I suspect I have a bad Jaguar, but thought I’d post here just in case.
I’m experiencing a “fault” condition on one Jag that seems to be hapenning too often. By “Fault” I mean that the Jag is shutting down and the LED blinks slow red.
I’ve got two Jags side/side driving two Swerve drive wheels (in a six wheel setup with the Swerves in the middle and two omni’s at each end). Wheels spin freely, robot is only loaded at about 70 lbs, and there is no perceivable different between the two wheels.
While test driving I noticed that when I changed direction sometimes the robot would go into a spin.
I nailed it down to the fact that one Jag is not liking switching from Reverse to Forward suddenly. When the direction changes from R-F the error trips and the wheel shuts down until I stop the robot. The weird thing is that the error NEVER trips when I replicate the action in the opposite direction (rapid Fwd to Rev change.) It also doesn’rt seem to happen when I accelerate from a stop.
I considered battery voltage so I swapped in a new battery with a solid 12.75V No change.
It doesn’t always happen, but seems most common when doing the rev-fwd change after a SHORT period of driving.
It’s as if one set of FETS (1/4 of H-bridge) are less capable than the others, and are causing an error condition when they engage.
Tonight I’ll probaly have the elec. student swap out the Jag, but is there anything else I can try to fix this?
Slow flashing red indicates one of the following fault conditions:
Power supply under-voltage
Over temperature
Over current
Limit switch activated in the current direction of motion
also,
When a fault condition occurs, the motor shuts down and the LED indicates a fault state during the fault condition and for 3 seconds after the fault cause is cleared (except for the limit switch fault, which is cleared instantly). A slow flashing Yellow LED indicates that the MDL-BDC24 is not receiving a valid control signal.
So, lets walk through it… you replaced the battery with a known good one, so it’s likely not under-voltage. You said it usually happens after a short period of driving, which makes me guess it’s not over temperature.
Can you check the limit switch jumpers? Is it possible one of them is loose, and jarring the robot in that direction is causing it to slip out of place? i figure this is probably unlikely, but you never know…
That leaves us with over current. Switching directions from full forward to full reverse (or full reverse to full forward) makes your drive train have to work incredibly hard - you have a lot of mass moving in one direction, and now your trying to pull it in the opposite direction. If you have high traction wheels, or some slop in your drive train, this can cause the motors to stall - pulling up to 133A for a CIM motor. I would take a close look at your drive train and ensure there’s no slop or any way for the motor to stall on that side. Depending on your drive train setup, it may only be binding in one direction, which is why you can go fwd-rev just fine, but rev-fwd is a problem.
I’ll double check the jumpers and drive wheels, but I just can’t accept that the driving forces I’m putting the robot through should be sufficient to trip the over current fault. If so, then other teams are also going to have a lot of trouble. It just seems to happen to fast, too easilly.
I don’t suppose the current limit is configurable (and mine got set too low).
Going along with the over current thing, The jaguars will hold 60Amps for 2 seconds, its possible that you are exceeding the current limit on the jaguars and they are shutting off. If that is the case, then they should start responding to inputs, and the LED should stop blinking after 3 seconds.
I wondered what the hold time was.
The fault is occuring well before 2 seconds elapses.
eg: I’ll be driving along at 50% power (very low load) and then do a snap reversal. The left motor cuts out immediately, causing the robot to spin, rather than reversing.
Hmmm.
The LED does stop blinking after about 3 seconds, but that doesn’t indicate the specific cause as a current limit.
The current limit is not configurable, but the fault shut-off time is. You can set the fault shutoff time as low as 500 ms.
I am not familiar with the proper WPI Library calls to set this parameter, maybe someone else can comment.
The over current protection uses a timer to count how long the Jaguar is operating above the limit. This timer follows an exponential equation that intersects at 60A and 2 seconds. Above 60A the Jaguar will shut off at exponentially smaller time periods. For >100A the time until shutoff is quite short.
