mesh bed compensation not applying correct corrections

[Phaedrux]

@necrorat said in mesh bed compensation not applying correct corrections:

The image seems to show a bowed gantry. What areas are not touching the bed?

[necrorat]

not the best pictures but best i can do at the moment.

the main area is the region 3 threes what appears to be a island where the print head is too far from the bed

[Phaedrux]

Where are those locations in relation to the heightmap?

[Phaedrux]

I see you're using the BLTouch. You can attempt to get a bit more accuracy and repeatability out of it by modifying your M558 a bit.

M558 P9 H5 F60 T6000 A10 R0.75 S0.003 B1

Slower probe speed, with repeated attempts if 2 consecutive probes don't match to within a tighter tolerance. Probing will take longer because of the slower dive height and it will do at least 2 probes per point. But you will be far less likely to get outliers.

[necrorat]

@Phaedrux so 1 is the rear right of the height map and 3 is the front centre on the height map. sweet thanks for the tip ill change that out and give it a try probably wont be till tomorrow till ill get to try it now though.

[Veti]

can you measure the z trigger heigh at different positions?

it is possible that you have different trigger heights based on position.

[necrorat]

@Veti so i measured the z trigger height and it didn't vary by more than 0.005mm at each of the 9 points i measured at, it did vary from point to point but thats kinda the point of bed levelling isn't it ?

[Veti]

@necrorat said in mesh bed compensation not applying correct corrections:

so i measured the z trigger height and it didn't vary by more than 0.005mm at each of the 9 points i measured at, it did vary from point to point but thats kinda the point of bed levelling isn't it ?

no the z trigger height has to be constant across the entire bed otherwise you have the effects that you are seeing.
to clarify i am not refereing the the result of a G30 but the valude of of Z for G31 when calibrating it.

[necrorat]

@Veti just ran though a 3x3 grid to check and i am consistently getting 2.31 for the trigger height

01/02/2020, 15:16:27 M98 P"0:/macros/probe test"
Current reading 0, threshold 25, trigger height 2.31, offsets X-29.5 Y11.4

[Veti]

@necrorat said in mesh bed compensation not applying correct corrections:

01/02/2020, 15:16:27 M98 P"0:/macros/probe test"

does that macro do this?
https://duet3d.dozuki.com/Wiki/Test_and_calibrate_the_Z_probe#Section_Calibrate_the_Z_probe_trigger_height

[necrorat]

@Veti ive wrote it as this to probe 3 times per spot:
G1 Z10
G30 S-1
G1 Z10
G30 S-1
G1 Z10
G30 S-1

[necrorat]

so have just run it again and i am getting quite a large variance now in the region of 0.11mm is this a mechanical issue? the machine is set up on linear rails and i have ensured that they are as square and parallel as possible

[Veti]

@necrorat said in mesh bed compensation not applying correct corrections:

G1 Z10
G30 S-1
G1 Z10
G30 S-1
G1 Z10
G30 S-1

that determines repeatability, but does not determine z probe trigger height as you do not determine z=0 before probing as described in the above linked document

[necrorat]

@Veti so do i have to go though the full calibration procedure for each point that i probe on the bed?

[Veti]

@necrorat
do one at the point where the lines where ok and one where the lines where not ok

[necrorat]

@Veti so just did a test over 5 point and theses are the results :

2.064
2.068
2.066
2.059
2.046

[Veti]

thats a lot different than your original trigger height of 2.31

[necrorat]

@Veti i have been tinkering with the frame and alignment all day. gonna probe the bed one last time then try another print

[Stuffi]

Sry to bring up that old post again.

@necrorat Did you find a solution or the reason why the meshbed compensation didn't apply?

I have the same issue.