RE: VFA Artifacts on X+Y Straight Edges

My mockup accidentally removed the parts that keep the original pulley centred.  Fixed now, with an enlarged spindle hole to allow it to move and generate the required friction.

RE: VFA Artifacts on X+Y Straight Edges

I think friction is a bad idea. It won't work continuously. 

I don't understand your design either.

You're isolating the bearing mounting. The bearing runs around the inner ring. Where are you creating friction?

I could imagine a bearing with a soft material coating. 

Some people have had success wrapping pulleys with tape.

That would dampen the ribs of the belt.

You could test different materials and thicknesses to achieve the best effect.

RE:

Posted by: @rainer-2

I think friction is a bad idea. It won't work continuously. 

I don't understand your design either.

You're isolating the bearing mounting. The bearing runs around the inner ring. Where are you creating friction?

I have the same worries about friction, so this isn't really about a long term fix, more about seeing if somehow damping the energy on the meshing will prevent the resonance that leads to VFAs at certain speeds.  It needs to be an approach that doesn't simply store and release that energy, but kills it.

In my mockup, the friction would come from passing a longer bolt through the centre and placing a nyloc nut on the other end - there would be washers under the bolt head and the nut to span the enlarged opening and clamp onto the carrier.  So there'd be no rotational friction, only lateral, in the y-direction.  In fact I've now elongated the central spindle hole in that direction.

RE: VFA Artifacts on X+Y Straight Edges

I prepared a couple of idler carriers to try out the silicone bushing approach.  The silicone tube should arrive tomorrow, but in the meantime I think a TPU bush may serve the same purpose.  I think this arrangement is a bit like a spring (the yellow TPU bush) and damper (the white friction washers) setup.  The friction is coarsely adjustable via the nyloc nut.

The spindle is able to move in the lateral y-direction, due to the elongated hole.  The requirement for the bush is that it's stiff enough to keep the spindle away from the end of this elongated hole under normal tension, but flexible enough to move due to the irregularities in the meshing between the motor pulley and the belt.  This movement of the spindle will be tiny, and this should be resisted by the friction washers.

RE: VFA Artifacts on X+Y Straight Edges

I asked the same of Prusa tech support and they replied with the new firmware, the belts need to be between 96-98hz instead of 85hz. I will try the out later and see if it helps. 

RE: VFA Artifacts on X+Y Straight Edges

Posted by: @chris-hill

The spindle is able to move in the lateral y-direction, due to the elongated hole.  The requirement for the bush is that it's stiff enough to keep the spindle away from the end of this elongated hole under normal tension, but flexible enough to move due to the irregularities in the meshing between the motor pulley and the belt.  This movement of the spindle will be tiny, and this should be resisted by the friction washers.

In your last picture it looks like the screw has already moved all the way to the end of the hole, and is making contact with the plastic body? I would be concerned that this will happen over time even with a stiffer elastic bushing, due to creep under the constant pull from the belt. 

RE: VFA Artifacts on X+Y Straight Edges

Yes, everything is loose in the photo, just to illustrate the range of movement. You are probably right, but this is only an experimental setup, to try to reproduce @malpan’s result and confirm the theory. If it works then a better ‘spring and damper’ would be needed for a long term solution. 

RE: VFA Artifacts on X+Y Straight Edges

Hmm - now I see why Prusa went with a spindle screw that cuts its own thread into one side of the idler carrier - the belt on the right side, the lower of the two belts, runs too close to the gantry plate, so there isn't room for a nut or bolt head on the underside.  That complicates things.  I'll have to see if I can squeeze a small captive nut into a recessed friction washer and still fit a bush in there.

The right side, which coincidentally is the side for which I have a test mounting block on my desk, has plenty of room above and below.

RE: VFA Artifacts on X+Y Straight Edges

This is what I found when I asked GROK regarding the increase in belt frequency settings the following question, "For the new 6.3.4 Core One firmware, is it true that Prusa now recommends that the belts need to be between 96-98hz instead of 85hz?"

It appears that with the release of firmware version 6.3.4 for the Prusa Core One, Prusa has updated their belt tension recommendations to target a frequency range of 96-98 Hz, as opposed to the previously recommended 85 Hz. This change is reflected in user discussions and posts on platforms like Reddit, where community members have noted that the updated belt tuning instructions on Prusa's official belt tuner website ( https://belt.connect.prusa3d.com/) now suggest plucking the upper belt to achieve 98 Hz and the lower belt to achieve 92 Hz. Users have reported that these higher tension values align with Prusa's efforts to improve print quality, particularly to address issues like vertical fine artifacts (VFA) and gantry alignment problems. For example, one user mentioned achieving 92 Hz and 98 Hz with the new method, which resulted in improved calibration, while another noted that the new procedure corresponds to higher frequencies (around 108 Hz when measured with the older method) to potentially force the gantry into better alignment.

However, there is no direct confirmation from Prusa's official documentation in the provided search results explicitly stating that firmware 6.3.4 itself mandates this change. The updated recommendation seems to stem from Prusa's revised belt tuning guidelines rather than a firmware-specific change. Some users have expressed that their printers, when tuned to 85 Hz, experienced issues like homing failures or layer shifting, and Prusa support has advised increasing tension slightly to resolve these. Others have noted that belts tuned to 96-98 Hz under the new method have helped with calibration and print quality, though some still face challenges with gantry squareness.

