Can DAQmx regeneration cause 200019 error

[mfitzsimons]

I am using a 6230 DAQ card to output a continuous analog (AO_0) tone of varying levels.  The problem is that I am trying to use the same card (AI_0) to sample the data and keep getting error -200019 (Measurements: ADC conversion attempted before the prior conversion was complete).  I have tried to adjust the sample rates and can't resolve the issue.  Can the DAQ card sample while outputting a continuous tone with regeneration disabled?

NI DAQmx: Continuously Generate Waveform - Non Regeneration - Rapidly Updating - WEB Link: http://zone.ni.com/devzone/cda/epd/p/id/370

[Chris_D]

Are you using an external clock source? I understand that you have tried adjusting the sample rate, but may be this KB can shed some light. Let me know if this is not the case and we can investigate further.

What Causes Error -200019 When Increasing the Sampling Rate of an Analog Input Task with an External Clock?
http://digital.ni.com/public.nsf/allkb/A01AA72779C4604986257047007344EC?OpenDocument

[mfitzsimons]

I am not using an external clock source.  Still trying to figure it out.

Matt

[Chris_D]

I have tried testing this on a different DAQ card with regeneration disabled and was unable to reproduce the results. I would like to try testing this on a 6230 just to make sure, but before doing this I would like to get more information about your system setup. Is the only thing connected to the device the AO to the AI? Also what generation and sampling rates have you tried to reproduce the error?

[Kevin_Price]

Is the AI being hw-clocked at all or is it a software-timed on-demand reading?  Can you post a screenshot of the task config and write/read code?

I've used M-series boards for multiple simultaneous tasks before and they have typically behaved fine. 

Odd question:  any chance that the AO task is producing an uncaught error that ends up causing the AI error?  Your setup sounds like one where

*AO* errors would be more likely due to the timing needs of updating the non-regenerating buffer.  It's otherwise hard for me to picture how you'd

get the opportunity to attempt a 2nd AI Read before the 1st has finished.  You don't have multiple parallel attempts to do on-demand AI Reads, do you?

-Kevin P.

[CAS_MattF]

Here is the code in question.  I had another developer try it out and has the same error.  Stuped on the cause and looking forward to hearing the solution.

[Kevin_Price]

I don't have hw available for testing, but have you tried:

1. Separating the AO and AI into their own loops?

2. Using other values for parameters like sampling rate, # samples to generate, # to read?  The default value of 500 kHz is pretty fast and if you only read 1000 values at a time, you'd need really fast loop timing.  The presence of the "Basic Function Generator" block and the graphs could be an issue.
   Also, based on default values, you write 25000 samples to the AO buffer each loop while reading only 1000 from the AI buffer.  Both tasks run at the same sample rate though, so you're going to eventually run into a problem with that discrepancy.

Also, you seem to configure the AO to be triggered by the AI start trigger, but you haven't confitgured a  start trigger for your AI task.

-Kevin P.

[Chris_D]

I tested the code and was unable to produce any 200019 errors. I was, however, able to produce both 200279 and 200292 buffer errors. I believe Kevin has made some good suggestions to try.

[mfitzsimons]

All,

Breaking the code into two independent loops fixes the problem.  Not sure why I would get a sample rate error when the code it in one loop.  Thanks for everyones help!