setting z tilt probe points/bed mesh min max. vzprinthead/bltouch/vcore 400
I am trying to figure out the probe points for z tilt, and the min and max for bed mesh. Any assistance would be greatly appreciated
18 Replies
what error are you getting with the current ratos setup?
quickest-silver•2y ago
Im not getting a error, the z tilt points just don’t line up that well. The front left appears to be 20-30mm from the edge of the pei sheet while the front right side is right in the edge.
you probably want to change the probe offsets then
https://www.klipper3d.org/Config_Reference.html#probe
quickest-silver•2y ago
I’m just afraid it’s affecting the accuracy.
quickest-silver•2y ago
Oh I know how to change them, just trying to work out what to.
https://www.klipper3d.org/Probe_Calibrate.html#calibrating-probe-x-and-y-offsets gives you the steps to do
quickest-silver•2y ago
Good deal, thank you.
It's not 🙂 z_tilt probe points can be anywhere on the bed.
quickest-silver•2y ago
Just making sure I understand what you mean, if the front left point is say 20mm from the edge, and the front right is 40mm from the edge it’s not gonna affect anything? I’m just trying to get it to where the front probe points are both the same distance from the edge if thst makes sense.
What you said does make sense cause it’s trying to@make the bed flat within a certain tolerance.
Just making sure I understand what you mean, if the front left point is say 20mm from the edge, and the front right is 40mm from the edge it’s not gonna affect anythingIt'll select a different subplane of the plate than a different point layout would. In general if you want to min/max your bed mesh variance, it's a good idea to look at your mesh to determine which 3 points would make for the flattest adjustment. It's very rarely right next to the lead screws. Sometimes you want a very weird shaped triangle to make sure the subplane best matches the shape of your bed.
I’m just trying to get it to where the front probe points are both the same distance from the edge if thst makes sense.That doesn't necessarily make anything better, it's mostly an OCD thing. If you're moving them anyway, move them to where they actually benefit your mesh.
crazy Q here...(inquiring minds want to know) ref'ing your 'any three points' comment above @miklschmidt ... for the larger build plates... with many people having X0/Ymax seeming lower most of the time.... would it be safe to even EXPERIMENT with setting the three Z_tilt points to say 50, 450 250, 50 450 , 450 ...compare to a norm of 50,50 250,450, 450, 50, (rotating the triangle 180 degrees from the z screws) I'm wondering how Klipper would handle adjustments. I'm will to test...just don't want the F up my printer 🙂
would just be intersting to see what klipper wourl adjust and by how much
Yes, i've often done this and wonder if maybe that should be the default.
It works well
Do it
mathamatically, i'd be inclined to believe Klipper would correct Z in the opposite direction needed. If the 'new' left rear corner were lower (while thinking it's the right front corner, it would adjust the 'old right right corner' up making the real left rear corner even lower. I think the Z motor positions would have to rotate on the Z axiz as well. THAT IS unless you are suggesting, the system would pick which motor to adjust based on the 'positions' data in printer.cfg.?
it doesn't "think it's the right front corner" you're probing. It knows exactly where you're probing and where the pivot points are, so it can extrapolate, that's why any arbitratry Z probing points work.
THAT IS unless you are suggesting, the system would pick which motor to adjust based on the 'positions' data in printer.cfg.?Exactly Also it adjusts all motors at once to converge on the plane you've probed. So like "given the plane defined by these X points (which are out of level), pivot point a / b / c should be at a1 / b1 /c1 (to be level) but they're are at a2 / b2 / c2, adjustments = a2 - a1, b2 - b1, c2 - c1".
I guess we're kind of saying the same thing? .... the probing points set in printer.cfg ....irrespective of the physical arm mount points...Klipper still will know which arms to adjust based on the probed data. (stated correctly?)
If so, that creates a quandary.... (that I'm assuming is not currently in klipper.)...what would (if any) be an advantage/disadvantage of using a four corner probing like a voron? sure They have four Z's to adjust....I'm more thinking about the leveling algorithm taking into account a bad/weird corner reading...and shifting the table a bit to bring more of it into level. I guess in reality the leveling routine would really only be spreading that bad corners data further into the field of probed points in the array. Again..perhaps way overboard.
Interesting...just tested the reverse 3 point z_tilt just discussed. Each test run = Home_all Z_tilt 9,9 mesh cals Normal FL RC FR
Was surprised so ran it three time alternating
What I find most interesting was the repeatability. Yet a ~ 0.02 diff in the two sets of tests. z_tilt 3 pt postions Test1 - 6 (T1-T6)
FL RC FR T1 0.149 T3 0.151 T5 0.150 RL FC RR T2 0.172 T4 0.170 T6 0.171
What I find most interesting was the repeatability. Yet a ~ 0.02 diff in the two sets of tests. z_tilt 3 pt postions Test1 - 6 (T1-T6)
FL RC FR T1 0.149 T3 0.151 T5 0.150 RL FC RR T2 0.172 T4 0.170 T6 0.171
what would (if any) be an advantage/disadvantage of using a four corner probing like a voron?They use quad gantry levelling which works a bit differently. But if you define a fourth (or fifth, sixth, seventh) z_tilt point, i assume, that it constructs an average plane from them, but i actually don't know what happens, been curious about this myself but never tried.
looks like play time for me. Since i'm about to reconstruct some of this printer anyway, If i F it up, no big deal ha!