Extreme XY Shifting for Large Square Prints
When printing left to right on the diagonal, I am getting repeated large (5-10 cm) offset errors in a sawtooth pattern.
The closest I've come to observing it so far (although I'm planning on observing it now) is a high pitched sound and some crunchy, gritty noises (making me suspect the Z-axis not going down a tick...only it happens in the middle of a layer), whereupon the head is off by 10 or so cm.
The error manifests in identical places on distinct prints. However, it will randomly print successfully on occasion.
So that rules out a problem with the GCode. I've tuned the belts to 80hz several times, so the belts are tuned.
I've plugged the printer directly into the outlet, so I've ruled out my batter my backup/power supply.
I am printing with a maximum acceleration of 2000 mm/s2, and a default of 1250, so I've ruled out acceleration.
Now, I am researching the topic and am running into troubleshooting steps related to "grub screws" ("set screws"), couplers, shafts, pulleys and a lot of motor jargon. I imagine this language amounts to one or two screws near the motor assembly.
However, it's unclear where those screws are in the XL enclosure.
Another potential option is to upgrade the firmware, and check whether the motors are getting too hot (seems unlikely, as the error occurs only 2 hrs in, and I have accel dialed down to 40% of the max).
Before I started blunderbuss approach of fixing many things in the hope of fixing my one problem, I'm posting this image of a partial print (my son canceled it accidentally), plus the failures. All subsequent prints so far have all failed at the same point.
Has anyone else seen this problem, or have any recommendations for a fix?
RE: Extreme XY Shifting for Large Square Prints
The initial offsets are all virtually identical.
So, from my perspective, if this were a mechanical problem, I'd expect them to be noisy, even if they start at the same spot.
But hoping to catch the issue in real time today.
RE:
Seems the solution was either to regenerate the GCODE, or move the print a centimeter out (away from the extruder dock). Unfortunately I did both at the same time.
I am leaning towards moving it being the final solution, as I found one of the orange clips unclipped.
It seems as if the extruder wires from extruder 1 were getting caught on the orange clip of extruder 2 when making a diagonal move from the back left corner to the front right corner.
All of the wires are clipped in correctly now, so this could be a quirk of the machine, or the orange clip became unclipped from the bodum tube incidentally, and this was all a symptom of that.