I/O problem requires cycle power to clear

[ortec]

We are using a PCI-6503 using traditional Ni-DAQ. We are simply reading an input to see if a physical device is in position. It works as expected, except that around every 25 days or so, the input will suddenly "freeze" and show the same state (not in position) regardless if the device is in position.
 
We use another input (in the same bank) that continues to work as expected. Seemed to us that the problem would be with our coding, but restarting the software does not correct the problem. Restarting Windows does not correct the problem. Cycling power to the PC and the machine (literally pulling the plug from the wall) does correct the problem. The problem will vanish for another 25-days or so regardless of how often it is used.
 
Of course we monitored the O/S side of resource usages and such (i.e. memory leaks), but found not difference from day 1 to day 25. Regardless, restarting Windows should clear that anyway.
 
The custom software we are using is a Visual Basic .NET application. Our first thought was that maybe the software was not releasing the resources to the card or something, but the other input works and since restarting Windows does not correct the problem, it appears to be something else.
 
Oh yeah, another clue. When the input fails to change state, we can start the NI-DAQ test panel (from NI) and can see the input change state. Could the NI-DAQ interface we are using communicate via a different path to the card than the test panel? Again though, we can get it to return to normal with a complete power cycle only, which seems to elimate coding problems.
 
Any ideas?
 
Thanks,
Benson

[Nathan_R]

Hi Benson, If I understand what is happening you have a PCI-6503 being used with Traditional NI-DAQ drivers. You are periodically reading a digital input. After about 25 days you quit reading changes on that one input. Once this happens you can still read the input correctly using a test panel, but your code does not work again without unplugging the computer.If you are using the Traditional NI-DAQ test panel in Measurment and AUtomation then the test panel is communicating with the card the same way your program is. I'd be interested to know if once the input state locks up can you run one of the NI DAQ examples succesfully. NI-DAQ examples can be found at C:\Documents and Settings\All Users\Documents\National Instruments\NI-DAQ\Examples. Also what version of Traditional NI-DAQ are you using?I agree that having to unplug the computer to fix the problem should eliminate coding from being the cause. However the card is still working in the test panels. If it was a hardware problem it should show up in the test panel as well. Have you tried resetting the Traditional NI-DAQ drivers in Measurement and Automation once it locks up. You can reset the drivers by right clicking on My System>>Device and Interfaces>>Traditional NI-DAQ (Legacy) Devices and selecting "Reset driver for traditional NI-DAQ".Let me know if you have any questions. Have a great week and take care.Thanks,

[ortec]

Hello Nathan,
Just to clarify, does the Ni-DAQ Test Panel use the same communication path/interface as third party software using Ni-DAQ would use?
Thanks,
Benson

[Nathan_R]

Hi Benson,


Yes, the NI-DAQ test panel uses the same communication path
as any software using the NI-DAQ drivers. We generally use the test panels to
make sure the driver and hardware are working properly. If the test panel works
and the software doesn?t, it is almost always a problem in the software. Were
you able to try running an example program after the line state locks up?
Please let me know if you have any questions.


Thanks,