Klipper restart borks A and B steppers

[worthy gull]

Upon restarting klipper, the A and B steppers start acting up. It's already caused one crash on the build plate while homing (basically it went negative Y instead of X when moving to the "Safe Z" location), and now every time I adjust the printer.cfg whether automatically or manually, I have to shut the machine down and wait for cap discharge before it acts fine again.

And by fine, I mean it can print a test cube without a single issue.

I'm thinking that I need to reinstall the firmware, but if there's another option I'm all ears. 🙂

[woeful hull]

can you define/expand upon "acting up"?

what's changed since this was last working properly?

if anything it sounds more like an electronic issue rather than firmware

[wraith sun]

have u seen the reply above?

[worthy gull]

Sure. Acting up: A and B motors are fighting each other. They sound like they're missing steps, and the X and Y axis movements are not as I'd expect them to be (jagged diagonal movements while homing, not properly leaving the homed position for the second click, moving toward the front of the machine when it should be moving to the left - towards the B motor).

That said, I did see something that I need to check, and I'll feel a bit dumb if it's the case. I moved the printer, and there may be something pressing against the back plexiglass panel of the enclosure (like a power cable thats shifting occasionally). And that may be interfering with the belts since they run so closely to the back panel. I'll report back.

@wraith sun Thanks for the ping. 🙂

[gray ridge]

so here's the thing: it's literally impossible for A & B to fight each other. There's no combination of motions where they are actually in opposition. Any possible combination of motor actions just results in motion in some given direction

[worthy gull]

Yes, I'm aware of that - and fighting was probably the wrong word. One motor is not moving consistently to move the carriage horizontally, creating jagged diagonal movement. This is not the kind of movement I'd expect when homing the X axis. Also, when in the X and Y home position, if one motor is not moving, it could explain why the carriage would move on the Y axis in a rough manner (skipping on the opposite motor), not the X axis.

[worthy gull]

It wasn't the enclosure plexi; I removed that entirely. Running through more checks since its happening more often and not just after resets. My next hypothesis is a bad servo. I'm hoping it's not the stepper driver, since it's a single-piece Leviathan board.

[woeful hull]

can you upload a klippy.log for us to look at?

[worthy gull]

I was hoping to get a clean log, but it stopped doing it, so here's a... not clean log.

[woeful hull]

nothing obvious in there

[worthy gull]

Well, I'll keep plugging with it, and if I can reproduce it again I'll try and get a video. Thanks to everyone for your help and time.

[gray ridge]

what main controller is this?

[worthy gull]

Leviathan, with a Raspberry Pi 4.

I got the 300mm LDO kit from West3D and assembled that.

[gray ridge]

given the fact that the problem seems to go away when you let it sit & discharge, I'd be tempted to change up drivers. but that's not so easy on the leviathan. I guess you could at least switch A vs B (wiring & config) & see if the problem migrates to the other side...

[worthy gull]

Yeah, if it kicks back up that's not a bad idea. At least they're fairly easy to swap out... 😄

[grand ember]

This sounds like when I had one of my coil wires back out of the crimp housing. Have you looked closely at all motor connections?

[worthy gull]

As much as I can, yes. I've also dis- and re-connected all of the wiring to look at them and verify them. I did not tear into the motors or board.