My first thought would be rather than just “swap out” the Jaguar with one off of the shelf, I would be inclined to physically remove the suspect Jaguar from the robot (disconnecting all of the wiring first) and then move the suspect Jaguar to the position of the other swerve Jaguar that is working correctly, swapping the “good side Jaguar” with the “bad side Jaguar.” When doing this, the idea is that all of the cabling (power input, motor outputs, CAN/PWM, and encoder/limit switches, if relevant) do NOT move with the Jaguar. You want to move the “suspect Jaguar” to the position of the “good Jaguar” and use all of the cabling for the “good Jaguar.” Similarly, put the “good Jaguar” into the position previously held by the “suspect Jaguar” using all of the cabling from the “suspect Jaguar.” You can then retest the robot with the “suspect Jaguar” and the “good Jaguar” swapped.
I see four possible outcomes from this, each of which would tell you something about what is wrong:
A - the problem moves to the other side of the robot: this is almost surely a bad Jaguar. Contact TI to RMA the Jaguar (assuming it is a black Jaguar).
B - the problem stays on the same side of the robot: the problem isn’t with the Jaguar at all, but is being induced by mechanical or cabling issues
C - the problem goes away completely and both Jaguars work fine in their new positions: there was likely a mechanical or electrical cabling issue (loose connection) that was fixed by the act of unconnecting and reconnecting everything
D - neither Jaguar works: either the disconnect / reconnect was done sloppily and there is now a loose wire/cable on both sides, or there was some sort of mechanical/electrical issue on the “bad side” which had induced a fault in the suspect Jaguar and has now induced a fault in the previously good Jaguar, too.
In any case, I’m curious to hear the results of your troubleshooting, whether or not you end up trying the swap I describe
I’ve been experiencing similar symptoms and am trying to isolate it to the JAG. Strange symptom goes away when new JAG installed. Has happened once over the summer and am now isolating two different JAG problems in different functions in robot…both fixed when JAGs replaced…very frustrating. Will confirm the fault and RMA soon.
Phil,
What motor is this Jag feeding? Since you mention swerve, I thought that maybe it could be window motors controlling the steering. Window motors have a known issue with the locking pins that may be removed this year.
If not is one side throwing a loop in the chain that gets jammed momentarily and locks the transmission? This would run the current to stall on that motor.
Yes, in hindsight that would have been a usefull diagnostic test. Unfortunately I didn’t see this in time.
I did check all the wiring for obvious problems but didn’t see any (I re-wago’s all the connections).
We didn’t replace the bad Jag with a new one, since we didn’t have a spare black one handy, and we may want to try using the CAM at some point, so we moved the adjacent one to the failing position, and replaced the moved one with a grey Jag that we had around.
Prev: BJ1, BJ2, BJ3, BJ4
New: BJ2, GJ1, BJ3, BJ4
This did in fact eliminate the problem.
We drove it hard for a while and did not see the fault re-occur.
(note: this entire multi day process was performed with the wheels locked straight, with no actual means to rotate the swerve mechanism, so this was not a factor)
We’ll keep and eye on this (and may in fact put the “bad” Jag back in pos 2 to verify), but I’m 95% convinced at this point that it’s a Jag fault.
Glad to hear that your swap worked and that you’re off and running again. I would highly suggest that you follow the RMA procedure to return the black Jaguar to TI, as described at http://forums.usfirst.org/showthread.php?p=44255 I’m curious to hear what the problem ends up having been.
I am trying to get this info on the jaguar RMA. I tried the above site and the providedlink seems to be broken. is there any other way to request an RMA?
bad jaguars we lost three out of four in the last few days and last night we realized the crio was recieving twice has much power has it suposed too could are jaguars been because they were recieving to much power?
hydro,
There is no way for you to supply twice as much power to the Crio. It should be connected to the +24 volt output connector on the PD. This is the highest voltage on the robot. Everything else is either 12 volts or below. The Jaguars receive no power from the Crio.
