Faulty AI channels on NI PCIe-6321?

[thoult]

I'm developing a test system with a PCIe-6321 acquiring analogue voltages. I've noticed that the data from one of my differentially received channels (AI 0) was consistently falling outside tolerance for my test for a variety of devices.

After plenty of fault finding, I determined that AI 0- (AI 8 when RSE) was recording spurious voltages, as if it were a floating analogue input. AI 0+ was still working absolutely fine. My initial suspicion was that there was a fault in my electronics, as there's a fair bit going on there. A few hours of probing later and I could confirm that the expected voltages are present at the 68 pin receptacle on my PCB.

After disconnecting the device from my custom breakout interface, and connecting it to a SCB-68, I was able to determine that all the analogue inputs were working as expected (by shorting AI GND to D GND and shorting 5 V to the desired input terminal and configuring for RSE) with the exception of AI 8. I repeated this for a brand new cable from a sealed bag, but still the same.

The peak voltage measured by that analogue input does not exceed 3v3. The device runs self-test fine and there are no connector pins obviously damaged.

Any other ideas of what I could do to troubleshoot before I pull the card out and send it back to NI?

[JB]

Have a look at the DAQ Diagnostic Utility from NI.

[thoult]

I've just had a look at the connector under a microscope, and found that the pin in question is ever-so-slightly bent. A little tweezering later and it's apparent that that was the cause of the problem.

If in doubt, it's bound to be a cable or connector 😕

Many thanks though for the suggestion of the DAQ diagnostic tool. I'd not used that before, so it was good to have been introduced to it. Hopefully I never have actual recourse to use it in anger 🙂

[thoult]

Related to this, does anyone know of somewhere I can find a VHDCI port savers?