FFSAKE I put so much effort into my belt alignment and now this
I put so much effort into getting my belts evenly tensioned or so I thought. I was confident I had eliminated all friction as much as possible and now this WTF. I guess I should replace the belts now presumably. My understanding is that you can get gates ones which are better quality, anyone know where from? Also any tips on how to ensure its doesn't happen again? I scrutinized the pulleys and belt motion so this baffles me.
37 Replies
multiple-amethyst•2y ago
I have a feeling that the bearing stack design is not so good since the flange is not big enough. Also the belt coming from RatRig is not of amazing quality.
I agree. My belts were fine until I upgraded to 3.1 I absolutely think the flanges on the bearings are too small to keep the belts in place.
If all your pulley shafts are truly vertical, and pulleys are properly spaced, there is no reason for them to 'walk' off the pulleys. Rim to rim Looking closely at that pic..... it looks like the toothed side is what is deteriorating. The bearing stack is 10mm flange to flange. You belt is 9mm (actually 8.80mm) wide. 'Technically' the belt should never even touch the flanges... yes I know...in a perfect world.
QUESTION.... is that fiber stuff building up anywhere else besides what you pictured here? If not, you've found what is rubbing/wearing the belt! Looking at your graph, tells me your belts are a bit on the loose side, (freq range) so could easily be wearing against the aluminum rails (belt return route) There is negligible clearance there. The belts are stiff enough to not bend perfectly around the pulleys as they exit pulleys heading towards and away from the extrusions. You should be able to look down the length of the Y extrusions and see space between belt and extrusion. (albeit small space) If not, it just further supports the idea of belts not tight enough.
QUESTION.... is that fiber stuff building up anywhere else besides what you pictured here? If not, you've found what is rubbing/wearing the belt! Looking at your graph, tells me your belts are a bit on the loose side, (freq range) so could easily be wearing against the aluminum rails (belt return route) There is negligible clearance there. The belts are stiff enough to not bend perfectly around the pulleys as they exit pulleys heading towards and away from the extrusions. You should be able to look down the length of the Y extrusions and see space between belt and extrusion. (albeit small space) If not, it just further supports the idea of belts not tight enough.
metropolitan-bronze•2y ago
Thanks for your input guys I will do a postmortem over the weekend
multiple-amethyst•2y ago
In the ideal world it would indeed stay in place. However, it looks like my belt has some uneven local tensions inside.
I recommend doing the dual motor plate mod from ffb and get rid of the printed parts of the motor cage
much better visibility, alignment and access
adverse-sapphire•2y ago
Ill probbaly do the dual plate mod too as I had the same issue. I moved my drive gear on the motor down some though and it stopped the rubbing. But still the bet is bad looking now.
I personally never ran any of the oem RR supplied belts. I went right to the high temp real Gates EPDM belts right from the start. Much stronger, can handle temps better heavier duty, let alone not a knock-off.
to answer your orig 'source' Q...I got mine from
https://west3d.com/search?type=product&options%5Bprefix%5D=last&options%5Bunavailable_products%5D=last&q=GATES+GT2+EPDM+High+Temperature+Open+Belt+LL-2GT-E+%286mm%2F9mm+wide%29
West3D Printing
Search: 1 result found for "GATES GT2 EPDM High Temperature Open Be...
West3D specializes in sourcing and selling the best 3D printing products at affordable prices to inspire the next generation of makers.
There is no such thing as perfect alignment. Everything made has a range of tolerance. Even the aluminum extrusions are not perfectly straight. A good corexy design is one that can handle an expected level of small imperfection in the build without belts eating themselves. Because the 3.1 flanged bearing approach results in guides with much smaller grooves to retain the belt position, the result is a printer that requires near perfection in the build to prevent premature belt wear. I built my 3.0 with the same level of care that I did the 3.1 upgrade. I had no substantial belt wear on the 3.0 machine after many hundreds of hours of printing. The 3.1 machine started eating the belt after maybe 3 days and I can't see where my build is bad.
metropolitan-bronze•2y ago
Does anyone make proper pulleys for the 3.1 with soft contours at the edges?
You should be able to find a common frame reference point, and validate that all the pulleys are on the same plane (level) I did not rely on just the oem instructions for spacers (maybe because I mod'ed stuff as I went) but belt rubbing is not an issue that is not correctable by moving spacers around and making sure bearing stack pivot rods are all parallel to each other across all corners of the machine on the same plane.
metropolitan-bronze•2y ago
WTAF! £116 for 5.8 metres
??? oh.... s/h costs? 60x100 was 89.40 USD (84.85 EUR) for me $7 in s/h
optimistic-gold•2y ago
@elcoj where can I find some details about this mod? Are you talking about awd?
Printables.com
V-Core 3 motor mounts (aka FMMM) by RaYa | Download free STL model ...
Full Metal Motor Mount (FMMM) replace printed motor cage parts by metal plate | Download free 3D printable STL models
metropolitan-bronze•2y ago
metropolitan-bronze•2y ago
did you select 6mm perhaps mate?
If that MOD resolves belt rubbing then it suggests there is an issue with the design of the part.
I think the clearances are a bit small with the printed part, but at least with this mod you can see your belts better
and it is much easier to work on
optimistic-gold•2y ago
thanks
metropolitan-bronze•2y ago
Yea I 100% agree that would be a huge benefit. I would be a little concerned I would lose heat from the enclosure however
metropolitan-bronze•2y ago
Hmm... Look what I found https://www.printables.com/model/314241-ratrig-v-core-31-anti-rub-xy-motor-cages
Printables.com
RatRig V-Core 3.1 anti-rub XY motor cages by Arthur_C | Download fr...
Increased belt clearance for the XY motor cages to prevent belt rubbing | Download free 3D printable STL models
The bottom plate closes the enclosure,the top plate is fully enclosed
So i don't understand which problem you see
metropolitan-bronze•2y ago
Oh I see
Didn't have the back panel in at the time of the photo
And didn't have the right spacers, fixed that later, hence the gap between bearings
Too lazy to walk to the printer so cropping an old photo 🙂
multiple-amethyst•2y ago
It is easier to route the belt and align the pulleys. Besides that it doesn't really solve the problem. Every second V-Core seems to have the same belt problem at exactly the same spot.
I would like to see the frame designed in such a way that it naturally assembles square. The current approach with loose fitting t nuts and corner brackets that aren't even square to begin with make it a lot tougher. Of course saying it and doing it are two different things. It could get very expensive.
multiple-amethyst•2y ago
I don't think that's really possible and I like the current frame design. There are always tolerances in manufacturing. The current design allows you to compensate the curvature of the extrusions, length deviations and so on.
It's more possible for them to machine corner plates to very tight tolerances for end users to use as handy references during assembly than it is for end users to blindly square up a frame in 3 dimensions with 6 degrees of freedom at every corner lol.
multiple-amethyst•2y ago
It even has been done by one member of this server. But yes, this is expensive.
I used L squares, and long carpenters clamps. As well as motorcycle ty-down straps to pull diagonals worked great
metropolitan-bronze•2y ago
me too exactly the same procedure
I did too. It's still quite difficult because things are often in the way of getting a good measurement.
100%, its so difficult to make this frame square