Max V2 & Duet auto-calibration deviation

[Reikal]

Hi Folks,

I've been printing for a while on my V2 and its been OK, not great, but OK, however now its worse. So i'm looking for some ideas on what's causing the issue.

The auto calibration has been giving a steady 0.8 (+-0.1) since i fitted the FSRs
I want to get this down to 0.5 at least, ideally 0.3 as i want to do some fine detail work with a 0.25 nozzle, and use the whole bed with the 0.4 standard nozzle.

I did think the problem was the FSR mounts, they were printed in a bit of a rush and warped slightly due to poor bed adhesion. I decided yesterday to remake them and see if that would improve things. Basically the new prints came out really well, fitted perfectly, and made my problem worse.

The calibration now runs a deviation of 0.12 pretty constantly.
After hours of trying i'm running out of ideas on what to check.

The machine frame and mechanical parts are stock v2 (ball cup arms, molded carriages etc)
Electronics are Duet 0.8.5, PanelDue and an FSR board.
FSR connection via a JohnSL? board
Bed is the standard glass sheet covered with 0.2mm PEI (i'd love to just stick a 3mm PEI sheet on it, but its stupidly expensive in the UK)
Hotend is an E3D v6

Checks so far:

Clean bed
FSR Triggering - all the sensors work easily with a tap of a fingernail on the bed
Belt tension - seems even on all towers
Hotend - its tight, there's no movement on it anywhere.
Endstop screws - no movement, all are still a 4mm
Endstop switches - one seemed slightly loose; made no difference when tight (ive missed these in the past)

I'm open to any and all suggestions - with the exception of buying another printer, I need to save money for a laser build!

[dc42]

Have you measured the FSR trigger height at each probe point? You may need to use trigger height corrections in the bed.g file. See https://duet3d.com/wiki/Calibrating_a_d ... bed.g_file.

[Reikal]

Thanks DC, ill try that as soon as i get more time off work.