Probing has +0.1mm increase between samples

General Discussion
1

Basic Information:

Printer Model: Funmat HT - Using Euclid style detachable probe
MCU / Printerboard: Bigtreetech Manta E3EZ with CB1 (Minimal v2.3.2 image / Debian 11)
klippy.log (731.8 KB)

Describe your issue:

I have gutted the stock Funmat HT printer and installed new hotend (E3D Volcano) and replaced the stock MCU with a Bigtreetech Manta E3EZ with CB1. The Z axis motor uses a ballscrew and I verified the rotation distance with digital calipers. I used the latest v2.3.2 minimal image which is Debian 11. Klipper / Moonraker / Mainsail installed with KIAUH and all are up to date as of writing this post.

The original printer only had manual bed leveling so I added a probe. Originally this was an old Creality capacitive probe but I was getting bad bed mesh results. I switched it out for a Euclid style magnetic detachable probe (using an Omron D2F-5 plunger microswitch).

It seems that when I run a PROBE_ACCURACY test each sample is roughly 0.1mm higher tan the prior sample. Homing Z axis resets this increase.

probe accuracy
probe accuracy1660×1532 237 KB

SCREWS_TILT_CALCULATE probes all 3 screw locations without this 0.1mm increase. But running it again introduces the increase. Homing Z in between running the command twice yields probe points on the same heights.

screws tilt calculate
screws tilt calculate1650×875 133 KB

double screws tilt calculate
double screws tilt calculate1590×1382 234 KB

When doing a 5x5 bed mesh the initial probe is at 2.965 and last probe is at 4.90375

bed mesh
bed mesh550×1417 159 KB

Have manually verified the bed is level with the gantry with a BED_SCREWS_ADJUST.

Just need a sanity check. It was doing this before the recent Klipper commits with bed_mesh modifications.

2

You may increase run_current for the Z-axes

and

decrease lift_speed in [probe] (and also hop_speed in [safe_z_home])

Probing should be slow for accuracy.

Also the acceleration on the Z-axes is quite high. Particularly with that low motor current.

1 Like
3

Thanks knew i was overlooking something! You’re right the Z acceleration was quite high and the run_current very low. Slight adjustment these and the lift_speed did the trick.

Was so ready to throw this printer in the bin I’ve been pulling hair out for a few weeks

1 Like