Alright, found the issue. I tried a bunch of things, contacted Russ Sadler for some troubleshooting ideas, and in the end I stumbled upon this thread:
https://sawmillcreek.org/showthread.php?269029-Laser-Slowly-Loses-Steps-in-X-Axis-throughout-project-or-quot-drifts-quot
Basically my X and Y were set differently for the PWM edges, one was set true and one false for rising edge valid. Once I set X to false like Y everything works great.
Interesting - I see that semi-often with lasers where people replace a motor driver, or installed the controller themselves, but it almost always shows up as an issue with the scans, not cuts. Thanks for following up - one more thing I can stash in the vault. 
I suspect now my scan backlash settings are probably forked and I’ll have to find new ones, which is no problem.
I’ve not replaced anything like that, but when I did get my laser it was miswired with laser on going to water protect, so its not that surprising there was a setting mistake in the firmware.
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.