Labview PID

[nathand]

Can you post your "Edited NI subvi" too? Seems that's where the actual PID work is happening.

[Benvalet]

Hi Nathand,

Attached please find the edited NI subvi as requested. Thank you.

Best Regards,

Opuk

[Benvalet]

Hi ZX81,

Thank you so much for your elaborate troubleshooting to of my problem. I have just one of the tests suggested by you and it seems that is where the problem lies. Now I skipped over the PID and outputted a copy of the input and lo the spikes are still present.

So I think the problem is with data acquisition and not PID? Just thinking. Attached please find my connections to target 7833R and my code for data acquistion. what could the be the problem?

Kind Regards,

Opuk

[Benvalet]

Hii All,

Wha would make the simple Analog in....analog out fpga code attached cause spikes on the analog output signal?

Best Regards,

Opuk

[ZX81]

I see one problem; maybe in documentation; maybe in wiring.  In one place, you have AI0 (connections image), another AI1 and AO1 (short-cut program).  It is common to have a little cross-talk across analog inputs that aren't connected to anything, so having the input on the wrong place may still seem to give you a valid input, but may have glitches (like what appear to be happening).

The connection image only showed two connections.  There should be 4, one pair at AI1(+ and -), and another at AO1 (+ and -).

I need to go to a long meeting... may not be available till tomorrow.

[Benvalet]

Hi ZX81,

Thanks for your response and for making me realize the exact source of my problem. I am so sorry the problem was on documentation  but the wiring is OK. I have used used 4 connections. AI0 (+ and -) and AO0 (+ and -).

I have used different types of function generators to test for my hardwares and the problem persists. I have also tried the same in a different set up of target and host and the problem is the same. I bet this is a SOFTWARE problem.

Any further clues?

Thank you.

Kind Regards,

Opuk

[ZX81]

(meeting postponed)

The short-cut program is on AI1 and AO1, not AI0 and AO0, so that test wasn't valid.  Please change it to match wiring & try again.

The main code reads AI0 (which you said is connected), but writes AI1 (should be AI0).  It is less common to get cross-talk on analog outputs, but still possible.  Since your first posting showed +/- 0.1V, that seems high for cross-talk on an output.  Another shows much higher (off scale; maybe +/- 0.8V), which is much less likely cross-talk.  But it makes me suspicious why the difference would be so large.

Again, a graph of the signal in LabView would help.  If you see glitches on the input (AI0 or AI1, or both), then the input has an issue.  If you see glitches on the output (AO0 or AO1, or both), then it's the PID or software.  But I have a hard time accepting the PID could be at fault if the short-cut case has the same issue (we don't know till you change it to the wired I/O).

[Benvalet]

Hi ZX81,

Thanks for the response. Just confirmed the problem is not PID at all. I have also used different setups for input and output (AI1 and AO1) and the glitches are still present. 

Best Regards,

Opuk 

[nathand]

The first thing I would check is grounding. Do the signal generator and the NI board share a common ground?

What happens if you generate a signal other than a sine wave?

[Benvalet]

Hi Nathand,

Thank you so much for your response.

Now the signal generator and the NI board do not share a common ground.

When I generate even a ramp wave the problem persists. See the attached.

Thank you.

Kind Regards,

Opuk