PCM behaving strangely: compressor short, sticky fault, intermittent functioning

[beerobber]

I need some tips on how to diagnose potential PCM issues.

Our PCM operates intermittently. When we enable the robot, the compressor will give a little sputter, then the PCM will blink orange. The NI console reports a compressor short. But if we wiggle the wires, or ignore that issue and focus on other things, typically the compressor will kick in a few minutes later. It won't run smoothly, just 4-5 seconds at a time until it finally builds up to 120 on the high side. Might take 5-7 minutes to get there.

The solenoids won't fire reliably through the PCM. They work great manually, but during the times the compressor isn't running and the light blinks orange, they won't respond to code control.

We have replaced the compressor. We have replaced the PCM. We have run new power wiring from the compressor to the PCM. We've isolated the PCM on the CAN bus. None of these things made it reliable. The new PCM seemed to get itself into a working state a bit more frequently than the old one.

This is a tense moment for our team. Our first district event loads in Thursday night. We spent 4 of our 6 out-of-bag hours tonight, and we have 2 left. So with the robot in the bag, I can't take pictures and I can't try your suggestions until tomorrow night. But please load me up with ideas so we can get this issue cleared!

Here are some things I've read on other threads:
- Try moving PCM further from radio (not sure this will be possible, but we can try some shielding material to block RF temporarily for troubleshooting)
- Reapply firmware
- Use Weidmuller ferrules like http://www.ctr-electronics.com/cabli...le-16-awg.html
- Make sure CAN isn't running parallel to power

Thanks in advance for your suggestions.
--Chris

[JeremyLansing]

We had exactly this issue during our week 0 scrimmage. We changed compressors, PCM, wires, re-imaged the PCM, same problem, just like yours. We had our router mounted on top of the RoboRio and the problem went away when we moved it about a foot away from the rest of the electronics. We did another 4-5 hours of testing after the fix and never had the issue again. Best of luck and I hope you can fix the issue.

[Peyton Yeung]

We had that issue and it went away as soon as we moved our radio to the other side of our robot.

[beerobber]

Interesting. We will give that a try. Finding a new home for our radio will be a challenge in this year's robot designed to go under the low bar, but we'll do it if it means working pneumatics! I'll post tomorrow night with updates.

Any further suggestions are welcome.

[beerobber]

Success! Thanks for the tips. I don't know when we would have ever thought to create separation between the radio and the PCM, but that did it.

Our PCM was less than 1 inch from the radio. Our first validation was to slide a triple-thickness sheet of aluminum foil between those components, and that was enough to allow the compressor start immediately and run until the pressure switch cut it off. Solenoids are also reliable. The PCM is signaling green and happy.

After a quick debate about the appropriateness of aluminum foil as a crucial component in a competition robot, we moved the radio (only 2 wires) and used zip ties to give it a new home about 9 inches from the PCM.

Some said the aluminum foil gave the robot a NASA-like quality, but others pointed out that blanketing components with a highly-conductive material that blocked air flow was probably not going to help get us to Einstein. So it came off before the robot went back in the bag.

[DonRotolo]

Aluminum is almost OK as a shield, except that it only conducts electricity and not magnetism. An electromagnetic field has both electrical and magnetic components, and you should want to shield both when possible.

Of course, moving the radio was the better solution.

[scca229]

Something just got added to my list of things to look for if a team's bot is acting erratically and they don't have a clue as to why...

Thanks for the follow-up from an appreciative FTA.