So I’ve done the tape mod - appears to have reduced the amplitude of the VFAs (2mm ones) but will need to do more testing to be sure.
1 Like
I retensioned the belts on my X1 last night and then did the hardware calibration. I then printed with PLA and standard profile 0.2mm layer height, 2 walls, as a starting point, at 225°C. I set the speed for the outer walls to a constant 65mm/s. The fill was adaptive cubic at 15%, all other speeds >165mm/s. I have irregularities in the layers when I hold the print at the appropriate angle against the light; so I don’t have a 100% smooth outer skin. But I can’t see any artefacts, like ripples at 2mm intervals.
It was completely different with the prints before. There I had larger parts in PETG and clearly these waves on the outer walls, held at the right angle against the light; the printing speeds here were: outside 25mm/s, inside 150mm/s, filling 120mm/s.
1 Like
Below 70mm/sec or so speeds you’ll see a different type of artefact that looks like noise. This is the stepper motor cogging as it moves.
2 Likes
Your case obviously is not what is being talked about here. I have probably retensioned the belts >30 times, have tried high and low tension, have tried all speeds ranging from 40-220mm/s and the 2mm VFA do not go away with any standard configuration. You are wrong that these issues are inherent to the way these are designed as I have also demonstrated that SOME machines have it bad and SOME do not.
The main linking factor I have found is that the Kickstarter units are more likely to be faulty than the later versions, corroborated by 3 separate pairs of machines. Had another friend with both a Kickstarter unit and later unit do a side-by-side comparison.
I have made a huge improvement in VFA reduction by modifying the pulleys closest to the toolhead with the teeth side rolling against the flat surface. Either the teeth striking is causing the problem or the teeth rubbing the flange is causing the problem. Why this happens with certain printers more is unclear, but it could be that the bearings are defective, that the pin they ride on was installed crooked, or that the belt track is out of level through some level of early-production QC limitation.
I removed all my electrical tape to do another control test and the issue was definitely improved with tape. I tried to do just the flanges, and just the center of the roller, but couldn’t get it very precise. That said, the one attempting to do the flanges provided more benefit than just the roller, leading me to think it’s more about the flange rub than the surface rub.
Given that my ideal solution would be to slip a thin rubber ring over the existing pulley or install a toothed pulley but those are problematic, I worried about the adhesive from the electrical tape wearing out or transferring over time. I was thinking maybe Flex Tape or another silicone tape, but a friend suggested Teflon tape which I have started using. It takes a lot more wraps because it is so thin, but so far it is holding up really well through 4 hours of actual model printing and a couple hours of test prints. I am debating adding more tape, but the 5-6 layers I’ve applied already really helped. A benefit of its thinness is that unlike electrical tape, accidentally overlapping does not create humps to the pulley diameter.
3 Likes
@Slarm
I believe it affects the X-axis primarily because only travels along X require the tooth side of the belt to roll across the smooth idler. There is still some of this noise in Y travels but it could be the varying inflexibility of belt even smooth to smooth or the teeth of the belt lightly snagging the drive pulleys.
That was kinda my thinking on it too. The back of the belt travels against the smooth idlers when it’s Y motion right? Sorry since my old diy printer died and I got the X1C, I’ve become incredibly lazy at looking at the printer lol…
I think I have an idea though, brb
This is a GT2 belt right?
Bambulab Wiki says:
The ringing effect shows on the model walls:
Solution - The belts need to be retightened and the vibration compensation recalibrated. Please follow these instructions to do this - Decrease the acceleration value at Speed → Acceleration → Normal Pressure.
Please read the thread and stop parroting the wiki. It does not apply here. It is not a solution. The default canned response of Bambu support to “clean the rods and retension” fixes nothing.
5 Likes
So? What else does it say, already tried? Does it help or not? Just telling people not to parrot the wiki is not a good idea! What did I do to you? Because I’m trying to help?
That’s why I would try it out. Acceleration is not so far-fetched, don’t you think?
