RE: VFA Artifacts on X+Y Straight Edges
WHAAAAAAT! I didn't read correctly, I ordered 6 pairs instead of 6 idlers...
So you are the one to blame that they are now sold out! 😉
(You didn't order idlers but pulleys, right?)
Oh no sorry! I think I will ask a refund for 3 pairs as soon as I get them, I don't plan to use as much pulleys (yes it was pulleys indeed, not idlers)!
Crazy, the shop keeper must have faint out seeing this much orders in a short period hahaha!
I'm gonna buy loads of them and scalp people on ebay 😂
RE: VFA Artifacts on X+Y Straight Edges
Little off-topic:
It is so funny, that the first post in this topic was sent 4 months ago, the similiar XL topic is older (i think), mk4s also had similiar issues. And Prusa still try to get some funny cute figures clients, instead of helping their prosumers to get good quality.
RE: VFA Artifacts on X+Y Straight Edges
4 months ago Prusa said they would fix it in software a few weeks after I returned it ... still no software fixes
Bambu have leaked a new printer of H2D variety so that might be my replacement!
RE: VFA Artifacts on X+Y Straight Edges
4 months ago Prusa said they would fix it in software a few weeks after I returned it ... still no software fixes
Bambu have leaked a new printer of H2D variety so that might be my replacement!
They also released a new firmware and the dev behind Orca Slicer said he won't make it compatible for the foreseeable future out of principle.
If you don't care about that, the new printer might be awesome though.
RE: VFA Artifacts on X+Y Straight Edges
I'm still waiting on my Core One conversion kit but I have ordered a few pulleys and idlers from Mellow on AliExpress. Earlier I ordered some pulleys from a local webshop but I'm not sure of the brand.
It would be nice to actually test the pulleys objectively. I am very interested in measurements of the extruder or belts, to see how constant the motion is. I would expect to see the VFA effect in a variation of the speed like in this video:
In the video it is even mentioned that belt effects cause a variation in speed and thus extrusion width. They measured the predominant pitch to be 2 mm, with the strength varying with the belt tension.
RE:
Hello, I am experiencing VFAs on my Core one (Assembled Batch 4) aswell.
I experienced them since day 1. It took me very very long to diagnose the Issue since this is my first Printer.
It has been a few months since I got mine.
I would like to have a solution.
What was the Last Statement from the Company toward VFAs?
Has the Company made any fixes of the Core One VFA Issue? A hardware fix or a Software fix?
I have just watched this Youtube video made by an ex dev at Prusa.
It delves deeply into the Problems of VFAs on the Core One.
It led me to this Forum page.
If the Company has not made any fixes, is there a guide made by the community that fixes the Issues?
Input Shaping did nothing for me...
Thanks for reading.
RE: VFA Artifacts on X+Y Straight Edges
Joseph said they were looking i to it as they have some returned printers with the issue. This was on a reddit post linking that video.
So far the community are doing a lot also, investigating potential issues with bearing and idler quality and also belt pitch. No definitive results yet though. What is typical with this issue is that someone will find something that fixes it for them, but it won't work for someone else. That's what is happening with the forums efforts so far anyway.
A few of us are in various stages of waiting for parts or time to test out mods.
I believe the guy who posted the linked video ordered the belts and pulleys but ordered the wrong length of belt.
Someone on the core one users group on Facebook replaced their belts and pulleys with 1.5mm belts and it fixed the issue, though they reverted because they couldn't get the m92 gcode to take the new steps/mm which is potentially a showstopper. Might be able to fix it with shrinkage in the slicer, but not sure as the extruder will still extruder assuming it is moving (for example) 100mm when in reality it is moving 98.5mm
RE: VFA Artifacts on X+Y Straight Edges
Joseph said they were looking i to it as they have some returned printers with the issue. This was on a reddit post linking that video.
So far the community are doing a lot also, investigating potential issues with bearing and idler quality and also belt pitch. No definitive results yet though. What is typical with this issue is that someone will find something that fixes it for them, but it won't work for someone else. That's what is happening with the forums efforts so far anyway.
A few of us are in various stages of waiting for parts or time to test out mods.
I believe the guy who posted the linked video ordered the belts and pulleys but ordered the wrong length of belt.
