I'm trying to acquire the tic signal of a french power meter. From the datasheet available here of the TIC signal we can find out that the signal is (page 7/38): 6 Vrms +/- 10% at 50 kHz

It comes from two terminals and it can be converted with an optocoupler circuit.

I found out a device doing this and converting the signal through USB but I would like to do my own to use it directly on GPIO of a raspberry pi (seen as a serial of 1200 bauds).

Here is the initial device.

This is my circuit : Optocoupler circuit with NAND component

To convert this signal, there is also a NAND component. I think it is used to power up the signal and to make it "squared".

My problem: I tried to simulate the power meter signal with a low frequency generator sinusoïd, but the circuit only work until the 16kHz. After the signal just go down. For now, I can't test it on a real power meter so I'm afraid it won't work if it's already not working with a low frequency generator.

Any idea why? I thought using an existing circuit will be safer to make this acquirement but apparently it's not :-/

Any advice or ideas are welcome !


  1. I tested the circuit on prototyping board and I can now go to 30kHz before it goes down. I did the exact same circuit but it's just not with CMS component. Maybe is it related to my capacitors?
  2. I tested the initial circuit also with the low frequency generator and it can go to 32kHz. Maybe my test with the lfg is not representative enough of the real signal emitted by a meter. I find it strange that a product sold on the market reacts this way without respecting the official documentation.
  3. I finally had access to a power meter and the circuit worked :-D I still don't know how it's possible. There is the informations I get from the serial at 1200 bauds :


DATECOUR 28/01/19 16/22/50 X

EA 0Wh 5

ERP 0varh H


DATEPA1 28/01/19 16/20/00 Z

PA1 0kW 4

DATEPA2 28/01/19 16/10/00 Z

PA2 0kW 5

DATEPA3 28/01/19 16/00/00 Z

PA3 0kW 6

DATEPA4 28/01/19 15/50/00 _

PA4 0kW 7

DATEPA5 28/01/19 08/06/35 +

PA5 0C.kW )

DATEPA6 28/01/19 08/00/00 ^

PA6 0kW 9

DEBUTp 24/01/19 10/10/43 Z

FINp 01/01/92 00/00/00 6

CAFp 5 /

EApP 0kWh


EApHCH 0kWh #

EApHCD 0kWh _

EApJA 0kWh [

EApHPE 0kWh -

EApHPH 0kWh 0

EApHPD 0kWh ,

ERNpP 0kvarh 1

ERNpHCE 0kvarh 1

ERNpHCH 0kvarh 4

ERNpHCD 0kvarh 0

ERNpJA 0kvarh ,

ERNpHPE 0kvarh >

ERNpHPH 0kvarh A

ERNpHPD 0kvarh =

KDC 90%

KDCD 80% #

PSP 1kW &







PSJA 1kW !


PA10MN 0kW ?

PREA1MN 0kvar X

PREA10MN 0kvar H

TGPHI 0,00 8

U10MN 401V <

These informations can be next decoded with the datasheet each line corresponding to one header and its value.

  1. I found these in the datasheet (my translation of the page 9/38) with the schematic curve of the signal (just below). I think the 50 kHz might be in fact the "carrier" frequency. It would explain the filter outside the otocoupler and why my circuit is working. If I understand correctly, the carrier is only here to carry the real signal. We don't need to preserve this carrier but only the signal carried. I think I just confused the signal frequency with the carrier frequency. If anyone can confirm or deny my suppositions?

The signals present on the information circuit are represented by Figure 2 and are characterized by the following parameters: a)Vevh1 is the maximum level of the envelope for the transmission of a "1"

b)Vevl0 is the minimum level of the envelope for the transmission of a "0"

c)Vevh0 is the maximum level of the envelope for the transmission of a "0"

d)Tev1 is the minimum guaranteed time during which the envelope has a level lower than Vevh1

e)Tev0 is the minimum guaranteed time during which the envelope has a level between Vevl0 and Vevh0

f)Vevl0 and Vevh0 are not the extreme values of the envelope, but rather the limits "low" and "high" guaranting a correct functioning

g)during the duration Tev0 the level of the envelope must not vary by more than 20%

h)during the time intervals between Tev0 and Tev1, the increasing or decreasing evolution of the enveloppe of exponential type, or sinusoidal damped with addition of low frequency transients
i) the harmonic distortion rate during a continuous emission of the carrier on a resistance of 100 Ω is lower to 15% (also applicable to historical mode)

j)all voltages are specified in peak values.

Characteristic of the carrier envelope Table of the TIC signals carracteristics

  • \$\begingroup\$ R4 and C1 will form a sort of low pass filter. Cutoff around 7.5kHz. So, anything above that will be mangled. The higher the frequency, the worse the mangling. The falling edge will be sharp (shorted by the optoisolator,) but the rising edge will be limited by the RC time constant of R4 and C1. Are you sure of their values? \$\endgroup\$
    – JRE
    Jan 28, 2019 at 12:26
  • \$\begingroup\$ Do you have any information on this TIC signal for those of us who don't speak French? \$\endgroup\$
    – Hearth
    Jan 28, 2019 at 13:30
  • \$\begingroup\$ I'm pretty sure of the values, but maybe I didn't correctly understand the meaning of these values. I can't find any english documentation about it so I will try to translate some of it. Do you want informations about the physical signal or about the informations carried on ? \$\endgroup\$
    – A.Girafe
    Jan 29, 2019 at 14:47
  • \$\begingroup\$ I just add more informations in my initial question (edit number 3 and 4). I think it was a misunderstanding linked to the carrier frequency and to the signal frequency itself. \$\endgroup\$
    – A.Girafe
    Jan 29, 2019 at 16:31


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.