Jags resetting in CAN
 

In speed mode using the USD E4P-250-N-S-D-D-B encoder we are tripping the JAG reset when we switch from +max to -max or the opposite, using a joystick controller. This happens in any P value , we are not an I or D value. No matter how we tune the PID, we see a choppy transition from + to -. In % votage mode everything works fine.

Re: Jags resetting in CAN
 

We had this issue three years ago and tracked it down to the Jaguar browning out. The PID tries to drive max current to achieve the large change in commanded velocity. We measured peak currents well over 100A - the Jaguar is rated at 60A for short (<5sec ?) periods and 92A instantaneous. The self resetting breaker in the PDB is only rated at 40A but I seem to remember someone on CD testing it and reporting 2x tripping in ~15sec.

You need to limit the peak current to <60A. There are several ways to do this, the best being to add a slew limit or ramp. There is a 0.125sec ramp in the Jaguar code but it is not used in PID mode. You can/should add a ramp in your code before you send the command to the PID.

Also, the Jaguar uses an integrating velocity PID which is best tuned by using the I term first and then adding the P term, keep D term zero. There is a CD thread from 2011 "jaguar speed control only reaches 50%" - start with reply 34.

Re: Jags resetting in CAN
 

One thing thing that bit us the first year of using Jaguars was letting the battery get run down. If the battery is getting low, loading it can drop the voltage below where the Jags will reset.

If you are using CAN and settings he Jag to non default settings, be aware that browning out the Jag may cause some or all of the settings to be lost.

In speed mouse you should enable ramp rate limiting. It's been a while since I've looked at it, but you can do this through some combination of jumpers on the jaguar itself.

Re: Jags resetting in CAN
 

I believe that the Jaguar ramp only works in %volt mode - not in velocity PID mode. I suspect that is why they have an issue with PID but not in %volts. The user needs to add their own ramp before sending the value to the PID in the Jaguar. I confirmed this directly with the Jaguar programmer back when TI was providing the firmware. I suppose IFI could have updated that when they took over, but I have not seen any indication in the release notes that suggest they did. Certainly, there is no harm in adding a ramp on the cRIO even if there is one on the Jaguar.