G29 / 3 point bed compensation - always tilted slightly
-
that make zero sense.
G28 Z should establish Z=0 using G30.
G29 getting a different offset would mean that the trigger point is different.
-
@Martin1454 said in G29 / 3 point bed compensation - always tilted slightly:
G1 X30 Y80 F6000 ; G30 ;
@Martin1454 said in G29 / 3 point bed compensation - always tilted slightly:
G1 X160 Y170 F6000 ; G30
Your homeall and homez place the probe at different locations. Make them the same.
-
@Phaedrux Hmm....
That got me thinking - I Use the X30 Y80 because that places the probe not that far from my first lead screw and that saves time since it doesnt have to move when it starts the 3 point leveling - But! since it is not level when it does the first zeroring of Z, will placeing it off center, make the axis it rotates about not centered also? Or since its a plane, it doesnt matter?
But I'll change it when I get home
-
Are you certain that you dont have a dropped corner on one of your axis?
-
You also need to remember there is a huge difference between a surface being level and it being flat...
if it ispossible to use a digital inclinometer you can check the level, first zero the inclinometer to the surface the printer sits on (firstly on the X axis) then carry out 3 point leveling and check how level it is, then zero the inclinometer on the surface on the Y axis orientation and carry out the same 3 point leveling and see what it says. and you will have your answer, which i think might be that your bed it level but not completely flat.
you can also check that your rails are level too.
-
@sinned6915 said in G29 / 3 point bed compensation - always tilted slightly:
Are you certain that you dont have a dropped corner on one of your axis?
Im pretty sure non of the corners is dropped - But even if one corner is lower, shouldent the whole system stay "parallel"?
Im wondering if the M561 in my HomeZ.g have anything with the offset to do?
-
@CaLviNx I dont have one of those tools (yet) but the surface of the bed is a cast aluminium plate that is bought with a really flat surface, so Im suprised if its not flat. As I see it in the first picture, it looks like a flat planar that is just tilted, and not bend.
I'm gonna get the offset working again, and then I'll try to place a bit around with the Z-leadscrews placement and see the impact on the leveling.
-
@Phaedrux said in G29 / 3 point bed compensation - always tilted slightly:
@Martin1454 said in G29 / 3 point bed compensation - always tilted slightly:
G1 X30 Y80 F6000 ; G30 ;
@Martin1454 said in G29 / 3 point bed compensation - always tilted slightly:
G1 X160 Y170 F6000 ; G30
Your homeall and homez place the probe at different locations. Make them the same.
Did this make any difference?
-
@Phaedrux said in G29 / 3 point bed compensation - always tilted slightly:
@Phaedrux said in G29 / 3 point bed compensation - always tilted slightly:
@Martin1454 said in G29 / 3 point bed compensation - always tilted slightly:
G1 X30 Y80 F6000 ; G30 ;
@Martin1454 said in G29 / 3 point bed compensation - always tilted slightly:
G1 X160 Y170 F6000 ; G30
Your homeall and homez place the probe at different locations. Make them the same.
Did this make any difference?
No not really
-
Unless you spent a lot of money the tolerance for flatness and parallelism in cast aluminium tool plate can vary significantly, many suppliers don't list their tolerances other than the cut dimension. Even in cast plates with a milled finish some twist or unevenness can't be excluded, you also cannot assume that the two faces of the plate are parallel, it also isn't clear how you have mounted / constrained the plate, an over constrained plate can twist and bow as it expands when heated.
I don't understand your commands either - Z screws aren't calibrated with G28 ???
Either way a Z screw can only adjust if it is free to do so, if it is over constrained you will be over stressing the plate and probably twisting it to some degree depending on your mounting method - don't underestimate the strength of steppers with lead screws. If your mount to the bed is rigid there are limits to how much a z axis can help overcome errors, they may even introduce some.
You need to make sure it is as close to mechanically level as possible in relation to the X and Y axis before trying to take out any error with the Z screws. You then need to ensure that when you do calibrate Z using the steppers that you don't cause binding on the lead screws for one or more steppers.
-
@Martin1454 said in
Im pretty sure non of the corners is dropped - But even if one corner is lower, shouldent the whole system stay "parallel"?
Im wondering if the M561 in my HomeZ.g have anything with the offset to do?
I don't believe so.
Level is the average of the 3 points that you are probing.
Flatness is what mesh bed leveling is trying to improve.
If your motion system is not parallel to the level plane, then nothing you can do to fix it. Imagine if you have sag in your gantry, you will have a hollow that the controller will try to fox by filling it in wich makes the problem worse.
Not knowing your machine or machanics, I am imagining rods with holders, try loosening and then retightneing the rod holders, or if you can, loosen and them shim them 'down' with a shim of known thickness just to see what the effect is. Even the thickness of a piece of paper will show you clear results.
