X and Y dimension difference on prints

Hello all,

I have a sort of diy cartesian printer, frame made from 2mm bent sheet metal, 8mm steel guide rods with linear bearings, trapezoidal screw and bronze bushing for z, a heated 5mm alu sheet with painters tape on top as the bed, 3d printed parts holding everything together.

Now the problem: basically it prints ovals instead of circles and rectangles instead of squares so to say. The x dimension always comes out slightly bigger than y. It is not entirely constant. But about 0.2-0.4 mm depending on the size of the print (0.2 on a test cube 10x10, then 0.4 on a circle of diameter 80mm and about 0.2-0.3 difference on the same test cube only 20x20 with same settings to see if the difference is scaling with print size..not sure if it is. If yes, then itsbnot much)

What ive done: checked settings, steps per mm, made sure my pulleys are really with same teeth numbers, tentioned the belts as much as i can, basically to the point where the motors have a chance of stalling during prints..not liking that. No changes.

I am stumped. If i change steps per mm the prints will scale and only be correct-ish size for parts of a size that the steps per mm were calibrated for.

Any help appreciated a lot.


  • Seems like you suffer backlash. Every time you switch direction it takes a bit until the motor movement is translated in extruder movement. So some part has some play from your number 0.2-0.4mm.In circles you normally see it best.
    In repetier-firmware you can define that you have it and compensate, but that might cause speed changes and is really only a last solution. Quality will be much better if you can remove the backlash. It might be just something simple like belt has some play between tooth or pulley is not 100% fixed by a screw on the flat shaft and has some play when reversing directions, or extruder fixture has some play. Something like that has play and causes backlash.

    As a good test print blocks with different sizes in same direction. Say 50 and 100 mm. Difference is 50mm and if not 50mm then axis steps per mm is wrong with that amount. The remaining difference is caused by extrusion multiplier just extruding too much so all sides get a bit wider or backlash. Comparing x and y direction gives then more hints as the wider extrusion should be nearly equal on all sides while backlash must not.
  • Hey, thanks for the answer. I actually did suspect backlash, thats why I tightened my belts. Then I thought maybe my extruder is the culprit, since it was really not rigiddly situated (a ghetto solution for the start), I printed a new structure, mounted it, and still the same results. Set screws are tight, belts too, I just dont see where the backlash can come from.
  • edited February 2018
    sometimes pulleys get loose on motor shaft , that´s hard to see ,but as it´s a small number it might be the pulleys are not centered, had some of those cheap chin***  with this issue a couple of weeks ago.
    if you swap pulleys from x to y and vice versa the error should move also
Sign In or Register to comment.