RE: Prusa CORE One kit assembly instructions
Now if we could only figure out the underlying difference between a Core One with and without VFAs...
I wonder if moving the belt position on one stepper some number of teeth so its rotor tooth relation to the other stepper syncs up. Is your gantry trammed (runs square)?
Also, in the VFA thread I mentioned an odd behavior I noticed:
- After homing, running the y-axis into the front stops is as one would expect (no x-axis movement)
- Then running the y-axis into the back stops causes the print head to jump to the left (-x)
- Now running the y-axis into the front stops causes the print head to jump to the right (+x)
This is just using the y-axis move function and turning the control knob, and one doesn't need to bump the stops hard (a single notch of movement in the control knob will cause the jump). I can write up some g-code to automate it and post a video if that would help.
But otherwise no VFA's to recognize so far.
Would you mind trying the above procedure? If touching the y-axis stops does not make your print head jump in the x-direction, perhaps it's a useful test.
This shouldn't be necessary. When you tension the belts the printer should be off or the motors disabled. So when you adjust the tension, the steppers are free to turn therefore creating equal tension throughout the system.
If you attempt to tighten the belts with the motors engaged then you are not properly tightening the belts.
RE: Prusa CORE One kit assembly instructions
No problem. I'll test it when I get to the printer later and report how mine behaves in this situation.
RE: Prusa CORE One kit assembly instructions
This shouldn't be necessary. When you tension the belts the printer should be off or the motors disabled. So when you adjust the tension, the steppers are free to turn therefore creating equal tension throughout the system.
If you attempt to tighten the belts with the motors engaged then you are not properly tightening the belts.
I wasn't talking about belt tensioning. Steppers have teeth on their rotor which is why they "step" so I was postulating that if the rotors on the two motors are out of sync with each other perhaps they contribute to VFAs. Anyway, after thinking about that a bit more, the steppers rotors will shift in relation to each other at different positions during a print, so VFAs wouldn't be consistent, and since they are consistent, it's not that.
RE: Prusa CORE One kit assembly instructions
I wonder if moving the belt position on one stepper some number of teeth so its rotor tooth relation to the other stepper syncs up. Is your gantry trammed (runs square)?
Also, in the VFA thread I mentioned an odd behavior I noticed:
- After homing, running the y-axis into the front stops is as one would expect (no x-axis movement)
- Then running the y-axis into the back stops causes the print head to jump to the left (-x)
- Now running the y-axis into the front stops causes the print head to jump to the right (+x)
This is just using the y-axis move function and turning the control knob, and one doesn't need to bump the stops hard (a single notch of movement in the control knob will cause the jump). I can write up some g-code to automate it and post a video if that would help.
But otherwise no VFA's to recognize so far.
Would you mind trying the above procedure? If touching the y-axis stops does not make your print head jump in the x-direction, perhaps it's a useful test.
I moved the X-axis up to 252 mm (already against the mechanical stop) and then slowly moved it backwards in the Y-direction until it touched the end stop. There was no movement whatsoever in the X-axis, neither in the millimeter values shown on the display nor mechanically on the printer. What I forgot to answer in your last question, my gantry seems to be very well square.
RE: Prusa CORE One kit assembly instructions
- After homing, running the y-axis into the front stops is as one would expect (no x-axis movement)
- Then running the y-axis into the back stops causes the print head to jump to the left (-x)
- Now running the y-axis into the front stops causes the print head to jump to the right (+x)
This is just using the y-axis move function and turning the control knob, and one doesn't need to bump the stops hard (a single notch of movement in the control knob will cause the jump).
I just tried the same on my Core One build (which does produce belt ripple artifacts). There is no X movement in either of the three steps above.
RE: Prusa CORE One kit assembly instructions
The ghosting at the seam is surprising to me. Doesn't the seam essentially come from excessive filament being squeezed out where the beginning and end of the trace overlap? I.e. there is no real radial movement of the print head, right? So why would ringing occur?
I can only explain the ringing by the deceleration and acceleration when changing to the outer perimeter.
I then calibrated the input shaper with my accelerometer. Interestingly, the printer now uses completely different shapers to before.
The default values are for X: EI 60Hz and for Y: EI 50Hz.
After the calibration, the printer now uses for X: ZVD 54Hz and for Y: ZVD 59 Hz. The new shaper should therefore smooth the acceleration/deceleration not as strongly.
I then ran a classic ringing test (complete with the new shaper values) in vase mode, 0.2mm speed profile, cooling fan at 100% and a minimum layer time of 1s. The results are completely flawless.
Neither ringing nor VFA's are visible.
And thanks for the congratulations!
RE: Prusa CORE One kit assembly instructions
Nice!
Could you try to print this file with a glossy filament please?
https://www.printables.com/model/1189256-vfa-test-prusa-core-one
RE:
I then calibrated the input shaper with my accelerometer. Interestingly, the printer now uses completely different shapers to before.
[...]
I then ran a classic ringing test (complete with the new shaper values) in vase mode, 0.2mm speed profile, cooling fan at 100% and a minimum layer time of 1s. The results are completely flawless.
Neither ringing nor VFA's are visible.
Could you share a link to the VFA test print you used? I still need to run the input shaper calibration (got an accelerometer with the Core One since I was half-expecting the need to get into this rabbit hole...) While I did not see any major ringing with the stock settings and don't expect the input shaper to have an impact on the belt ripple, I might as well run another test and use the same test print as you did, for direct comparison.
