extra z-offset every new print G29?

o_lampe

@IndeX4D said in extra z-offset every new print G29?:

the amount off extra offset I need, is exactly the amount of the ´´Maximum deviations: -1.612´´

If that's correct, it's a bug.
Do you allow you z-axis to dive below zero during prints? If not, it's worth a try.
But such a high deviation is an issue you should take a look at, too. ( if it's a not a random number you picked )

IndeX4D

@o_lampe
ehhmm I´m not sure, if I´m allowed to bo under z0.

To make it clear.... the nozzle is by z=0 higher as it should be.

I have that since rrf 3.5... I was waiting, if somebody is posting something. I have it on all machines and I surprised, that nobody else has a problem but yes the deviation is much.
But also with my smaller printer I have 0.2 deviation. No fun to make readjusting babysteps EVERY new print.

I have a 1030mm x 1030mm glasbed. So it´s not easy to make a real flat.

what, now?

o_lampe

@IndeX4D said in extra z-offset every new print G29?:

I´m not sure, if I´m allowed to bo under z0.

Just send M208 from the DWC console and see the axis min/max limits.
Allowing to go below Z-zero is the best way to make meshbed compensation work. There are other tricks like setting the probe- and nozzle height wrong, but I don't like that. Because the heightmap looks totally off, especially with such a deviation. I bet you only see dark red zones?

Phaedrux

@IndeX4D said in extra z-offset every new print G29?:

G30 P2 X700 Y700 Z-99999 ; probe near a leadscrew
G30 P2 X700 Y150 Z-99999 ; probe near a leadscrew

Your bed.g has conflicting P2 points.

@o_lampe said in extra z-offset every new print G29?:

Allowing to go below Z-zero is the best way to make meshbed compensation work.

Allowing move below zero is only needed to allow baby stepping to go lower than the first layer height. The mesh will follow the contours of the heightmap already as it is.

What kind of a probe is it?

Can you post an image of your heightmap?

IndeX4D

ah sorry. It´s my 850mm machine. Not the 1030 like I said above.

m208
Axis limits (mm)s X0.0:860.0, Y-100.0:860.0, Z0.0:850.0, C0.0:200.0

but I can go further down with babysteps. So that I had z-3.3 after 4 prints.

Probe: tool changer switch E3D

@Phaedrux thanks. I corrected bed.

By the way.
I have the same problem with a smaller amount on my bigger machine.

EDIT: one time I applied the needed extra offset via babysteps BEFORE printing --> nozzle crash into heatbed.
Usally I did it always when printing the brim.

Phaedrux

A dish shape like that on a large printer like that makes me wonder if you're getting some sag on the XY axis rails. I would bet the bed is flatter and stiffer than the mesh indicates.

Can you also try changing your Z homing so that you're probing the enter point of your mesh? If you hover your cursor over the center point it will give you the XY coordinate. That will at least place the center of your mesh as the Z0 point.

IndeX4D

@Phaedrux said in extra z-offset every new print G29?:

if you're getting some sag on the XY axis rails

To be honest.
In this case the corners are moved up because of the isolation foam under the glass.
I have full carbon x gantry and not more than 400gramm as toolhead.

BUT without such massive deviation. It would have been much more complicated to find out the problem.
I change this at the moment at my bigger printer.... after that on this machine.
So or so I have this issue. Also ennoying when only a deviation at 0.2mm.

Let me try to use an exact point for g29 and homeZ...

droftarts

@IndeX4D said in extra z-offset every new print G29?:
This is your current bed.g:

G30 P0 X150 Y150 Z-99999 ; probe near a leadscrew
G30 P1 X150 Y700 Z-99999 ; probe near a leadscrew
G30 P2 X700 Y700 Z-99999 ; probe near a leadscrew
G30 P2 X700 Y150 Z-99999 ; probe near a leadscrew
G30 P3 X150 Y150 Z-99999 S4 ; probe near a leadscrew and calibrate 3 motors

As @Phaedrux pointed out, you have two P2 lines. You also have X150 Y150 probed twice. And then you have S4, when there are 5 points. I'm surprised it doesn't throw an error. Try this:

G30 P0 X150 Y150 Z-99999 ; probe near a leadscrew
G30 P1 X150 Y700 Z-99999 ; probe near a leadscrew
G30 P2 X700 Y700 Z-99999 ; probe near a leadscrew
G30 P3 X700 Y150 Z-99999 S4 ; probe near a leadscrew and calibrate 4 motors

With the bed mesh, it looks like you are homing Z at a corner. So you then can't go down any further in the middle. Home at the lowest point, ie the middle, then the bed mesh can correct it.

