OK, bit of a weird one here. So I'm creating rotary 4th axis spiral toolpaths, and they look fine. But when carved, I get some strange and very annoying artefacts showing up. Maybe easier to just show you. First in fusion, with a section of toolpath shown. Second, how it actually carves. Third, using Mach4 to show the actual toolpath. This is a carved wooden item (here in ebony) about 22mm diameter.
The toolpaths are done with 0.14mm step over, 0.02mm tolerance. Making tolerance smaller down to 0.005mm (5 microns) improves, but is not ideal obviously, and it does not fix the issue just makes the problem a little smaller. Also, I went through a gcode file by hand, comparing Z height on successive lines at the same A angle, and found the deviation was more like 0.1mm than 0.02mm. I also tried turning the model into a mesh (tesselating) and then creating new toolpaths, but same issue. Finally, I tested a very simple square, and again exported toolpath into mach4, with a similar (if not quite so random) result:
So I think the problem is in post. I'd try a different one, but can only find mach3 and uccnc posts that have rotary you can specify to an axis and inverse time feedrates. At first I actually thought my machine was at fault and ahh... I might have completely redesigned and rebuilt a new Z axis including a high end ball screw etc. Not that I'm salty or anything about all the time and money, of course....a popular phrase about assumptions spring to mind... Anyway!! Any ideas?! Right now my workaround is to 0.005mm tolerance, and two almost identical toolpaths but with +1 and -1mm offsets, as this offsets the position of the artefacts. But it doesnt actually fix the core issue, and I'd rather the carving just came out nicely the first time :).
Any devs who want to look at the files, please contact privately!