Someone on the core one users group on Facebook replaced their belts and pulleys with 1.5mm belts and it fixed the issue, though they reverted because they couldn't get the m92 gcode to take the new steps/mm which is potentially a showstopper. Might be able to fix it with shrinkage in the slicer, but not sure as the extruder will still extruder assuming it is moving (for example) 100mm when in reality it is moving 98.5mm
RE: VFA Artifacts on X+Y Straight Edges
I'm still waiting on my Core One conversion kit but I have ordered a few pulleys and idlers from Mellow on AliExpress. Earlier I ordered some pulleys from a local webshop but I'm not sure of the brand.
It would be nice to actually test the pulleys objectively. I am very interested in measurements of the extruder or belts, to see how constant the motion is. I would expect to see the VFA effect in a variation of the speed like in this video:
In the video it is even mentioned that belt effects cause a variation in speed and thus extrusion width. They measured the predominant pitch to be 2 mm, with the strength varying with the belt tension.
That was actually a really good video! I was worried by the ai voice lol.
Shame the project is abandoned.
RE: VFA Artifacts on X+Y Straight Edges
Hello, I am experiencing VFAs on my Core one (Assembled Batch 4) aswell.
I experienced them since day 1. It took me very very long to diagnose the Issue since this is my first Printer.
It has been a few months since I got mine.
I would like to have a solution.
What was the Last Statement from the Company toward VFAs?
Has the Company made any fixes of the Core One VFA Issue? A hardware fix or a Software fix?I have just watched this Youtube video made by an ex dev at Prusa.
It delves deeply into the Problems of VFAs on the Core One.
It led me to this Forum page.If the Company has not made any fixes, is there a guide made by the community that fixes the Issues?
Input Shaping did nothing for me...
Thanks for reading.
Yeah I posted this video a few weeks back.
Just to be clear, the guy who made the video isn't an ex-Prusa dev... he simply relays a text/email convo he had with an ex-Prusa dev (Nathan) near the end of the video.
RE:
Someone on the core one users group on Facebook replaced their belts and pulleys with 1.5mm belts and it fixed the issue, though they reverted because they couldn't get the m92 gcode to take the new steps/mm which is potentially a showstopper. Might be able to fix it with shrinkage in the slicer, but not sure as the extruder will still extruder assuming it is moving (for example) 100mm when in reality it is moving 98.5mm
I'm not on Facebook - presumably that user wasn't able to permanently set the required number of steps per millimetre into the firmware, but it's still possible do it on a print-by-print basis, which I think would be a good workaround. I've just done a quick experiment in which I included the necessary M92 command into the printer's start G-code. With the command
M92 X98.5 Y98.5
I was able to produce a print that is 1.5% bigger.
So by creating a new printer preset (which I've done anyway to create a 'hot start' printer without the 'absorbing heat stage') it should be possible to accommodate the 1.5mm belts and pulleys.
RE: VFA Artifacts on X+Y Straight Edges
Someone on the core one users group on Facebook replaced their belts and pulleys with 1.5mm belts and it fixed the issue, though they reverted because they couldn't get the m92 gcode to take the new steps/mm which is potentially a showstopper. Might be able to fix it with shrinkage in the slicer, but not sure as the extruder will still extruder assuming it is moving (for example) 100mm when in reality it is moving 98.5mm
In view of the recent finding that carefully selected 3rd party pulleys, together with the standard GT2 belts, significantly reduced the belt ripple too -- isn't it likely that it was not the transition to 1.5 mm belt pitch, but rather the transition to better-matched pulleys & belts, which brought the improvement?
If that is indeed the case, there would be no need to deal with 1.5 mm pulleys and the corresponding change in steps/mm. I'm really keen to see whether others can reproduce the improvement via the Mellow-brand pulleys!
RE: VFA Artifacts on X+Y Straight Edges
That could be true of course, but I've seen other discussions around belt pitch, and there seems to be a growing opinion that the 1.5mm pitch is better than 2mm. Since VFAs are worse at some speeds than others, could it be something to do with resonance? And could the 1.5mm belt pitch shift the resonant frequencies to speeds that aren't routinely encountered? Apparently one or two manufacturers are now using 1.5mm and getting good results - e.g. the QIDI Plus4.
