Unexplained intermittent CAN / 2CAN Jaguar problems at GSR

[techhelpbb]

We rebuilt the frame for our 2nd robot and once again, with an entirely rebuilt electrical board and new mounting for the cRIO (and some heavier wiring in some places)...we still have CAN timeout errors occasionally...though we can still manage in voltage mode as we have done so far. So this is now 3 entire reworks of that system all producing similar results. I should like to point out however, that we can literally stall this current robot and we can't demonstrate that the timeouts we see match that stall event.

After addressing some issues with crimps of the RJ connectors on this build, I'm adding to my list of projects a Jaguar CAN cable tester, maybe one that can produce a 1 or 2 MHz test signal for a testing of the wires. Maybe with a nice counter to see if the pulses are being dropped.

We have set up a collection of test equipment that can be mounted on our robot (while it drives around freely) and when we get done with competition hopefully we'll get some nice readings of the signals at various points in this system. Clear up some of possible issues that have been suggested. I'd really like to get a nice protracted recording of the CAN bus communications prior to the timeouts and an analog recording of the power supply voltages, the CAN bus and a few other things. At very least it'll be highly educational for the students and handy to have the tools (we'll share as we get it going). Hopefully what we'll work out will be budget conscious enough that it can be replicated by other teams without undue cost (so far I feel that's the case).