As for why it resets higher up at the end of each print, I'd guess there's a G92 Z... in your end Gcode somewhere that is messing it up. Post a (short) Gcode file that exhibits this behaviour and we'll have a look.

Ian

IndeX4D

@droftarts
bed.g is already corrected. same behaviour...

Probing in the middle is not an option. At the other printer there is low deviation and there is not really a deepest point. I don´t like the idea to look for the deepest point on several printers. Also there will be still any deviation, even it´s not much.
I mean the first print is fine.... then z-1.6 , 3.2, 4,8-

and I already was working with this commands without problems for months. I think it came with 3.5

When I stop a print, next print starts at same height.
--> when finishing a print. --> extra offset added.

EDIT: when Using a deeper point for homeZ, the extra offset is smaller, the higher the point is, the extra amount which is applying is bigger.

G1 X610.176 Y449.295 E0.10056
G1 X611.585 Y450.704 E0.19684
G1 X610.566 Y450.704 E0.10066
G1 X609.158 Y449.295 E0.19677
G1 X608.14 Y449.295 E0.10056
G1 X609.548 Y450.704 E0.19677
G1 X608.53 Y450.704 E0.10056
G1 X607.121 Y449.295 E0.19684
G1 X606.103 Y449.295 E0.10056
G1 X607.512 Y450.704 E0.19684
G1 X606.493 Y450.704 E0.10066
G1 X605.085 Y449.295 E0.19677
G1 X604.067 Y449.295 E0.10056
G1 X605.475 Y450.704 E0.19677
G1 X604.457 Y450.704 E0.10056
G1 X603.048 Y449.295 E0.19684
G1 X602.03 Y449.295 E0.10056
G1 X603.439 Y450.704 E0.19684
G1 X602.421 Y450.704 E0.10056
G1 X601.012 Y449.295 E0.19684
G1 X600.745 Y449.295 E0.02637
G1 X600.745 Y450.046 E0.07419
G1 X601.402 Y450.704 E0.09185
G1 X600.745 Y450.704 E0.0649
;TIME_ELAPSED:5734.465541
G1 F6000 E-0.2
M140 S0
M204 P4000
M204 T4000
M566 X1200 Y1200
M82 ;absolute extrusion mode
M107
G91 G1 Z20 F400 G90
M83 ;relative extrusion mode
M104 S0
;End of Gcode

Phaedrux

@IndeX4D said in extra z-offset every new print G29?:

M566 X600.00 Y600.00 Z2000.00 C3000 E600 ; set maximum instantaneous speed changes (mm/min)
M203 X24000.00 Y24000.00 Z300.00 C10000 E24000.00 ; set maximum speeds (mm/min)
M201 X1700.00 Y1700.00 Z2000.00 C550 E8000 ; set accelerations (mm/s^2)

Your Z axis accel and jerk values are very high. Are you sure you're not skipping steps?

@IndeX4D said in extra z-offset every new print G29?:

I don´t like the idea to look for the deepest point on several printers. Also there will be still any deviation, even it´s not much.

This isn't necessary. The compensation can go above and below the 0 point as needed, but using the center of the bed and more importantly using one of the mesh points will ensure that at least one point of the mesh matches Z0. At the very least this gives a more useful representation of what is going on.

What is in your slicer start gcode?
What is in your homeall.g?
Do you use start.g?

@IndeX4D said in extra z-offset every new print G29?:

and I already was working with this commands without problems for months. I think it came with 3.5

Have you tried rolling back to 3.4.5?

IndeX4D

@Phaedrux
I´m pretty sure not to loose steps
I print with z = 6mm/s.
Prints are looking good.
When I print 1 hour and skip at the last layer, I don´t have this extra amount.
When I print just 1 minute and finishing it, I have extra amount.
And. When I do not use mesh compensation, I don´t have this problem.

