Motors spinning while robot is disabled

[jambrown1397]

Our robot's wheels would randomly spin while the robot was on, but disabled. It obviously makes working with the robot significantly more dangerous. We would like to prevent this problem in the future. Any ideas on what caused this?

Additional details:
We were using Victor SP's attached to CIM motors. We tried to re-calibrate the speed controllers, but sometimes the calibration wouldn't take, while other times it would. We put our robot on a dolly and most of the time the wheel spinning could be activated by a slightly hard tap on the frame. Thanks.

[AdamHeard]

This shouldn't be a calibration issue, as when disabled no pwm signal is sent at all.

Is it both sides of drive or just one?

Any other systems?

Is there a chance metal shavings got inside where the signal pins are?

[jambrown1397]

Mainly the left drive system, but occasionally the right. We checked for metal shavings, using an air canister and vacuum, but it is possible that we missed some. I don't know if it was an issue on our other system (a lift powered by a CIM) because the gearbox the CIM was attached to was geared for high torque (was very hard to turn even manually).

[GeeTwo]

Another place to check for intermittent shorts is in the PWM cables. Look especially for pinch points, tight corners, and damaged insulation between the ground and signal lines.

[OccamzRazor]

Grab a multimeter and start checking for current and continuity. Something somewhere is making a short and supplying current when it should not be from the sound of it. Check if it stops when you pull out the pwm cables from the speed controllers, if it does then check the cable, or speed controller and replace it to see if the problem repeats.

If that does not work start testing wires and systematically checking the wires from the distribution board to the speed controllers. Comparing amperage from those wires should tell you which component or wire is the problem.

Stay safe and be careful with that problem!

[Jon Stratis]

What are the lights on the speed controllers doing when this is happening? That can tell you a lot about where the problem is - if the lights show movement, then the problem is likely somewhere in the control path - a shorted PWM or some shavings shorting something on the roboRIO pins. If the controllers show no signal (flashing orange), then the problem isn't in the control wiring - look for a short between the controller and the motor, compounded with another short elsewhere that's allowing the motor to get power around the speed controller.

[scca229]

Watching this thread with interest since the same type of thing was happening to us:
http://www.chiefdelphi.com/forums/sh...d.php?t=134694

Moving to different PWM channels was the only permanent cure. Position 0 causes all kinds of activity in Disabled mode, Position 1 about half the activity. Positions 2 up have no issues.

Reminds me that we still need to get that RoboRIO off to have it looked at.

[Greg McKaskle]

Are you using Y cables for the PWM connection? Either way, I'd follow the PWM cable from the roboRIO to the motor controller pinching and jiggling it to see if you can make this happen. I've generally seen it at the Y-merge of the Y cable, and rolling in your fingers will make some motor controller see false edges and do a short (100ms) or less glitch. When tracing the cable, be careful not to be near chain or other pinch points. And I'd pull breakers on all but one circuit.

Greg McKaskle

[page2067]

We also found it with Y cables at the junction - several of one style of Y cables have exhibited the issue -

[jambrown1397]

We didn't use any Y cables, and I don't think there were any kinks, but I'll check for the continuity between the wires and look for shorts. I don't remember what the lights were displaying at the time I'll also check for those (though theoretically they should blink showing that they have no signal right, since its disabled). Thanks for all the input so far though!

[mastachyra]

Could it be as simple as crossed polarity on the power to/from the motor controller? I recall this happening before.

[Tom Line]

Originally Posted by mastachyra

Could it be as simple as crossed polarity on the power to/from the motor controller? I recall this happening before.

Yes. We've found that even when disabled, if you short a motor controller output together or to ground on the frame it can cause spontaneous motion on any of the other connected speed controllers. That was a rather scary 'find' as our head mentor was walking by a can whip when it came down in front of him at full speed.