RE:
The weird thing is, i have 3 coreXY printers:
- Bambu P1S - has VFA
- Core One - has VFA
- Voron Trident - does not has VFA (or at least alot less)
The VFA on P1S and Core One are very similar.
RE: VFA Artifacts on X+Y Straight Edges
From the kit instructions thread:
Posted by: @jurgen-7 Posted by: @n3xt3d
I am currently running the VFA test print again. Turns out that the only glossy PLA I have (besides the Galaxy Black I ordered with the printer) is white -- probably less than ideal, but I am using that at the moment. Also got the two small samples which came with the printer kit (Jet Black and "Viva La Bronze" PLA). Tjose are only 25g each, but that may be just enough for the test print, or at least for the more interesting (s)lower half. I'll try one of those next if the white comes out inconclusive.
Just to point out that the VFA test file has a .3mf file as well as a .bgcode file, so you could change the filament type to PETG, to possibly have a bigger range of glossy filaments available to you.
RE:
Just ran the VFA test print again on my kit-built Core One. Confusing results...
- In general, the 60° and 90° direction are most affected by VFAs (or rather belt ripple).
- At 60°, VFAs occur up to ~2/3 of the speed gradient.
- At 90°, VFAs get better at ~1/2 of the speed range, but then reappear at faster speeds.
- At 90°, there are minor phase shifts in the ripple at each speed step.
- At 0° and 90°, one side of the arm is noticeably smoother than the back side, i.e. VFAs depend on the travel direction.
- The most striking ripple is actually in the curve between the 30° and 60° arms, at higher speeds
I don't think that tells me anything about the root cause of VFAs/belt ripple... But I will definitely keep and label the test print, so I can do before/after comparisons if we come up with any ideas what to tweak.
(The relatively low resonance frequencies of the tunes input shaper have me wondering... I might just try and tension the belts a bit more, or look for a different phone to double-check my frequency readings. Also, for those who have an accelerometer around and can run the input shaper calibration: Do you also experience the slightly "harsh" sound around the resonance frequencies? I am wondering whether something is loose and rattling in my Core One, or whether this is normal at resonance.)
RE: VFA Artifacts on X+Y Straight Edges
Just ran the VFA test print again on my kit-built Core One. Confusing results...
- In general, the 60° and 90° direction are most affected by VFAs (or rather belt ripple).
- At 60°, VFAs occur up to ~2/3 of the speed gradient.
- At 90°, VFAs get better at ~1/2 of the speed range, but then reappear at faster speeds.
- At 90°, there are minor phase shifts in the ripple at each speed step.
- At 0° and 90°, one side of the arm is noticeably smoother than the back side, i.e. VFAs depend on the travel direction.
- The most striking ripple is actually in the curve between the 30° and 60° arms, at higher speeds
I don't think that tells me anything about the root cause of VFAs/belt ripple... But I will definitely keep and label the test print, so I can do before/after comparisons if we come up with any ideas what to tweak.
(The relatively low resonance frequencies of the tunes input shaper have me wondering... I might just try and tension the belts a bit more, or look for a different phone to double-check my frequency readings. Also, for those who have an accelerometer around and can run the input shaper calibration: Do you also experience the slightly "harsh" sound around the resonance frequencies? I am wondering whether something is loose and rattling in my Core One, or whether this is normal at resonance.)
The VFA test increases by 10 mm/s from 40 to 200, with tiny tabs at the end of the arms every 5 mm at the transitions. When you say 1/2 or 2/3 of the gradient do you mean (200-40)/2=80 mm/s or (200-40)*2/3=106.667 mm/s?
RE: VFA Artifacts on X+Y Straight Edges
The VFA test increases by 10 mm/s from 40 to 200, with tiny tabs at the end of the arms every 5 mm at the transitions. When you say 1/2 or 2/3 of the gradient do you mean (200-40)/2=80 mm/s or (200-40)*2/3=106.667 mm/s?
Yep, sorry -- I was too lazy to look up the speed values of the test print. Yes in principle to what you wrote, but it's
40 + (200-40)/2=120 mm/s or
40 + (200-40)*2/3=146.667 mm/s, right?
RE:
Update:
Swapped my extruder motor with a brand new one from prusa: No effect on the "belt or whatever ripples", so its definately not the source of the problem or they sent me an equally broken one.
Support told me the next firmware-update is in the latest stage of internal testing. Hoping for the best.
Thank you for your patience. There is a firmware update on the way that should help with this.
RE: VFA Artifacts on X+Y Straight Edges
If it can help, below my results comparing the CORE One to XL and MK4S (click to enlarge). I used this test (only 40-90mm/s) because the STRUCTURAL profiles use an external perimeters speed of 45mm/s. To me this is the most critical speed to check as it should must give us the best quality possible.
