VC4 500mm Hybrid Belt Tensioning Trouble Shooting
Having some trouble getting a good clean input shaper now with the Hybrid Y installed. I'll start from the beginning:
First three photos are belt tension and the X and Y Axis input shaper. All three look good. CoreXY Y Axis MZV is 3,400.
Fourth photo shows the initial hybrid setup giving me two peaks. At this point, I was getting lots of 2 and 3 hump EI recommendations (not great).
Fifth photo - I loosened the hybrid belts from 4th photo, which improved the look of the graph and gave me more MZV recommendations, but resulted in significantly lower input shaper recommended values. This pattern continued until the hybrid belts were so loose that I regained my stock CoreXY profile.
From the fifth photo, I tightened, and got a photo similar to the fourth photo with minimal adjustment. Last last night, I took apart the hybrid system and ensure everything was built correctly.
Sixth photo was after rebuilding the hybrid system and running super, super, super low tension (barely attached). Unsurprisingly, graph has very low recommended acceleration.
Seventh photo - Tightened from #6 and observed the same pattern - eventually two peaks form again. I tried loosening from here, only to see the same pattern - higher frequency peak disappears but get a very meager improvement in recommended acceleration. Since things weren't working like, decided to even tighter. Eighth photo shows me going tighter. I get two very distinct peaks at that point.
9th photo is even more tension - two peaks start to equalize but I feel like I am really, really at a high tension 10th photo: continued to increase tension. First peak is reducing, but still present.
That's more or less where I sit right now. I feel like the Y hybrid motors are super high tension. Reducing tension from here results in two peaks returning.
Sixth photo was after rebuilding the hybrid system and running super, super, super low tension (barely attached). Unsurprisingly, graph has very low recommended acceleration.
Seventh photo - Tightened from #6 and observed the same pattern - eventually two peaks form again. I tried loosening from here, only to see the same pattern - higher frequency peak disappears but get a very meager improvement in recommended acceleration. Since things weren't working like, decided to even tighter. Eighth photo shows me going tighter. I get two very distinct peaks at that point.
9th photo is even more tension - two peaks start to equalize but I feel like I am really, really at a high tension 10th photo: continued to increase tension. First peak is reducing, but still present.
That's more or less where I sit right now. I feel like the Y hybrid motors are super high tension. Reducing tension from here results in two peaks returning.
12 Replies
Went back to non-hybrid and found a broken plastic piece on the hybrid assembly. Not sure that was the cause of my issue, but starting back over once more with hybrid to see where I get. Experimented with belt tension a bit more on the stock CoreXY and got some improvements.
now onto hybrid again
This was the best I could get for hybrid. Unfortauntely it looks like even this input shaper degraded after repeating it a few times and now I'm back to two distinct peaks:
Starting to look more correct, but now the 100hz peak is getting terrible.
100hz seems related to wiring ( @hhes pointed this out and holding the umbilical seems to help it, so thank you). Still working on getting the Y peak clean, but I'm at least starting to get input shaper results that resemble a functioning printer.
I think the primary issue was that I was running the electronics enclosure bolted directly to the frame. Now with the electronics enclosure removed, resonance has significantly improved. Don't want to definitively say that is the cause yet until I can truly get good clean graphs, but I'll keep updating here in case this helps someone in the future)
I had strange noise issues with my rear panel and the electronics enclosure. Ratrig shipped enough of the foam tape that I cover both the top and bottom of the back face of the horizontal rear cross member and the left and right side of the back face of the vertical rear cross member with it. I then used some cork to offset the electronics enclosure from the rear panel. You might need longer screws for that though.
100hz peak is addressed. Umbilical sheath is installed (upper two most zipties on toolhead plastic piece secure it to the toolhead) and wires are run through it. Removed all zipties hold electrical wires to the umbilical wire and just let the sheath hold the wires in. Seems to work well:
will start tensioning the hybrid motors now
Do you have filament loaded while doing the tests?
Yeah has been with filament in it
I need to get back to tuning it again, but haven't. Been a bit busy using the printer