RE: VFA Artifacts on X+Y Straight Edges
Interesting that Josef still thinks the VFAs can be resolved with tuning:
https://www.reddit.com/r/prusa3d/comments/1lp3eo7/psa_vfa_update_2/
I really really hope that a improved belt tuning app isn't their only solution for this VFA issue...
Like many here I tried all kinds of frequencies, from 70hz to 100hz and it made only very slight improvements on the VFA's but doesn't eliminate them completly. I am very doubtful that there is some kind of "magic tension" that will solve this issue and I hope they will not only release this new tuner and say they fixed it, so they are done with it.
RE: VFA Artifacts on X+Y Straight Edges
Interesting that Josef still thinks the VFAs can be resolved with tuning:
https://www.reddit.com/r/prusa3d/comments/1lp3eo7/psa_vfa_update_2/
I really really hope that a improved belt tuning app isn't their only solution for this VFA issue...
Like many here I tried all kinds of frequencies, from 70hz to 100hz and it made only very slight improvements on the VFA's but doesn't eliminate them completly. I am very doubtful that there is some kind of "magic tension" that will solve this issue and I hope they will not only release this new tuner and say they fixed it, so they are done with it.
I strongly suspect that Prusa wants to exhaust "no hardware" ideas first to improve the VFA issue. If an improved belt tension and/or a new firmware gets the VFAs into an acceptable range for most users, that's a win for Prusa.
And you have to read what Prusa says carefully, last month they didn't say "We aren't seeing VFAs, at all", they said (paraphasing slightly) "We aren't seeing unacceptable levels of VFA", so I fear that any changes Prusa comes out with will lessen VFAs on the average Core One. It likely won't fix the heavy VFAs some people are reporting nor will it get rid of less severe VFAs completely.
We might get pleasantly surprised, but I'm not getting my hopes up. I do very much hope some repeatable fix will come to light, either from the community or from Prusa, before my Core One ships 🙂
RE: VFA Artifacts on X+Y Straight Edges
To be fair, so far no company was able to completly eliminate VFA's on a FDM machine. There are always only work arounds to make them less visible, like Bambu Lab that print just very fast and slightly under extruded to make the material matte.
I really doubt that Prusa will eliminate them entirely, the only thing I want is to make them on the same level as on other machines where they are not visible at certain speeds. My machine, like many others, have them at all speeds from 40 to 200mm/s to some degree. If I am able to print the outer perimiter without VFA's at let's say 60mm/s, I will be completly happy.
RE: VFA Artifacts on X+Y Straight Edges
To be fair, so far no company was able to completly eliminate VFA's on a FDM machine. There are always only work arounds to make them less visible, like Bambu Lab that print just very fast and slightly under extruded to make the material matte.
I really doubt that Prusa will eliminate them entirely, the only thing I want is to make them on the same level as on other machines where they are not visible at certain speeds. My machine, like many others, have them at all speeds from 40 to 200mm/s to some degree. If I am able to print the outer perimiter without VFA's at let's say 60mm/s, I will be completly happy.
But what's the point of printing outer perimeters at 60mm/s when it can print them at the same quality and precision at 200 mm/s (which is still conservative value for Core One)?
RE:
To be fair, so far no company was able to completly eliminate VFA's on a FDM machine. There are always only work arounds to make them less visible, like Bambu Lab that print just very fast and slightly under extruded to make the material matte.
I really doubt that Prusa will eliminate them entirely, the only thing I want is to make them on the same level as on other machines where they are not visible at certain speeds. My machine, like many others, have them at all speeds from 40 to 200mm/s to some degree. If I am able to print the outer perimiter without VFA's at let's say 60mm/s, I will be completly happy.
But what's the point of printing outer perimeters at 60mm/s when it can print them at the same quality and precision at 200 mm/s (which is still conservative value for Core One)?
Mainly for the consistent outer surface quality.
Everything else can be printed as fast as possible, but the outer surface, which is the most important one for print quality, should be consistent.
There will always be layers where the printer has to slow down to meet the minium layer time, overhangs, or small segments that will not be possible to print at full speed because the printer can't reach them quickly enough because it has to accelerate first.
These spots will have a noticeable different surface, often a little shinier, and personally I think these differences in speed look worse than consistent VFA's.
RE: VFA Artifacts on X+Y Straight Edges
Exactly my opinion. As has been mentioned here many times before, if there were a simple method, manufacturers would have already done it. Various methods are used to try to conceal it.
