Skewed Prints on CoreXY

Gamefanatic3D

@Phaedrux said in Skewed Prints on CoreXY:

@Gamefanatic3D said in Skewed Prints:

double checked my X axis alignment. The X axis was off about 3mm

What exactly do you mean by this?

Based on your picture I adjusted the distance between P1 and P3 to be equal to P2 and P4. So that the X-Axis extrusion was parallel to the front and rear extrusions.

Are your belts parallel where they need to be?

Not exactly sure what you mean by this, but they are straight. My belts are aligned such that there are no angles in their runs. Where as in this picture Belt J icomes in at an angle between β and P3, mine would be aligned straight.

@mrehorstdmd said in Skewed Prints on CoreXY:

Maybe more importantly, are the X and Y rails square?

When you tension the belts, the X axis angle relative to the Y axis will change. The end point of the tension adjustment is when the axes are square.

If I understand you correctly, I adjusted the tension on the belts to align my X-Axis parallel to the front frame, but only after I ensured that the frame was square.

aidar

If mechanical adjustment does not help, try this : https://duet3d.dozuki.com/Wiki/Gcode#Section_M556_Axis_skew_compensation

Veti

can you post a picture of your corexy setup from above?

mrehorstdmd

@Gamefanatic3D Are the Y axis guide rails square with the frame?

If the rails aren't square, the prints won't be square. You can use software to compensate for skewed rails, but why wouldn't you just square up the rails instead?

Squareness can be verified by measuring diagonals of objects that are supposed to be square (prints or printers). If the diagonals match, the sides are square.

zapta

Can this be a configuration issue?

E.g. if one of the motors has higher steps/mm configuration than it should then the diagonal it controls will be printed shorter than the other.

mrehorstdmd

In the diagram posted above, belt segments labeled A-H have to be parallel to their respective guide rails. The other segments don't matter. See: https://drmrehorst.blogspot.com/2018/08/corexy-mechanism-layout-and-belt.html

If you want to be sure the guide rails are square and belts are properly arranged, measure the distance from P1 to P4 and from P2 to P3. If they are equal, things are probably square (assuming some symmetry in your build). That should be true no matter where the extruder carriage is positioned along the X axis and no matter where the X axis is positioned along the Y axis.

SIam

hi,

your X and Y stepps musst be the same if not you get your results

Gamefanatic3D

I did tweak my X-Axis a bit more and got a change that I am able to track the effects with a test cube and know I was adjusting in the wrong direction. Apparently measuring against the front frame seems logical, but the real issue may be with the Y brackets. I noticed it wasn't perpendicular to the setup, and that I am trying to correct that using tension on the belts. I don't currently have offset wheels to allow me to correct the Y mount geometry. Will get those if I must, but working towards installing side rails in the near future.

Good concept about the steps should be the same. The motors are the same 0.9° model from OMC, so it makes sense they shouldn't be all that much different if not identical. I know when printing my 100mm square that I get about ~100.3 mm in one direction and ~100.5mm in the other. So a little more tweaking needed there.

You have all given me some things to try. I will report back my findings. I am definitely learning something here.

Veti

@Gamefanatic3D said in Skewed Prints on CoreXY:

100mm square that I get about ~100.3 mm in one direction and ~100.5mm in the other.

that might be due to extrusion.
print the same cube scaled to 200% or 50% and measure again.

SIam

@Gamefanatic3D said in Skewed Prints on CoreXY:

I know when printing my 100mm square that I get about ~100.3 mm in one direction and ~100.5mm in the other. So a little more tweaking needed there.

then your belt' tension is not not exactly calibrated they must on both sites X/Y the same.
you can print a cube flip it 45° you can see when you print it that always only one stepper is working.
Bild Text

The right(R) side is for the right Motor and the left(L) side for the left motor

when you finished your print messure the qube

if R > L then then is the belt tension R stronger as the L side
if R < L then then is the belt tension L stronger as the R side

This you have to repeat until you are happy with the result (and this take time)

Gamefanatic3D

@SIam said in Skewed Prints on CoreXY:

hi,

your X and Y stepps musst be the same if not you get your results

Okay, so this had the greatest impact. Setting X/Y steps to equal has things almost perfect. Diagnal measurement of print in nearly negligible, and probably have to adjust the belt tension ever slightly to get this final dimension dialled in.

Thanks so far! I will post some pics when I get a moment this weekend.

zapta

CHEP had recently a video about belt tensioning. https://www.youtube.com/watch?v=_m9TTzEh5xM

"It [almost] can't be tight enough"

Gamefanatic3D

It's not perfect, but darn close! I just about have the steps dialed in, think I need to put a bit more into extrusion and the orthogonal is in the money...

Adamfilip

you need to adjust belt tension until gantry is square. too much tension on one stepper or the other will skew the gantry

mrehorstdmd

The motor shafts are only 5 mm diameter, and the bearings can only take so much load. If you're using the "standard" 3D printer pulleys with the microscopic bearings, excess belt tension will wear them out quickly. Set the tension so the belts don't flop around while the mechanism is moving and you'll be fine.

Adamfilip

you dont need excessive tension but the belt tension will effect the squareness of the gantry. on my corexy I need to set the one motor/belt to be noticeably tighter in order to get square prints out of my printer. Equal tightness does not mean a Square gantry. needs to be adjusted and checked for squareness