I have a Atomstack A70 max and lighburn is cutting ovals and not circles and also not cutting perfect squares.I have researched this issue and it says there is a issues with the machine having loose equipment. I have reassembled my Atomstack thoroughly inspecting that all the parts are fitting correctly and all bolts etc are tight but the issue still persist. I have also have a brand new A70 and have thoroughly set this one up and the same issue persists has anybody found a solution. I think there is an issue with Lightburn programming.
I, same as @thelmuth , see nothing wonky about your settings. @baricl1317 Questioned if you had the rotary enabled. Can you show us a screen shot (WinKey+SHFT+S) of the Rotary Settings window?
I have an A10 with frame extension and 40w head, it also cuts circles out of round, how did you calibrate sizes, squares and rectangles do appear to be correct for size.
I print or cut a square of a desired size (e.g.100mm x 100mm) then carefully measure the result with calipers. Under Laser Tools => Machine Settings => Calibrate Axis select each axis and enter the desired size and the measured size. Simply let Lightburn calculate the correction and save the result. I then print the desired square again to check. If the calibration was way out to start with, I’ll repeat the adjustment.
The cut result is smaller by the width of the kerf, so you must add the kerf to the measurement to get the true machine motion. A diode laser typically has a different kerf width along the X and Y axes, so each axis has a different adder. Worse, the kerf width depends on the material.
Which is why the recommended procedure uses lines / marks on the surface: the center-to-center distances are easier to measure, don’t depend on the material, and require no kerf adjustment.
Until now I’ve always used an open caliper with one end resting on a fixed point and the other end against the moving point (the axis I want to measure). I give a command for a movement of a certain distance (usually the longest possible) and compare the value of the command given with the actual distance traveled.
But the method you indicated also seems quite effective and I will test it.
Thanks!!!
