OK. The 8825s did NOT like that G0 speed, lol. Tweaked it all down to about 1000 and ran it on the Y test, dropped right back into the correct origin. I’m going to run the full square now, but I think there are only a couple explanations: either some sort of systematic noise on the signal wires between the control board and the drivers or bum drivers.
I guess I get some wire and start with the short runs. If that doesn’t fix it re-wire the whole thing. If that doesn’t fix it get new drivers. Mostly because that’s the order of cheapness.
I am almost positive I got a whole batch of drivers that had incorrect steps. I have no idea how that could happen, but I remember z steps needing to be like 786 instead of the usual 800. Other than that they worked perfect I just swapped them out ~5 years later.
The full square pattern finished, and went to the correct origin. I really hope it isn’t the drivers, because that’s gonna be tough to understand. Some pattern or combination of signals creating a meta signal of interference…I can wrap my head around that.
Probably won’t hear from again me until I get that squared away. In the meantime, thanks a TON for the help. I was ready to take a sledgehammer to it.
Maybe it is something like the uno doing step, step, switching the dir, step, step and the driver seeing either step, step, step, dir, step or step, step, dir, step. Is there anything you can do with timing in the drivers to slow it way down to see if that helps? I know marlin has a driver type, which ends up as some specific rates and delays.
Otherwise, remember that there is software (firmware) running on the drivers and there could be mistakes in that code…
The only thing I can think of is changing the microstepping. That did have an effect. I’m really hopeful that it isn’t drivers, because running just the X direction gave me no error, and running the Y gave me all the error. Like maybe one or both of the Y are interfering with the each other AND the X?
OK, I tried twisted pairs and a little more space, no effect. A friend came over today (who has experience with external drivers). He agreed that the machine is in good shape mechanically, and there is no way to explain it through any sort of slop and donated some shielded wire to the cause.
We re-wired the dir/step with that, hooked the shield into an unused endstop ground, and about 10% of the error disappeared.
I also swapped the y2 and z drivers, no extra change, then the x and y1, no extra change. So now neither of the drivers originally on the y is on the y axis. Indeed, one of them is now on the X and the other is Z, but I have the same error as before swapping them so it’s even more difficult to imagine the drivers being the problem.
My friend left me some more shielded cable for the steppers. I think that’s the only thing left.
On a happy note, the wife’s computer just had a whole mess of dust in the CPU fan and the heatsink was WAY hot when I got it to shut down. Easy fix, except I didn’t have any thermal paste. COLGATE TO THE RESCUE!!! Amazon got some thermal paste to me a couple days later and it’s been sitting on my desk because the toothpaste worked fine, so maybe I’ll go finish that up and pretend I did something good today.
@jeffeb3, @vicious1
Looks like it was garbage drivers. I got my new set installed and finally got the result I expected. Thanks a bunch for talking it through with me. It was the first thing I found that made sense, but because I’d gotten such strong suggestions that it was a wiring problem, it was tough to just ORDER them without trying the other things first.
I really do appreciate both of you taking the time to noodle through this. Really helped me to feel better about the other stuff not helping, even though I knew it should be done anyway. Kept my frustration way WAY down. I’m chalking these things up to faulty components or a bad firmware flash.