Top to bottom:
- MK4S PLA Eco Black Arianeplast
- MK4S PETG Black eSun
- XL PLA Eco Black Arianeplast
- XL PETG Black eSun
- CORE One PLA+ "Gold" Arianeplast
- CORE One PLA Galaxy Black Prusament
- CORE One PLA Eco Black Arianeplast
- CORE One PETG Black eSun
From what I can see:
- The four samples from the CORE One show very visible artifacts. A little bit less with Galaxy Black depending on the viewing angle.
- The XL also suffer of this issue, but less (phase stepping enabled and calibrated once when released). This is a thing I curiously never noticed before to specifically try to show this issue.
- The MK4S has no VFA, at all. Even though the surface finish is rougher, not mirror polished.
So now it's not obvious to me to tell if it's "normal" and up to which level it is acceptable. To my standards, I won't upgrade my MK4S for the moment.
However, when printing "non-flat vertical surfaces", the CORE One does a very good job regarding the smoothness, really frustrating.
RE: VFA Artifacts on X+Y Straight Edges
To add something else is to the mix, is it possible that some idlers running the belts are not concentric? I'll be honest, reading through most of this thread I've sort of been following g along, but without a build in front of me, I'm struggling to work out what's what. I did take away that in one instance the belts run through a chain of idlers to make the X-axis? move. Could badly centred idler spindles/bearings contribute to the VFA's, or am I barking up the wrong 3D printed tree?
Scott
RE: VFA Artifacts on X+Y Straight Edges
Yep, sorry -- I was too lazy to look up the speed values of the test print. Yes in principle to what you wrote, but it's
40 + (200-40)/2=120 mm/s or
40 + (200-40)*2/3=146.667 mm/s, right?
Ah, yes, I messed up the math. Sorry about that and thanks for clarifying.
My preassembled C1 has bad VFAs from 40-120 mm/s at all angles, and it’s worst at 80, which is pretty much the entire range of normal filament print speeds. Clearly I need to rebuild it, which is why these threads are so interesting since I’ll have something specific to look for.
RE: VFA Artifacts on X+Y Straight Edges
Tbf if rebuilding the whole thing is the only thing that helps, no thanks. I sold nearly all of my DIY printers because I don't want to fiddle that much anymore to get everything running as I want. Minor repairs or maintenance is okay and should be counted in, but rebulding the gantry? Nah...
RE: VFA Artifacts on X+Y Straight Edges
I have now printed the VFA test in silky gold.
It looks very yellowish in the pictures. However, this is due to the image.
In the lower speeds, especially at 40mm/s, a clear resonance of the motors can be heard and felt.
You can recognize a pattern in the different angles, which mainly occur at speeds between 40-100 mm/s.
Unfortunately it was very difficult to capture with the camera in the correct light and angle, but I think you can see the VFAs. These are partly on one surface at different distances. So I would rule out belt ripple (for now), especially since you can clearly hear the resonance of the motors.
RE: VFA Artifacts on X+Y Straight Edges
I don't want to say I am relieved that there are some VFAs in your machine too -- but at least my belief in the principle of causality has been restored to some extent... 🙄
Maybe we should all take a deep breath and wait for the upcoming firmware release, presumably with phase stepping calibration enabled. According to what support told @rontaboga, it should help with the VFAs and should be coming out very soon. (Unless they run into more problems during the internal testing, I guess.)
If your observation is correct that the spatial frequency of the ripple changes with speed (an a given side of the VFA "star"), and that we are seeing motor resonances rather than belt ripple, that would make it even more plausible that smoother stepping can improve upon this.
I'm not sure I see changes of the spatial frequency in my test print though; it looks more like phase shifts to me. In which direction(s) do you see the frequency changes, and how pronounced are they? In proportion to the speed, and changing over the whole range of speeds where VFAs are seen?
RE: VFA Artifacts on X+Y Straight Edges
I have now printed the VFA test in silky gold.
It looks very yellowish in the pictures. However, this is due to the image.
In the lower speeds, especially at 40mm/s, a clear resonance of the motors can be heard and felt.
You can recognize a pattern in the different angles, which mainly occur at speeds between 40-100 mm/s.
Unfortunately it was very difficult to capture with the camera in the correct light and angle, but I think you can see the VFAs. These are partly on one surface at different distances. So I would rule out belt ripple (for now), especially since you can clearly hear the resonance of the motors.
Thanks a lot for the test, and indeed your printer seems to suffer from VFA too.
But, regarding how clean matte filament looks like when printer that mean I have to switch from standard filament to matte ones ahah.
Can’t wait to see improvements the next firmware will bring.
