RE: Wavy mystery, unacceptable quality
Wonder what the date is on these kits. The bad batch of pulleys theory sounds plausible.
RE: Wavy mystery, unacceptable quality
I just bought mine in the Black Friday special. Being on the old age pension I wasn't sure about buying a Prusa but the brand reviews have all been so positive I thought go for it. I think that was a mistake as I've been in contact with support over the linear carriage binding on the rail. I got to chat with a Bot, and it went on a meal break, go figure.
RE: Wavy mystery, unacceptable quality
I bought the Black Friday special (kit) and am not having those sorts of problems. I think very slight vertical artifacts are the nature of CoreXY systems and we get them on the Bambu printers at work too. There are two screws I'd check. Be sure the THK X-axis rail is tightened to the crossbar. The kit instructions do the final tightening at a future step and I almost missed it. It tightens like a cylinder head, back and forth from the middle. Also be sure the vertical guide rails in front (next to the top of the Z screws) are solidly clamped in the upper aluminum piece.
The case flexes slightly during printing and it's important that the CoreXY frame is tightly coupled to the bed so they vibrate as a unit. That's why having the rods (and probably everything else) tightly clamped is important.
RE: Wavy mystery, unacceptable quality
Thanks Conrad I will check all those things as I put it back together. Still waiting on the new pulleys to arrive from Melbourne.
RE: Wavy mystery, unacceptable quality
Somebody noticed this thread.
Just maybe a fix.
https://www.printables.com/model/1521515-no-more-vfa-prusa-core-one-20t-idler-mod/files
RE: Wavy mystery, unacceptable quality
Somebody noticed this thread.
Just maybe a fix.
https://www.printables.com/model/1521515-no-more-vfa-prusa-core-one-20t-idler-mod/files
I think the consensus was that a toothed idler makes no difference. If that were the case Prusa would have switched to them on the Core One L or Plus. I believe only switching to 1.5mm belts and pulleys on the motor may help. VFA are a feature of the CoreXY design and it seems most printers have this. One obvious work-around is to switch to matte filaments.
RE:
a lot of "should", "untested", etc in this link...
Somebody noticed this thread.
Just maybe a fix.
https://www.printables.com/model/1521515-no-more-vfa-prusa-core-one-20t-idler-mod/files
RE: Wavy mystery, unacceptable quality
Are there any news on how to solve this problem? I'm experiencing similar issues with my newly built kit.
RE: Wavy mystery, unacceptable quality
@TazaParts any news about this?
RE: Wavy mystery, unacceptable quality
This is worth a read if you haven't- https://blog.prusa3d.com/new-in-prusaslicer-consistent-surface-finish-and-nerfing-vfas_120400/
I don't think VFAs are gone, but if I keep my belts at the recommended tension and the gantry square and use the "balanced" profiles, they aren't much of an issue. Remember that the Core One is sold with the assumption that you'll put it on a solid surface. If the support is flimsy or you do some non-factory damping, either with fancy feet, or loading the top, the whole dynamic of the thing will change. All bets are off then. IMO, on a solid surface the bottom of the frame should be pretty much still (it's on a solid surface), but the top of the frame will vibrate on the order of 1 mm or so; that appears to be normal and expected.
RE: Wavy mystery, unacceptable quality
Can't prove it yet, but I'm currently experimenting with some GT2 pulleys and deflection rollers.
I suspect my VFAs are caused by poorly manufactured drive rollers.
They probably run slightly out of round.#
I'll let you know how it goes.
RE: Wavy mystery, unacceptable quality
I wouldn't rule that out. I wasn't too happy that some of the pulleys looked slightly rusty on some parts. Not too much, but slightly brown-ish.
RE: Wavy mystery, unacceptable quality
@TazaParts any news about this?
I have news but not in the capacity I'd like. Long story short: 1.5 mm belts have acceptable quality, but I ran out of the filament I used with my ender and 2 mm belts.
On new years day I swapped out the gt-2 belts and replaced them with 1.5 mm belts from aliexpress. This took a few hours to do and basically meant disassembling the entire motion system. I took it as an opportunity to also investigate the resonance noises. I didn't find anything that was out of place, not tightened or damaged. All the pulleys were turning smoothly and the belts were straight. Only the x-motor raised concerns, as it was noticeably more difficult to turn than the y-motor and the steps were clearly pronounced. I would estimate it was twice as rough as the other motor. But since prusa support claims no issue and refuses a replacement, I kept it in. The motor also makes more noise, so I might replace it with a genuine LDO at a later date.
