RE: VFA Artifacts on X+Y Straight Edges
I've just ordered 5 replacement timing pulleys. When I build my Core ONE upgrade, I'm going to choose the two pulleys with the least amount of runout (using a dial gauge) with hopes that perhaps non-concentric center holes might be the problem. That's my theory for why picking the best pulleys might be making VFAs better.
Sometime in June, according to the current ship schedule.
RE: VFA Artifacts on X+Y Straight Edges
I'm going to choose the two pulleys with the least amount of runout (using a dial gauge) with hopes that perhaps non-concentric center holes might be the problem. That's my theory for why picking the best pulleys might be making VFAs better.
As discussed a few posts ago, that seems an unlikely cause. If runout were to blame, you would expect some pattern on the prints with the period of a pulley revolution, right? But that has not been observed. The ripple pattern is constant over large lengths of straight printed structures, with a period that seems to match the tooth distance.
RE: VFA Artifacts on X+Y Straight Edges
I agree Jurgen. Not sure if this applies to the Core but I've used a dual gauge to check concentricity for the mk3s+ pulley. It was quite concentric but slightly off on opposite end of grub screw. I used foil to make it better. Didn't make much a difference. I also changed many different e3d/gates/aliexpress pulleys and it made a difference. I can almost eliminate the ripple with tighter belt tension but I don't like the idea of bending my shafts. You can always use a support bearing but Im pretty sure I'm overconstraining it and making it worse.
RE: VFA Artifacts on X+Y Straight Edges
Given the thread by @Shushuda saying that choosing the "best" set of pulleys greatly improved their VFAs, there's something here. It cost me less than USD $20, and will take very little time to go through them. If it's not any better, then it's a data point.
I agree Jurgen. Not sure if this applies to the Core but I've used a dual gauge to check concentricity for the mk3s+ pulley. It was quite concentric but slightly off on opposite end of grub screw. I used foil to make it better. Didn't make much a difference. I also changed many different e3d/gates/aliexpress pulleys and it made a difference. I can almost eliminate the ripple with tighter belt tension but I don't like the idea of bending my shafts. You can always use a support bearing but Im pretty sure I'm overconstraining it and making it worse.
RE: VFA Artifacts on X+Y Straight Edges
I'm agreeing with Shushuda that different pulleys made a difference with belt ripple.
RE: VFA Artifacts on X+Y Straight Edges
I'm agreeing with Shushuda that different pulleys made a difference with belt ripple.
I don't think there is a debate on that aspect. The question was whether selecting the pulleys based on minimal runout is likely to identify the best ones; that's the part I doubt.
RE: VFA Artifacts on X+Y Straight Edges
Thanks. I'm willing to try this, and if (as I suspect) the pulleys will turn out to be pretty consistent, then as I said it's a data point. Then I'll start mix-n-match to try to find the best out of my set.
I'm agreeing with Shushuda that different pulleys made a difference with belt ripple.
I don't think there is a debate on that aspect. The question was whether selecting the pulleys based on minimal runout is likely to identify the best ones; that's the part I doubt.
RE: VFA Artifacts on X+Y Straight Edges
I think I can point out a possible source of the problem.
I’ve been testing a lot of printers in a specific scenario: low-speed, low-acceleration printing. The best printers I own are the Bambulab A1 and the Creality K1C.
I bought the K1C, and it only took me one day to dial in the settings for clean surfaces without artifacts. When comparing the K1C to the Core One, there’s no way anyone can convince me that the belt alignment, pulley quality, or overall build quality is better in the Creality. Core One clearly wins there.
I also don’t think the issue is related to audible vibrations. Both my A1 and K1C sound like 1980s Russian motorbike engines 😅 when printing at speeds between 20–40 mm/s.
The mechanical parts in the K1C and Core One are quite similar, and I’d even say the Prusa has better quality components. The main difference is the firmware: Marlin vs Klipper. Even my stepper drivers run at the same Vref values as Prusa’s — I’ve lowered Creality’s default (which are crazy high) by half.
So, in my opinion, something is messed up in the software.
RE:
I think I can point out a possible source of the problem.
I’ve been testing a lot of printers in a specific scenario: low-speed, low-acceleration printing. The best printers I own are the Bambulab A1 and the Creality K1C.
