Mesh Levelling Question/Issue.

[Syko_Symatic]

I have used mesh levelling for years but I have noticed a certain issue. I probe the bed (in the centre) then run the mesh levelling, but I’ve noticed that none of the points are ever 0 or near 0, surely the height should be close to zero when near the probes point for Z0?

For example, sometimes all measurements will be over 0, this shouldn’t be possible should it?

Do I have something configured wrong, perhaps a macro or something??

[oliof]

@Syko_Symatic for the reason you state, I try to make sure that the G28 Z coordinates co-incides with a mesh point, either bed center, or bottom left -- either is fine.

All probes have a certain swing in repeatability, which is a combination of how precisely the probe will trigger on repeat attempts, how precisely the Z axis is driven, and how fast the probe moves.

For example, cheap inductive probes have a resolution of 80Hz, so at 5mm dive speed you can hit differences of 5/80=0.0625mm just by missing one beat at 80Hz and triggering earlier or later. Probing slower will immediately improve the precision and repeatability. More expensive probes like the SuperPINDA derivatives have a resolution of 320HZ IIRC so they will immediately have resolution gains.

Another factor is ambient temperature. Many inductive probes trigger earlier if the bed is heated; so if you home on a cold bed and run the mesh on a heated bed, that would be sufficient to explain the difference. RepRapFirmware has a facility to adjust the probe measurement based on a thermistor reading; while it is meant to use a thermistor in the probe (like in the PINDAv2), it can also make use of the bed thermistor -- but you will need to prepare the temperature coefficient for adjusting probe trigger points for this.

I hope this helps you investigate further.

[Syko_Symatic]

@oliof thanks for the information! How would I make sure one of the points is aligned with my centre point? Is it just using the math and dividing the mesh correctly?

[Phaedrux]

@Syko_Symatic said in Mesh Levelling Question/Issue.:

Is it just using the math and dividing the mesh correctly?

yes, exactly

[dc42]

@Syko_Symatic also bear in md that a G30 probe (which you are probably using to establish ZX=0 at the centre of the bed) probed wherever the head currently is, whereas G29 takes account of the probe offset. So the XY coordinates of the initial centre probe need to correspond with the XY coordinates of one of the G29 probe point minus the XY offsets of the probe hat you declared in your G31 command.