This is something I’m discussing with support, so help is in progress, hopefully, but I wanted to post it here as well in case someone else has encountered this issue.
When I started using my HDM and after I tried to just cut something from CutRocket as a test, I noticed some inaccuracies. The test was a box and the lid didn’t fit well, and depth on Y was not consistent. This was resolved by simply squaring the machine manually to the front support.
This problem remains though: I noticed that the distance over X was always less than Y by about 0.1mm. This was measured repeatedly on a circle, diamond, square test, and it also happens on tiny 10mm cuts. Yesterday, I checked tramming. It was off, and I though I found the issue. Trammed and got it square on all sides with both the spindle body and a 1/4 inch rod. Now, this seems to have shifted the error where X is longer than Y by about 0.2mm (see photo). All these are huge inaccuracies for my usage. On my old Shapeoko 3, this is something I had tweaked with those ratios in firmware. How about the HDM? Has someone dealt with this?
I had also tried the diamond square test using HDPE, with same result.
I’m not at the machine right now to try it, but does Carbide Motion allow you to display and set those params or do I have to use something like CNCJS for that? (that’s what I was using with the SO3)
I’m curious why it should be necessary to tweak steps/mm on a machine with ballscrews ? I could imagine its needed on a SO3 where the belts might be stretching - or when first setting up your own DIY machine. But with ballscrews that value should be specific (and never change) to the motors and ballscrew pitch.
I’d be worried that by adjusting that value to account for OPs inaccuracies is just going to mask the real issue.
Even if the ballscrew is (relatively) perfect, the stepper motors are not. The other choice would be to adjust the stepper motor/driver. They are precise (repeatable), but not perfectly accurate.
Well it looks like a 160.42 steps/mm did the trick on Y axis, testing with an 80mm square, getting to 79.98 or thereabouts in both directions, good enough for my calipers.
A weird thing happened on the first test. At some point, X and Y seemed to drift a little bit, then it went to a grind while retracting Z, stopping after a few mm. Maybe the weird cuts on X and Y were due to that. Jogging was no problem after I re-homed. The next test went fine, then the third test did the same grind on Z. I cut this one shallower and there was no X/Y problem. I oiled everything up just in case (I have Vactra 2 oil for my lathes). The next 3 tests went fine so it might have been the problem but weird that it would be already so dry after relatively little use. Anyway, I think the dimensions are working now, I’ll see what happens in real use after my replacement 1/16 end mill arrives, since I broke mine yesterday doing something stupid.
For the Z grinding - check your wiring and connectors. An intermittent connection could drop one of the phases of the motor, which could cause the behavior you are seeing.
