Ideas for skew correction on a Core One
Using the Califlower I have measured a XY skew of -0.82° on my Core One. Doesn't seem like a lot but I can imagine it becoming significant on very large prints. It's not enough to cause problems with Y calibration. I can possibly correct it by unevenly tensioning the belts but it would be a tedious trial-and-error process.
From the software side, it seems everyone (except Prusa) has a way to calibrate and make the correction in firmware. In Marlin this is M852 which is not implemented by Prusa. One alternative is to use something like this: goskew to post-process the g-code.
Does anyone have a process they recommend? I may give the goskew script a try.
RE: Ideas for skew correction on a Core One
If my trigonometry is not failing me, 0.8° skew amounts to a 3.5 mm position error in Y over the 250 mm bed width in X. That seems excessive -- I am sure you do not have that big a deviation (gap) at the gantry end stops in the front. What's going on there?
RE:
If my trigonometry is not failing me, 0.8° skew amounts to a 3.5 mm position error in Y over the 250 mm bed width in X. That seems excessive -- I am sure you do not have that big a deviation (gap) at the gantry end stops in the front. What's going on there?
Don't know to be honest. No gap when the belts are slack, no gap under tension. X,Y and homing calibration pass. I would not have found it had I not explicitly measured it. I may have introduced it by very slightly unevenly tensioning the belts. IDK, maybe the front of the frame is not orthogonal to the sides?
I'm starting to get annoyed again with this printer.
RE: Ideas for skew correction on a Core One
@Hyiger: is it actually messing up prints, or is just a measurement you stumbled into? I mean, if prints are OK, and calibrations pass, how do we know there is a real problem? Just curious.
Cheers,
Neal
RE: Ideas for skew correction on a Core One
I printed various geometries then did several measurements over X and Y and calculated the skew. Then I found: https://vector3d.shop/products/califlower-calibration-tool-mk2 which gives a more accurate assessment.
Easiest way to check is just print a large square. If X and Y are different then there is skew.
RE: Ideas for skew correction on a Core One
Can't edit my posts now.... @!$#%@!
I also wanted to add that I discovered this after a rather large L shaped print had a screw hole misaligned.