[3.6.0 RC1] MultiMotor Z bed leveling

[dc42]

@cichaczech you say that the rear motor is adjusted by a larger amount using RRF 3.6; however I see this for all 3 leadscrews. This is because the magnitude of the initial mean value is much greater in the data you post for 3.6rc1, i.e. 0.073 to 0.191mm (ignoring the first adjustment), compared to 0.002 to 0.008mm with 3.5.4.

Are you definitely using the same config.g file and bed.g file in both cases?

[cichaczech]

@dc42 Yes, first I did 3.5.4 -> 5x auto adjust and than G29. immediately after I upload 3.6.0 and run same 5x auto adjust and than G29. I am using same printer, Duet2, config and all gcodes in both examples.

[dc42]

@cichaczech please can you run levelling with 3.6.0-rc.1 again and this time observe the leadscrews to check whether they actually move when they are supposed to make the corrections. Also please try firmware 3.6.0-rc.1+3 available at https://www.dropbox.com/scl/fo/geqmn8gbn97n6b587mkbk/AJ2hBIqO-L57_QpR_uqHJ9c?rlkey=fw37wycbp2gil8rvxhe7aopy7&dl=0.

[cichaczech]

@dc42 leadscrews moves after last G30 P..... S command normaly in 3.6.0 RC1.

Here are my test first 3.6.0 RC1:

then I run 3.5.4 just to test it again:

and then 3.6.0RC1+3:

same printer, gcodes and same conditions - bed 70C hotend 230C.

[dc42]

@cichaczech please can you add a G4 S1 command after the G30 command with S parameter and see if it makes any difference. Unfortunately I don't have a machine with multiple Z leadscrews to test this on.

[cichaczech]

@dc42 still the same with G4 S1 after last G30. I have another printer with 3x Z motors and bed 240x240mm, but it still printing. I will post update when it finishes.

[cichaczech]

@dc42 so on 2nd printer both firmwares (3.5.4 and 3.6.0 RC1+3) works same, 3.6. mayby bit better. I suspect some problem with Duet2 board on 1st printer. Tomorow I will try flash it with BOSSA.

Sorry for bothering you whole week.

[dc42]

@cichaczech do both printers use Duet 2 ?

[cichaczech]

@dc42 yes and same configuration. One is just bigger and one smaller. I remember that one time I had some problems with Duet 2 in bigger printer and flash FW with bossa helped me. (I completly forgot abou it)

Update:
No change after BOSSA. Bed is still drifting away.

May there be problem with Bltouch? I do not know.

[Phaedrux]

@cichaczech said in [3.6.0 RC1] MultiMotor Z bed leveling:

May there be problem with Bltouch? I do not know.

One way to check that would be to switch to probe type 0 (M558 P0) to use manual probing where you get prompted to jog the nozzle down to touch the bed. This would isolate any bltouch issues.

[cichaczech]

@dc42 Today, I noticed an issue when using G30 the print head with the probe shifts a few millimeters to the right (+X axis) between multiple probing actions and after probing the reported X position in DWC is lower.

Example:

G28 X Y
G0 X113 Y200 ;move to bed center
G30 ;probe

1. The first probe is at X113 Y200 as expected.
2. When the bed moves away from the probe, the toolhead slightly shifts to +X before probing again.
   3.After probing, DWC updates the X position incorrectly—showing X111.3, even though physically, the tool is still past X113.

2 problems - probe moves between multiple probing and DWC reports wrong axis movment.

This happens on both of my CoreXY printers running firmware 3.6.0.

@Phaedrux I didn't have time to probe it manually today, but I'll try it tomorrow.

[dc42]

@cichaczech are you using M556 axis skew compensation?

How much does the X position shift after probing? It may help if you post a video.

[cichaczech]

@dc42 Yes, I do use M556. It should be in the config I posted earlier, but I just realized I copied the wrong one, sorry! I'll update it right now.

Desktop 2025.03.31 - 20.00.53.04.mp4