IndeX4D

@Phaedrux

What is in your slicer start gcode? -nothing, don´t have a homeall.g or so
What is in your homeall.g? -nothing / don´t use that
Do you use start.g? - no

Can´t go back to 4.... cause of closed loop and other issues.
With readjusting I can work for a moment

and BIG EDIT: If i would loose steps, I would loose steps in the upper direction?
Is it clear, that the nozzle is to far away from the heatbed and not the heatbed is moving?
So I think loosing steps in this direction is not possible.
Sorry when I forgot to describe my axis.

Phaedrux

@IndeX4D said in extra z-offset every new print G29?:

What is in your slicer start gcode? -nothing

So how are you preparing for printing between power on and the print starting? When are you homing? When are you leveling? When are you doing the mesh?

IndeX4D

@Phaedrux

Power on ---> homing x, y, z, c manually. --> select tool --> start print

I Do the mesh probe after homing ... not before every print.

I tried to apply G29 S1 before and after homing z . Does not made any differences.
When would be the correct time to apply G29 S1?

I just finished a 2 hours print and cancelled the print a few seconds before finishing. After that the z position was still correct. I´m wondering why finishing prints can produce this Issue in any way.

Phaedrux

Do you have a stop.g present in the sys folder?

When do you run G32 for bed.g?

When do you run G29 S1 to load the heightmap. It should be loaded after homing Z.

Can you add M561 to the start of your homez.g?

IndeX4D

@Phaedrux
Do you have a stop.g present in the sys folder? - no, should I?

When do you run G32 for bed.g? - also when homed manually.not every print. I´m not sure if before or after G29. What would be correct?

When do you run G29 S1 to load the heightmap. It should be loaded after homing Z. ----- after homing Z ( I tried before only one time but of course I got an error)

droftarts

@IndeX4D as I said earlier

As for why it resets higher up at the end of each print, I'd guess there's a G92 Z... in your end Gcode somewhere that is messing it up. Post a (short) Gcode file that exhibits this behaviour and we'll have a look.

Or post whatever is the last 10 or 20 lines of your Gcode (actual Gcode, not the comments).

Ian

IndeX4D

This post is deleted!

IndeX4D

@droftarts
I did already above..

G1 X610.176 Y449.295 E0.10056
G1 X611.585 Y450.704 E0.19684
G1 X610.566 Y450.704 E0.10066
G1 X609.158 Y449.295 E0.19677
G1 X608.14 Y449.295 E0.10056
G1 X609.548 Y450.704 E0.19677
G1 X608.53 Y450.704 E0.10056
G1 X607.121 Y449.295 E0.19684
G1 X606.103 Y449.295 E0.10056
G1 X607.512 Y450.704 E0.19684
G1 X606.493 Y450.704 E0.10066
G1 X605.085 Y449.295 E0.19677
G1 X604.067 Y449.295 E0.10056
G1 X605.475 Y450.704 E0.19677
G1 X604.457 Y450.704 E0.10056
G1 X603.048 Y449.295 E0.19684
G1 X602.03 Y449.295 E0.10056
G1 X603.439 Y450.704 E0.19684
G1 X602.421 Y450.704 E0.10056
G1 X601.012 Y449.295 E0.19684
G1 X600.745 Y449.295 E0.02637
G1 X600.745 Y450.046 E0.07419
G1 X601.402 Y450.704 E0.09185
G1 X600.745 Y450.704 E0.0649
;TIME_ELAPSED:5734.465541
G1 F6000 E-0.2
M140 S0
M204 P4000
M204 T4000
M566 X1200 Y1200
M82 ;absolute extrusion mode
M107
G91 G1 Z20 F400 G90
M83 ;relative extrusion mode
M104 S0
;End of Gcode

droftarts

@IndeX4D said in extra z-offset every new print G29?:

@droftarts
I did already above..

Sorry, yes you did, missed that. I’d say it’s this last Z move:

G91 G1 Z20 F400 G90

It’s at 6.66mm/s over a longer distance than usual Z moves. Try halving the speed to F200 and see if it still does it.

Ian