Thanks everyone for you effort by the way, that’s nice to see we are at least several people to care about quality and noises ahah 😀
RE: VFA Artifacts on X+Y Straight Edges
If it can help, below my results comparing the CORE One to XL and MK4S (click to enlarge). I used this test (only 40-90mm/s) because the STRUCTURAL profiles use an external perimeters speed of 45mm/s. To me this is the most critical speed to check as it
shouldmust give us the best quality possible.Top to bottom:
- MK4S PLA Eco Black Arianeplast
- MK4S PETG Black eSun
- XL PLA Eco Black Arianeplast
- XL PETG Black eSun
- CORE One PLA+ "Gold" Arianeplast
- CORE One PLA Galaxy Black Prusament
- CORE One PLA Eco Black Arianeplast
- CORE One PETG Black eSun
From what I can see:
- The four samples from the CORE One show very visible artifacts. A little bit less with Galaxy Black depending on the viewing angle.
- The XL also suffer of this issue, but less (phase stepping enabled and calibrated once when released). This is a thing I curiously never noticed before to specifically try to show this issue.
- The MK4S has no VFA, at all. Even though the surface finish is rougher, not mirror polished.
So now it's not obvious to me to tell if it's "normal" and up to which level it is acceptable. To my standards, I won't upgrade my MK4S for the moment.
However, when printing "non-flat vertical surfaces", the CORE One does a very good job regarding the smoothness, really frustrating.
First, nice choice of filament, Arianeplast is a good brand I buy from since 6 years! (I'm french) 🙂
Second: what surprise me a lot and tend to worry me is that XL with phase stepping seems to show at least as VFA as CORE One without. This would mean that Phase Steeping doesn't work well to reduce/suppress VFA...
RE: VFA Artifacts on X+Y Straight Edges
I wonder if whatever Prusa’s working on firmware-wise will help improve things on the XL too, or if it’s mainly just about enabling phase stepping on the Core One.
I’ve already got phase stepping enabled and calibrated on my XL, but the VFAs still look just as pronounced as what I’ve seen from the Core One.
RE: VFA Artifacts on X+Y Straight Edges
I don't want to say I am relieved that there are some VFAs in your machine too -- but at least my belief in the principle of causality has been restored to some extent... 🙄
=) I can well understand that hopefully everyone has the same problem, which will hopefully be solved with the phase stepping calibration.
I agree with you, it's best to take a deep breath first and if the phase stepping calibration doesn't lead to success, we can still rack our brains =)
Maybe it just looks like the distance is changing. Measured with the caliper, everything seems to be evenly spaced per speed. But the distance between the “waves” definitely changes at different speeds.
RE: VFA Artifacts on X+Y Straight Edges
If it can help, below my results comparing the CORE One to XL and MK4S (click to enlarge). I used this test (only 40-90mm/s) because the STRUCTURAL profiles use an external perimeters speed of 45mm/s. To me this is the most critical speed to check as it
shouldmust give us the best quality possible.
I disagree a bit. From STRUCTURAL profiles I expect high strength and accuracy for printing mechanical parts where look does not matter that much. If it prints nicer with SPEED profiles, let's use those for looks.
RE: VFA Artifacts on X+Y Straight Edges
I disagree a bit. From STRUCTURAL profiles I expect high strength and accuracy for printing mechanical parts where look does not matter that much. If it prints nicer with SPEED profiles, let's use those for looks.
That's assuming that the Speed profiles look as good as the Structural ones in other respects too -- especially fine detail, where the higher speed may push the input shaper to its limits.
Or one may want a glossy PLA surface, which can only be obtained at the slower speeds. (I have one project in the pipeline which needs a glossy glittery surface, and the choice of PETG filaments is very limited. Don't ask about my taste in materials, it's required for historical accuracy! 😉) So I would very much hope that, eventually, we can also get ripple-free prints at the slower speeds out of the Core One.
RE: VFA Artifacts on X+Y Straight Edges
First, nice choice of filament, Arianeplast is a good brand I buy from since 6 years! (I'm french) 🙂
I am living near the factory, that helps. It's by far not the best filaments I had, but it is the cheapest one especially during the sales days. So I can feed my printers like I eat spaghettis. 🤣
I disagree a bit. From STRUCTURAL profiles I expect high strength and accuracy for printing mechanical parts where look does not matter that much. If it prints nicer with SPEED profiles, let's use those for looks.
SPEED profiles doesn't care about the consistency of the surface finish, then you can get irregular matt/glossy horizontal banding depending on the MVS and layer time. STRUCTURAL profiles ensure the best result regarding the strength / accuracy like you said, but also regarding the surface finish.
Guess which is SPEED and which is STRUCTURAL. 😊
But I understand your thought.
RE: VFA Artifacts on X+Y Straight Edges
For the statistics my values for Input Shaper.
X: MZV 41hz
Y: ZVD 61hz
VFA are very acceptable. I made them in a blue color where you can really see everything.
The Nozzle Tool printed in Prusament Orange Petg looks really perfect. Perfect surface.
I would say that it prints a little bit better than my MK4S.
But the difference is marginal and will probably vary depending on the model.