I am getting a voltage spike on starting and stopping a daqmx task

[Robert_Hoffman]

Hello,

I am having a problem with a PXI-4461 DSA board that I can't explain.  When starting or stopping an analog output task, I am getting a voltage spike before the generation starts and after it stops.  We have put a speaker on the output, and the spike sounds like a "pop" when that happens.  However, the pop goes away when the card is also acquiring an analog input in the card.  This is a repeatable event, and very strange.  We have used the two examples "ContAcqGraph Voltage" for the analog input and the "Retriggering_Analog_Output_Waveform_Generation" to generate this phenomena.

Has anyone ever seen this before?  Is our card the culprit?

Thanks,

Rob

[MicrowaveDave]

Hey Robert,

I don't think I've seen this one before.  Occasionally, I have seen instances when an output voltage will spike before generation if an old waveform is still in memory on the device.  You mentioned that you are seeing a spike at the end of this waveform.  Does this happen consistently?  Does it happen have programatically resetting the device?

Regards,

David Hall | Applications Engineering | National Instruments

[Robert_Hoffman]

David,

Unfortunately, it is very consistent...let me see if I can describe it in detail.

When running the generation example, the spike or pop happens when the task just before the task starts and just after the task stops.  However, if we then run the acquisition example (while the generation is running), the spikes immediately stop.  Then if we stop the acquisition example (and leave the generation running), the spikes do NOT come back.  So, the only thing that causes the spikes are the starting and stopping of a task.  We are not doing any resetting of the device programmatically.  However, if we power off the controller and power it back on, the spikes come back, and the only way to stop them is to go through the same sequence described previously.  Unfortunately, we didn't have another card to try out to see if it was consistent across cards.

Is there anything in the start and stop task controls which could cause this?

Thanks,

Rob

[gaving]

Hey Rob,

     I'm sorry your having trouble with your board.  I'd like to suggest us coming to some reproducible situation.  Namely, can you let me know version of DAQmx you are using, and just for clarity's sake, the version of LabVIEW as well.  Now, can you run an AO shipping example and recreate this behavior?  These are located in c:\Program Files\National Instruments\LabVIEW #.#\examples\DAQmx\Analog Out.  The one I ran here and watched for this spike was "Cont Gen Voltage Wfm-Int Clk.vi".  I modified it to stop/restart the task when I pressed a button.  I could not reproduce the spike.  If you can reproduce it with this or another simple example, please respond and attach that VI.  Then we can see if it is an issue with your board or something else.  Thank you for using our NI DSA products!

Good Luck,

Gavin Goodrich

National Instruments

[Robert_Hoffman]

Gavin,

The application is a deployed program that we based off of the example that is provided with the sound and vibration toolkit (SVXMPL_Swept Sine FRF (DAQmx).vi).  If you don't run our application, there are no pops.  But, as soon as you run our app, then try to run the analog output example (Retriggering_Analog_Output_Waveform_Generation), pops appear.  As soon as you run the analog input example (ContAcqGraph Voltage), then the pops disappear.  It is very repeatable.  I am checking on the version of DAQmx.

Thanks,

Rob

[gaving]

Hey Rob,

     I found out some information that might be helpful to you.  There was a problem identified with that particular SVT example awhile ago where spikes would happen due to a Device Reset during the generation.  This device reset was performed by default in the example if there was an error with the AO generation.  Since your code was based off of this example, I believe that there's a Device Reset happening due to some error.  If you could, please go through your application and ensure proper error chaining so that any errors generated are seen by the user.  You can also go through and probe the error clusters coming out of the VIs in order to get a quick picture of what might be happening. 

     The specific VI in question is "SVT Close Swept Sine (DAQmx).vi".  You can set a breakpoint before the "DAQmx Reset Device.vi" on the block diagram and verify that this is happening(see attachment screenshot).  Let me know what you find out.

---

Good luck

Gavin Goodrich

National Instruments