And a CoreXY makes it even more difficult.
I'm not interested in what all the influencers post. In a photo, I can make everything look extremely good or bad.
When I look at pictures from real life, I like to look at make pictures, I recognize VFA across all brands and printers. Sometimes more, sometimes less.
Completely gone is an illusion that will not come true.
I have had a wide variety of experiences with my Core One. From really very good to you can't leave it like that.
Most of the time, it's the former.
There are so many parameters that come into play. Starting with the material, speeds, slicer settings, the model, etc.
Not just the individual parameters, but the combination.
That's why there are individual printers from every brand and model that are worse than average.
I don't expect miracles from Prusa. It will be a combination of a checklist for the hardware, optimized measurement and frequency of the belt tension, firmware, and slicer settings.
Depending on the original condition, this will yield more or less results.
The VFA will not disappear, but will be less visible.
It was clear to me that the Core One would not improve the VFA problem compared to my MK4S.
If I look closely, I can also detect VFA there.
But to be honest, I only discovered it now after being made aware of it by the Core One.
If you expect injection-molded parts, FDM may not be the right choice for you.
That's not meant to be an excuse for Prusa. I'm just being realistic.
There is certainly room for improvement.
But I think Prusa is continuously working on this.
To be fair, so far no company was able to completly eliminate VFA's on a FDM machine. There are always only work arounds to make them less visible, like Bambu Lab that print just very fast and slightly under extruded to make the material matte.
I really doubt that Prusa will eliminate them entirely, the only thing I want is to make them on the same level as on other machines where they are not visible at certain speeds. My machine, like many others, have them at all speeds from 40 to 200mm/s to some degree. If I am able to print the outer perimiter without VFA's at let's say 60mm/s, I will be completly happy.
RE: VFA Artifacts on X+Y Straight Edges
It was clear to me that the Core One would not improve the VFA problem compared to my MK4S.
If I look closely, I can also detect VFA there.
But to be honest, I only discovered it now after being made aware of it by the Core One.
If you expect injection-molded parts, FDM may not be the right choice for you.
Mhm, it's so impossible that my MK4S has literally NO VFA. Perfectly smooth walls. Telling people FDM is not for them because they expect their 1k euro printer to print with the same quality as their other 1k euro printer is ridiculous. "I've never seen a clean print so they don't exist"...
I want my Core One to print like my MK4S. With zero VFA whatsoever. Meanwhile it prints worse than all its competitors. And while competitors have speeds at which they don't artefact, Core One does not and leaves artefacts on every speed.
RE: VFA Artifacts on X+Y Straight Edges
it can print them at the same quality and precision at 200 mm/s
Wrong. It can't. The print will be matte (or a mix of gloss and matte depending on geometry) and the bond between layers will be weak. This is not the same quality. Also good luck printing Prusament PLA Blend at those speeds. Officially manufactured and supported filament, so it should print the best, right? It won't even utilize this speed due to flow restrictions.
I buy glossy filament directly from the printer manufacturer - I expect a glossy print as advertised. Both regular PLA and silk PLA. Printing at 200mm/s isn't solving anything, it will make these official filaments look ugly as sin, and it's definitely NOT the same quality.
RE: VFA Artifacts on X+Y Straight Edges
It was clear to me that the Core One would not improve the VFA problem compared to my MK4S.
If I look closely, I can also detect VFA there.
But to be honest, I only discovered it now after being made aware of it by the Core One.
If you expect injection-molded parts, FDM may not be the right choice for you.
Mhm, it's so impossible that my MK4S has literally NO VFA. Perfectly smooth walls. Telling people FDM is not for them because they expect their 1k euro printer to print with the same quality as their other 1k euro printer is ridiculous. "I've never seen a clean print so they don't exist"...
I want my Core One to print like my MK4S. With zero VFA whatsoever. Meanwhile it prints worse than all its competitors. And while competitors have speeds at which they don't artefact, Core One does not and leaves artefacts on every speed.
With every material? With every model? At all speeds?
When I illuminate an FDM print from above, it usually looks modest when viewed from the right angle.
It's just plastic sausages stacked on top of each other.
Under normal lighting conditions and at a realistic distance, none of this is visible.
But if I really want to find something, I can find something on the MK4S too.
And in general, I still think Bettslinger is a better alternative for many.
Unless I want to print a lot of ABS or similar materials.
