RE: VFA Artifacts on X+Y Straight Edges
Really looking forward to hear back from those trying a different belt.
Im also really looking forward to whatever Prusa discover, but I'm not holding my breath 😂
RE: VFA Artifacts on X+Y Straight Edges
Couple of questions, I have a set of Core One upgrade kits for my small farm that are supposed to be sent out soon, so I need to know if I need to cancel them.
1) Is the VFA issue visible when using fuzzy skin? I print large boxes with fuzzy skin enabled and silk PLA on a 0.6 nozzle.
2) Has Prusa officially responded with a confirmation that they acknowledge the issue and are working on it?
3) Has anyone found a consistent fix? I haven't seen one in my short time looking into this issue.
RE: VFA Artifacts on X+Y Straight Edges
1) I actually just printed a couple of fuzzy skin covered objects. Using a light 0.05 thickness and 0.1 distance. Since the VFAs are caused by some kind of tiny "disturbance" in the motors or belts, even the smallest fuzzy skin is a lot more movement than whatever causes VFA. You're no longer really moving in a straight line to deposit the "bumps". In any case VFAs aren't visible on my prints.
2. I don't know about an official announcement but J. Prusa himself acknowledged they are looking into it in a reddit comment.
Couple of questions, I have a set of Core One upgrade kits for my small farm that are supposed to be sent out soon, so I need to know if I need to cancel them.
1) Is the VFA issue visible when using fuzzy skin? I print large boxes with fuzzy skin enabled and silk PLA on a 0.6 nozzle.
2) Has Prusa officially responded with a confirmation that they acknowledge the issue and are working on it?
3) Has anyone found a consistent fix? I haven't seen one in my short time looking into this issue.
RE: VFA Artifacts on X+Y Straight Edges
Thanks @darksharpie, I cancelled my Core One kits for now, and I'll probably reorder sometime in the future from Printed Solid once this is all settled. I was also concerned about paying that 50% Europe -> USA tariff. No point in getting the kits if I'm going to wait for the VFA's to be fixed before I convert the printers.
Good to know that the VFA issue isn't apparent on objects with fuzzy skin, just sucks to go down a peg on print quality when the MK4S does such a good job with PLA.
I also reached out to support on this, they didn't have any official docs or timelines on the topic but it was clear that support is aware of the VFA issue.
RE: VFA Artifacts on X+Y Straight Edges
I think there will be some problems with correctly working of all systems.
I use 20t pulley in MK4S on the Y axis, with properly changed M92 and M350. Which sometimes causes some strange noises during the bed leveling.
In XL i couldn’t change pulley size because of not properly working M92 and hardcoded M350. And the code files behind new buddy firmware are so chaotic, so I decided to not make my own firmware for testing - I think that can be spaghetti code with need to check all calibration procedures for appropriate stepping.
Really looking forward to hear back from those trying a different belt.
Im also really looking forward to whatever Prusa discover, but I'm not holding my breath 😂
RE: VFA Artifacts on X+Y Straight Edges
Couple of questions, I have a set of Core One upgrade kits for my small farm that are supposed to be sent out soon, so I need to know if I need to cancel them.
1) Is the VFA issue visible when using fuzzy skin? I print large boxes with fuzzy skin enabled and silk PLA on a 0.6 nozzle.
2) Has Prusa officially responded with a confirmation that they acknowledge the issue and are working on it?
3) Has anyone found a consistent fix? I haven't seen one in my short time looking into this issue.
I have one Core One that I built from a kit, but I'm in a similar boat in that I'm holding off pulling the trigger on 2 more Core Ones pending the outcome of this VFA issue.
I was really lucky in that a slight adjustment with a pulley has made my VFAs much better, yet others had no luck adjusting the same pulley (it was completely un-scientific guesswork on my part.)
All Prusa have said is that they're looking into it, yet at the same time claiming none of their machines show any such problem (this is looking like PR at this point as the problem is really widely reported)
If they were to come out and admit a widespread issue I would guess sales of the Core One would fall off a cliff. It's a real shame as besides this issue I couldn't be happier with the printer, if they can resolve it the Core One has the potential to be a real work horse.
I doubt you would see VFAs with fuzzy skin, I don't even see them on matte filaments, or anything thats fibre infused (I've tried PETG-CF and ABS-GF, and they both look good)
RE: VFA Artifacts on X+Y Straight Edges
I think there will be some problems with correctly working of all systems.
I use 20t pulley in MK4S on the Y axis, with properly changed M92 and M350. Which sometimes causes some strange noises during the bed leveling.
