The weirdest thing ever.
 

We have a compressor on our robot and we need to charge it out in the match waiting line because it has a small leak and drops significantly. So we tether it out around the field and it charges the tanks. Well the weird thing is that it gives a strange time issue, the time starts and stops erratically, and displays a watchdog error. It wouldn't be that strange if it happened all the time. But it only happens out around the field. In the pits we set up the same conditions, no problem, on the field we set up, no problems. This isn't a severe problem the tanks still get charge up and everything its just strange.

Any ideas on why this is happening. Is the field surrounded by a forcefield that I should know about.

Re: The weirdest thing ever.
 

Probably a coincidence with it being on the field only, but I've seen this problem before. It's a problem with the camera and the WPI code. If the camera hasn't finished initializing, the console is flooded with errors until the camera finishes loading (ring turns green). The flooding slows down the rest of the system enough that watchdogs trip and halt the robot intermittently.

If you wait for the ring around the camera to be green before enabling the bot (and an image is showing on the DS if you have it referencing the camera in robot init), there shouldn't be any freezing

Re: The weirdest thing ever.
 

Mutually exclusive statements.

Fix the leak; 1 PSI loss per minute is easily achieved. Saves grief in queue, too.

Get a small cup of soapy water, pressurize the system, and paint the entire pneumatic system (with a 1/2" paintbrush) at every connection. Include the two connections on the compressor head, and the connections between parts of the pneumatic valves. Any bubbles that form and grow indicate leaks; tighten the connection/replace the teflon tape/cut the hose end square to stop the leak.

Re: The weirdest thing ever.
 

First, welcome to Galileo! I'm glad you could make it in as a late addition. I believe we will be pit neighbors (365 & 379 in Galileo) and I would be happy to take look at your code for potential watchdog issues (camera sounds like a likely one as mentioned). Also, I'm sure we have someone who could help find the leak.

Re: The weirdest thing ever.
 

On the leak part we found i think it was 3 leaks and fixed them (with the soapy water trick) but there still is a small leak somewhere. On the camera we don't even have one attached (would it be bad to leave the ethernet cable for the camera plugged into cRio port 2 to remove the possibility of error of plugging the radio into port 1).
Edit: Woot Galileo.

Re: The weirdest thing ever.
 

Well if it isn't the camera there still are a few things that can cause the same error (jag not communicating over CAN, printfs in the code, anything that prints to stdout)

Re: The weirdest thing ever.
 

We have no jags to communicate over the CAN with.

Re: The weirdest thing ever.
 

When the problem occurs near the field, and when it doesn't occur in your pit, are you using the same tether cable?