cwsint
Topic Author
Posts: 6
Joined: 05 Oct 2018, 22:21

Crosstalk between In1 and In3 on AIO module?

30 Jan 2019, 11:18

Hello everyone,

we are observing a weird thing in our test-setup. Our setup looks like this:
  • Out1 --> In1 --> In2
  • Out2 --> In3 --> In4
  • Output range for OutN is 0..20 mA
  • Inut range for InN is 0..20 mA
So unless no wire is broken, we should see the same signal on In1 and In2 as well as on In3 and In4. We are outputting a sawtooth and a sine-signal on Out1 and Out2 (or vice versa). Whenever the sawtooth drops, we will see irregularies in In1 or In3, respectively. In2 and In4 behave as expected (see attached files, where v1..v4 represent In1..In4, respectively). So our guess is that we have some kind of crosstalk / cross influence on channels In1 and In3.

Any ideas on how to solve this issue?

Regards
Cord
Attachments
Crosstalk V3-V1.png
Crosstalk V3-V1.png (413.66 KiB) Viewed 171 times
Crosstalk V1-V3.png
Crosstalk V1-V3.png (547.72 KiB) Viewed 171 times
 
User avatar
dirk
KUNBUS
Posts: 191
Joined: 15 Dec 2016, 13:19

Re: Crosstalk between In1 and In3 on AIO module?

06 Feb 2019, 16:11

Hi, with interest we have discussed your issue here and we have some questions:
1) Did you use an oscilloscope for the inputs
2) How did you generate the sample data
3) Could you describe your hardware setup and the connections
4) Could you make a plot without the sawtooth
5) Could you make a plot without using the inputs but with an oscilloscope instead
 
cwsint
Topic Author
Posts: 6
Joined: 05 Oct 2018, 22:21

Re: Crosstalk between In1 and In3 on AIO module?

11 Feb 2019, 09:35

Hi Dirk,

thanks for your reply. Let me try to answer your questions:

1) No, unfortunately the device is currently with out software developer who does not have access to an oscilloscope.
2) The waveforms are generated by our software for test-purposes. The process image is updated every 10 ms. The range of the signals is 0..20k.
3)
AIO Module.png
AIO Module.png (35.71 KiB) Viewed 82 times
Output 1 is driving Inputs 1 and 2, Output 2 is driving Inputs 3 and 4. Since we are using 0..20 mA here, Inputs 1 and 2 as well as Inputs 3 and 4 should measure the exact same signal. So (even without the osciloscope) I would assume the signal-Generation is working since we measure the correct signal on Input 2 and 4, respectively.
4) We generated three cases for you (please also see next post due to attachment limitations). The sinewave on Out2 is unchanged, only Out1 (formerly sawtooth) was set to constant values
a) Out1 = 10.000
Sine with Out1 at 10k.png
Sine with Out1 at 10k.png (164.31 KiB) Viewed 82 times
b) Out1 = 10
Sine with Out1 at 10.png
Sine with Out1 at 10.png (189.08 KiB) Viewed 82 times
5) Unfortunately not now, see 1).
Last edited by cwsint on 13 Feb 2019, 10:02, edited 1 time in total.
 
cwsint
Topic Author
Posts: 6
Joined: 05 Oct 2018, 22:21

Re: Crosstalk between In1 and In3 on AIO module?

11 Feb 2019, 09:40

… continued Question 4)
c) Out1 = 0 :shock:
Sine with Out1 at 0.png
Sine with Out1 at 0.png (169.37 KiB) Viewed 80 times

In addition, when zooming in on In1 and V2 (which should show a constant value of 10k or 10, respectively), we seem to get some crosstalk from the sine-wave of Out 2 (more pronounced on In1 than on In2):
Sine with Out1 at 10k In3 and 4 turned off.png
Sine with Out1 at 10k In3 and 4 turned off.png (173.52 KiB) Viewed 80 times
Sine with Out1 at 10 and In3 and 4 turned off.png
Sine with Out1 at 10 and In3 and 4 turned off.png (169.93 KiB) Viewed 80 times
 
User avatar
dirk
KUNBUS
Posts: 191
Joined: 15 Dec 2016, 13:19

Re: Crosstalk between In1 and In3 on AIO module?

15 Feb 2019, 13:50

We have reproduced your setup here. We have used an RevPi Core 3 and an AIO module that is configured using 0-20 mA for Output1 and Output2.
The wiring is setup as follows: Output1 is connected to Input1 and Input2; Output2 is connected to Input3 and Input4.
We have created the attached program "aiotest.py" that creates a sinus signal and prints out the values. You may run and sample the data by the command
chmod +x runme.sh; ./runme.sh
Our result is that no outliers apparent so we have the guess that the wiring could be the root cause of the suspicious behavior.
Probably the GND circuit could be enhanced. GND input and output must not be connected because differential inputs can not work otherwise.
Idea: Separate the two outputs from each other
Attachments
aiotest.zip
(623 Bytes) Downloaded 3 times

Who is online

Users browsing this forum: dirk and 1 guest