CAN being wierd

[bardd]

We're using CAN for the first time. We have two jaguars chained and connected to the cRIO's serial connection.
When we try to activate both jaguars simultaneously, one of three things happen:

1. Both jags work correctly,
2. Only one jag works (always the same one), the other's signal light shows it gets an "enabled" signal but no "move" signal,
3. None of the jags work.

One thing we noticed is when you change the voltage you want the jags to send as output drastically over a short period of time, both jags don't work.

We have no idea why that happens... Help would be very much appreciated.

[Levansic]

I certainly could be wrong on this, but I'm assuming that you are using the jaguars to control single CIM motors on a tank drive setup. What is your gear ratio, wheel size, and wheel configuration?

I ask, because it sounds like you are inducing over current situations. The one that cuts out more often likely has a chain that is a bit too tight, but both sides are likely dealing with gear ratios that are too low for the torque that they are asked to provide. Another possibility is that the one that cuts out more is not the one directly connected to the cRIO, and you have a CAN bus connectivity or termination problem.

An attempted quick fix for the overcurrent condition, is to use a bigger number for the voltage ramp rate parameter. By default, the parameter is set to ~500, which corresponds to a 1/8 second ramp from 0-100% output. This is too quick for heavily loaded CIM motors, causing an over-current cutout. If you aren't already using voltage ramping on the jaguar, it sounds like you should be. This will throttle the response to "drastic" changes of input.