Crazy Speed Controller Action

[Fred Agnir]

Here's a first for me. I've got a wild speed controller. It runs in reverse if I run both driver Jaguars in reverse. You'd think it was a programming error, but the program wasn't changed (and it worked yesterday), and it doesn't matter if I plug the PWM cable into any open PWM output on the Sidecar. In other words, if it's on PWM 3, it runs wild when the Jaguars are in reverse. If I plug it in PWM4, it runs wild when the Jaguars in reverse. As long as that PWM is open in the program and that Speed Controller is plugged into it, it runs wild when the drive Jaguars are in reverse.

To make it worse, an hour earlier, it was a different speed controller behaving similarly (it would run in reverse if either one of the drive Jaguars was running in reverse), so I yanked it and replaced it. Everything seemed to be going fine until at the time we all had to go, another Speed Controller started behaving the same.

Differences? We swapped/put the entire cRIO, sidecar, Jaguars onto the robot from the mock up. We've run the mockup with no issues for several weeks now. We no longer have all the cards in as we pulled everything that wasn't in the program. We don't need to keep all the cards in (like the Analog Card 1) do we?

More symptoms:

The run light isn't turning on even though we are seeing the LED blink on the Sidecar where it's plugged in.

The Speed Controller problems came after the cRIO disabled itself. It seemed to disable going into reverse with the drive Jaguars.

I suspect a short somewhere in the new wiring, but I just can't find anything. Are there crazy issues about this cRIO that I just don't know about?

As a 10 year veteran and the Controls Lead, and a career automation engineer, I can't believe how stumped I am.

Regards,

Fred Agnir
Team 126 Gael Force

[Mike Betts]

...As a 10 year veteran and the Controls Lead, and a career automation engineer, I can't believe how stumped I am...

Fred,

If the cRIO starts spinning or you see green soup oozing from the driver's station, leave the room immediately...

Seriously, there's not much I can think of telling you that you have not already tried but I'll look forward to hearing the end of the story...

Regards,

Mike

Postscript: I pulled all but the AIO in slot 1 and the DIO in slot 4 and ours works just fine...

[Alan Anderson]

All these weird symptoms can be explained by the Digital Sidecar not having a functioning 12 volt input. If it looks like everything is wired and plugged in properly, check the WAGO connections to make sure they're not grabbing the insulation of a wire that was stripped too short.

[Fred Agnir]

All these weird symptoms can be explained by the Digital Sidecar not having a functioning 12 volt input. If it looks like everything is wired and plugged in properly, check the WAGO connections to make sure they're not grabbing the insulation of a wire that was stripped too short.

It does seem like that. 12.95 volts at the SideCar, however. I'll put the SillyScope to it to see if it's clean.

Hopefully, we see something so obvious today (like when someone dropped a standoff and it rolled and tucked itself onto Dig I/O 11).

Fred

[Russ Beavis]

You might want to check out the FRC Diagnostic Window application from Think Tank. It's a very easy-to-use cRIO app with PC "front panel" to provide low-level control of each I/O.

Russ

[EVanWyk]

By speed controller, do you mean victor or jaguar?

It sounds like you know what you are doing: can you check with your oscope the signal going to the wayward speed controller? It would be a nice simple way to see if the fault is up stream or down stream.

[Fred Agnir]

By speed controller, do you mean victor or jaguar?

It's the Victor. And as of this moment, it's all Victors. We've got four wayward Victors doing very repeatable motions, regardless of how which PWM signals we plug them into!!!

It's the same related issue that you've responded to on e-mail to our team today.

It sounds like you know what you are doing: can you check with your oscope the signal going to the wayward speed controller? It would be a nice simple way to see if the fault is up stream or down stream.

I could check, but it doesn't seem to be PWM dependent. It's Speed Controller dependent. In other words, if Roller One - Speed Controller plugged into PWM 6 goes forward when I run the drive Jaguars in reverse, that same Roller One - Speed Controller will run forward when I run the drive Jaguars in reverse if it's plugged into PWM 7, PWM 4, PMW 3, etc.... Same with the other three Speed Controllers! They will run (forward, backward, or slow speed) the same regardless of what PWM signal they're plugged into!

Fred <~~~ wits have ended

[Russ Beavis]

FRC Diagnostic Window

[Fred Agnir]

FRC Diagnostic Window

Just to give an update..... Everything is running fine as of right now. In all of our testing, one of the Jaguars decided that it had had enough. It smoke up and honestly, I can't find why it would have. We replaced it and as of this moment, all systems are behaving normally.

It doesn't make me very happy, since at this point, I didn't know what caused it, and I don't know what we did to fix it.

Fred

[EVanWyk]

Just to give an update..... Everything is running fine as of right now. In all of our testing, one of the Jaguars decided that it had had enough. It smoke up and honestly, I can't find why it would have. We replaced it and as of this moment, all systems are behaving normally.

It doesn't make me very happy, since at this point, I didn't know what caused it, and I don't know what we did to fix it.

Fred

Fred -

I was talking with Mike Copiolli yesterday about a nearly identical issue - are you on the same team?

He had 7 victors working perfectly, but adding an 8th cause them all to fail if driven backwards. Removing any single one fixed it, and it didn't matter what PWM outputs were being used.

It turns out that he had plugged in the DSC's power backwards, which is the equivalent of not plugging in power at all. This means that the amplifiers were getting ghost powered by the signals from the FPGA leaking through protection diodes.

Here is the cool part: Since reverse is a short pulse (and forwards is a long pulse), going in reverse ghosts less power to the 5V rail on the DSC. This explains the directional dependency.

Double check that the power is actually getting to the DSC. It is possible that you have it reverse biased OR not properly stripped.

Cool, eh?