I bought the K1C, and it only took me one day to dial in the settings for clean surfaces without artifacts. When comparing the K1C to the Core One, there’s no way anyone can convince me that the belt alignment, pulley quality, or overall build quality is better in the Creality. Core One clearly wins there.
I also don’t think the issue is related to audible vibrations. Both my A1 and K1C sound like 1980s Russian motorbike engines 😅 when printing at speeds between 20–40 mm/s.
The mechanical parts in the K1C and Core One are quite similar, and I’d even say the Prusa has better quality components. The main difference is the firmware: Marlin vs Klipper. Even my stepper drivers run at the same Vref values as Prusa’s — I’ve lowered Creality’s default (which are crazy high) by half.
So, in my opinion, something is messed up in the software.
Interesting stuff, but doesn’t explain how @shushuda fixed it with pulleys. I managed to reduce mine quite significantly just by a very slight adjustment to one of the pulleys.
Obviously nothing is concrete yet, but there seems to be quite a lot of smoke originating from the pulleys direction.
RE: VFA Artifacts on X+Y Straight Edges
Obviously nothing is concrete yet, but there seems to be quite a lot of smoke originating from the pulleys direction.
You should probably get off the internet and attend to that smoke as a matter of urgency 🤣 🤣
RE: VFA Artifacts on X+Y Straight Edges
Obviously nothing is concrete yet, but there seems to be quite a lot of smoke originating from the pulleys direction.
You should probably get off the internet and attend to that smoke as a matter of urgency 🤣 🤣
Starting to think tightening the belts to 200 Hz was a bad idea 😂
RE: VFA Artifacts on X+Y Straight Edges
Firmware/stepper driver configs can really only affect low speed artifacts coming from the motors themselves independent of the pulley/tooth geometry, in the 0 to ~50mm/s range. The 2mm belt ripple being discussed here, which is proven to be coming from the pulley/belt teeth geometry occurs on all corexy printers in the ~70-160mm/s range.
I think I can point out a possible source of the problem.
I’ve been testing a lot of printers in a specific scenario: low-speed, low-acceleration printing. The best printers I own are the Bambulab A1 and the Creality K1C.
I bought the K1C, and it only took me one day to dial in the settings for clean surfaces without artifacts. When comparing the K1C to the Core One, there’s no way anyone can convince me that the belt alignment, pulley quality, or overall build quality is better in the Creality. Core One clearly wins there.
I also don’t think the issue is related to audible vibrations. Both my A1 and K1C sound like 1980s Russian motorbike engines 😅 when printing at speeds between 20–40 mm/s.
The mechanical parts in the K1C and Core One are quite similar, and I’d even say the Prusa has better quality components. The main difference is the firmware: Marlin vs Klipper. Even my stepper drivers run at the same Vref values as Prusa’s — I’ve lowered Creality’s default (which are crazy high) by half.
So, in my opinion, something is messed up in the software.
RE:
https://www.reddit.com/r/prusa3d/comments/1lf9mjv/psa_lets_address_the_vfa/
Josef posted on Reddit about their take on the causes of VFAs on the Core One. Seems to mostly match what the community has concluded here.
RE: VFA Artifacts on X+Y Straight Edges
Yep. Nothing new here beside the facts that it means they are working on it (which is nice).
Now they need to work on the resonance most of users get at 80mm/s
RE:
Slightly sceptical it can be fixed by merely tuning the belts.
RE: VFA Artifacts on X+Y Straight Edges
Slightly sceptical it can be fixed by merely tuning the belts.
I agree. I am a bit nervous that all we might get from Prusa eventually is something along the lines of "Make sure your belts are tuned to 97 Hz, and only print in the speed range from 123 to 131 mm/s".
RE:
Yep. Nothing new here beside the facts that it means they are working on it (which is nice).
Now they need to work on the resonance most of users get at 80mm/s
It’s weird how some issues are affecting some machines and not others. The only time I’ve been slightly bothered by the noise of mine was this weird buzzing sound it would make at times…after about a week I deduced it was my hygrometer I had attached to the side with magnets😂
I took it off and it’s never buzzed since.
RE: VFA Artifacts on X+Y Straight Edges
I just received my pulleys and idlers from Mellow 3D. Must say that to the eye they seem very high quality. Normally there is always some form of milling marks visible on the teeth. With these parts, such marks are not clearly visible.