RE: VFA Artifacts on X+Y Straight Edges
isn't it likely that it was not the transition to 1.5 mm belt pitch, but rather the transition to better-matched pulleys & belts, which brought the improvement?
This is exactly what I've been wondering about. I do have the 1.5GT belt and pulleys from POWGE on hand, but since a 2GT pulley swap fixed it for me, I didn't try them, didn't even unpack them.
I'm very curious if the Mellow pulley swap will help others or if it's a more complicated issue. In that case I'd like to see if a 1.5GT swap would help instead. Otherwise it's hard to tell if it's the pulley that did the trick or the 1.5GT as a whole. Qidi+4 does have better surface quality, but it also has some outliers with VFA here and there.
If there's an easy way to tell whether the issue is the pulley or the pitch - I'd like to know, I'm out of ideas lol.
RE: VFA Artifacts on X+Y Straight Edges
Someone on the core one users group on Facebook replaced their belts and pulleys with 1.5mm belts and it fixed the issue, though they reverted because they couldn't get the m92 gcode to take the new steps/mm which is potentially a showstopper. Might be able to fix it with shrinkage in the slicer, but not sure as the extruder will still extruder assuming it is moving (for example) 100mm when in reality it is moving 98.5mm
I'm not on Facebook - presumably that user wasn't able to permanently set the required number of steps per millimetre into the firmware, but it's still possible do it on a print-by-print basis, which I think would be a good workaround. I've just done a quick experiment in which I included the necessary M92 command into the printer's start G-code. With the command
M92 X98.5 Y98.5I was able to produce a print that is 1.5% bigger.
So by creating a new printer preset (which I've done anyway to create a 'hot start' printer without the 'absorbing heat stage') it should be possible to accommodate the 1.5mm belts and pulleys.
Went back and checked his post
Just a quick update, I have been trying to adjust the settings in GCODE and the printer went mental, I have feeling that the accelerator might be needed to accurately calibrate the input shaping after this change. I don’t want to be doing lots of mods and want to keep it simple so anyone can do this mod not just me.The GCODE:M92 X81.27 Y81.27 ; To adjust the steps/mmM500 ; to store to EEPROM permanentlyNot sure why, but these settings had no effect.I’m not planning going Klipper route yet and my next step will be changing the value in firmware, although i’m still thinking that I will have to add the accelerometer to recalculate the input shaping
Possibly thuse of m500 to try and store it?
Anyways, I need to print an octopus for my kid so I'll try your advice so I can hit the ground running when I do my testing.
RE:
isn't it likely that it was not the transition to 1.5 mm belt pitch, but rather the transition to better-matched pulleys & belts, which brought the improvement?
This is exactly what I've been wondering about. I do have the 1.5GT belt and pulleys from POWGE on hand, but since a 2GT pulley swap fixed it for me, I didn't try them, didn't even unpack them.
I'm very curious if the Mellow pulley swap will help others or if it's a more complicated issue. In that case I'd like to see if a 1.5GT swap would help instead. Otherwise it's hard to tell if it's the pulley that did the trick or the 1.5GT as a whole. Qidi+4 does have better surface quality, but it also has some outliers with VFA here and there.
If there's an easy way to tell whether the issue is the pulley or the pitch - I'd like to know, I'm out of ideas lol.
Great minds think alike. With your testing i ordered some replacement gt2 pulleys from powge on top of the 1.5mm pitch belt and pulleys.
My plan is to do the pulley testing with the existing belts and 5 replacement pulleys and then try the 1.5mm belts and pulleys although I only have 2 pulleys to test with there.
That should give a spread of results with swapping out one motor to find the best pulleys on 2mm followed by taking the best 2 and using on both motors then comparing that to 1.5mm with 2 random pulleys. All from the same manufacturer, not mellow in this case though the pulleys do look machined.
RE:
Went back and checked his post
Just a quick update, I have been trying to adjust the settings in GCODE and the printer went mental, I have feeling that the accelerator might be needed to accurately calibrate the input shaping after this change. I don’t want to be doing lots of mods and want to keep it simple so anyone can do this mod not just me.The GCODE:M92 X81.27 Y81.27 ; To adjust the steps/mmM500 ; to store to EEPROM permanentlyNot sure why, but these settings had no effect.I’m not planning going Klipper route yet and my next step will be changing the value in firmware, although i’m still thinking that I will have to add the accelerometer to recalculate the input shaping
Possibly thuse of m500 to try and store it?
