How to correct slight/severe skew
After making some changes/upgrades, my xyz calibration went from being perpendicular to being "slightly skewed".
As I understand it, the printer should be able to automatically correct for this skew, but I'd still prefer to get it as physically perfect as possible.
So how do I actually go about adjusting or correcting the skew? I've seen others mention "adjusting the z axis" but I don't know how to do that.
Re: How to correct slight/severe skew
great question I would also like to know sinc it has been impossible to build the frame without a wobble prua chat was unable to help
The Latest Firmware can be found here https://github.com/prusa3d/Prusa-Firmware/releases
Open Firmware Issues https://github.com/prusa3d/Prusa-Firmware/issues
Re: How to correct slight/severe skew
Have you tried switching it to the opposite end or rotating it 90 degrees either way with another side up? Remember to tighten in pairs and not to over tighten as the directions talk about.
Re: How to correct slight/severe skew
Have you tried switching it to the opposite end or rotating it 90 degrees either way with another side up? Remember to tighten in pairs and not to over tighten as the directions talk about.
Switching what?
Re: How to correct slight/severe skew
David means that you can turn around the Y-axis front and rear plates to see if it helps.
When you screw the YZ body together, it is extremely important to do this on a flat surface.
Load the body with weights or ask someone to squeeze down the four corners when you tighten the screws. The frame must not tilt when it is on a level surface (see the manual).
Press all plastic parts downward to mount so that the right and left ends do not end up at different heights.
Be sure to remove excess filament on details that may affect the location. Then calibrate XYZ again.
Bear MK3 with Bondtech extruder
RE: How to correct slight/severe skew
David means that you can turn around the Y-axis front and rear plates to see if it helps.
When you screw the YZ body together, it is extremely important to do this on a flat surface.
Load the body with weights or ask someone to squeeze down the four corners when you tighten the screws. The frame must not tilt when it is on a level surface (see the manual).
I did all that. I'm assembling the kit now. In spite of doing everything carefully according to the instructions, assembling the frame on a granite countertop polished to an optical mirror finish (can't get much more flat than that), tightening gradually the diagonal screws, my frame ends up so that one corner is 0.6mm off the table. I have tried reversing the end plates and swapping around the extruded aluminum, checking for burrs, and so on. I don't know what else I can do.
RE: How to correct slight/severe skew
A better question is what are the skew numbers in the LCD display telling you?. They don't have much to do with any tilt issue introduced by the feet not being true and level.
RE: How to correct slight/severe skew
I've still just assembled the frame only, have not gotten to putting together the electronics yet. I'm just concerned that the frame wobbles because all four corners don't rest on the table. If it was a tiny fraction of a millimeter, that's probably OK, but this is over half a millimeter. Even when I put on the rubber feet, the wobble is the same. I guess there's nothing to do but continue with the build and see how the calibration works out.
RE: How to correct slight/severe skew
My frame wiggled. Even a strong arm didn't help. Yes - I tried to bend it flat. My skew numbers are 0.03 or some such... but that has nothing to do with the frame not being flat. If anything, the frame flatness affects bed mesh level. One corner being high means the bed in that corner will be "low" ... but isn't that what the auto-level mesh code is supposed to deal with?
If you want perfect flatness, there are 8 different positions for each of the frame extrusion pieces. If its only the one "arm" of the frame that is low, try all 8 orientations. If they all result in the same offset, then swap section with the other side and try the now 16 variations you have to play with.
If still not satisfied, begin with the rear arms. In all, there are only 64 front arm variations and 64 rear arm variations, times two; or about 8,100 possible combinations to try. It's doable to try them all.
RE: How to correct slight/severe skew
@tim-m30
Yeah, I did that calculation too, and came up with 4096 variations (8^4) so I settled for turning one bar 90 degrees and flipping another end to end, and reassembling, to find the exact same result, convincing me that it's the frame and end plates, not the extruded bars (the end plates seemed to make the wobbling worse).
Oddly, as my son and I are progressing in building this printer, the frame seems to be straightening out as we screw more stuff together. It no longer wobbles. We have the X,Y,Z axes all done, about to start on the extruder assembly. It's slow going because my kid isn't accustomed to using tools and I'm letting him do as much as he can while supervising and teaching (and tightening after him, since he lacks strength). I think I could get the build done all by myself in about 4-5 hours, but so far I've spent about 8 hours with him on this and we're barely half done.
We have been consuming the gummi bears strictly as instructed, though. 🙂
RE: How to correct slight/severe skew
The frame is not intended to sit flat...
The body of the frame is the lowest point. The arms are up a bit.
What you want is for the arms to be parallel. Have them rest on a flat surface with the body off the edge.
Tighten the screws with the arms flat and parallel. Then do the other side the same way.
RE: How to correct slight/severe skew
@robert-rmm200 - the frame should sit flat with the end plates on, with the feet on, or with equal size blocks under it. Without the end plates, two legs should touch the table; you shouldn't be able to wobble it.
Good tip about resting the legs on a flat surface while tightening. Thanks. I've finished with all the mechanical assembly, so I'm not inclined to go back now that the wobble seems to have settled out as the assembly progressed. The real test will be in the skew measurements during calibration.