Welcome to the SeeMeCNC Community

I have checked my belts, cheapskate tension, just replaced steppers with no improvement, and have adjusted stepper driver current to ~1amp (should be more than sufficient for x/y/z). I continue to get what appears to be skipped steps on random prints, but sometimes if the print continues for any length of time it appears to self correct. The offset is typically about 1mm, not much more, sometimes less. Initially I thought there was some issue with the nozzle colliding with warped overhangs, but the issue occasionally happens on parts with perfectly straight walls (cubes, cylinders). I am running conservative speeds (30mm/s, 1200mm/s^2 accel, 35mm/s^3 jerk). I have the TrickLaser arms which I have confirmed have no friction in their joints, my cheapskates fall freely when released. The only issue I can imagine is belt tension. Right now I have the belt tension fairly loose since if I make it any tighter a single tower gets an uneven friction with every rotation of the pulley. I believe the pulley is possibly warped. This symptom was what led me to believe I had bad bearings in a stepper motor when in fact they were fine. When I initially built my machine I had no such issues, I occasionally had the nozzle collide with warped overhangs which resulted in an unsightly print but never any issues like these skipped steps. The fact the issue also occasionally appears on prints with nothing but perfectly vertical walls leads me to believe there are other issues here. I just ordered replacement pulleys as well, hopefully that will help somewhat but I am not extremely hopeful for that solution either. I have checked everything I know to check and continue to get issues. The steppers are new so new crimps and ends on them, no extensions were required. The stepper drivers are DRV8825 drivers (CoolDrv) on RAMPS 1.4 adjusted to 1.1amps per motor, the steppers are Kysan 5.5kg/cm NEMA17, the same thing that was removed except these have flats ground in them. One of my pulleys lost the threads on a setscrew the first time I removed one of the set screws when assembling the kit (they shipped with screws already in, threads galled when removing set screw for first time).

I remember reading about another individual who had a nearly identical issue although much worse here:
http://forum.seemecnc.com/viewtopic.php ... teps#p2730

There was also another thread I read that had a similar problem that had no solution in the thread.

EDIT:
This thread along with graphics most closely describes my problem: http://forum.seemecnc.com/viewtopic.php ... 602#p25594

I typically only get an offset towards the right when facing the front of the machine, though. The right tower (Y?) is also the one with the questionable pulley installed.

Also just realized, I didn't specify that I have a Rostock Max v1.

I thought I had fixed the issue. Made some beautiful prints the last few days, but now I printed a part, no overhangs or anything to warp and the print randomly shifted away from the x-tower in the middle of the print. Last night I had a print fail with what looked like an extreme x-tower shift (big ball of filament shifted directly away from the x-tower), but the machine had homed at the completion of the print so I don't really know what happened. Stepper current is set at ~1.1 amps for X,Y,Z running new Kysan steppers on DRV8825 drivers on RAMPS 1.4. My endstop wires are now running independently in their own cable down the side of the machine, no longer through an extrusion. Belt tension is certainly not too tight, cheapskates fall freely with belts loosened and arms removed. The carriages move very freely even with the belts tight and the effector attached. I'm almost certain this is an electronics issue now, not a mechanical one.

Has anyone had a microswitch fail intermittently? Perhaps the short bit of cable inside the base of the Rostock Max where the stepper wires are near the endstop wires is still inducing enough current to trigger the endstop? Perhaps the driver is failing intermittently?

You didn't mention if you checked the pulleys attached to the stepper shafts. Those little setscrews can loosen and cause exactly the type of shift you describe. My money is on this or another mechanical issue over an electrical issue. You'll usually hear skipped steps, especially if its that much.

I did in fact have a loose pulley. I had checked them but either I forgot the one in the back or where it was on a flat it felt tight but wasn't actually tight. This time I removed it along with the motor and made sure it was lined up on the flat and is tight. It wasn't on the tower I thought it was. It was on the back (z?) tower and it was loose enough to just shift slightly back and forth on the flat, I thought I'd got it completely tight when I replaced replaced the motors. Either I did and it loosened or perhaps I didn't get it completely tight. I used offbrand Loctite on them before which was a mistake, it only half-cured but did keep them tight, made for a mess when I took them off though and made threading them back in difficult. I didn't use anything this time, but if I have another problem I have genuine blue loctite I'll use on them.

Thanks for the help mhackney. I also appreciate you posts in the PEI thread, I just got my PEI sheet in and taped down a couple days ago, it is quite possibly the second best investment I've made in 3d printing (RoMax is best investment).

I realize you already found a problem, but this is a key statement.

NETio wrote:... it appears to self correct. ...

Skipped steps due to electronics issues almost never correct themselves. Oscillating or self-correcting symptoms are virtually always mechanical in nature.

NETio, great, glad it was simple. The "self correct" comment is what clued me in too. Belt slippage can also self correct but more rarely.