Anyways, I need to print an octopus for my kid so I'll try your advice so I can hit the ground running when I do my testing.
Those step values (M92 X81.27 Y81.27) are way off the original values (X and Y are both 100 steps per mm I believe). When I did my test above I also first tried dramatically different values like that, figuring that the effect would be easier to see. Indeed the print head banged around and went crazy, and the printer crashed with a blue screen that, from memory, implied that the print head was unable to reach the commanded position.
Sticking to values within a few percent of the original '100' values didn't present a problem. However, since smaller values lead to a bigger print, it's possible that a large print would again lead to a situation where the head couldn't reach the required position. Of course, with a different belt, and where the intention is to restore the correct scale, then this wouldn't be a problem.
I'm puzzled by the scaling though, and haven't managed to resolve it in my own mind yet. I think you said that the user with the 1.5mm belts is obtaining prints that are 1.5% smaller (hence my attempt to scale up the prints with the M92 command). And I think you also said in an earlier post that the 1.5mm pulleys have a very slightly bigger diameter than the 2mm pulleys. So if the pulleys are bigger, then surely a given number of steps will move the belt further, so wouldn't that make for a bigger print? Maybe I've misremembered the figures, but if not, where's the fault in my logic?
RE: VFA Artifacts on X+Y Straight Edges
Ahem - flattered as I am that this is considered by @baztm to be the Best Answer, I can only imagine that was a slip of the finger/mouse on his part, since it doesn't answer the VFA issue at all in my opinion, it's just part of the ongoing community testing effort.
RE: VFA Artifacts on X+Y Straight Edges
Went back and checked his post
Just a quick update, I have been trying to adjust the settings in GCODE and the printer went mental, I have feeling that the accelerator might be needed to accurately calibrate the input shaping after this change. I don’t want to be doing lots of mods and want to keep it simple so anyone can do this mod not just me.The GCODE:M92 X81.27 Y81.27 ; To adjust the steps/mmM500 ; to store to EEPROM permanentlyNot sure why, but these settings had no effect.I’m not planning going Klipper route yet and my next step will be changing the value in firmware, although i’m still thinking that I will have to add the accelerometer to recalculate the input shaping
Possibly thuse of m500 to try and store it?
Anyways, I need to print an octopus for my kid so I'll try your advice so I can hit the ground running when I do my testing.
Those step values (M92 X81.27 Y81.27) are way off the original values (X and Y are both 100 steps per mm I believe). When I did my test above I also first tried dramatically different values like that, figuring that the effect would be easier to see. Indeed the print head banged around and went crazy, and the printer crashed with a blue screen that, from memory, implied that the print head was unable to reach the commanded position.
Sticking to values within a few percent of the original '100' values didn't present a problem. However, since smaller values lead to a bigger print, it's possible that a large print would again lead to a situation where the head couldn't reach the required position. Of course, with a different belt, and where the intention is to restore the correct scale, then this wouldn't be a problem.
I'm puzzled by the scaling though, and haven't managed to resolve it in my own mind yet. I think you said that the user with the 1.5mm belts is obtaining prints that are 1.5% smaller (hence my attempt to scale up the prints with the M92 command). And I think you also said in an earlier post that the 1.5mm pulleys have a very slightly bigger diameter than the 2mm pulleys. So if the pulleys are bigger, then surely a given number of steps will move the belt further, so wouldn't that make for a bigger print? Maybe I've misremembered the figures, but if not, where's the fault in my logic?
Thats a good spot. I didn't read the numbers as he said he had stopped and rolled back. He did originally say his parts were only 1.5% smaller elsewhere, so he's made a mistake somewhere. The parts being slightly smaller is correct as the maths is basically [number of teeth on pulley] x [belt pitch] = distance travelled
RE: VFA Artifacts on X+Y Straight Edges
The parts being slightly smaller is correct as the maths is basically [number of teeth on pulley] x [belt pitch] = distance travelled
Ah, of course, thank you. So one turn of the original 16T 2.0 pulley moves the belt 32mm, and one turn of the 21T 1.5 pulley moves it only 31.5mm. That's 98.44% of the original, hence the ~1.5% reduction in scale.