Counter output edges visible on other channels

[Flumen]

Hi everyone,

I am using a PCIe-6353 DAQ card and a BNC connector module (do not know which one) to generate different trigger outputs. The triggers are generated using the counter output, which I unfortunately is not displayed in the attached snippet.

When checking the output on an oscilloscope I can see small spikes at the edges of the trigger as you can see in the attached picture. This is obviously no big problem for the TTL trigger signals but it can also be seen on the analog out (AO) channel. In the snippet, the AO is set to 1V an connected to the blue channel on the oscilloscope.

I will use the AO to control galvo mirrors where the voltage controls the mirror position. I am afraid that the small spikes might have a negative influence on the stability of the mirror position. 

Is there anything I could do to prevent that from happening? I tried searching for crosstalk etc but could not find anything similar.

Thank you!

[Henrik_Volkers]

Looks like signal cross talk.

General causes:

The high dU/dt causes dI/dt ...

-capacitive/inductive coupling (close traces/cables)

-shared current traces (GND?)

The use of the corresponding AGND and DGND lines is mandatory! 

Try a shoke (ferrite core with n windings of both (AO+/AGND) lines ) see picture

What's maybe worth a try: Try different DO lines for the counter output.

Have a look at AO signals at the DAQ output connector , with a 10kOhm load . don't connect the DO lines to the scope. 

If you still find the spikes, and can't use a low pass filter for the galvos I would switch to seperate units.

If the signal looks fine, go up the line.. NI connector cable, BNC box ...

[Flumen]

Hi Henrik,

Thanks for your answer. I might have used the wrong word for the BNC module so for clarification here are the right names:

NI DAQ card:   PCIe-6353 

BNC box:         BNC-2110

Using the BNC box the 'mandatory' use of the ground is automatically guaranteed, right? For the digital out channels you would still have to connect the DGND, I get that, but using the BNC connectors (AO, PFI0, PFI12) the GND is connected, correct?

I simplified the program and thereby the problem quite a bit. I see the crosstalk even when only the AO channel is connected to the oscilloscope, while a single counter output is generated on a line but not connected to the oscilloscope. Also, it does not matter on which line I generate the counter output.

After resetting and using the Tristate VI (which seems to be almost mandatory when using CO ?!) I got the crosstalk down quite a bit. It is about +/- 50mV now.

I will try the ferrite core if I can find one.

Thanks,

Lukas (German, too!)

[Flumen]

... actually the magnitude of the interference might just have been higher before because of the BNC cable. Resetting and tristate.vi did not have an influence on it.

[TimothyDug]

Hey Flumen,

I have a few questions about this behavior.

1. To be clear, this behavior goes away when you stop using the counter, correct? The counter is clearly responsible for this?

2. Does changing the AO line affect it?

3. Are the two grounds shared? The Digital and Analog?

Hope we can get this figured out. Thanks!

[Flumen]

Hi Timothy,

1. Yes. As you can see in the picture of the oscilloscope, the counter(s) edges can be seen on all other signals (digital and analog). 

2. No. It is the same on all the lines. I can see it on the DO and AO lines.

3. I am not sure if I understand the question. For DO I am using the PFI0 & PFI12 BNC connectors on the BNC-2110 module. For the AO I am obviously using AO0 and/or AO1 with BNC connectors, too. 

Does this answer your question?

The frequency of this small change is probably so high that it will not have any effect on the galvo mirrors.

Thanks for the help,

Lukas