IS not working or what's causing the ringing?
I've been at this for days now and just can't figure it out. What ever I do I just can't get rid of the ringing.
These four parts are printed with exactly the same settings beside external perimeter speed which is 80, 180, 190 & 200mm/sec.
I was trying lower speeds when someone said the steppers worked best in the 180-200 range, hence the big gap in speed.
From the left: 200, 190, 180, 80 on all pictures.
To provoke the issue I manually put a seam in the middle of a smooth face and left the middle open.
My current (RRF) config is attached, the printer is a VCore 3.1 400, Hiwin Z1 X-rail, Vzbot head/extruder/hotend and a 6HC/1LC running v3.4.6.
I activate IS with this M593 P"zvddd" F61 in config.g and PA with this M572 D0 S0.04 from my filament config in PrusaSlicer 2.6.
The two IS graphs are on without IS but I still left the zvddd graph in there so you can see if it's a good fit or not. The other (lower amplitude) is with zvddd active at 61Hz.
I hope someone can help me fix these issues.
19 Replies
i assume this is a limitation of RRFs IS implementation
afaik it cant handle two axes at once
you can test this by rotating that piece
so that you have a y-move only
conventional-tanOP•2y ago
Like this?
well, i would rotate it so that you have the two holes in the y-movement
but yes
just to check if the y IS works at all
conventional-tanOP•2y ago
Isn't the seam a better way to see it? At least it seams very clear on the picture above?
only one way to find out
conventional-tanOP•2y ago
🙂
Printing now, I'll report back as soon as it's completed.
conventional-tanOP•2y ago
Here we go. Not sure why the seam changed after just rotating the model though. (New on the right)
conventional-tanOP•2y ago
Printed one more, this time with Input Shaping disabled (and the seam fixed)..
Input Shaper might do something but it's so minor it's neglectable.
https://imgur.com/a/2TGuFIu
Imgur
Untitled Album
@oliof maybe you can help since you know a lot about RRF?
we already have discussed this at length in #rrf-duet
I am currently caught up in other projects so my V-Core is not ready. I can say that IS works on my RRF running V-Minion (to the extent it does in RRF)
@Helge Keck RRF does IS on both axes, it just doesnt support different algos and frequencies.
well, i have never seen a IS result with the same frequencies on X and Y
that s just how its implemented in RRF ¯\_(ツ)_/¯
(doesn't mean I agree with it)
yeah, this could be problematic
thats what i meant actually
well, this is a point of contention between the RRF implementation and the Marlin/Klipper one. But @miklschmidt already pointed out ... two years ago(?) that RRF's IS is probably ineffective at lengths shorter than half the IS frequency in millimeters. There are other bugs around IS interacting badly with mesh bed compensation that require mesh bed taper -- @zyber.se I asked you to try this https://docs.duet3d.com/User_manual/Reference/Gcodes#m376-set-bed-compensation-taper
conventional-tanOP•2y ago
Interesting information, didn't know that. So as I have mine at about 60Hz that would be it won't help for anything < 30mm. If that's so it feels like the implementation is a bit flawed. I did hear v3.5 have an improved Input Shaper, not sure exactly what has been done though.
What do you want me to try with M376? I have it set to 10mm. I also tried disabling the mesh completely with no apparent improvement.
It would be interesting to see how the same model prints on some other printer, wither with RRF or Klipper or preferably both to get a reference point.
If someone would be up for that, I've attached the model (little parts of a cat iglo) 😄
conventional-tanOP•2y ago
@oliof latest attempt, better but not perfect. Wdyt?
https://imgur.com/a/gahSFnm
Imgur
Untitled Album
bulging corners indicate there are issues with PA which you need to retune after IS. I would set compensation taper to 20x line height, so 4mm fpr 0.2layer height.
what is your current jerk and acceleration set to?
I havent tested IS on 3.5 yet and I am not aware that there have been improvements made. I will need to go and check the commit history
conventional-tanOP•2y ago
I actually had the taper at 4mm already and I'm printing 0.2mm layer heights.
conventional-tanOP•17mo ago
I printed a ringing test which was configured to increase speed six times. From bottom we have:
30, 50, 100, 150, 200 & 250
No other settings were changed since earlier examples.
Jerk is still at 2.5 on both x & y, acceleration 3k.
https://imgur.com/a/ilIudF4
Imgur
Untitled Album
