RE: VFA Artifacts on X+Y Straight Edges
. I reaaly hope next updates will improve those resonance noises
About your resonances. On my MK4S it was a loose screw from a bearing cover, on others it was one from the heating bed.
Of course it's not great to start looking for faults on a new printer. But I would check all the screws to see if they are tight.
But too tight is also bad. Especially with the plastic parts.
Last time I saw how someone tightened the screws for fastening the nozzle so tightly that the filament was sticking.
RE: VFA Artifacts on X+Y Straight Edges
That could be possible if CORE One would have a built in accelerator I think.
I guess they are aware of such thing but it may be interesting to submit it to GitHub?
I've been researching the issue of resonances and VFAs and came across an interesting system used by the RatRig V-Core 4.0 printer. This printer includes a real-time resonance analysis system that automatically adjusts print parameters during the printing process, significantly reducing these types of artifacts.
This might be an effective solution for the Prusa Core One as well. It would be interesting if the Prusa team considered implementing similar functionality in future firmware updates.
Here's the link for those who want to explore it further: Rat Rig V-Core 4.0 (All Sizes) - Pre-configured kit
I hope this can help us find a definitive solution to the problem.
RE: VFA Artifacts on X+Y Straight Edges
Isn't this just Input Shaping? Core Ones are tuned with a factory tune for reducing resonances. I bought a Prusa accelerometer and ran input shaping tests to see if it improved VFA or sounds. It did not.
RE: VFA Artifacts on X+Y Straight Edges
I'll add that my MK4 has zero VFA'S and makes the nicest prints I've ever personally seen. Much better than the 30K 3d printer at work.
RE: VFA Artifacts on X+Y Straight Edges
That’s more like step phase stepping than input shaping. 🙂
Input shaping just change how the printer react when accelerating and decelerating, not how it behaves at any time and at constant speed.
Isn't this just Input Shaping? Core Ones are tuned with a factory tune for reducing resonances. I bought a Prusa accelerometer and ran input shaping tests to see if it improved VFA or sounds. It did not.
RE: VFA Artifacts on X+Y Straight Edges
[...] an interesting system used by the RatRig V-Core 4.0 printer. This printer includes a real-time resonance analysis system that automatically adjusts print parameters during the printing process, significantly reducing these types of artifacts.
I didn't understand that RatRig performs any automatic parameter adjustment based on these measurements. The web page just advertises "A comprehensive toolset to help you identify unwanted resonances on your machine and isolate their origin". I assume that means you can drive the printer through speed ramps or frequency scans automatically, see nice amplitude (resonance) plots, and then manually tension, damp, or otherwise tweak things and observe the effect.
RE: VFA Artifacts on X+Y Straight Edges
Thanks, Jürgen!
You're absolutely right — after rechecking the documentation, it seems the system doesn't automatically adjust print parameters during the print job. Rather, it's a diagnostic tool that helps users identify resonances through frequency sweeps and amplitude plots.
Still, I think this kind of tool could be very useful. If we can measure and know the specific resonant frequencies of our machine, it opens up the possibility of implementing something to counteract or minimize those effects. Maybe not in real time, but even manually adjusting speeds or implementing frequency cancellation techniques could have a big impact.
In signal processing, for instance, it's common to use notch filters to suppress unwanted frequencies. A similar concept could be applied in firmware to dampen known resonances, or at least avoid exciting them by adjusting motion parameters.
Even though this wouldn't eliminate VFAs entirely, it feels like a solid step forward — especially for high-end printers like the Core One. Having visibility into the machine’s resonant behavior gives both users and developers a powerful tool to diagnose and tune print quality more precisely.
RE: VFA Artifacts on X+Y Straight Edges
I did a temperature test this morning and it made no difference at all to the VFAs.
I would like to know what exactly are we seeing?
is it a deformity in the surface ( I can’t feel mine)
is it a alteration of extrusion - slow - fast - slow - fast (slower = matter faster = glossier)
is it more and less extrusion- thicker and thinner.
anyone know?
RE: VFA Artifacts on X+Y Straight Edges
Thanks, Jürgen!
You're absolutely right — after rechecking the documentation, it seems the system doesn't automatically adjust print parameters during the print job. Rather, it's a diagnostic tool that helps users identify resonances through frequency sweeps and amplitude plots.
