LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

PCI6229 card - DIsplays NaN on all channels

The setup is a test equipment running on Windows 11 + 2 Nos PCI-6229 and 1 No PCI-6514 cards + LV2024. 

 

All going good for a few months and suddenly there is a power tripping and the UPS also crashed simultaneously. The PC just went through a forced power down in the middle of a test. 

 

Once the power resumed and we started the PC, nothing worked and we tried to open MAX which responded with a message " NI-MAX could not launch because the required library mxrmcfg.dll failed to load " 

 

We then did a fresh install of DAQMX 24 drivers and LV2024 run time and started again.  This time it worked and the test started - but our happiness was short lived.  Randomly after a few hours of running suddenly the system crashed reading "NaN" on all 25 analog  channels and the only way to recover was a reboot. 

 

System is working like this as we are unable to resolve the issue whatever it is . 

 

In such situations is there is any supervising software or tool kit available to log events , so that we can later find out what went wrong ?  Any ideas ??

 

( Something like  Windows ProcMon help ? If yes I need to ask the System Administrator permission to install and run the same !) 

 

Thanks 

Raghunathan
LabVIEW to Automate Hydraulic Test rigs.
0 Kudos
Message 1 of 3
(180 Views)

I'm not completely sure what you're saying the problem currently is.

 

Is it

 

"Every time we reboot, it works for a few hours, then changes to all NaN outputs"

 

or

 

"It worked for a few hours one time, then changed to NaN on all outputs permanently, even after a reboot"

 

?

 

Also, are you really running PCI (not PCIe) DAQ cards?  If those are PCI, then it would indicate that your hardware is possibly quite old and it might be related to general hardware fatigue, exacerbated by the power down incident.

0 Kudos
Message 2 of 3
(131 Views)

It would be possibly useful to tell us roughly what the program is supposed to be doing. 

  • It appears you are sampling 25 channels of single-ended A/D.
    • What is the sampling rate?
    • What is the Buffer size?  Are you using "Continuous" acquisition?
    • Are you using a Producer/Consumer design where an acquisition (Producer) loop does a DAQmx Read of 25 channels and "exports" the (say, 1000-point) buffers to a Consumer loop running asynchronously in parallel, conveying the data from Producer to Consumer by a Queue or a Stream Channel Wire?  Explain how "Acquisition" and "Storage" are linked.
    • How have you tested your system?  Something to try is to write a simple "Simulate System" routine where you do nothing but simulate data acquisition.  For example, if you are sampling 25 channels of A/D at 1 kHz using 1000-point buffers, use a Timed Loop to generate data as though it came from 1000 A/D samples in the form of a 16-bit integer (Loop Index * 100 + Chan #) then sent it to the Consumer and see if it can run OK for 2-3 hours.  If so, then the problem may be faulty hardware (the DAQ boards).
  • Is there other processing going on that could interfere with either the Producer (acquiring data from the DAQ boards) or Consumer (writing/saving the data) to go astray?  Can you monitor the memory use of your code while it is running?  In particular, do you see memory use growing as the program runs, suggesting you have a "bottlenick" somewhere?
  • Have you considered showing us (or your "hired consultant") the LabVIEW code (the entire Project, saved as a "Compressed (zipped) Folder")?

Bob Schor  

0 Kudos
Message 3 of 3
(82 Views)