Unfortunately I am still waiting for my conversion kit so I cannot do any tests yet.
RE:
Hey everyone,
Apologies for the late answer on the VFA discussion. Jo Prusa just posted a really detailed deep-dive on Reddit about it. You can read it here:
https://www.reddit.com/r/prusa3d/comments/1lf9mjv/psa_lets_address_the_vfa/
Martin (CORE One product manager)
RE: VFA Artifacts on X+Y Straight Edges
Post on Reddit:
PSA: Let’s address the VFA topic that has appeared on some social media and forums in recent weeks. There were several user reports regarding visible VFA on the models printed with the CORE One. First of all: Sorry it took us so long to respond, but I wanted to be sure that I brought a good insight instead of speculations, which meant analyzing the issue in-house first.
The short version is that this has been a particularly tricky topic for our devs, as the artifacts appear consistently only on some machines, which makes diagnosing the root cause a serious challenge. We now have printers from the community exhibiting this issue here at HQ for analysis. We believe that this issue can be solved via an improved belt tuning process, its suggested values and by optimizing print settings. And for the severe cases, an extended troubleshooting checklist.
We’ll be sharing these over the next couple of weeks. We want to make this process as easy as possible, without the need for exhausting troubleshooting on the users’ side.
And now, technical VFA deep dive.
VFA (Vertical Fine Artifacts) are microscopic imperfections in the extrusion that repeat at the same point in each layer, creating vertical lines on the model. These lines are tiny, visible from 10um, and the worst you have seen are around 40um from peak to valley. This is too small to even show up reliably on a lot of metrology equipment. However, the combination of plastics with a highly reflective surface and a spot light makes it very easy to see.
Because things are rarely easy, there are three main causes:
Motors
Print head resonance, resonance caused by accelerations (this is what Input Shaping is trying to mitigate)
Improper belt meshing
Our internal research shows that the type of VFA we’re seeing on reported prints is not caused by motor artifacts. How could a motor cause artifacts? The theory says that stepper motors aren't actually rotating smoothly - they jump from one position to another, quickly accelerating and decelerating while the extruder is constantly pushing out plastic, creating places with ever so slightly more material. This happens with micro-stepping, too, just softens it a little bit. These would have a very high pitch of 0.16 and 0.08 mm on CORE One. The motors, drivers, and voltage are well tuned from the times of MK4, and the CORE One uses the same stack. This is why we have ruled out the motors as the cause.
Moving on: Resonance artifacts are caused by the elasticity of the belts and the printer constructions. This affects any machine, 3d printer, mill, robotic arms, and so on. They cannot be fully removed, just reduced and compensated with technologies like input shaping. They start at a sharp turn and fade out after a few millimeters of travel. On vertical walls, these overshoots can also stack into VFA.
Now, “improper belt meshing”. As the belt teeth mesh into the pulley, there is a small transition period when a new tooth engages from pushing to pulling, creating a minuscule variance in speed. This applies to every belt under the sun with straight parallel teeth. A good read on this phenomenon is a paper from Masanori Kagotani (Influence of Idler on Transmission Error in Synchronous Belt Drives (Under Transmission Force)). We identified this part as the most likely culprit on the severely affected printers. A side note: This happens only on motor pulleys. As there is no force transfer on idlers, it makes no difference whether the pulley is smooth or teethed
It is primarily influenced by two factors: speed and belt tension. The effect intensifies significantly at specific motor rotational speeds, with smaller pitch belts having said speeds lower and larger pitch belts having higher speeds. It's important to note that this speed refers to the motor's rotation, not necessarily the print speed. Increasing print speed can often mitigate the issue by moving outside the problematic speed range. However, lines printed at certain angles can still hit the resonant rotational speed.
Belt tension is important, as it has an impact on the meshing - the belt can exert pressure on the teeth. There is a sweet spot. We are reviewing the belt tensioning in the app on a larger sample of smartphones now.
And then there are CoreXY specifics - CoreXY is unique because every movement is a compound of the movement of both motors. So, in some cases, this configuration can mask the effect, and in some cases it can amplify it. On cartesian machines, the effect is still there, just more predictable.
And all three causes mix and stack at different intensities during just a single print. So it is never “simple” to get rid of.