Mk4 x-axis motor firmware bug? Lost steps?
I've been printing with my Mk4 for 3 weeks now. Most prints are perfect. Last night I loaded up the tray with some gridfinity boxes and noticed this morning that the entire print had shifted to the left. It's a satin sheet w/ Overture PETG - which I've used for all my prints just fine so far.
Is this a firmware or slicer bug? Lost steps on the x-axis motor? Mk4 is running firmware 6.0.0+14794. Nothing in the 6.0.1 firmware release notes sounds like it addresses this.
At 3:03:53am on the burned in time code in the upper right of this video, you can watch the x-axis motor just freaks out and shudders a bit. It isn't on an extreme end of travel when it does this - its right in the middle of the axis. And then from that point on, the x-axis is printing everything about 1 cm to the left of where it should be.
This happened once before, but I didn't have video and it was just a gridfinity grid which is all connected and I just assumed I had layer adhesion issues or something that caused the shift.
RE:
I guess I can't edit the post anymore for some reason.
Note that after the x-axis freak out event, everything behaves perfect again for another 7+ hours until I discover the issue and stop the print.
Control > Calibrations & Tests is all green check marks. I did an XY axis test after stopping the print and it all passes.
Across ~50 prints, this has happened 3 times now (and it was only this last time I had video to figure out what happened). All 3 times the print bed has been pretty full of stuff so its really annoying to lose all that filament.
RE:
It looks like the nozzle collided with one of the printed objects, and the offset caused by lost steps from the collision.
The video resolution is not good enough for me to see clearly, but it looks like the corner of that object (where the collision occurred) is either raised slightly (adhesion or warping issue?), or there is a "blob" there that was deposited earlier in the print. I can see a bit of a shadow at the left-front corner of that object, where the collision occurred, that makes me think it is raised slightly off the bed.
It *could* also be a corrupted gcode issue, but that would likely result in more than that single collision. How are you transferring gcode to the printer?
Inspect the parts (particularly that front object, second from right) after removing from the bed. See if there are signs of that corner lifting, or signs of exactly where the nozzle collided. Look for any kind of deformation or "blob" there as well, if it hasn't been fully covered up by the successive layers.