In XL i couldn’t change pulley size because of not properly working M92 and hardcoded M350. And the code files behind new buddy firmware are so chaotic, so I decided to not make my own firmware for testing - I think that can be spaghetti code with need to check all calibration procedures for appropriate stepping.
Really looking forward to hear back from those trying a different belt.
Im also really looking forward to whatever Prusa discover, but I'm not holding my breath 😂
Yeah I did think that some software changes would be necessary alongside changing the belts.
RE: VFA Artifacts on X+Y Straight Edges
Hey, so, tldr I fixed it.
What I did:
- I went through multiple Mellow motor pulleys, placing them on the left motor, then running a test, then placing a new one etc. I picked the best 2, marking which one was the absolute best.
- I undid the belts and put in modified parts with toothed idlers.
- While doing that, I replaced the pulleys with those best 2. I put the absolute best onto the right motor.
And that's it. Here are the results. The remaining rippling is most likely the typical coreXY ripple since it actually changes depending on speed and angle. I will tension the belts a bit more, maybe 95Hz, then run phase stepping, IS and redo the test. Then pick the nicest looking speed.
RE: VFA Artifacts on X+Y Straight Edges
Nice job! Thanks for the testing and feedback.
Few questions, where did you get the Mellow pulleys? AliExpress? Can't find much about them.
Also, do you feel it would be interesting to replace the smooth idlers as well?
RE: VFA Artifacts on X+Y Straight Edges
Yes, Aliexpress. They have an official store there.
Pulley, pick 16 teeth:
https://www.aliexpress.com/item/33002989677.html
Idler, pick 20T With T:
https://www.aliexpress.com/item/4000040551804.html
You can also go with genuine Gates 2GT. I can't easily source them in EU so I had to go with Mellow.
I don't see the reason to replace smooth idlers personally. I had one that wasn't spinning as nicely as the rest, but it's been replaced with toothed anyway. Keep in mind Mellow idlers are thicker than stock Prusa idlers. You will need modified parts to fit them. Not worth the effort imo, stock are fine.
Modded printed part for toothed idlers:
https://www.printables.com/model/1283661-prusa-core-one-xy-carriage-toothed-idler
Nice job! Thanks for the testing and feedback.
Few questions, where did you get the Mellow pulleys? AliExpress? Can't find much about them.
Also, do you feel it would be interesting to replace the smooth idlers as well?
RE: VFA Artifacts on X+Y Straight Edges
Just something I have been thinking:
Due to axis x and y being mutually connected by belts (Not independent) and stepper motors having very real magnetic lands (Just turn one by hand and you will feel the magnetic steps with no power on), unless the magnetic lands are absolutely perfect between the x and y motors, you will get variable tension between lands as the motors rotate differently to one another, so surely this is a motor price point issue, a cheaper stepper will still step but be less accurate between lands, as in the lands spacing is not exact.
At the end of the day, these are no feedback stepper motors so will always be prone to error, the only way to fully get rid of VFA's might be a closed loop control system or servo drives (Motors with resolvers), all we can hope for is a reduction in VFA's but elimination might not be possible.
Normal people believe that if it is not broke, do not fix it. Engineers believe that if it is not broke, it does not have enough features yet.
RE: VFA Artifacts on X+Y Straight Edges
The motors use microstepping in operation, probably 1/256 of the steps you feel when manually turning the motors. Due to motor tolerances, these microsteps are indeed not necessarily equidistant. The "calibrated phase stepping" aims to compensate for the motor tolerances to smooth out the motion even further. Hope it will be fully released soon!
RE: VFA Artifacts on X+Y Straight Edges
Ich würde sagen, dass ich die „Fehler“ fast eliminieren konnte. Dazu haben ich Lager der y und y Achse entfettet und neu geschmiert. Der aber wesentliche Unterschied kam bei einer Riemenspannung von 115Hz.
Probiert es gern auch aus, über ein Feedback ob es bei euch geklappt hat würde ich mich freuen.
ps. An den schrecklichen Geräuschen die ich mit dem Core One habe hat dies nichts geändert.
RE: VFA Artifacts on X+Y Straight Edges
Hey, so, tldr I fixed it.
What I did:
- I went through multiple Mellow motor pulleys, placing them on the left motor, then running a test, then placing a new one etc. I picked the best 2, marking which one was the absolute best.
- I undid the belts and put in modified parts with toothed idlers.
- While doing that, I replaced the pulleys with those best 2. I put the absolute best onto the right motor.
And that's it. Here are the results. The remaining rippling is most likely the typical coreXY ripple since it actually changes depending on speed and angle. I will tension the belts a bit more, maybe 95Hz, then run phase stepping, IS and redo the test. Then pick the nicest looking speed.
