RE: VFA Artifacts on X+Y Straight Edges
Are you swapping just the front 2, Chris.
Yes. @Malpan showed some impressive results by just stuffing a bit of silicone behind the tension nut, so as I understand it the theory is that the belt system just needs a bit of compliance somewhere, anywhere, to damp the resonance.
EDIT: VFA tower is printing now.
I'm really excited to see the results. If it's successful, maybe you could do a tolerance test to see if it has any negative effects on accuracy?
In any case, I think your willingness to experiment is great 👍
No matter what the outcome, you always learn something new 😀
RE:
Hmm - my SE(TM) (Silicone Enhanced) idlers have had an effect, but not an entirely beneficial one!
In all the photos below, the print on the build plate is with the SE idlers, the one balancing on top of it is with the standard idlers. Belt tension was set to 97Hz using the old method for the original idlers print, and to the recommended frequencies using the new method for the SE idlers. However, I note that after the print had finished, both belts has dropped by about 10Hz, so I've readjusted them and am reprinting the VFA speed test. Note also that the SE print is using an ordinary brass nozzle, whereas the earlier print used a high-flow nozzle, which may explain the differences in shine - maybe the ordinary nozzle doesn't heat the PETG as efficiently.
There are bits of the photos that don't show unless you click on them.
So firstly there IS an angle where the VFA's seem to have improved slightly:
But mostly the SE idlers seem to have introduced some extra resonance bands:
In that last photo the resonance speeds seem to have reduced compared to the original idlers, while the faster speeds have tidied up.
I've no idea what this tells us. If resonance is setting in at new, lower speeds, maybe that's consistent with a less highly strung system, i.e the extra compliance makes the whole setup 'softer'. Maybe the extra compliance would work better with a different belt tension, but which way to go? I'll let the re-adjusted print finish, and then try one with much lower tension, just to see what happens.
RE: VFA Artifacts on X+Y Straight Edges
Well that is disappointing. I think that I may have gotten lucky with what I did.
Although that also led me to just realise the way I have the whole idler block being slightly compliant will mean that as the idler block slides over it's housing it'll dump a lot of energy into friction. I thought it was the silicone absorbing the energy but now I realise it's far more likely to be sliding friction.
RE: VFA Artifacts on X+Y Straight Edges
Yes, disappointing, but as Rainer says, every experiment gives a bit more information to learn from.
Keep in mind that my setup uses the GT1.5 belts, so will likely contribute to a difference between what you found and what I see. So your approach may be entirely valid with GT2.
RE: VFA Artifacts on X+Y Straight Edges
Just had the thought that if you extend the raised section of your guides/flanges so that it can be clamped onto the inner races of the bearing you should be able to get the same effect. Even more so because you will be able to tune the amount of friction added.
Furthermore I think your result does however show that the problem is caused by the motion system resonating because when you added another spring to it then the VFA occurred at an additional speed.
RE: VFA Artifacts on X+Y Straight Edges
Yes, I get that - you've allowed the idler mount to move (through compressing the silicone), whereas previously it couldn't. Is there an obvious way to make the silicone alone dump the same amount of energy? Thicker section? Or is it better to just have less energy in the system, e.g. by lowering the tension? Is that even what would happen?
RE: VFA Artifacts on X+Y Straight Edges
Just had the thought that if you extend the raised section of your guides/flanges so that it can be clamped onto the inner races of the bearing you should be able to get the same effect. Even more so because you will be able to tune the amount of friction added.
Furthermore I think your result does however show that the problem is caused by the motion system resonating because when you added another spring to it then the VFA occurred at an additional speed.
Your post arrived while I was typing.
I don't understand your suggestion. Even if the shoulder clamps fully onto the inner races (which I think it does already when the spindle is screwed up tight), the outer races are virtually friction-free. Are you suggesting that the shoulder should extend out to the outer races? Just to be clear over terminology - I'm using 'inner' to mean the race with the smaller diameter, and 'outer' to mean the one with the larger diameter.
RE: VFA Artifacts on X+Y Straight Edges
Hm, I have tried new method for belt tensioning method.
Before, while I was troubleshooting my printer, I've found that 93HZ on both left and right gives me the least amount of vibration from X and Y motors.
