.2.2 Additional corrections

3.2.1 Corrections Example

01.02.2007 5.3 Additional corrections for temp. measurement

26.05.2010 Various smaller error ﬁxes. General layout change

acam-messelectronic gmbh - Am Hasenbiel 27 - D-76297 Stutensee-Blankenloch - Germany - www.acam.de 7-3

TDC-GP2

Data Sheet

acam-messelectronic gmbh

Am Hasenbiel 27

76297 Stutensee-Blankenloch

Germany / Allemagne

ph. +49 7244 7419 - 0

fax +49 7244 7419 - 29

e-mail: support@acam.de