Z Probe mesh compensation issues

evanok

Using the Duet Maestro system with a delta system,
Firmware Name: RepRapFirmware for Duet 2 Maestro
Firmware Electronics: Duet Maestro 1.0
Firmware Version: 2.02RC5(RTOS) (2018-11-28b1)
Web Interface Version: 1.22.3

I'm using a simple levelling probe setup from https://www.thingiverse.com/thing:3039808
as the z probe. I've installed everything correctly to this point but the mesh compensation looks quite weird. The probe works fairly reliably during the G30 S-1 probe the values it gives are between -0.164mm to -0.177mm, which is quite good, but then during a G29 S0 operation it seems to be showing these massive deviations in the heightmap. First time dealing with a active z probe, I've made sure that each time the head is clean, to make sure that hasn't been playing a role in the height deviations.

Any ideas where I could be getting this wrong?

Phaedrux

Effector tilt perhaps?

When you test repeatability of G30 S-1, are you using the same spot on the bed? What about if you test it at different locations?

evanok

@phaedrux
There's no effector tilt that I can see.
When using G30 S-1 I'm using just the centre of the bed, except for using G29
I haven't test G30 S-1 at any of the other points on the bed

Phaedrux

Have you gone through this?

https://duet3d.dozuki.com/Wiki/Calibrating_a_delta_printer

evanok

I've gone through the docs a few times, I can't see what might be out of place.
From another post made, it might be a case of head acceleration during the probe descent that could be causing the issue

dc42

Did you run auto calibration before you ran G29? It looks like your Z endstop switch offset could be a long way out.

evanok

So I ran G32 but had issues with the probe not activating, had originally been using a heavily modified config.g and bed.g from someone else. I found the reprap configuration system (https://configurator.reprapfirmware.org) and found my issues, a few commands had been put out of order by the looks of it. I ran G32 with 9 point 6 factor calibration and modified the M666 and M665 commands.
I ran G29 S0 to get the new mesh and it looks a hell of a lot better, I'll attach a screenshot later when I get back home
https://configurator.reprapfirmware.org
Thank you both for the help

evanok

handy document for anyone doing delta calibrations

http://boim.com/DeltaUtil/CalDoc/Calibration.html

has some good information on what different mesh outputs look like