Bed leveling compensation error

cvolpi

I set up the bed leveling correction and it works fine till it is correcting the Z height during a print.
The direction of the correction is in the opposite direction of what is needed.
I mean Z is increasing when decreasing is needed.
Is there a feature dedicated to adjust the correction direction?
Thank you
Claudio

Repetier

Do you have only G32 or also G33 used? Always start with G32 an check the output. It return the plane parameter measured and the rotation matrix. Especially the plane parameter are easy to understand. a is correction per x going right and b is same for y. So you can directly see if the measured orientation matches what you expect.

cvolpi

What is the right sequence of g32 and g33.
now I am just sending g32 s2
thank yuo

cvolpi

I am sorry but I don't understand the results of the transformation matrix.
Are they correct to move down Z with X going from min to max?

My printer is an Eustathios spider V2, similar to the ultimaker. Is it correct the type I choose Cartesian printer ?

15:43:27.745 : Z-probe:10.685 X:100.00 Y:100.00

15:43:39.039 : X:0.00 Y:0.00 Z:0.000 E:0.0000

15:43:50.600 : Z-probe:6.174 X:10.00 Y:10.00

15:43:55.632 : Z-probe:6.467 X:10.00 Y:130.00

15:44:00.665 : Z-probe:6.741 X:10.00 Y:240.00

15:44:06.622 : Z-probe:6.707 X:115.00 Y:10.00

15:44:11.798 : Z-probe:7.007 X:115.00 Y:130.00

15:44:17.006 : Z-probe:7.319 X:115.00 Y:240.00

15:44:22.976 : Z-probe:6.704 X:220.00 Y:10.00

15:44:28.179 : Z-probe:7.076 X:220.00 Y:130.00

15:44:33.432 : Z-probe:7.487 X:220.00 Y:240.00

15:44:33.463 : plane: a = 0.0030 b = 0.0027 c = 6.1550

15:44:33.463 : Transformation matrix: 0.999996 0.000000 -0.002992 -0.000008 0.999996 -0.002726 0.002992 0.002726 0.999992

15:44:33.463 : Printer height:279.96

15:44:33.463 : CurrentZ:6.16 atZ:6.20

15:44:33.463 : Info:Autoleveling enabled

15:44:34.262 : X:219.98 Y:239.98 Z:7.467 E:0.0000

Thank you
Claudio

Repetier

At probing height the bed follows the equation
z(x,y) = a * x + b * y + c

a=0.003 means it will change 220 * 0.003 = 0.66mm while traveling from x=0 to x=220. So it is really rotated. For better results you should correct it manually to be more close. If that is not the error you see at the end the correction is working good.

What you should consider is that the bed is not planar also deviation is not too big. If it is mounted on 4 screws that might bend bed a bit. Firmwrae computes an average ideal bed that reduces error to all 9 points. So it will not match perfectly the distance, only reduce it.

G32 S2
G33

is the order for plane + distortion correction. On cartesian you can normally omit the distortion if you watch out when fasteining the bed to not bend it.