Print quality is improved with the 1.5 mm belts, yet not "VFA free" as claimed by some people. I have some comparision pictures for you but I ran out of the same PETG filament I used previously. On a side note, I can not recommend Elegoo "pro" PETG, do not buy. I have halted my side business for a short while and it will take some time until I have a need to purchase more of the PETG I used previously. So the comparisions are a little lacking and perhaps do not give the best idea of quality. The part I used in first post is also included, but I changed the speed during printing and towards the top the waves appear again, which could have been avoided. Also the infill on that part is showing through, but since I ran out of the filament I can't make a better example at the moment.
Overall I am happy with the new belts and I have been shipping parts that were printed on the Core ONE+. It's not the best but acceptable. If you are going to do the mod, beware that the scaling will be off afterwards. The scaling can be corrected with a g-code command at the start of each print, I have added a line to the start g-code that handles that. I find it weird there is no "e-step adjustment" or similar on the printer, perhaps because everything is perfect from the factory, but I would expect that to be a standard feature on any printer. Another noteworthy finding is that the pulleys required for the mod have different height, so using the standard tool to locate them on the motor shaft will not work.
RE: Wavy mystery, unacceptable quality
@TazaParts any news about this?
I have news but not in the capacity I'd like. Long story short: 1.5 mm belts have acceptable quality, but I ran out of the filament I used with my ender and 2 mm belts.
On new years day I swapped out the gt-2 belts and replaced them with 1.5 mm belts from aliexpress. This took a few hours to do and basically meant disassembling the entire motion system. I took it as an opportunity to also investigate the resonance noises. I didn't find anything that was out of place, not tightened or damaged. All the pulleys were turning smoothly and the belts were straight. Only the x-motor raised concerns, as it was noticeably more difficult to turn than the y-motor and the steps were clearly pronounced. I would estimate it was twice as rough as the other motor. But since prusa support claims no issue and refuses a replacement, I kept it in. The motor also makes more noise, so I might replace it with a genuine LDO at a later date.
Print quality is improved with the 1.5 mm belts, yet not "VFA free" as claimed by some people. I have some comparision pictures for you but I ran out of the same PETG filament I used previously. On a side note, I can not recommend Elegoo "pro" PETG, do not buy. I have halted my side business for a short while and it will take some time until I have a need to purchase more of the PETG I used previously. So the comparisions are a little lacking and perhaps do not give the best idea of quality. The part I used in first post is also included, but I changed the speed during printing and towards the top the waves appear again, which could have been avoided. Also the infill on that part is showing through, but since I ran out of the filament I can't make a better example at the moment.
Overall I am happy with the new belts and I have been shipping parts that were printed on the Core ONE+. It's not the best but acceptable. If you are going to do the mod, beware that the scaling will be off afterwards. The scaling can be corrected with a g-code command at the start of each print, I have added a line to the start g-code that handles that. I find it weird there is no "e-step adjustment" or similar on the printer, perhaps because everything is perfect from the factory, but I would expect that to be a standard feature on any printer. Another noteworthy finding is that the pulleys required for the mod have different height, so using the standard tool to locate them on the motor shaft will not work.
RE: Wavy mystery, unacceptable quality
This is worth a read if you haven't- https://blog.prusa3d.com/new-in-prusaslicer-consistent-surface-finish-and-nerfing-vfas_120400/
I don't think VFAs are gone, but if I keep my belts at the recommended tension and the gantry square and use the "balanced" profiles, they aren't much of an issue. Remember that the Core One is sold with the assumption that you'll put it on a solid surface. If the support is flimsy or you do some non-factory damping, either with fancy feet, or loading the top, the whole dynamic of the thing will change. All bets are off then. IMO, on a solid surface the bottom of the frame should be pretty much still (it's on a solid surface), but the top of the frame will vibrate on the order of 1 mm or so; that appears to be normal and expected.
I have read everything I could find about the issue and the Core One VFAs. Absolutely nothing helped.
The belts are tight, the gantry is square, the motor pulleys are installed correctly, the motors are fixed on the mounts, there's nothing rubbing, binding or grinding. I've checked. A LOT of times. My printer is on the floor, so there's no wobbly table involved, and I'm using the latest firmware and latest Prusaslicer.
Prints are absolutely horrid when using speed or balanced profiles, and somewhat acceptable with structural.
I'm considering replacing the belts and motor pulleys to see if I can get acceptable results when printing faster than 50 mm/s. But honestly, right now I'm regretting choosing this printer over another RatRig.
RE: Wavy mystery, unacceptable quality
I have read everything I could find about the issue and the Core One VFAs. Absolutely nothing helped.
The belts are tight, the gantry is square, the motor pulleys are installed correctly, the motors are fixed on the mounts, there's nothing rubbing, binding or grinding. I've checked. A LOT of times. My printer is on the floor, so there's no wobbly table involved, and I'm using the latest firmware and latest Prusaslicer.