My Core One has just finished a PCCF print. The part looks amazing and printed without any problems.
I've done this on the MK4S before, but with much more effort.
That's why I bought the Core One instead of upgrading the MK4S.
Every printer has advantages and disadvantages, which I try to exploit to the fullest.
RE: VFA Artifacts on X+Y Straight Edges
it can print them at the same quality and precision at 200 mm/s
Wrong. It can't. The print will be matte (or a mix of gloss and matte depending on geometry) and the bond between layers will be weak. This is not the same quality. Also good luck printing Prusament PLA Blend at those speeds. Officially manufactured and supported filament, so it should print the best, right? It won't even utilize this speed due to flow restrictions.
I buy glossy filament directly from the printer manufacturer - I expect a glossy print as advertised. Both regular PLA and silk PLA. Printing at 200mm/s isn't solving anything, it will make these official filaments look ugly as sin, and it's definitely NOT the same quality.
You have to adjust temperature or lower layer height to get strong shiny prints at higher speed. It's described a few posts back.
PLA Blend or Silk has very low volumetric flow rate limitations. You would usually print them at very low speeds where there are no VFAs yet.
From my experience with regular filaments the flow rate issues starts with speeds above 300 mm/s, or lower speeds with bigger nozzle. Where you have to test it a bit for particular filament how it behaves at higher volumetric flow rate. Or switch to high flow filaments as easier option.
RE: VFA Artifacts on X+Y Straight Edges
When I illuminate an FDM print from above, it usually looks modest when viewed from the right angle.
It's just plastic sausages stacked on top of each other.
This thread nor discussion isn't about layer stacking. Stay on topic.
With every material? With every model? At all speeds?
Every material. Every model.
And no, not every speed, why would it? As with near every other printer on the market, there are speeds that are free from artefacts and speeds that are not. Core One has the issue of not having such speeds, it artefacts on all of them.
NO ONE wants Core One to print cleanly on ALL speeds here, it's unrealistic. We want at least one speed where it doesn't artefact. Just like my MK4S, just like other brands. Yet it doesn't have that, it artefacts on all of them.
My Core One has just finished a PCCF print. The part looks amazing and printed without any problems.
PCCF prints fuzzy, which hides all artefacts. This isn't an argument. Every modern printer prints PCCF with high quality.
Core One is advertised as compatible with Prusa-manufactured silk filament, so it should print it in high quality. It does not. It's unnacceptable. End of story.
I see no reason to constantly defend them and try to convince people that VFA are impossible to escape. It hides the issue and gives Prusa ammunition to look at severely subpar quality, way below the competing brands, and say "every printer has them, stop bothering us, no refunds". It also slows down progress in FDM printing.
These artefacts absolutely are possible to escape. Core One is just a bad printer for quality surface finish.
I've done this on the MK4S before, but with much more effort.
Then there's a problem with your MK4S, mine prints PCCF flawlessly with zero effort.
You would usually print them at very low speeds where there are no VFAs yet.
This entire thread exists because Core One does not have "very low speeds where there are no VFAs yet". They are on all speeds. That's the issue discussed in this thread.
RE: VFA Artifacts on X+Y Straight Edges
This entire thread exists because Core One does not have "very low speeds where there are no VFAs yet". They are on all speeds. That's the issue discussed in this thread.
Without disagreeing: my Core One, which has the usual (20/)40/80 mm/s resonances and has VFAs for almost all speeds, interestingly doesn't have them at all for 25 mm/s and 45° moves.
I need to do a test print with more angles, but I was positively surprised, that a test print came out beautifully at 25 mm/s for external perimeters.
Elegoo PLA+ black, 0.6 mm Phaetus SiC nozzle.
I didn't take a picture, it has VFAs at all speeds, but going really slow, but above the resonance at 20 mm/s seems to work great, if you're fine with shinier prints.
RE:
Without disagreeing: my Core One, which has the usual (20/)40/80 mm/s resonances and has VFAs for almost all speeds, interestingly doesn't have them at all for 25 mm/s and 45° moves.
I need to do a test print with more angles, but I was positively surprised, that a test print came out beautifully at 25 mm/s for external perimeters.
That's a big part of the problem I see with the "just pick a speed which does not produce VFAs" approach: That safe speed must refer to the speed of the belt, right? Hence at different angles of the printed surface, the linear speed of the print head would have to be different to always achieve that "safe belt speed". Which is not something the slicer lets us control, since it focuses on the speed at which the print head moves and deposits filament.
