2.03RC5 leveling issue

Veti

there were some changes to leveling since the rc. please update to 2.03 release and try again.

also why 0.1?
M208 Z0.1 S1 ; Set minimum Z

SPAX

@dc42 ok, I'll try

SPAX

@veti coz I print with 0.1 layer step
If I set 0.2, first layer (with 0.1 layer step) very poor and don't catch on bed....

Veti

no you adjust the offset of your probe in the M558 command not with M208

SPAX

Still not work!

G32
Calibrated 7 factors using 16 points, deviation before 9.727 after 0.077

G32
Calibrated 7 factors using 16 points, deviation before 0.436 after 0.055

G32
Calibrated 7 factors using 16 points, deviation before 7.144 after 0.063

G32
Calibrated 7 factors using 16 points, deviation before 6.261 after 0.092

G32
Calibrated 7 factors using 16 points, deviation before 11.004 after 0.046

G32
Calibrated 7 factors using 16 points, deviation before 15.613 after 0.078

dc42

I suspect you are getting inconsistent results from probing. Try replacing S7 on your final G30 command by S-1 so that it prints the errors without making corrections. Then run G32 several times and see how consistent the results are.

Also, we don't normally advise using 7-factor calibration unless the height map is a particular shape. 6-factor usually gives better results.

SPAX

@dc42

02:39:01G32
G32 bed probe heights: -2.687 -2.702 -2.807 -2.825 -2.948 -3.058 -2.749 -2.889 -3.022 -2.864 -3.056 -2.895 -2.875 -2.700 -2.974 -3.005, mean -2.879, deviation from mean 0.122

02:36:56G32
G32 bed probe heights: -2.693 -2.696 -2.826 -2.850 -2.954 -3.071 -2.761 -2.895 -3.022 -2.858 -3.056 -2.913 -2.882 -2.706 -2.968 -3.005, mean -2.885, deviation from mean 0.121

dc42

Those results show differences of around 0.02mm which isn't ideal, but not too bad. Try S6 instead of S7.

SPAX

@dc42
Did s6.
Now G32 repeats little good....
G32
Calibrated 6 factors using 16 points, deviation before 0.132 after 0.106

But before on 2.02 was be after 0.03-0.08(maybe because I did matte on my glass today)

Thank you. I will try to print something tomorrow....

SPAX

@dc42
Now better, but not ideal (((

Calibrated 6 factors using 16 points, deviation before 0.450 after 0.109
Calibrated 6 factors using 16 points, deviation before 0.260 after 0.174
Calibrated 6 factors using 16 points, deviation before 0.207 after 0.156
Calibrated 6 factors using 16 points, deviation before 0.346 after 0.113
Why this happening?
First layer poor....some place the nozzle touching the bed some plase have pit...
If I run G29 the nozzle starting print like on wave... and first layer much more poor .

dc42

@spax, what type of Z probe do you have on your delta printer?

After run G32 a few times, if you then run G29, what does the height map look like?

SPAX

@dc42, I use your IR probe.
22:32:26G29
109 points probed, min error -0.158, max error 0.575, mean 0.214, deviation 0.150
Height map saved to file heightmap.csv
22:26:15G32
Calibrated 6 factors using 16 points, deviation before 0.178 after 0.100
22:25:04G32
Calibrated 6 factors using 16 points, deviation before 0.239 after 0.095
22:23:55G32
Calibrated 6 factors using 16 points, deviation before 0.371 after 0.126

SPAX

I have 6mm aluminium plate + 0.5 heatsink pad + 4mm glass (back side painted in black).

SPAX

@dc42 any thoughts?

Phaedrux

Can you rerun the G29 heightmap with a piece of plain paper covering the bed surface? The idea being to eliminate any inconsistencies in the IR probe reading the surface.

dc42

Also please post a screen shot with the height map rotated so that we are looking along the X axis.

SPAX

@dc42 @Phaedrux

19:36:16G29
97 points probed, min error -0.587, max error 1.307, mean 0.383, deviation 0.415
Height map saved to file heightmap.csv

19:28:00M500

19:27:48G32
Calibrated 6 factors using 16 points, deviation before 0.254 after 0.091

19:26:43M500

19:26:37G32
Calibrated 6 factors using 16 points, deviation before 0.132 after 0.111

19:25:31M500

19:25:01G32
Calibrated 6 factors using 16 points, deviation before 0.209 after 0.080

19:23:54M500

19:23:44G32
Calibrated 6 factors using 16 points, deviation before 0.379 after 0.108

dc42

That height map looks terrible, as if you are probing off the edge of the bed or something like that.

SPAX

@dc42 Coz idea with paper is bad idea) Paper of sheet not 100% flat...

garyd9

@spax said in 2.03RC5 leveling issue:

@dc42 Coz idea with paper is bad idea) Paper of sheet not 100% flat...

That height map has high points "0.25mm of more" and low points "-0.25 or less." A sheet of paper isn't going to deviate in thickness by over 0.5mm!