Stripped Belts

[Polygonhell]

This is really a follow up to this http://forum.seemecnc.com/viewtopic.php ... Belt#p4399 post in the troubleshooting forum, but I thought it was worth point here.
The last couple of days I've be struggling with some print issues, and the fact every time I start a print it seems to be out of calibration, my first thought is I was having problems with limit switch repeatability, but I've been able to zero the machine to 0.05mm or better relatively easily and that wouldn't affect the doming which was also off.
I was moving one of the limit switches, so the contact point was at the base of the lever rather than the end, when I noticed this

Two of the belts are bad (pictured) in exactly the same place the 3rd is fine, it's hard to know what caused this, I haven't slammed the head into the glass at any point.
I did increase the tension on my belts recently to try and address some slight backlash I'd noticed, but it's hard to say if there was already an issue with the belt and increasing the belt tension made it worse, or increasing the belt tension broke fibers in the belt and that caused the pulley to eat it. There was some discussion of burs on the early drive gears, but I'd expect more even wear if that were the issue.

Interested in peoples theories on cause and effect here I have a fair few hours on the machine at this point, but nothing compared to my other machines.
I've ordered some replacement belt from SDP-SI, so we'll see if it recurs.

[geneb]

I would check the teeth on the drive gear to make sure they have no sharp edges. That's what got me.

g.

[Polygonhell]

I would check the teeth on the drive gear to make sure they have no sharp edges. That's what got me.

g.

But I can't see how that would cause this it's only one spot and very localized.

[MorbidSlowBurn]

Out of curiousity where on the belt was the wear? Was it near the drive gear when the print head was in the home position or at 0 Z height? Was it at a zero position for any of the towers when performing a bed leveling?

[Av8r RC]

I had a issue just like this. I attributed it to the belts being too tight, one of mine actually broke on about the 10th layer after I tightened them.

[Bill Havins]

Out of curiousity where on the belt was the wear? Was it near the drive gear when the print head was in the home position or at 0 Z height? Was it at a zero position for any of the towers when performing a bed leveling?

Tell us a little more about what you're thinking here, please.

[MorbidSlowBurn]

I am trying to think of the underlying root cause. Knowing what position the print head is in when the suspect section of the belt is engaged with the drive gear could indicate an impact with a surface or help narrow down where to look for additional information.

Another thing that could be present (just brainstorming) is that as the print head gets closer to a tower and therefore the arm is more vertical, more force could be exerted on the belt which could wear the belt quicker. This could be compounded by extra friction in the ujoints. This is one major difference between a linear driven 3D printer and a delta type.

It could also be as geneb suggested in that the drive gear is sharp.

[Polygonhell]

Out of curiousity where on the belt was the wear? Was it near the drive gear when the print head was in the home position or at 0 Z height? Was it at a zero position for any of the towers when performing a bed leveling?

Closer to the base than the top, it's possibly it's in the range of belt that's with the head on the glass plate depending on position, the belts aren't on the machine now so I can't be exact.
The obvious cause is grounding the head into the glass plate and the belt slipping on the gear, to do that though once it's calibrated you have to lose steps while printing a low layer, I guess that's possible the tooth profile on the GT2 belts is very shallow and it wouldn't take much.

[cambo3d]

how about a glitch in the software/firmware, marlin is known to have jerky movements from what I'm reading on this forum.

what if there was a slight delay in axis's, and another axis responding faster? Basically the axis fighting against another on certain moves or gcodes causing the belt to strip on one axis, then eventually breaking into the inner fibers of the belt which get caught on the pulley teeth and then causes this. Maybe this happened before you switched firmwares.

just an idea, to throw out there.

[aehM_Key]

Basically the axis fighting against another

This is not possible...or I don't understand what you try to say.

[MorbidSlowBurn]

I agree that the axis cannot fight one another. But what could happen is if one axis didn't move when commanded a collision with the build plate could occur. This is where the fine line between missed steps and belt damage is blurred.

[Polygonhell]

FWIW I am back up and running as of last night, when I got my replacement belt from SDP-SI.
Having just reinstalled the belts, there are a couple of things I observed.

1. You can absolutely have your belts too tight, after I tensioned mine, one of the carriages had more tension than the other two and was considerably harder to move by hand, given I knew they were all comparable without the belts I slightly released the tension on that belt and they were all pretty much equivalent again. I'd say the radial load on the stepper was increasing the required force to move the carriage, and I suspect it would adversely affect the print.

2 I'm now pretty sure that I know what happened, I think at some point I damaged the belt, probably a head crash of some description, that introduced the backlash I was seeing, when I over tightened the belts to compensate the belts failure became obvious as seen in the first picture.