Prints are absolutely horrid when using speed or balanced profiles, and somewhat acceptable with structural.
I'm considering replacing the belts and motor pulleys to see if I can get acceptable results when printing faster than 50 mm/s. But honestly, right now I'm regretting choosing this printer over another RatRig.
You have probably covered this, but since you did not mention it: Have you run the "phase stepping calibration", to improve smooth micro-stepping of the motors? (You will need the accelerometer accessory for this.)
If that still fails, I think replacing the motor pulleys is worth a try. It appears that complaints about VFAs/belt ripple have gone up recently, and a bad batch of pulleys could be a reason.
Before you embark on that: Have you compare the ripple on surfaces which are printed at the +- 45° diagonals (i.e. with only one motor moving)? Are both motors affected to the same extent, or is one notably worse? If it's the latter, you could try swapping the pulleys and see if the ripple moves to the other axis, to confirm or invalidate the "pulley hypothesis".
RE: Wavy mystery, unacceptable quality
Before you embark on that: Have you compare the ripple on surfaces which are printed at the +- 45° diagonals (i.e. with only one motor moving)? Are both motors affected to the same extent, or is one notably worse? If it's the latter, you could try swapping the pulleys and see if the ripple moves to the other axis, to confirm or invalidate the "pulley hypothesis".
I don't want to start a separate thread about this (for now).
Does anyone know of a good source where I can get high-quality GT2 pulleys and idlers? Price is secondary for now.
I would like to replace them all to find out if it really is a bad batch from Prusa.
I know that one user (source: Reddit) also had success with “1.5GT pulleys,” but I think that's more likely to be for the same reason I'm pursuing this. In that case, the bearings and the quality of the entire component were definitely higher than the original Prusa ones.
RE: Wavy mystery, unacceptable quality
I have read everything I could find about the issue and the Core One VFAs. Absolutely nothing helped.
The belts are tight, the gantry is square, the motor pulleys are installed correctly, the motors are fixed on the mounts, there's nothing rubbing, binding or grinding. I've checked. A LOT of times. My printer is on the floor, so there's no wobbly table involved, and I'm using the latest firmware and latest Prusaslicer.
Prints are absolutely horrid when using speed or balanced profiles, and somewhat acceptable with structural.
I'm considering replacing the belts and motor pulleys to see if I can get acceptable results when printing faster than 50 mm/s. But honestly, right now I'm regretting choosing this printer over another RatRig.
You have probably covered this, but since you did not mention it: Have you run the "phase stepping calibration", to improve smooth micro-stepping of the motors? (You will need the accelerometer accessory for this.)
If that still fails, I think replacing the motor pulleys is worth a try. It appears that complaints about VFAs/belt ripple have gone up recently, and a bad batch of pulleys could be a reason.
Before you embark on that: Have you compare the ripple on surfaces which are printed at the +- 45° diagonals (i.e. with only one motor moving)? Are both motors affected to the same extent, or is one notably worse? If it's the latter, you could try swapping the pulleys and see if the ripple moves to the other axis, to confirm or invalidate the "pulley hypothesis".
I have not! I made the mistake of not getting an accelerometer together with my kit, when I placed my order. I did not expect that an add-on would be an essential piece of hardware to make my printer produce acceptable quality prints. I have tried to manually adjust IS values, but I couldn't get anywhere.
When printing at 45°, it seems like the Y motor produces more intense VFAs. I have swapped the pulleys and the ripples seem less intense, but they're still there. If teeth meshing is behind this, the problem with my printer might be one of the belts.
My next step will be getting an accelerometer. I'm just waiting on a voucher approval to get some Prusament along with it. If phase stepping doesn't help, will swap the belts and motor pulleys. If that also doesn't help, I'll just accept that my expensive, fancy core xy is slower than my bed slinger.
RE: Wavy mystery, unacceptable quality
I fear adding more confusion than light, but I deal with similar belts at work, in an application of equal or more precision. Pulleys have been a problem because they're either junk or crazy expensive. Cost probably doesn't matter here. If you read the Gates catalog and/or design guide, they say the pulley profile is proprietary, and the belts won't work properly with any other pulleys than Gates pulleys. Most of the belt manufacturers say this! Still, Gates is a high-quality manufacturer and if they have a suitable pulley, it might be a good place to start.
If it were me, I'd make some effort to check the existing pulleys, plus the rest of the COREXY system. If you confirm the pulleys are concentric, there's likely something binding or a motor problem still undiscovered.
RE: Wavy mystery, unacceptable quality
Unless you really need to print showroom parts with glossy filaments I would just ignore the VFA's. I got obsessive over it a first but then I just choose to ignore them since 99% of the stuff I print is functional.