For a new experiment, I made a LabVIEW program for controlling LED boards. I used a PCI-6229 and a PCI-6723 for controlling the boards: on/off by using the digital outs of 6229 and intensity modulation by the analog outs of the 6723. With no actual hardware attached (but the DAQ cards in place) everything worked well and stable at my place.
Now the experiment is assembled with LED boards and 2 stages of drivers behind the DAQ cards it functions properly but is not stable. After a random number of hours (1..24h) it ends up in a permanent on state of the LEDS while both digital out and intensity are programmed to be off and zero. I'm really surprised this can happen. I do not know the ins and outs of the driver electronics.
Can external signals and noise change the settings of these boards? I did not measure yet the signals, I'm going there next week. Hopefully some of you can offer some ideas that I can test next week. I know the environment is very noisy (pumps, heating, cooling, etc.) and also the mains quality is bad. A test with putting the PC/DAQ cards on a separate (nobreak) mains inlet improved a litte, but not much. Also the last addition to the experiment, sending commands by RS232 to a gas controller, made things worse. I have no clue what is going on. Can it be grounding? Bad PC power supply? I do not do any analog in.
Any ideas welcome! Then I can try/test them next week.
I would need to see the LabVIEW code to be certain But I can give you some broad statements and guesses.
It sounds like your devices are resetting. I've had this happen to me in one case years ago. I was continually opening and closing my DAQmx Tasks and other resouces. I had my LabVIEW.ini set to automatically clean up VISA sessions and forgot that this meant VISA sessions would accumulate untill I Closed LabVIEW. A Simillar thing happens with DAQmx tasks. You can only open and close so many task referances before the API hangs up (4096).
Take a look at how you are handeling the opens and closes.
You mean the devices are resetting by themselves? I cannot find information on that topic yet. Is there a way to find out in the program a reset has occured? I guess, we better try to prevent the device(s) from resetting. But good to know this can happen.
Have you verified with a meter the state of the digital outputs when the LED's are latched on?
How much current do the driver inputs circuits consume from the digital output lines?
Any sketches or schematics for the wiring of your setup?
Yes, more queries than answers 🙂