Still, I think this kind of tool could be very useful. If we can measure and know the specific resonant frequencies of our machine, it opens up the possibility of implementing something to counteract or minimize those effects. Maybe not in real time, but even manually adjusting speeds or implementing frequency cancellation techniques could have a big impact.
In signal processing, for instance, it's common to use notch filters to suppress unwanted frequencies. A similar concept could be applied in firmware to dampen known resonances, or at least avoid exciting them by adjusting motion parameters.
Even though this wouldn't eliminate VFAs entirely, it feels like a solid step forward — especially for high-end printers like the Core One. Having visibility into the machine’s resonant behavior gives both users and developers a powerful tool to diagnose and tune print quality more precisely.
It depends what frequency range and source one is looking at. In the Prusa XL, it seems that calibrating the phase stepping operation can help to reduce noise and VFAs. I have never found a clear definition and description of "calibrated phase stepping", but it seems to deal with irregularities of individual stepper motors on the single-step or microstep level. Hence a level "below" what can be addressed via acceleration ramps and input shaping.
I have hopes that we will get phase stepping calibration for the Core One, and that it will help to mitigate VFAs and noise. But it seems non-trivial to adapt the implementation from the XL's 1.8° stepper motors to the Core One's 0.9° motors? I have seen quotes from a Jo Prusa interview where he regretted that the 0.9° motors from the MK4S had to be re-used in the Core One; apparently 1.8° would actually have advantages in the Core XY setup.
(By the way, was that ChatGPT writing the post? Not a problem if it helps to get a point across, but I still feel that it is good style to state the attribution. Guess I'm old-fashioned...)
RE: VFA Artifacts on X+Y Straight Edges
Thanks, Jürgen!
You're absolutely right — after rechecking the documentation, it seems the system doesn't automatically adjust print parameters during the print job. Rather, it's a diagnostic tool that helps users identify resonances through frequency sweeps and amplitude plots.
Still, I think this kind of tool could be very useful. If we can measure and know the specific resonant frequencies of our machine, it opens up the possibility of implementing something to counteract or minimize those effects. Maybe not in real time, but even manually adjusting speeds or implementing frequency cancellation techniques could have a big impact.
In signal processing, for instance, it's common to use notch filters to suppress unwanted frequencies. A similar concept could be applied in firmware to dampen known resonances, or at least avoid exciting them by adjusting motion parameters.
Even though this wouldn't eliminate VFAs entirely, it feels like a solid step forward — especially for high-end printers like the Core One. Having visibility into the machine’s resonant behavior gives both users and developers a powerful tool to diagnose and tune print quality more precisely.
It depends what frequency range and source one is looking at. In the Prusa XL, it seems that calibrating the phase stepping operation can help to reduce noise and VFAs. I have never found a clear definition and description of "calibrated phase stepping", but it seems to deal with irregularities of individual stepper motors on the single-step or microstep level. Hence a level "below" what can be addressed via acceleration ramps and input shaping.
I have hopes that we will get phase stepping calibration for the Core One, and that it will help to mitigate VFAs and noise. But it seems non-trivial to adapt the implementation from the XL's 1.8° stepper motors to the Core One's 0.9° motors? I have seen quotes from a Jo Prusa interview where he regretted that the 0.9° motors from the MK4S had to be re-used in the Core One; apparently 1.8° would actually have advantages in the Core XY setup.
(By the way, was that ChatGPT writing the post? Not a problem if it helps to get a point across, but I still feel that it is good style to state the attribution. Guess I'm old-fashioned...)
Nice catch! I wrote the text, but my English isn’t that good, so I ask it to fix the spelling. Maybe I used some weird-sounding phrases. Sorry!
RE: VFA Artifacts on X+Y Straight Edges
I think you're speaking of the super fine vertical artifacts which most people refer to as VFAs. The 2mm ones are referred to sometimes as belt ripple which this thread is about.
RE:
I think you're speaking of the super fine vertical artifacts which most people refer to as VFAs. The 2mm ones are referred to sometimes as belt ripple which this thread is about.
(Not entirely sure which post you were referring to, but assuming it's mine:)
I was indeed talking about the vertical artifacts with 2 mm pitch -- or 1 mm, when the print head moves diagonally. As mentioned earlier, I do assume the belt to be part of the root cause -- it's the only mechanism I can see which would reproducibly create that pitch, and always in the same XY positions, creating vertical bands. An interaction of the belt teeth with either the flanges of idlers, or the surface of smooth idlers, seems like a plausible cause.