But leave it alone, I’ll try it out myself when I get the chance.
Have a nice evening!
The pulleys for the X axis don’t move at all when there is only movement on the Y-axis. For X-axis 2 belts roll tooth-side on smooth and two are smooth on smooth. I tried adding some tape to the smooth side ones and there was literally no change. I stopped at 6 or 7 wraps of various lengths (to avoid overlapping the same spot too many times) and even at the last additional wrap I saw a subtle improvement. I haven’t wanted to chase the upper limit because there will be a point where too much tape causes new issues, and removing the tape is very tedious.
Yeah, the belt is GT2 - I 3D printed a GT2 pulley for one of the motors and bought a permanent metal one and the fit is perfect. I am curious about your idea!
I have many times, including directly to you, described that I have used the default troubleshooting steps repeatedly. I have clarified this is not motor rippling, and it is not due to a bad frequency compensation nor poor or uneven belt tensioning.
This is a novel issue that is not covered by the Bambu wiki, nor their calibration system, and has to be dealt with new as a defect in the mechanical motion system. Repeating things that have been covered many times does not facilitate finding a cause and solution for this issue that plagues many people so I ask that you read the thread before posting wiki links as they’ve already been crossed off as a potential resolution.
I appreciate the desire and intent to help, but it needs to be in consideration of all the testing and experimentation that has come before.
This was originally not in English so I missed this one. Acceleration is not related either - the issue where I see it most occurs on flat faces. It is prominent on both diagonals and lines oriented along X and Y, but it is the most visible on X. Since it appears the most on long faces, I safely rule out acceleration as a factor, especially when taping the idler pulleys the teeth ride against improved the situation.
Not entirely sure if it would work, maybe with some super glue, I was curious if I could actually print it. I used a .2 nozzle and it printed, but obviously it would take more time to perfect, took me 2 days to get it to print at first! 
Talk about a super band-aid to not have to disassemble the printer. I also guessed on measurements and gave it 21 teeth since it’s slightly larger than the smooth idler. No idea if this would work, perhaps a base.
That is quite an interesting idea. I wonder how you’d get it in - unless you’ve printed it in flexible material like TPU or flex PLA? Did you use a standard GT2 profile. McMaster has Step and SolidWorks files available - that is what I based my drive pulley on.
There are the dimensions I took when I had one of the tensioner pulleys fully removed. Normally I think GT2 always comes in multiples of 2 teeth, but since this is a bit of a weird case adding teeth to a 12mm diameter pulley it might need to be an odd number like you did.
Did you actually try it yet?
We have a new theory that a friend of mine is putting to the test and if it makes a big difference it’ll be another solution. I am fairly convinced that the 2mm VFA at least on mine is more related to flange rub than teeth slapping, but there’s no reason it can’t be both!
1 Like
I just eyeballed the belt that’s on there and took rough measurements of one I had lying around, hoping it was the same idler lol. I could see flange rub being a culprit. I noticed the back right stepper gear, the belt is not “in the middle” of the gear as it’s slightly off center.
Examination of a P1P/S would be easier. I almost thought about taking off my enclosure since it’s in a pop tent.
Is this just standard plumbing teflon tape? And it stays on there without any adhesive?
I’ve done it too. Yes it is standard Teflon tape and yes indeed it sticks on! Move the tool head all the way to one side, stretch out some Teflon tape near the pulley and move the tool head the other way to roll the tape on.
Standard tape! I just used some from a cheap 3-pack I have had around. Once it is on it is very hard to remove without slicing, which will damage the pulley surface. Since it would probably stay wrapped forever, not really a problem if it gets damaged. Ask how I know!
One thing to be careful of IMO is to make sure the layers of wraps do not all start/stop same place or you could get mounds. I think I did a total of 6-7 wraps of varied lengths which would ultimately depend on the thickness of your teflon tape. I hope it is working for you!
Are you guys retensioning the belts afterwards? And then again after the belt sinks into the tape a bit?
No, I didn’t on mine. Just rerun resonance calibration.