Resonances on y axis (v-core 3.1 300) are horrible, it's driving me nuts.
I've been trying to tune my belt tension and get rid of resonances for a week, and I feel like I'm getting nowhere. I managed to get the belt tension to a point that I thought was "good" , and then ran the resonances graphs for input shaper. X doesn't look too bad to me, but y is an absolute mess. I've checked for loose screws, and anything that may be vibrating, but I can't find anything and I'm getting desperate. I would appreciate it if anyone more knowledable than I would lend me a hand, as I obviously have no idea what I'm doing 👼
19 Replies
ratty-blush•16mo ago
So, I've been obsessing about this, and I read on another thread someone saying that made me think my toolhead was resonating. I checked, and indeed the toolhead was a bit loose, so I retightened the mgn12c print part (which seemed to be what's lose), and immediately the "lower" belt response went up without any other changes. I've been trying to look for other issues such as binding or other loose bolts and couldn't find anything obvious, but I got to this point.
I think this is still quite bad, but at least input shaper seems to be able to eliminate most of the vibrations. If I understand the graphs correctly (which I probably don't despite having watched the video on the pinned post like 5 times), there's something causing resonance on the 125Hz frequency, but I have no idea what it is. Any clues as to what could cause this?
secure-lavender•16mo ago
I have the same problem as you ... I have eva 3 with heat inserts, all tightened very well, bearings and linear guides lubricated
ratty-blush•16mo ago
Did you print the parts for your printer yourself? I'm thinking it could be that my parts are pretty bad quality because I printed them myself and have visible defects, and one of the things I have yet to try is to reprint everything, disassemble the whole thing and check again.
The other possible theory I have is the "wire loom" creating the problem, but according to the video on the FAQ that'd look different
secure-lavender•16mo ago
I have the same problems both with the parts printed by RR and with the parts printed by a guy in the group, which are done very well (asa)
as soon as I installed the 3.1 kit I didn't have these problems, they arose over time, now I would like to try replacing the RR belts with the gates
as soon as I installed the 3.1 kit I didn't have these problems, they arose over time, now I would like to try replacing the RR belts with the gates
ratty-blush•16mo ago
Ah, my printer is barely a month old, so it's not something that happened over time for me. It might be the belts as you said, but I'd expect everyone to have the same resonance at that frequency if that was the case?
If you don't have a large peak, but the entire graph is low, something is binding
Your second set of graphs is better
ratty-blush•16mo ago
Isn't that second peak on the 125hz range an issue? I've seen other people's graphs in this channel, where they only have the one peak near the 75hz, I thought that was what it is supposed to be?
Yes, that should not be there
Check that the duct lower screw and the plastic around it is okay
That could be a source of resonance
ratty-blush•16mo ago
Apologies, I thought I had responded to this. I checked the duck screw and it seemed to be okay, and I was still unable to find the source of the resonance. So I ended up reprinting all my parts, disassembling everything but the frame and electronics, and reassembling everything with new printed parts. After retensioning the belts, I got to this:
ratty-blush•16mo ago
That's considerably better, isn't it?
definitely better, but the Y peak seems a bit low compared to the x peak in the belts graph
ideally you'd have 2 sharp peaks
but if it prints well, just print
ratty-blush•16mo ago
Should I tighten one of the belts (y) a little more then? Print quality is already noticeably better than it was before, but I guess I can always come back to this point 😂
What I have been doing is homing the printer, then sending this command ' M18 X Y' to disable both motors, and opening a guitar tuner app. Put the phone display facing up and place your phone againts the bottom idlers and pull the string on either side. On my vcore I end up at e^2, if its too low ill push the grantry down, if its too high then i pull it up. Once you have a single side hitting it just right, move to the next side. Here youll either loosen or tighten your belt to match the same note. That has gotten me 80% there instantly. However like you i have some resonance im trying to figure where its coming from.
I use GuitarTuna : Different tuners use slightly different formulas to tune, this one seemed the most consitent and reliable, being that its a belt and not a string.
ratty-blush•16mo ago
I messed up my graphs, badly, trying to bring that Y peak a bit higher 😠.
What I'm struggling with is knowing what belt I need to tighten/loosen when I'm fiddling with this. too often I think I'm handling the right belt and end up with a completely messed up graph. After many attempts, I just loosened the belts completely, and tightened them with a tunner app as suggested by drops up to this point.
ratty-blush•16mo ago
So my theory goes that means blue needs more tension, and I believe that is the lower belt, to the right if I look at the printer from the front. So I tightened that belt by half a turn, run the belt tension graph again and end up with this:
ratty-blush•16mo ago
That kinda seems better? So I tightened the same belt by another half turn. Which results in this:
ratty-blush•16mo ago
At this point my logic says continuously tightening the same belt is only going to make me build up ridiculous tension, so I start to wonder if I should just loosen the other belt a bit. But I've messed this up so many times I have no idea if I'm on the right track anymore.
Well you graphs are definitely looking better in comparison to your initial ones posted. Im on the same boat when I said it gets you 80% there because i too am stuck in a similar graph , and after trying to chase perfection I end up fucking it up after a while.
My understanding is this , if you made the vcore and have no mods then the upper/lower are correct on graph. The blue is the left spike, the yellow is the right spike. The belts have a natural range of resonance, so when you loose and tighten usually they stay around the same. frequency unless there too loose, then the whole graphs tens to shift.
Additionally when you tighten one side it will also affect the other side slightly, because you increase/decrease tension on that belt too.
I would also like to mention something about the tuner, when you tune both belts to same note, that doesnt mean there tension the same. So dont use it to keep tuning after every time you build the tension graph. Because as you can see they should slightly resonate at different frequency.
This one has looked the best, myabe loosen the upper would of been the sweet spot.Honestly I would have ran the input shaper graph and if i got resonable numbers and at least on zv , i would have left it alone lol
ratty-blush•16mo ago
The thing that bothers me is that I had a beautiful graph last week and something I did messed it up, but I have no idea what. So I'm obsessing about this. To the point that spent half of today disassembling the whole x/y gantry and idlers trying to find any sources of binding or misalignment and couldn't find anything obvious.
I have now reassembled everything again and I have a very similar graph to the one you just mentioned (after 4 or 5 hours fiddling with the belts), for which I ran an input shaper graph and decided to settle on it for now as it wasn't too bad.
I'm thinking part of the problem may be my wire loom, as it feels like it's pulling on the toolhead at certain positions (when the toolhead is not near the center of the bed) and is quite phat. So I'm considering replacing some fat wires I had in there with single awg 28 wires in an attempt to reduce this and hope for the best.
When you say tightening one side affects the other, how does it affect it? I've read this before, and my intuition/understanding says it should "loosen" the other side (as there is a stronger force pulling on the opposite side), but I've not been able to confirm it from the graphs.