(And even if a "constant belt speed" mode were implemented, it might have undesirable effects on filament deposition because the print head speed would then be all over the place, depending on the surface orientation.)
RE:
A little more 'real world' experience with my GT1.5 CORE One. I have an electronics housing that prints in two halves, so I printed one half on my 'best' MK4S (I have two - the other one shows very slightly more belt ripple), and the other half on my GT1.5 CORE One. I'm starting to think My CORE One is now the best printer I have by any measure - it prints faster, copes better with ABS and ASA, and it shows fewer surface imperfections than either of my MK4Ss.
The two halves of my housing fit together via a joint that is very tight, and hence the fit is very sensitive to any errors in the dimensions. In this test they fit together perfectly, proving that the M92 command completely corrects for the belt/pulley differences.
I printed the two halves in Overture grey ASA, so it's not the shiniest filament, but it is more representative of what I normally print than the earlier silk PLA prints that I showed. In the following photos, keep in mind that the MK4S printed the larger half, which includes the band around the joint, and the CORE One printed the smaller half. In all but one photo of each of the four sides the surface is indistinguishable by eye (the photos are zoomed-in crops and are showing a bit of a moiré pattern, but the main features are still visible I think).
The fourth side reveals a faint 2mm belt ripple on the MK4S, of which there's no sign on the CORE One.
Once my 1m MMU cable arrives, hopefully next week, from Blurolls I'm going to transfer the MMU to my CORE One, and then I don't think I have any reason to keep both MK4Ss any more.
EDIT: I do have some spare GT1.5 pulleys and some left over 1.5mm belt. I've modified the X-carriage and the Y-tensioner parts for the MK4S to work with 1.5mm pitch belts, so I may have a go at swapping one over to 1.5mm pitch too. I need to check that the M92 commands will work to correct the scale first though.
RE:
I printed these with Prusament PC Blend, it is shiny enough, but I think I tuned belts well enough this last time, and also run Phase stepping.
I was doing it a few times alrdy, and I would always get weird sounds from motors and vibrations, after PS run. Until last time, I loosened belts completely, squared the gantry, and started slowly tightening the screws. Now when I flick both left and right, they are steady 91Hz. And also, when I bring gantry to the front, it is perfectly aligned.
Motors sound perfect, and don't pruduce vibrations like before. (They usualy produced a lot of vibrations when both motors were working)
After that I run P. Stepping again, and, it reduced motor vibrations approx. 60% both X and Y (before that, it would either be 90% X 40% Y or vice versa)
I don't have pictures from before right now.
Any opinions on print quaity here?
RE: VFA Artifacts on X+Y Straight Edges
I reprinted the smaller half of my electronics enclosure with the 0.2mm SPEED profile (the earlier photos were with the STRUCTURAL profile). I noticed a little bit of 1.5mm belt ripple on the upper part of the model where it's only printing the wall of the enclosure, but elsewhere, where there are other details to print inside the housing, it was clean. I figured this is probably because it was slowing down to meet the minimum layer time, and was dropping below the 200mm/s perimeter speed. So I set a minimum layer time of 0s, to disable that setting, and sure enough the print is clean all the way up.
Obviously it won't always be appropriate to disable that setting, but for ASA with a purely vertical wall it worked fine.
Default 10s on the left, 0s on the right.
RE: VFA Artifacts on X+Y Straight Edges
I reprinted the smaller half of my electronics enclosure
Would you mind printing this model in silk PLA with the structural profile, please?
https://www.printables.com/model/1250394-noisy-star-test-for-core-one
I find that the amount of angles show rippling etc. a lot better than the other VFA test print. I'm very curious how 1.5GT handles those angles. With silk + structural to truly show every issue there is clearly.
RE: VFA Artifacts on X+Y Straight Edges
Sure - will post photos later.
RE: VFA Artifacts on X+Y Straight Edges
Some photos. There are 13 arms, each with two sides, so it would be a lot of photos. Most of them look like this:
These are the worst two:
RE: VFA Artifacts on X+Y Straight Edges
Brilliant. Thank you very much. Looks like 1.5GT can look a lot better than the 2GT pulley lottery. Despite replacing mine, I still have more bad looking angles on this print than you.
I will do this mod as well, assuming my replacement printer won't have EMI issues like my current one (long story, connecting Advanced Filtration or MMU3 causes severe changing diagonal artefacting...).