RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Huh, I'm now constantly getting Y crashes on my printer. Weird. Maybe I should just re-flash the firmware.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
For what it's worth, I've gotten a series of crashes and layer shifts over the past month or so. Every time, I've been able to track it down to one mechanical problem or another:
1. Belt tension
2. X-Axis idler went bad, deflecting to the side and not rotating smoothly.
3. Grub screw worked itself loose, letting the pulley slip. Happened on both X and Y, plus the Bondtech gear on the extruder.
4. X-Axis idler holder cracked and was slowly working itself loose. (I printed that part in PC, too!)
5. Y-Axis motor holder started warping, deflecting the belt so that it rubbed into the side of the idler's flange.
6. Did I mention belt tension yet?
I've convinced myself it had to be the firmware several times, but nope, every time it was mechanical.
In the meantime, I'm still impressed by what the 0.9° steppers can do. I'm having a blast experimenting with 0.25mm and 0.15mm nozzles this week!
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
...oh, and one more (because edit timed out):
7. Insufficient z-hop with a design that warped upwards, leading to actual for real collisions with the model.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
I don't doubt it could be something mechanical, it's just frustrating when there's no obvious sign what it might be. I suppose the Y pulley is the place to start, that's the main thing that got messed with during the motor swap; everything else was left alone.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
... I may have just stumbled on something rather interesting. The tower I just printed had VFAs.
... But only on ONE side of Y.
(and a tiny, tiny bit on the leading edge of the other side).
This points to something else as a very interesting cause of VFAs: Motor load. If the motor is unevenly loaded, it stands to reason any nonuniformity in the cycles/steps will be amplified. (These were not 2mm "tooth" bands, they were at the very fine-pitch VFA spacing).
Looking at the X pulley, I noticed it appearead to be further from the body of the motor... and this was true, because the OMCs do not have the flats ground all the way to the base of the motor shaft like the LDO ones from Prusa do. So I'd seated it as far left as it would go, but it wasn't far enough and the belt was still rubbing on the side of the drive pulley in one direction, and when the direction changed it'd work its way back and stop rubbing, hence the minor VFAs at the leading side. I've flipped it around and repositioned it best I could but I may need to either file down the motor shaft more (blech, done enough of that in my Skelestruder build!) or alter the motor mount slightly (Also somewhat blech, I hate modifying STLs - I'm using a more solid mount already than the stock one).
Will report back when the calicat finishes printing.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Not a VFA to be found after I'd shifted the pulley.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
I've convinced myself it had to be the firmware several times, but nope, every time it was mechanical.
For me it never was the firmware directly but the way my printer handles the firmware in general.
First homing would not work anymore. Flashed the same firmware again and it worked again. Every time with gcode reset and reset button reset. After that it worked again. Then I had layer shifts which were my fault because I designed a part without considering the print temperature but at the same time the printer began to begin printing at a higher layer and X/Y would not pass self test anymore. So again, flashing and resetting. Hopefully it stays this way now and it's not the board itself which starts to degrade somehow.
----
I also like to share some "real world" prints with the 0.9 motors. First three are with direct lightning and one with flash which does hide any inconsistency very well. The others are with lightning from above to show all inconsistencies (filament allows diameter change of 0.05mm).
Right now I'm very pleased with the quality. I must admit that the VFA did it's job in hiding those inconsistencies. 😀
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Doing a full reset worked. Thanks.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Not a VFA to be found after I'd shifted the pulley.
That's almost funny (in a sad way). My X-Axis belt doesn't rub, and my Y-Axis rubs like it's a polishing wheel (no amount of adjustment helps, thinking the motor is not aligned); and guess which axis has artifacts and which has none?
