Verticle belt movements on idlers

I am beating my head against the wall trying to figure this issue out. My A and B belts are moving up and down on the idlers when I move the x carriage left and right. When I move the carriage to the left the A belt (top) moves up on the idler, the B belt (bottom) moves down on the idler. It reverses when moving the carriage to the right. This also happens when moving the carriage back and forth. The A belt moves down when moving forward and the B belt moves up when moving forward.

I am using 9mm gt2 belts, the idlers have an 11mm body within the flanges. The design should have the belts riding in the middle but they move all they way up and down, so 1 full mm of travel. I have double, triple, quadruple checked the idler heights and they are all on the same plane, well within 1mm of tolerance.

Is there some other flaw that could cause this? I can attach a video or images of the build if needed.

Tilted pulley axles can cause this. Are the pulley axles supported at the top and bottom?
It's probably not going to affect print quality unless the belts start climbing onto the pulley flanges.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

All Idlers are on shoulder bolts, supported on the top and bottom. I will double check that none are leaning. I did move some components from heat set inserts to captive nuts to help with leaning idlers. I'm not as concerned about the print quality as belt life. The belts are touching the flanges and I am worried that will wear on the belts quickly.

Well, I have all the shoulder bolts updated to use captive nuts and they are not leaning but I'm still getting vertical belt movement on the idler. Could this be caused by my belt holding? The belts attach to the xy carriage by pressing a small plate against them using a grub screw. So the belt is compressed a little bit, far less than a mm.

Is there a good way to troubleshoot idler by idler? It's so hard to tell where it initially gets off track.

Edited 1 time(s). Last edit at 08/23/2022 04:05PM by AgentNoise.

t might be very hard to detect the tiny amount of tilt that would be required to cause the belt to move up/down. As I said, it probably isn't a problem as long as the belt doesn't climb into the pulley flanges.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]