Z_TILT for 4 stepper_z

General Discussion
1

Basic Information:

Printer Model: Home Made
MCU / Printerboard: BTT Octopus + EBB 42

klippy.log (6.1 MB)

Hi,
I have a problem and I don’t know where the error is. It works on one printer and the other one I’m building doesn’t.

I used STEPPER_BUZZ STEPPER to verify the correct positioning of the motors and direction of rotation.

After starting QUAD_GANTRY_LEVEL, everything goes as it should, but after 3 measurements it ends with an error:

Probed points range is increasing. Possibly Z engine numbering is wrong

If I do PROBE_ACCURACY over each QGL point they are
results like this:

[stepper_z]
probe accuracy results: maximum 1.511250, minimum 1.506250, range 0.005000, average 1.509375, median 1.509375, standard deviation 0.001790

[stepper_z1]
probe accuracy results: maximum 1.643750, minimum 1.638750, range 0.005000, average 1.642000, median 1.642500, standard deviation 0.001601

[stepper_z2]
probe accuracy results: maximum 1.231250, minimum 1.227500, range 0.003750, average 1.229750, median 1.230000, standard deviation 0.000935

[stepper_z3]
probe accuracy results: maximum 1.553750, minimum 1.551250, range 0.002500, average 1.552375, median 1.552500, standard deviation 0.000673

configuration for QGL:

QUAD_GANTRY_LEVEL_TMC2226.cfg (4.2 KB)

2

Make sure you are following the Z-Stepper layout as per [quad_gantry_level]

3

sorry i made a mistake i used QUAD_GANTRY_LEVEL for bed instead of Z_TILT with 4 motors. I’ll redo it tomorrow and see.

4

Is it possible to use Z_TILT for 4 stepper z motors?

5

As for this, yes:

https://www.klipper3d.org/Config_Reference.html#z_tilt

6

But it doesn’t work for me for 4 engines.
I don’t know why anymore?

7

Please attach a new klippy.log to your next post.

8

Hi klippy.log probably won’t be needed because the alignment of 4-motors using Z_TILT doesn’t work due to crossing height adjustments.

I removed BED and tested just the points themselves and it always ended up with an error. “Measurement interrupted increases tolerance”.

I only tried it for 3-motors and the points evened out in 3 Z-TILT cycles.

IMG_20240707_103049
IMG_20240707_1030491920×1440 254 KB

That’s why I’m asking if anyone has tried Z_TILT for the Z 4-motors.

Thanks a lot.

9

Why don’t you check us your settings?

10

Ok, configuration for 3 stepper_z which works great.

klippy.log (238.4 KB)

11

Ok, configuration for 4 stepper_z which ends with an error.

klippy (1).log (617.9 KB)

image
image1008×393 31.1 KB

12

How have you confirmed your Z Axis Stepper Motors wiring?

When you are doing QGL and the tilt is increasing, you most likely have a wiring error.

Also, 50 retries for your QGL is excessive. I run with 5 without any problems.

1 Like
13

Hi, the wiring is correct. I performed STEPPER_BUZZ STEPPER for each stepper Z and found that the sequence according to the drawing and the direction of rotation are also the same.

##  ----------------
##  |Z1          Z2|
##  |  ---------   |         
##  |  |       |   |         
##  |  |       |   |
##  |  x--------   |
##  |Z           Z3|
##  ----------------

## Where x is the 0, 0 point on the bed

50 is just for testing, otherwise I’ll put 5 there.

For BED alignment using 4 stepper_z neither QGL nor Z_TILT works

14

When I think about it, using 4 motors for a BED with a thickness of 8 mm is a bad solution, because the motors will “pull” each other.

I will use Z_TILT with only 3 Z motors, that will be a more sensible solution.

##  ----------------
##  |Z1            |
##  |  ---------   |        
##  |  |       |   |        
##  |  |       | Z2| 
##  |  |       |   | 
##  |  x--------   |
##  |Z             |
##  ----------------

Thanks.

15

There is a bit of give in the system that will keep everything in line - when you are doing your QGL, the final adjustments will be in the range of tens of microns.

I tried three steppers on a printer I built a couple of years ago and found that I had problems with lead screw slippage because of the weight of the gantry (the printer has a stationary build surface).

Regardless, if you’re having problems with four steppers and an increasing range on QGL, there’s a good chance you’ll have the same problem or worse with three.

I recommend figuring out what the problem is with four before tearing things up and going with three.

What kind of Z-Axis sensor are you using?

16

I also tested the Z-TILT with 3 Z motors and the point comparison
was within 0.010mm during the 3 measurements.

I use a Microprobe from BIQU to sense the Z position.

I’m ready to test before I take it apart and convert it to 3 motors, but now I’m at a loss because everything seems correct but the Z-TILT with 4 motors doesn’t work.

note:
I do not lift the XY gantry - it is stationary, I move it from the BED

17

I mentioned my problems with three steppers because it was with the gantry and it masses a lot less than the 8mm aluminum bed on my Voron. From your photograph it looked like you were going to move the print bed.

How often do you get the increasing distance on your QGL? Are your steppers in line with the corners of the build plate (I built a printer where they weren’t and it took more steps to get QGL working, with the occasional increasing error)?

18

Have you tried reducing the number of probed points to 3, e.g. first one at Z1, 2nd one at Z2 and third one between Z3 and Z4?
Not sure but from looking at the function it seems to try calculating a plane and then determining the adjustment needed for “balancing” this plane.
Calculating a plane from 4 points seems like a bad idea.

19

Did you rebuild the printer to test that from 4 to 3 z axes?

20

No it didn’t rebuild, I just disabled the Z3 stepper in the [z_tilt] section.
(point 4 in the photo).

And here the alignment succeeded on the 3rd attempt with a tolerance of 0.010 mm.
For now, I’m testing it without the 8mm thick BED.

Is that bad too?