Thread | Brutkey

I've got a really weird issue I need help with:

My modified 3D printer got a Z-Probe (BIQU Microprobe) for calibration. However, for some weird reason, it always falsely measures the left side of the print bed causing that side of prints to get this weird texture-related elephant foot while the right side lacks adhesion (and texture).
I do not understand why the probe would falsely measure on the left, but not the right. It's a classic i3 bedslinger.
#3DPrinting #Klipper #diy #Biqu #3DDruck

3D printed object's underside on the left: there's a distinct "bumpy" texture to it, however the outlines are rough. The plastic was squeezed through the textured build plate because the nozzle was too low, causing the corners of the object to not be a straight line. Also obviously a too low nozzle night crash into the build plate, which fortunately did not happen here.

The other end of the object, still underside. The texture is almost completely gone, the lines of plastic the printer drew are very well visible. Although the corners don't have any bumps anymore the surface does neither, indicating the nozzle was too high. This also leads to the whole 3D object to lift off the build plate during prints which can cause damages, a warped object and eventually the whole object print to fail. It also looks and feels bad to have this discrepancy from one end of the object to another.

Salearlyman
@duk3luk3@cloudisland.nz

@Natanox@chaos.social I don't really have a great idea but I suppose it could be backlash related or X-axis droop related (if you only have one Z axis lead screw, the X-axis gantry will be drooping a little). In theory both of these should apply equally to the probe and the print head.

Salearlyman
@duk3luk3@cloudisland.nz

@Natanox@chaos.social Can you clarify how the calibration works - is it just levelling compensation, or is it full ABL mesh probing? There's a whole bunch of different algorithms related to bed levelling in Marlin and I don't know which one your printer uses and whether it's modified - maybe knowing that could jog my memory.

Natasha Nox 🇺🇦

🇵🇸

@Natanox@chaos.social

@duk3luk3@cloudisland.nz It's a full mesh calibration every time. I do have two lead screws (it's a modified Anycubic i3 Mega), so there's no sacking. Below is my microprobe config.

Also a picture of the printer's inside so you get a picture of how it looks. The photo happens to show the issue once more, if you zoom in you can see how bad the infill of the left object looks because the nozzle is too low.

[probe]
pin: ^!PJ0 # NOTE FOR V2 users: Set this to ^!D6 to set the low level trigger
deactivate_on_each_sample: False
x_offset: 2.2
y_offset: -5.9
z_offset: 0.410
samples: 2
samples_tolerance: 0.03
samples_tolerance_retries: 3
activate_gcode:
PROBE_DOWN
G4 P500
deactivate_gcode:
PROBE_UP

[safe_z_home]
home_xy_position: 90, 90 # Change coordinates to the center of your print bed
speed: 50
z_hop: 10 # Move up 10mm
z_hop_speed: 5

[bed_mesh]
speed: 250
horizontal_move_z: 3
mesh_min: 6, 10
mesh_max: 190, 190
probe_count: 5, 5
mesh_pps: 2, 3
algorithm: bicubic
adaptive_margin: 5

Photo of the inside of a 3D printer, there currently are three objects being printed.

Salearlyman
@duk3luk3@cloudisland.nz

@Natanox@chaos.social ah, sorry, I missed that you're using Klipper.

The only thing I can think of is that your probe is hitting a small dip / rise that distorts the mesh.

Here's what I would do:

1) Get a visualization of the bed mesh to see if there's an obvious problem
2) Play around with the mesh_min parameter to move the probe points to try and avoid localized faults in the bed
3) If you can identify the faults, increase the number of probe points to make their influence smaller, or use faulty_region