Keep in mind that achieving these frequencies requires careful adjustment, and the reliability of Prusa's belt tuner app and web tool has been criticized for inconsistent readings. Some users recommend alternative apps like the Gates Carbon Drive or Spectroid for more consistent measurements. Additionally, ensuring the gantry is square is critical, as belt tension alone cannot correct alignment issues, and over-tightening may lead to long-term wear concerns.

For the most accurate guidance, check the latest instructions on Prusa’s official belt tuner page ( https://belt.connect.prusa3d.com/) or contact Prusa support directly, as the community discussions suggest ongoing updates to the tuning process that may not yet be fully reflected in all documentation.

RE: VFA Artifacts on X+Y Straight Edges

Posted by: @bartgalok

This is what I found when I asked GROK regarding the increase in belt frequency settings the following question, "For the new 6.3.4 Core One firmware, is it true that Prusa now recommends that the belts need to be between 96-98hz instead of 85hz?".

Assuming the subsequent text was verbatim from Grok's response, it is a pretty good summary of the recent discussions here on the forum.

Too bad that the LLM capabilities are probably reaching their peak right now, as they are increasingly learning not from humans, but from regurgitated AI output which is now all over the internet, even including discussion forums. 🙄  

RE: VFA Artifacts on X+Y Straight Edges

I can't find the source, but didn't someone post that the new tensioning manual got released too early and that the belts should still have 85 Hz via the old method?

As long as not all official Prusa apps/website/manuals are identical again, I'd be very careful to tune the belts massively higher to the new tuning, which apparently equals over 100 Hz in the old position. 

RE: VFA Artifacts on X+Y Straight Edges

Agreed. I put this up to provide a summary verbatim from Grok. In earlier posts someone asked if there could be a summary of the posts given the now 59 page thread. Might be something to review for accuracy but if so, might be helpful to those who join in later to have a recap of the exciting events that have transpired. 🙂 

I have no plans to increase the the belt tension until there is an official recommendation from Prusa.

Cheers and I must say it is impressive to see all the engineering discussion and knowledge of forum members. Great to learn new concepts and approaches to problem solving.

RE: VFA Artifacts on X+Y Straight Edges

Yeah, 59 pages makes it tough to find things.  I'm printing components for the Core One on an H2D so that they are VFA freeish, we'll see if the Universe implodes.

RE:

Well this is a bummer. A very thorough video about the replacement of pulleys and idlers with Mellow variant. He even replaces the M3 bolts with smooth pins. There is an improvement but it's only a slight one. On the other hand, there have definitely been cases where Mellow pulleys and idlers have greatly reduced VFA's.

Video with replacement of pulleys and idlers on a Core One

RE: VFA Artifacts on X+Y Straight Edges

Posted by: @chris-hill

I'll have to see if I can squeeze a small captive nut into a recessed friction washer and still fit a bush in there.

I've rearranged things a bit so that I can recess a captive nut within the original width of the idler carrier.  It means that the diameter of the bush mounting had to increase, which I don't think is a problem, and I've had to wrap the TPU bush around the recessed friction washer.  This means TPU is my only option now, because I don't have any silicone tube with that ID.  It also means that the movement under compression will be less, but I don't think that should affect whether it works or not, since the movement due to belt/pulley meshing should be tiny.

This is all a bit of a stab in the dark, regarding the 'spring rate' (i.e. how much the TPU bush will compress and how much compression energy is released) and the 'damper rate' (i.e. how tight the friction washers should be).  But hopefully some change will be visible if this idea has any merit.  I think I need to reduce the size of the VFA speed tower to concentrate on the mid-range speeds where the VFAs are worst in my setup - that should cut down the print time usefully, so that I can try various adjustments of the friction.

RE: VFA Artifacts on X+Y Straight Edges

Thanks for the summary! But you have marked the topic as "answered" now -- was that the intent?

RE: VFA Artifacts on X+Y Straight Edges

I cannot edit the OP so the only option I have to pin messages is with best answer - the thread remains unsolved.

RE: VFA Artifacts on X+Y Straight Edges

Yep, print as fast as possible... funny enough, I'm printing a part for a Bambu printer on my Core One (it's like crossing the steams in Ghost Busters)... and there is no (zero) VFA when you print with PETG / .25mm layers / and SPEED.  Too funny.

There are some issues on the straight edges... (my homage to the OP)

RE: VFA Artifacts on X+Y Straight Edges

I'm about done with my shock absorber idler mounts.  I've done a couple of VFA test prints, one with what I felt was a loose friction setting, and one that was a bit higher, and they haven't made a blind bit of difference.  If there's merit in @malpan's theory, which still seems a strong possibility, then it needs a better implementation than my attempts.  The main problem I have is knowing what the right spring and damper rates are, and how to set those more methodically than just turning the screw a bit and waiting another 80 minutes for another VFA tower.

I noticed while I was assembling my mounts that the printed TPU bushes are likely to match a rubber o-ring size.  I have a set of various o-rings, so may see if I can get some to fit - theory being that they'd be more compliant than my TPU bushes.  But I'm fast running out of enthusiasm for this, because actually my GT1.5 setup prints really well.