I've tested new tensioning method, I get them to 98Hz and 92Hz on point, and it gets me to 103Hz when I move nextruder to back and center. Print quality is nearly itentical.
No reduction in VFA's.
*Only, I won't be using new belt tension, since, my motors sound like dying donkeys even after PS.
When I do Phase stepping on my old 93Hz, it reduces motor vibrations by 93% and 95% after few runs, and afterwards, my C1 loudest sound is ventilation. When printing Polycarbonate, filtration fans are off, so You dont even hear the machine.
On VFA tower test, they look like on every single picture here, both 95% identical on old belt and new belt tuning method, so I won't post it.
**Also, what I noticed, when printing both technical parts, and Robo Lama, is that I got most VFA's when nextruder is traveling from right to left, in my case, when it decelerates.
I've previously posted few images on how my vfa's look. I deem them satisfactory to my needs (But it sucks that they are here). I will post them again, just to hear Your opinions, this is structural speeds, with Solid infil/Infil reduced to 90mm/s
Btw, When I use speed profile on this part, and with PLA, surface finish sucks completely. I will try to post a pic, for comparison when I get chance.
RE: VFA Artifacts on X+Y Straight Edges
Furthermore I think your result does however show that the problem is caused by the motion system resonating because when you added another spring to it then the VFA occurred at an additional speed.
Makes sense - so the silicone is acting as a spring, and not a damper. What's needed is a way to take energy out of the system (e.g. friction), rather than storing and releasing it.
RE: VFA Artifacts on X+Y Straight Edges
Isn't it a bad idea to change the belt tension in addition to swapping the idlers? For me, the tension alone changes to position of the artifacts. Keeping the tension the same should give more conclusive results, IMO.
Hmm - my SE(TM) (Silicone Enhanced) idlers have had an effect, but not an entirely beneficial one!
In all the photos below, the print on the build plate is with the SE idlers, the one balancing on top of it is with the standard idlers. Belt tension was set to 97Hz using the old method for the original idlers print, and to the recommended frequencies using the new method for the SE idlers. However, I note that after the print had finished, both belts has dropped by about 10Hz, so I've readjusted them and am reprinting the VFA speed test. Note also that the SE print is using an ordinary brass nozzle, whereas the earlier print used a high-flow nozzle, which may explain the differences in shine - maybe the ordinary nozzle doesn't heat the PETG as efficiently.
There are bits of the photos that don't show unless you click on them.
So firstly there IS an angle where the VFA's seem to have improved slightly:
But mostly the SE idlers seem to have introduced some extra resonance bands:
In that last photo the resonance speeds seem to have reduced compared to the original idlers, while the faster speeds have tidied up.
I've no idea what this tells us. If resonance is setting in at new, lower speeds, maybe that's consistent with a less highly strung system, i.e the extra compliance makes the whole setup 'softer'. Maybe the extra compliance would work better with a different belt tension, but which way to go? I'll let the re-adjusted print finish, and then try one with much lower tension, just to see what happens.
RE: VFA Artifacts on X+Y Straight Edges
Isn't it a bad idea to change the belt tension in addition to swapping the idlers? For me, the tension alone changes to position of the artifacts. Keeping the tension the same should give more conclusive results, IMO.
Agreed, although in practice there's very little difference in this case. I found that the new method of setting 98Hz for the upper belt and 92Hz for the lower belt resulted in 103Hz if measured at the old positions. And that's not very far from the 97Hz that I set in the earlier test.
RE: VFA Artifacts on X+Y Straight Edges
If you would like a material that does both, you are after low resilience material: https://en.wikipedia.org/wiki/Resilience_(materials_science)
I'm not really sure where to find a table of such things, but I know silicone rubber isn't too terrible at returning energy it absorbs.
Looking at your current setup as I understand it you put those gray plastic parts around the bolt. The silicone sits over these and then the bearings on the silicone. For the bearings to move on the silicone there must be a bit of play between the sides of the bearing inner races and the gray plastic parts. If you are to adjust how much the outer grey sections clamp onto the inner races then you can adjust the amount of friction that the bearings moving on the silicone have.
