I’ve been running my MPCNC pretty well for the last couple of months and then today all of a sudden circles were not circles anymore!
I cut a small circle 2 days ago and its bang on (ish) at 34.2mm any way I measure it. Then today I found I was getting ovals, stretched on the x-axis. I cut a 80mm bore (using fusion and flyfisher604 processor) and it measured 79.9mm on Y and 82.3mm on X. I checked all the belt tensions and truck tensions and it all seemed okay, but the back X belt seemed a little slack.
This is where I have to come clean, I bought my parts from a super shitty local distributor here in South Africa, and had to completely rebuy all of the bearings from the V1 shop. The other issue was the belts were steel reinforced, but not being able to find an easy local replacement I trucked on.
My guess is that I f***ed a belt up and its gotten stretched out, causing the weird stretching on the x-axis.
I also ran some calibration tests (the vector76 rulers, and one 100mm circle) and found something weird. All the tests were done without moving the paper. You can see my measurements in the picture below, it seems like there is a section where the x-axis is stretched, which my guess is where the issue on the belt is and then later the error goes back down.
Anyway, can someone potentially confirm that this is likely a buggered belt and I should get fibre reinforced ones? Or potentially advise if you think its something else?
I second MakerJim. Steel reinforced belts will fracture over time due to the tight turns around the pulley. This will result in uneven streatching of the belts. We’ve seen this issue before on the forum.
The beginning and end of your circles match, ruling out most other causes for a difference in the X and Y movement.
Definitely want Neoprene Rubber With Fiberglass Core like these…
Curious if anyone’s able to recommend a parts supplier more local to South Africa? Even if local means a reliable AliExpress supplier.
In the short term, depending how immediately you need to finish the job, you could try chasing the stretch by adapting your steps per mm. Do you have or want guidance for this? Even this workaround may not help if belt stretch is uneven, and localized to specific segments .
This will be a losing battle, but am only suggesting incase you have a time crunch for a current project, and can’t wait for delivery of some nice Fiberglass Core belts. Consider lower feedrate so less force until new belts arrive.
I like the calibration circle and ruler test small enough for calipers.
How big is your usable XY area? How far off is actual motion, versus expected motion, when you move 2’ or 600mm or so in each direction?
Ah thank you! Yeah luckily I managed to get the time dependent project out before the stretching became noticable
I have just decided to get some locally, there are two options (incase any other South Africans see this), Netram in Cape Town and diyelectronics.co.za online.
And avoid Stelltron like the plague (had to replace all the bearings and now all the belts, even though I asked in advance if they were steel cored before ordering…)
.