It's interesting that you ended up with significantly different input shaping values. But I assume you have set the belt tension to the nominal values, i.e. tuned to 85 Hz with the Nextruder centered in X, all the way back in Y? (Fond hope that retuning the belts will remove my belt ripple... 😉)
RE: Prusa CORE One kit assembly instructions
Sure. Here is the link.
https://www.printables.com/model/472432-klipper-ringing-tower
I set both belts exactly to 85hz and positioned the Nextruder exactly as you described.
I hope that the belt ripple disappears for you too.
RE: Prusa CORE One kit assembly instructions
Input shaping values won't have any impact on VFA, I'm 100% sure of it, it will only impact ringing on sharp corners.
RE:
Sure. Here is the link.
https://www.printables.com/model/472432-klipper-ringing-tower
I set both belts exactly to 85hz and positioned the Nextruder exactly as you described.
I hope that the belt ripple disappears for you too.
That's not a VFA test but ringing test, it changes IS frequency or acceleration values over the height of print, but speed doesn't change.
Coule you try to print this VFA test with a colored glossy PLA (not matte) please?
https://www.printables.com/model/1189256-vfa-test-prusa-core-one
RE: Prusa CORE One kit assembly instructions
Input shaping values won't have any impact on VFA, I'm 100% sure of it, it will only impact ringing on sharp corners.
Yes, that's my understanding too. As mentioned I don't really expect them to have an impact. Just trying to make everything equal with N3XT3D's build (to the extent we are aware of potential differences).
It's strange: N3XT3D and I have both built our printers following the instructions, adding some extra diligence (aligning the Y rods and Z rods/lead screws), setting the same belt tensions -- and yet, one of the printers shows ringing with the stock settings, the other doesn't. But in turn, only the other printer has belt ripple. What gives?
RE: Prusa CORE One kit assembly instructions
I suspect that he used quite matte filament, on a color that's good to hide print defects (grey).
That's why I asked to print a real VFA test with a shiny colored PLA.
Let's compare apple to apple.
RE: Prusa CORE One kit assembly instructions
Input shaping values won't have any impact on VFA, I'm 100% sure of it, it will only impact ringing on sharp corners.
No, I am aware that this is a ringing test.
That may have come across a bit wrong the way I wrote it. Because of my previous post where the ringing occurred at the Z-seam,
I calibrated the input shaper and then printed the ringing test.
The VFAs were just information, independent of the test part. The VFAs usually occur on long linear movements. So far I have not been able to recognize anything.
Nevertheless, I hope that the phase stepping calibration will be released soon.
RE:
I suspect that he used quite matte filament, on a color that's good to hide print defects (grey).
That's why I asked to print a real VFA test with a shiny colored PLA.Let's compare apple to apple.
I thought matte gray filament would be best suited for this, as the galaxy balck, white or silky filaments, for example, would tend to hide problems. Which filament would you recommend? If I have such a filament, I will of course be willing to print the test again in this filament. After all, I want to help and not gloss over things =)
RE:
I suspect that he used quite matte filament, on a color that's good to hide print defects (grey).
That's why I asked to print a real VFA test with a shiny colored PLA.Let's compare apple to apple.
I thought matte gray filament would be best suited for this, as the galaxy balck, white or silky filaments, for example, would tend to hide problems. Which filament would you recommend? If I have such a filament, I will of course be willing to print the test again in this filament. After all, I want to help and not gloss over things =)
Ahah, matte filament will hide VFA as it's much visible with light reflecting (very small deviations).
Silk or galaxy will hide layer inconsistencies, but will show VFA better.
Any classic colored or black PLA (that became glossy at low speed, so not matte) would do the trick, but silk would be the best. 🙂
RE: Prusa CORE One kit assembly instructions
Any classic colored or black PLA (that became glossy at low speed, so not matte) would do the trick, but silk would be the best. 🙂
Ah ok. Learned something again =)
I would have the Galaxy black, normal matte black filament or silky gold filament, which I could use for this case.
Should I print the same test part again, or better something else? You wrote that VFA's occur at low speed, I have kept the default speeds of the profiles on the Core One so far. Should I change them for the test?
RE: Prusa CORE One kit assembly instructions
This test:
Coule you try to print this VFA test with a colored glossy PLA (not matte) please?
https://www.printables.com/model/1189256-vfa-test-prusa-core-one
Gold silk would be perfect!
Any classic colored or black PLA (that became glossy at low speed, so not matte) would do the trick, but silk would be the best. 🙂
Ah ok. Learned something again =)
I would have the Galaxy black, normal matte black filament or silky gold filament, which I could use for this case.
Should I print the same test part again, or better something else? You wrote that VFA's occur at low speed, I have kept the default speeds of the profiles on the Core One so far. Should I change them for the test?
RE: Prusa CORE One kit assembly instructions
Ok, I'll report back later when the part is printed =)
RE:
I thought matte gray filament would be best suited for this, as the galaxy balck, white or silky filaments, for example, would tend to hide problems. Which filament would you recommend? If I have such a filament, I will of course be willing to print the test again in this filament. After all, I want to help and not gloss over things =)
Thank you so much for helping us narrow down VFA causes! It helps immensely to have an attainable target.
If you could print the VFA test model that @jurgen-7 posted above, that would be wonderful. If it shows no VFAs (or ripple or whatever) that in the range of 50-120 mm/s, we’ll know that most materials are perfectly printable once we tune the printer properly.
Edit: Just saw that there were other posts while I was writing. That VFA test will be really helpful, thanks again!