RE:
Unless I did the new or old method wildly different from you, I'm not sure what's going on. 98Hz upper and 92Hz lower belt for the new method gives me 112/113Hz using the old method (using the same, new web page for measurement). Where does the difference of 10Hz come from? Could it be that you new idlers make the belt behave differently when tuning the tension?
Isn't it a bad idea to change the belt tension in addition to swapping the idlers? For me, the tension alone changes to position of the artifacts. Keeping the tension the same should give more conclusive results, IMO.
Agreed, although in practice there's very little difference in this case. I found that the new method of setting 98Hz for the upper belt and 92Hz for the lower belt resulted in 103Hz if measured at the old positions. And that's not very far from the 97Hz that I set in the earlier test.
RE: VFA Artifacts on X+Y Straight Edges
Oh dear - my idler gave up the fight on the second VFA print - it suddenly started making a screeching noise so I stopped the print. This is what I found.
The print had already got past a key speed, and the extra resonance that was visible in the first print has gone.
So now then - is that due to the extra tension (i.e. the 'correct' tension that had evidently slackened off after the first print), or is it due to the friction caused by the broken part?!
Too many variables again, and now I'm wary of this whole printed idler. That built-in spindle sleeve is very thin, and unless it's the perfect fit over the spindle it's always going to have some stress. Maybe I need to adapt the idler holder to use an M4 spindle, and do away with the sleeve.
RE: VFA Artifacts on X+Y Straight Edges
Unless I did the new or old method wildly different from you, I'm not sure what's going on. 98Hz upper and 92Hz lower belt for the new method gives me 112/113Hz using the old method (using the same, new web page for measurement). Where does the difference of 10Hz come from? Could it be that you new idlers make the belt behave differently when tuning the tension?
Hmm - no, my comparison of old and new methods was with the original idlers. Curious.
RE: VFA Artifacts on X+Y Straight Edges
Looking at your current setup as I understand it you put those gray plastic parts around the bolt. The silicone sits over these and then the bearings on the silicone. For the bearings to move on the silicone there must be a bit of play between the sides of the bearing inner races and the gray plastic parts. If you are to adjust how much the outer grey sections clamp onto the inner races then you can adjust the amount of friction that the bearings moving on the silicone have.
The silicone is a tight fit inside the bearings, and once fitted inside the bearings the grey sleeve is then a tight fit inside the inside diameter of the silicone. So I think the inner races are virtually locked to the grey pieces. The outer grey flanges can turn by hand, but with the belt running on the bearing outer races there's no moment applied to the inner races, assuming the bearings are friction-free. I think my assembly failed because the belt tension snapped the inner grey sleeve off of the flange, and then the belt ran off-centre.
RE: VFA Artifacts on X+Y Straight Edges
I meant to add this photo of the partially completed VFA test above.
RE: VFA Artifacts on X+Y Straight Edges
@chris-hill Is the 'flange' separate from the gray bit that the silicone sits on? I may be misunderstanding but the belt will catch on the flange and then start rotating that on the bolt rather than rotating the bearings.
What I mean by allowing the inner races to move is it should have some lateral, not rotational, motion. Tension changes due to oscillations should cause the inner race to move away from the front of the printer, that is along the y axis of the printer. You want to allow that sort of movement with a bit of friction.
RE: VFA Artifacts on X+Y Straight Edges
@chris-hill Is the 'flange' separate from the gray bit that the silicone sits on? I may be misunderstanding but the belt will catch on the flange and then start rotating that on the bolt rather than rotating the bearings.
What I mean by allowing the inner races to move is it should have some lateral, not rotational, motion. Tension changes due to oscillations should cause the inner race to move away from the front of the printer, that is along the y axis of the printer. You want to allow that sort of movement with a bit of friction.
Ah, ok - understood.
Meanwhile, while replacing the original idlers and readjusting the tension, I've got the dreaded 'spinning tensioner nut'. In the left tensioner 🙄
Time to step away and tackle it fresh in the morning.
RE: VFA Artifacts on X+Y Straight Edges
Do we know the print settings that they use for the PC-CF tensioners? Should they be reprinted in 100% infill to minimize the chance of that square nut spinning?
-J