Axis twist comp

General Discussion
1

Basic Information:

Printer Model: neptune 4 max
MCU / Printerboard:
Host / SBC
klippy.log

Describe your issue:

hello. n4 max openneptune. x axis perfectly parallel to base. bed is warped on y axis. when i try axis twist comp (atc) with paper method then save then bed mesh, mesh is still same. is it normal? shouldnt i see the bed mesh according to atc? in printer.cfg atc’s measurements are at the end of the page. i know that paper method is not the best way but it has to change the mesh even a little. and the top side’s z offset and deep area z offset’s are very different when i do paper method.

2

Please attach the klippy.log to your next post.
Please attach images here to the forum, not to (another) file server.

3

1
11080×2400 131 KB

4

i will add klippy.log when i get home.

5

klippy.log (1.4 MB)

6

Did you create a completely new mesh after performing the twist compensation?

7

screw tilt adjust and full bed mesh yes. but i realized that if i try to catch same z offsel during atc it doesnt work. if i make it depend on bed mesh i mean much lower where that hill or mush higher where the hollow it works. is it normal?

8

This seems to be a chicken-and-egg problem:

  • If your bed is extremely crooked, then you “bake” this into the ATC.
  • If your axes are crooked, then the same applies.
  • If your bed is not perfectly levelled, this also becomes an issue.

I guess this will always remain an issue if you need to work in an environment where you only have a “relative truth” but never an “absolute truth.”

Generally, I’d also use only 3 points since you want to counteract on a twist or bend of the axes and if you look at your values:

zy_compensations = -0.043833, -0.071333, 0.034917, -0.022583, 0.019917, -0.035083, -0.022583, 0.032417, 0.002417, -0.032583, 0.012417, 0.044917, -0.008833, 0.079917, 0.009917

It looks more like a bumpy road, and it is quite unlikely that your axis really looks like this.

9

bed is leveled with screw tilt calc. im sure x axes are parallel to base.
for neptune 4 max, maybe more than 5 sample is unnecesarry but im just trying different solutions..

10

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.