Voron 2.4 bed levelling and first layer squish

[wennbom]

 

I have an issue with uneven fist layer squish that I have been battling forever. I have used a workaround (180% flow) but now I feel I just have to sort things out.
At first I had inconsistencies between and within prints. This was solved by making sure the z-belts were same tension, and more importantly that the z-loops are tight too.

With that sorted I now get consistent but bad results.
Basically I get too much squish at the front and not enough at the rear. This is well reflected in the mesh below as well.
What I do not get is why my calibrated mesh doesn’t compensate this deflection.

I went from a 5,5 mesh to 7,7, reduced the tension to 0.1 and moved the outer probe points closer to the edge of the bed. 
Didn't make any difference.

I get the same result wether I have a mesh calibration or not in my config.
Clearly I’m doing something wrong here but I just can’t figure out what it is.

[bed_mesh]
speed: 300
horizontal_move_z: 10
##    350mm build
mesh_min: 30, 30
mesh_max: 320,320
fade_start: 0.6
fade_end: 5.0
probe_count: 7,7
algorithm: bicubic
bicubic_tension: 0.1
relative_reference_index: 24

[gcode_macro PRINT_START]
gcode:
    # Parameters
    {% set bedtemp = params.BED|int %}
    {% set hotendtemp = params.HOTEND|int %}
   # {% set chambertemp = params.CHAMBER|default(0)|int %}
    BED_MESH_CLEAR
    G28
    # <insert your routines here>
    QUAD_GANTRY_LEVEL
    G0 X175 Y175 Z30 F3600                                                       ; Go to centre
    M190 S{bedtemp}                                                              ; set & wait for bed temp
    G4 P1200000                                                                   ; wait 20 minutes for bed to stabilize
   # TEMPERATURE_WAIT SENSOR="temperature_sensor chamber" MINIMUM={chambertemp}   ; wait for chamber temp
    # <insert your routines here>
    BED_MESH_CALIBRATE                                                           ; Calibrate mesh
    M109 S{hotendtemp}                                                           ; set & wait for hotend temp
    # <insert your routines here>
    CLEAN_NOZZLE                                                                 ; Clean nozzle in purge bucket
    G28 Z      
    G0 X175 Y175 Z30 F3600                                                       ; Go to centre

 

[VPP_Team]

First, your variance is minimal and should be fine to print with. Second, bed mesh issue is a known problem with Klipper, try printing without it, my guess is bed mesh is making it worse.

Edited April 13, 2022 by VPP_Team

[claudermilk]

0.084 on a 350 isn't bad at all. My 250 has 0.076 and runs fine.

[wennbom]

Thanks. I'll make some more tests without bed mesh.

It's a real problem since parts on the low squish areas come loose during printing

With a loaded bed mesh I should be able to by hand feel the z-motors moving while traversing, right?

[Wanderer22]

I have similar problem, not too much issue within the print but 1st layer on next print is always a problem for me with cloned Omron inductive probe. I always have to adjust Z offset at start especially with ABS. Once the print is completed, I saved config to Z endstop, next print is same problem nozzle is too far from bed.

Part of my problem is I don't understand relationship between probe, nozzle offset and bed mesh clear/calibrate and how/when to use them.

Your print start routine is useful.

Yesterday I decided to buy Euclid probe kit after watching Kapman's Basement Workshop on YouTube. I'll post an update if someone is interested on findings.

[wennbom]

In a std configuration the inductive probe has no link to z-endstop value. Z-endstop is the z distance between your z-endstop switch and your bed. It should not change print-to-print.

Make sure you do a "clean nozzle", to ensure that there is no material left on the nozzle, and G28 Z (homing Z) at the end of your print start. If not you will introduce variation.

There's good purge bucket mod with a wire brush that really does clean the nozzle.

The inductive probe is used as a relative measurement.

For QGL it is used to get all 4 corners of the gantry within a certain range -gantry level to bed. Absolute reading is not that important, just that it's repeatable. Probe thermal drift is also not important.

For Bed Mesh however a good accurate temperature compensated probe is important since the absolute readings from the inductive probe are used to build the compensation mesh that is applied to z-values. I still use a cheap clone btw:)

[Wanderer22]

Thank you for the explanation. That explains the variations I've been getting.

I have printed purge bucket + wire brush, still learning how + where to modify the g codes, also how 'upload & print' work.

[Simon2.4]

15 hours ago, wennbom said:

hand feel the z-motors moving while traversing, right

yes and hear a buzz since motors are 'right there'