RE: VFA Artifacts on X+Y Straight Edges
maybe a barrel-shaped idler with no flanges would do the job. If i remember right, from old beltdriven machines, the barel-shape keeps the belt centered.
Could be printed easily, even with spokes to act as adsorbers ? (i think i shop for some 5mm bearings 🙂
RE:
Another approach - a printable idler 'assembly' that could be printed in TPU without supports. Space for an ID 3mm x OD 8mm x 4mm bearing in each half. Dimensions can be adjusted to make sure it's a tight friction fit over the bearings and the central bearing spacer.
RE: VFA Artifacts on X+Y Straight Edges
Prusa is already a niche product.
Sales in 2025: 100 million compared to Bambulab with 1,100 million.
Those numbers can’t be correct. There’s only 8 billion people on earth (half of whom are poor) …No way BBL have sold a printer to 12.5% of the worlds population…in 2025 😂
RE: VFA Artifacts on X+Y Straight Edges
I think there are two types of users: those who go back and produce beautiful 3D prints, and those who drive recklessly at crazy speeds to go faster. As they say in Italy, “Que va piano va sano...”
MK3S + MMU2 migrée vers une MK4 +MMU3 , puis vers MK4S MMU3, en attente du kit upgrade Core One, Elegoo Mars, bricole sur Home assistant, arduino, 8266, Esp32, kicad, Freecad et Fusion 360
RE: VFA Artifacts on X+Y Straight Edges
Prusa is already a niche product.
Sales in 2025: 100 million compared to Bambulab with 1,100 million.
Those numbers can’t be correct. There’s only 8 billion people on earth (half of whom are poor) …No way BBL have sold a printer to 12.5% of the worlds population…in 2025 😂
Currency units, rather than printers?
RE: VFA Artifacts on X+Y Straight Edges
I assumed 100 million (dollars?) in sales vs. 1.1 billons in sales...
RE: VFA Artifacts on X+Y Straight Edges
Prusa is already a niche product.
Sales in 2025: 100 million compared to Bambulab with 1,100 million.
Those numbers can’t be correct. There’s only 8 billion people on earth (half of whom are poor) …No way BBL have sold a printer to 12.5% of the worlds population…in 2025 😂
$ Dollar
RE: VFA Artifacts on X+Y Straight Edges
Yeah that makes more sense 😂
RE: VFA Artifacts on X+Y Straight Edges
Nice work.
When the Bambu users were trying to find a solution to this problem on the P1S and X1C, a few found improvements by wrapping the idlers in Tessa tape or plumbers teflon tape. This was either creating a slightly compliant surface or creating a quasi toothed idler as the material compressed, or both.
I'm finding some great improvements with slightly modifying the front idlers to have a piece of silicone in front of the square nut to absorb vibrations. I still have some slight VFA but they are now much smaller than before.
I strongly suspect the problem pulleys have a very slightly larger or smaller diameter causing the shoulder of the belt to pinch during engagement and disengagement.
@shushuda Or anyone else who has a good pulley. If you have a micrometer could you check the diameters of those pulleys that are problematic compared to the good ones.I think the best solution may actually be to use polymer pulleys overmolded onto metal allowing a bit of give to each tooth engagement. Maybe someone could print a flexible resin outer to put in a smaller pulley.
RE:
I am thinking of making a pulley with high durometer silicone rubber overmolded on a ppa-cf core. The top left of the image would be the stiff core and the bottom right would be the overmolded silicone. This way it still would have good rotational stiffness, thus accurate positioning, while being compliant to tooth engagement and absorb belt vibrations.
RE: VFA Artifacts on X+Y Straight Edges
You must have access to some very good facilities. CNC machining is one thing, but overmoulding is rather a specialist technique isn't it?
RE: VFA Artifacts on X+Y Straight Edges
I wish I had access to a materials lab. I don't have my resin printers in the country I'm living unfortunately but I'm going to order some resin prints to cast it into as the mould and just use a two part silicone with some additives such as ptfe powder to modify the properties. The best would be graphene but you need ultrasonic homogenisers to get that into anything really and it scares me. Will need to coat the resin prints so the silicone cures though which might change the geometries more than I'd like.
RE: VFA Artifacts on X+Y Straight Edges
It'll be very interesting to see your results. I have some bearings arriving today, so I'm going to experiment with the TPU idler. Although my printer is now running the 1.5mm belt conversion, I do still have some speeds with rather pronounced belt ripple, so if I concentrate on those speeds it'll be easy to see if the TPU idler makes any difference - good or bad. I image there'll be some impact on the ideal belt tension. Fun to experiment though.
RE: VFA Artifacts on X+Y Straight Edges
I had the thought of moving to a larger diameter idler bearing and putting a TPU bushing inside the bearing and over the bolt, however I think an elastomer pulley will be the best combination of absorbing vibrations while retaining positioning precision. Accuracy may suffer though.
RE: VFA Artifacts on X+Y Straight Edges
I think dialing back the corner velocities/acc/decel etc will help maintain accuracy with the compliant idlers. The downside is more time spent in potentially belt ripple problem speed zones after corners etc until desired speed is reached.
I had the thought of moving to a larger diameter idler bearing and putting a TPU bushing inside the bearing and over the bolt, however I think an elastomer pulley will be the best combination of absorbing vibrations while retaining positioning precision. Accuracy may suffer though.
RE:
I think dialing back the corner velocities/acc/decel etc will help maintain accuracy with the compliant idlers. The downside is more time spent in potentially belt ripple problem speed zones after corners etc until desired speed is reached.
I had the thought of moving to a larger diameter idler bearing and putting a TPU bushing inside the bearing and over the bolt, however I think an elastomer pulley will be the best combination of absorbing vibrations while retaining positioning precision. Accuracy may suffer though.
Problematic speed zones have gone away pretty much completely, now it's just a little bad everywhere. As it no longer seems to have different frequencies I think it's just the teeth pinching and deforming my compliant idler. Attached is the only bad face of a VFA tower, I can just feel the vfa on the slower speeds but the top feels like a smooth surface.
RE: VFA Artifacts on X+Y Straight Edges
That's really impressive. So if the frequency-dependence has gone, that suggests you've solved the resonance problem through damping, and now it's just a mechanical/meshing issue. But your compliant idler runs on the smooth side of the belt. So apart from the meshing compliance that you're tackling with the elastomer pulley, is it possible that the two idlers that run on the toothed side of the belt are now having an effect?
RE: VFA Artifacts on X+Y Straight Edges
I suspect they won't be an issue since the EPDM of the gates belts deforms onto them and from the literature smooth idlers don't seem to add much energy, imprecise toothed idlers can be far worse. That does however make me wonder about how much smooth idlers will deform the belt teeth over time and lead to engagement issues, so maybe a stiff self lubricating elastomer idler will improve service life.
That's really impressive. So if the frequency-dependence has gone, that suggests you've solved the resonance problem through damping, and now it's just a mechanical/meshing issue. But your compliant idler runs on the smooth side of the belt. So apart from the meshing compliance that you're tackling with the elastomer pulley, is it possible that the two idlers that run on the toothed side of the belt are now having an effect?
RE: VFA Artifacts on X+Y Straight Edges
Just want to get the 1000th post in this thread 😀
Maybe someone should look into patents for idlers/pulleys and check if there's one for soft-surface-anti-vibration coatings or bushings.
Sounds very interesting and the silicone washers for the tensioner's screw sounds quite easy and efficient.