06-08-2021 11:13 AM - edited 06-08-2021 12:10 PM
I have an FPGA program that is set up like the "NI 9467 Datalogging" project in the Examples.
It has worked fine for the past couple of months, and a couple of days ago it suddenly stopped working. What happens is that the loop will run until the Wait for PPS doesn't Timeout (ie when it receives a PPS), and then the loop freezes.
I tried swapping out my antenna, and I tried swapping out the 9467 module, and neither helped. So maybe there is an issue with the FPGA?
The Wait for PPS function doesn't generate any errors, so that's no help
What would cause this? How can I troubleshoot/fix this?
edit: I have also tried recompiling it, and running it on a different chassis. Neither helped
Thanks
06-11-2021 01:30 AM
Hi, You could look for the status of the Status LED. This will give you some indication on what's going with the module - Refer to page 8. You can also check if if the current consumed is close to the detection threshold the LED may flicker. You try removing module from project and adding it again and rerun the program.
06-11-2021 02:49 AM - edited 06-11-2021 03:02 AM
Another thing I had in the past. Check your power supply! It may seem ridiculous but that was actually the reason an application behaved erratically. According to the spec the power supply should have been big enough to deliver the required current, yet the system didn't start up properly or just stopped midway. Usually Ethernet connection was impossible if it started up at all.
After half a day of debugging we moved the whole thing from the intended installation location to a test bench and everything just worked fine. Took us a few more hours to finally pinpoint the cause to the power supply. After replacing it, everything was fine and no unexplainable stopping anymore.
This was with an sbRIO-9651 on a custom PCB carrier, but thought I would mention it here. And the power supply in question was an old lab power supply. My guess is that its capacitors had aged according to their natural life expectancy and therefore it wasn't able to provide the transient current surges that can happen when modern high frequency clocked digital hardware is massively changing logic states across entire fabrics, resulting in very short voltage drops on the power line that can impact operation.