Tracking down first layer inconsistencies

Exerqtor

I'm trying to track down some first layer inconsistencies i've been having for quite some time now. And i'm starting to think the issue might be in my mesh bed leveling routine.

I'm now running a direct nozzle probe (Voron TAP with an optical switch), and parts of my mesh.g is from the days when i ran a "klicky" and nozzle tip switch.

Anywho, this is the current "event tree" once mesh.g is called:

Checks if the bed is trammed or not, if it ain't it runs bed.g.
Check if the mesh probing is part of a print job (determines if the printer has been newly homed, and if it should load the print area data for doing and adaptive mesh).
Moves the printhead to the center of the bed (or print area), and does a G30 S-3.
Proceed to probe the mesh.
Mesh probing finished, end of mesh.g.

The part I think is causing inconsistencies is the G30 S-3 at the start of each mesh probe (which i'm quite sure is old code from "the klicky days").

Whats best practice? To have the trigger height set before each mesh probe, or rely on the "persistent" trigger height (assuming that it's been tunned correctly in beforehand of course).

gloomyandy

@Exerqtor What are the two trigger heights that you end up with (so the one you may have set from your config and the one that results with G30 S-3)?

Also how are you setting Z=0? I've always found that it is important to ensure that it is set using the same mechanism that you are using to create the mesh. Ideally the mesh should contain the same point that you "probe" to set Z=0 and you should be able to then check that this point in the mesh is indeed 0 (or very, very close to it).

Exerqtor

@gloomyandy said in Tracking down first layer inconsistencies:

@Exerqtor What are the two trigger heights that you end up with (so the one you may have set from your config and the one that results with G30 S-3)?

I haven't got any real numbers for that atm. but they're usually NOT the same. (I've got the M501 at the end of config.g and have the probe offsets stored in config-override.g). So that's why i'm suspecting this to be the cause of issue.

Also how are you setting Z=0? I've always found that it is important to ensure that it is set using the same mechanism that you are using to create the mesh. Ideally the mesh should contain the same point that you "probe" to set Z=0 and you should be able to then check that this point in the mesh is indeed 0 (or very, very close to it).

I'm setting Z=0 with the homez.g as part of G28 before mesh.g gets ran. And homez.g allways probes the bed center,

Maybe i should change the routine in mesh.g to probe Z=0 in the middle of the probe/print area instead of the trigger height lol. So that it's referenced of that instead.

EDIT:
I just found an mechanical issue too, the magnets for the TAP carrier had come undone.

They're now been glued in place with CA-glue and the Z-offset adjusted.

Best practice regarding the previous question still stands though. I feel like setting Z=0 in X/Y center of the area to probe sounds like a better idea than setting the trigger height.

Exerqtor

Some more opinions on setting Z=0 & / or trigger distance in print/probe area center? Maybe both? (before I mark the thread as solved)