Skew Correction Issue
I got two parts that are the same. one is cnc machined and the other 3d printed. The 3d Printed one doesn't align with the cnc one. I suspect a skew issue I tried to set it up in RatOs, but it only improves a bit. Am i supposed to do it multiple times ? I did the Klipper method. If that helps. Or could it also be something else ike speed, acce settings or belt tension?
19 Replies
could also be shrinkage related. Have you calibrated your filament shrinkage?
ok, no i havent i actually never heard of it. do you have a link?
This would be the optimal way: https://www.printables.com/model/174910-3d-filament-shrinkage-calibration-test
But you could just print a 10 or 20cm rectangle, measure it and calculate your shrinkage that way.
for example if you get 99.2 mm instead of 100 your shrinkage is 99.2%. Depending on the slicer you either put that in or scale your model by 1/0.992
Thats how I did it for ASA, just be sure to correctly calibrate your extrusion multiplier before!
However I only got a shrinkage of less than a %, so not sure if the problem your having is solely caused by shrinkage, since you have quite a lot mor deviation between the parts
Printables.com
3D filament shrinkage calibration test by sprior | Download free ST...
This is a 3D printer filament shrinkage calibration object that is inspired by the technique demonstrated by Clough42… | Download free 3D printable STL models
Does the left side of your part line up? its hard to tell in the picture.
Maybe double check you skew profile? not that you mislabled the distances or something
yeah left side pretty much lines up. the right side is messed up and doesn't lign up. i had that issue once before and I fixed it with the Skew adjustment ,but this time it didnt work. I'll check again if I fucked up something
hmm could be that the skew correction was applied in the wrong direction.
Follow the guide again, I think you have to disable your skew profile, print the calibration part and then enter the values.
Maybe you printed the calibration print with the old skew profile and then overwrote it
sounds good I'll try it right now. thanks.
@oskait i checked again i printed the object again i meassure again and still the same it doesnt match
what calibration object are you using and which guide are you following?
Whats your printer?
@oskait im using the stl from the klipper documentation
Did you do the verification part?:
Verifying your correction¶
After skew_correction has been configured you may reprint the calibration part with correction enabled. Use the following gcode to check your skew on each plane. The results should be lower than those reported via GET_CURRENT_SKEW.
CALC_MEASURED_SKEW AC=<ac_length> BD=<bd_length> AD=<ad_length>
did you set up loading of the skew profile in the start gcode?i just tried it and it said malformed. it should be like this right CALC_MEASURED_SKEW AC=<11111> BD=<22222> AD=<3333> ? thats what i found a bit confusing do they mean the start gcode in my slicer or in klipper ?
just to be sure i also measured my frame and i got +/-1mm difference max from the distances
if i try to use it says this. Error on 'CALC_MEASURED_SKEW AC=140.86, BD=140.99, AD=99.7': unable to parse 140.86,
because that is not what klipper is expecting
https://www.klipper3d.org/Skew_Correction.html#configure-your-skew
there are no
, in the example thereoh man thanks. well the results are lower, but the part is still to small and doesnt match the cncd part. Any clue what else it could be?
could to high of a temp impact it that much ? i print pla at 225 for some reason.
gotta go back to basics in my opinion. As Oskait has pointed out, you need to do all the calibrations. Have you verified that when you print a 20x20x20 cube and it actually measures 20x20x20?
its 20.05x20.06x20.4 is that still fine ? butt ill probably start at the basics again anyway
AC=140.86, BD=140.99, AD=99.7 looks quite good. If those are the values you measured, maybe the cncd part is off?
Check with calipers at different points which of the both parts matches your CAD file.
ok, an earlier version of the printed part fits. maybe the gcode is corrupted. I'll calibrate the basics and maybe use a different slicer.
I think it has to be something about the stl