But I have seen various posts which describe how these artifacts were significantly reduced in the Prusa XL after phase stepping calibration was released. I have no idea what the mechanism would be... Belt ripple is sometimes reported to depend critically on belt tension; maybe longitudinal oscillations of the belt (due to not-so-smooth stepping of the motors) can cause similar effects due to the constantly varying tension?
In general, "VFAs" do not seem to be a well-defined term, and that probably contributes to the confusing information about their origins and mitigation. As you said, the present thread seems clearly focused on the 2 mm artifacts -- yet it has "VFA" in its title. 🙄
RE: VFA Artifacts on X+Y Straight Edges
Here is some fun data:
I ran these with Prusament PETG at 260C limited to 20mm3 max flow. I used the speed profile. I printed these all on the same plate and changed *only* the outer perimeter speed. Starting from the top is 45mm/s then incremented by 20mm/s, so the bottom is 145mm/s. You can clearly see VFAs on all the parts except the last (145mm/s). I printed something with the speed profile (as is, no chages) and I didn't see VFAs at all (second picture). Speed profile uses 200mm/s outer perimeter.
RE: VFA Artifacts on X+Y Straight Edges
Here is a video as well, you can *really* hear the resonance sounds on the 85mm/s print.
RE: VFA Artifacts on X+Y Straight Edges
Here is a video as well, you can *really* hear the resonance sounds on the 85mm/s print.
Thanks for sharing and good finds. This still leads me down the motor tuning path. I'll also experiment with toothed pulleys when I get my machine. I recently changed them on my Prusa XL and it didn't make any difference, but I bought a set for the core one too and will test it as well.
I also need Prusa to release the printable parts because the pulleys are 1mm thicker and I need to modify the parts to match. I haven't checked printables lately but last time I checked there wasn't really anything there yet.
RE: VFA Artifacts on X+Y Straight Edges
I also need Prusa to release the printable parts because the pulleys are 1mm thicker and I need to modify the parts to match.
Which pulley did you buy? I'm also considering changing all the pulleys (even the smooth ones) and getting 1mm more for the belt to no grind the sledge is a nice thing!
RE:
By the way, next time someone wants to test VFA here is a gcode you can use that do all the job in one file:
RE: VFA Artifacts on X+Y Straight Edges
I don’t know how in Core One, but in XL motors are silent when turning clockwise, and loud when turning anti-clockwise. It’s very noticeable between 30 and 60mms. I have tried old motors, and newer ones. The same effect on both, and in my case most of VFA are noticeable on walls where one or two stepper go anti-clockwise. I think that is strange but maybe it is a result of high speed phase stepping calibration.
RE: VFA Artifacts on X+Y Straight Edges
I also need Prusa to release the printable parts because the pulleys are 1mm thicker and I need to modify the parts to match.
Which pulley did you buy? I'm also considering changing all the pulleys (even the smooth ones) and getting 1mm more for the belt to no grind the sledge is a nice thing!
This is the one I got. There's not 1mm more for the belt, the flanges are. 5mm thicker.
RE: VFA Artifacts on X+Y Straight Edges
I was indeed talking about the vertical artifacts with 2 mm pitch -- or 1 mm, when the print head moves diagonally. As mentioned earlier, I do assume the belt to be part of the root cause -- it's the only mechanism I can see which would reproducibly create that pitch, and always in the same XY positions, creating vertical bands. An interaction of the belt teeth with either the flanges of idlers, or the surface of smooth idlers, seems like a plausible cause.
In an earlier post I said the difference between a pure x or y movement and a diagonal movement is 2x but after thinking about it a little more and then doing the math it should be sqrt(2)x. I counted the number of VFAs on x movement vs a diagonal movement of the same length and got 1.4x difference which matches. The other thing I calculated is that the idlers attached to the linear rail actually move slower on a diagonal movement by a factor of 1/sqrt(2).
The other thing I noticed which is visible in the photos Scott posted was that if you increase/decrease the print speed mid-print, while the VFAs don't change in amplitude or frequency they do have a noticeable phase shift. That should tell us something, but I'm not quite sure what it is at the moment.