Hi all.
I’ve noticed for a while that things I carve out are slightly longer in the X dimension than it should be. It’s not really noticeable at 10cm (maybe 0.2mm) but at 30cm it is off by 1mm. The Y axis doesn’t seem to have this problem so a circle is slightly oval.
Listening to a 2mm deep cut of a 5cm diameter circle with a fresh short 1/8" bit the finish pass in the x-travel direction is audible (so cutting the bottom and top of the circle) but in the y travel direction (cutting left and right sides) it sounds like it’s cutting air. So it seems there is some stretch in x making a shallower cut in the x dimension when cutting in the y direction (does that make sense?).
Anyways, thoughts on what could be happening? Too much tension on the x belt making the stepper math incorrect? Going Home seems accurate so it seems like a scaling thing. Any thoughts on how to debug this?
Thanks!
I am fighting a similar issues, I noticed I could see my x belt move ever so slightly when cutting x direction and I have been trying to really tighten the x belt. It seems to help but I have not completely eliminated it yet.
How much are you leaving for a finishing pass?
I suggest putting a pen mount back on. Draw large squares and measure X, Y and the diagonal. If the problem occurs with the pen and the machine is otherwise cutting correctly, then I would suspect the belts. There have been a couple of belt problems posted to this forum that end the end was caused by using steel reinforced belts, and having the steel break due to the tight turning radius resulting in belts with more stretch.
You can adjust the steps per mm to compensate. It can be done in firmware it in this line in configuration.h (your initial values may be different):
#define DEFAULT_AXIS_STEPS_PER_UNIT { 200, 200, 800, 200 }
It can also be done in g-code using an M92. Once done using g-code, changes to the firmware will be ignored unless a factory reset is done (M502).
If the problem does not occur with the pen, then I would suspect some sort of deflection, though usually deflection does not scale uniformly.
If you are having to do this, there is some underlying issue.
My issue was not so much the belt but the tie straps. I could see them stretch ever so slightly which is where the tension helped
Thanks for the suggestions.
I’ll set up this test next.
I’ve been running the machine for a while, I may also do the test squares with a v-bit to see what happens.