Skew compensation doesn't work

[dagov42]

Printer type - MINI+ with SuperPinda

Printer firmware version- v1.0.7 (based on Prusa Buddy 4.3.3-final)

**Custom firmware - Llama

Optional upgrades - Filament Runout Sensor

USB drive or USB/Octoprint USB and Octoprint

Describe the bug I printed more than 20 test to compensate XY skew. And increase XY coefficient from 0.02 to 0.1. And diagonals length didn't change. I think this function isn't working. A - 142.04 B - 140.65 I restore settings to default after install Llama firmware after original. Tried to print via OctoPrint and USB.

[SiegelDaniel]

Did you store settings to EEPROM ?

[dagov42]

Yes, skew compensation works only in bed leveling mode, after that printer start printing - it doesnt work.

[SiegelDaniel]

I've noticed the same problem with mine. I did the skew test print, logged in the values the spreadsheet gave me and afterwards the results remained the same. @matthewlloyd Is this known yet?

[matthewlloyd]

It's not known. I don't have a Prusa Mini any more so it's impossible for me to test and debug personally. I haven't made any changes to the skew correction code but it's possible Prusa changed something in the latest firmware which broke it when I rebased the changes. Certainly it was working on the previous version. I did take a quick look through the source code and see no reason why it would be broken. So this will need to be figured out by someone else.

Some debugging ideas:

• Have you tried setting a high correction value with e.g. "M852 I0.20" then verifying that a square comes out as a parallelogram, either in a real print or manually with LCD-menu moves or "G1 X... Y..." commands?
• Are your skew settings showing up in M503?
• Do you have anything in your slicer's start G-code which could reset the settings?

[SiegelDaniel]

Thank you for replying that quickly.

Regarding the slicer's start G-CODE:

`M862.3 P "[printer_model]" ; printer model check
G90 ; use absolute coordinates
M92 E415
M83 ; extruder relative mode
M104 S170 ; set extruder temp for bed leveling
M140 S[first_layer_bed_temperature] ; set bed temp
M109 R170 ; wait for bed leveling temp
M190 S[first_layer_bed_temperature] ; wait for bed temp
M204 T1250 ; set travel acceleration
G28 ; home all without mesh bed level
G29 ; mesh bed leveling 
M204 T[machine_max_acceleration_travel] ; restore travel acceleration
M104 S[first_layer_temperature] ; set extruder temp
G92 E0
G1 Y-2 X179 F2400
G1 Z3 F720
M109 S[first_layer_temperature] ; wait for extruder temp

; intro line
G1 X170 F1000
G1 Z0.2 F720
G1 X110 E8 F900
G1 X40 E10 F700
G92 E0

M221 S95 ; set flow`

Using pr0nterface, the skew settings do show up in M503. However, there is no reply from the printer on the M500 command.

I will try the high correction value now and report back in an hour.

[dagov42]

Have you tried setting a high correction value with e.g. "M852 I0.20" then verifying that a square comes out as a parallelogram, either in a real print or manually with LCD-menu moves or "G1 X... Y..." commands?

• Yes, I set 0.9 koefficient and probing failed cause probing mesh has distorsion and head goes away. I used lesser koeffcients around 0.2, mesh deformates but square diagonals doesn't change. Are your skew settings showing up in M503?
• Yes, its the same settings as I set from display. Do you have anything in your slicer's start G-code which could reset the settings?
• No, I use start codes but it's not correct this settings.

[SiegelDaniel]

I've verified with I=0.2 that the probe does indeed run off of the bed, which is why probing fails. I'm now trying with 0.1 again. So it's safe to say that during probing / bed leveling the skew correction is certainly taking place.

[matthewlloyd]

Ok, so the skew correction is clearly working in general, but might not be working when actually printing. Printing with a high correction factor to make the effect obvious, but not so high as to break mesh leveling, may be a good idea. Maybe the skew correction is being reset or turned off somehow by the Prusa firmware at the start of a print. Some more debugging ideas:

• Try printing (a) from SD card, and compare to (b) printing via OctoPrint. Skew correction might be working in one but not the other.
• Put an M503 in the G-Code you are printing, and check the output - i.e. after the print has already started.
• Put the M852 I command in the G-Code to see if that makes a difference.

[SiegelDaniel]

Hi, I've printed the skew correction print with 0.1 for XY skew and it did not change a thing. I'll try the other two hints tommorrow, that are great ideas. Thank you very much.

[SiegelDaniel]

Hi, I've recently tried setting the skew correction again AFTER bed leveling in the printer's start GCODE. However, there seems to be some calculation error as a minor correction of 0,0015 on XY results in the print being placed at a whole different position on the bed.

Do you have any ideas why that might be?