Could you post some pictures showing before and after. Im struggling to see what the difference between the VFA ripple and the normal core xy ripple?
RE: VFA Artifacts on X+Y Straight Edges
VFA from motors is definitely a thing, but generally only shows up at lower speeds, like 10-40mm/s, and is much finer than the 2mm belt ripple VFA.
Just something I have been thinking:
Due to axis x and y being mutually connected by belts (Not independent) and stepper motors having very real magnetic lands (Just turn one by hand and you will feel the magnetic steps with no power on), unless the magnetic lands are absolutely perfect between the x and y motors, you will get variable tension between lands as the motors rotate differently to one another, so surely this is a motor price point issue, a cheaper stepper will still step but be less accurate between lands, as in the lands spacing is not exact.
At the end of the day, these are no feedback stepper motors so will always be prone to error, the only way to fully get rid of VFA's might be a closed loop control system or servo drives (Motors with resolvers), all we can hope for is a reduction in VFA's but elimination might not be possible.
RE: VFA Artifacts on X+Y Straight Edges
Could you post some pictures showing before and after. Im struggling to see what the difference between the VFA ripple and the normal core xy ripple?
I'm not sure what you mean with "VFA ripple". Technically this entire thread is using wrong terminology from the start. This isn't VFA (motor resonance), it's belt ripple. VFA from the motor have tiny spacing, even smaller on 0.9 motors. The 2mm spaced ripples we experience is mechanical, from the belt path, the spacing equals to belt teeth spacing in 2GT.
The ripple before my fixes was strange as in it was constant no matter the speed. Typically coreXY are plagued with ripples, BUT they change in intensity depending on speed. Part of tuning is finding the least rippled speed to use for external walls. Core One has the issue that all speeds are affected, so you can't find the best one - there is none. That's the distinction I've pointed out.
After my fixes, aside from reducing it to almost nothing, the ripple behaves as a typical coreXY ripple, meaning it changes in intensity depending on speed. I can notice 2 speeds on my test print already that are smooth on all angles, for example. So now they can be avoided by choosing the right speed, like you'd normally do on a different coreXY printer.
As for before, I don't have this exact print in this exact filament, but I have a different one that shows multiple different angles:
RE: VFA Artifacts on X+Y Straight Edges
I can notice 2 speeds on my test print already that are smooth on all angles, for example
Can you tell us the "clean speeds" for your machine? Ofc they aren't all identical, but I'd be interested to know the values and to test them for my Core One.
RE: VFA Artifacts on X+Y Straight Edges
Can you tell us the "clean speeds" for your machine? Ofc they aren't all identical, but I'd be interested to know the values and to test them for my Core One.
They will most likely be different for your motors. I can tell them once I finish testing, I'm currently checking out phase stepping and it actually introduces a lot of strange diagonal artefacts lol. I still need to tension the belts a bit more and rerun everything before I can decide on a speed.
RE: VFA Artifacts on X+Y Straight Edges
Posted by: @ratlet
Could you post some pictures showing before and after. Im struggling to see what the difference between the VFA ripple and the normal core xy ripple?
I'm not sure what you mean with "VFA ripple". Technically this entire thread is using wrong terminology from the start. [...]
As for before, I don't have this exact print in this exact filament, but I have a different one that shows multiple different angles:
I think what @ratlet meant was that he could not see an obvious difference between the ripple shown in your "after improvement" pictures, and the one seen in many typical Core One printers. I would concur.
It is difficult to assess VFA/belt ripple just from photos, which is part of what makes this whole troubleshooting effort tricky. Is my printer's ripple "par for the course" or is there something wrong with it? Did that guy's tweak bring his printer to previously unseen quality which I should aim to reproduce -- or did he just fix some major misalignment which initially made his particular unit print with stronger-than-usual ripple?
Any chance you could reprint your "before improvement" print, with the same filament you used for it back then? Then take some comparative shots with the two parts side by side, under the same lighting and camera angles. That should really bring out the improvement you obtained. Many thanks!
RE: VFA Artifacts on X+Y Straight Edges
Jurgen really captured everything i wanted to say.
The VFA tower is good, but also bad because it needs to be tuned also because the temperature needs to be increased as the speed increases otherwise the gloss goes to matte and the vfa effect appears to be lessened.
I took videos of my tests under the same lighting and I've got some violently green silk pla on route. I have a need for a large number of gridfinity boxes so I'll be using them instead of the vfa tower (i think). I can orientate them at the important angles and apply per object settings. Should work nicely. I hope.