But I'm fighting a .32 mm pattern in X, not VFA's related to microsteps. But based on this belt idea, maybe I need to "adjust" my X motor to an off-angle ... lol.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Tim, I would also recheck the bolt tensions on the x-carriage cover. If those cover bolts are too tight (or there is a burr in the printed bearing pocket) going tighter will put the bearings out of alignment. Try a bit tighter/loose on the cover bolts and see if it makes a difference. I have noticed this effect more with carriages printed in stiffer than PETG, but it could also happen in PETG. (You wouldn't believe how many spools worth of x-carriage and extruders I have printed and tested)
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Not a VFA to be found after I'd shifted the pulley.
That's almost funny (in a sad way). My X-Axis belt doesn't rub, and my Y-Axis rubs like it's a polishing wheel (no amount of adjustment helps, thinking the motor is not aligned); and guess which axis has artifacts and which has none?
But I'm fighting a .32 mm pattern in X, not VFA's related to microsteps. But based on this belt idea, maybe I need to "adjust" my X motor to an off-angle ... lol.
Are you running a smooth idler or a toothed idler? The former really aren't meant for toothed belts' toothed side as you'll get small variations in tension as a result. Another of PR's "little white lies"...
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
My artifact is 2 full steps of the motor and unlikely to be associated with belt teeth or microstepping. It sort-of fits the normal diagnosis of an unbalanced winding (extra turns). aka poor QC.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Does the 0.9-Degree-Stepper-Support branch have support for MMU2S?
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Does the 0.9-Degree-Stepper-Support branch have support for MMU2S?
My 0.9 firmware only supports 0.9 degree motors on X Y Z and E. It does not support 0.9 motors in the MMU2S. I doubt it would be beneficial to use 0.9 degree motors in the MMU2S as it is a relative crude motion filament moving device.
If you meant does can you connect an MMU2S to a printer that is running my firmware, then yes. It has all the normal Prusa firmware bits including MMU2S support.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Does the 0.9-Degree-Stepper-Support branch have support for MMU2S?
If you have a stock extruder, or a modded one with stock filament path lengths, it should be fine.
If you have a BMG, it's been added by Guy (he updated the lengths).
If you have a Skelestruder, then I'm working on getting those changes together.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
i use it with 0.9 deg X and Y motors and a BNBSX extruder and MMU2 - works fine
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
note: it looks like Prusa is getting closer to new releases of their software - prusaslicer, firmware and mmu control are all getting changes
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Yes. It is going to be an awful mess to merge back with the current Prusa/MK3 branch. I suspect I will have to manually add all my edits back into a fresh copy of the new MK3 firmware. Then somehow add the skelestruder, LA15 stuff back in. I don't think those changes can be auto-merged into the new MK3 files.
Anybody know if the new MK3 files already have LA15 support in them?
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
The way it's often for custom changes to the Linux kernel that can't be upstreamed is to maintain a patchset. By patching with this set, almost any version can be modified to gain its features. And modifying it for newer or different versions is not so difficult either.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
FYI MM-control 1.0.6 RC2 is on github
Summary:
- Improved stealth mode
- Unload speed decreased
- Filament cut feature
- Disabled unwanted selector movement
- Automatic idler disengagement
Detailed description of changes:
Improved stealth mode
The value of the idler holding current was increased in order to ensure its proper motion. Low idler holding current value might lead to an idler misalignment. For MK3S MMU2S the idler holding current values are the same for Stealth and Normal mode. For MK2.5S MMU2S only stealth mode is available.
During acceleration and deceleration, the current value drops or increases. This is directly affecting the value of the torque, which might be low in some situations and thus the motor doesn't have enough "power" to load/unload the filament. To avoid this situation we are changing the value of constant in the Trinamic driver, which results in sufficient current and torque under any type of movement.
Similarly in Stealth mode, for some movements, the unit switches temporarily to the Normal mode, which provides higher current and torque. As soon as the torque is no longer needed, the mode is switched back to the Stealth.
Unload speed decreased
As explained in the paragraphs above, there were some scenarios, where the motors might not have enough torque to load/unload the filament properly. To avoid any issues, unload speed is slightly decreased to maintain a higher torque value.
Filament cut feature
This feature is utilizing the included blade to cut the unwanted tips of the filament. The current release contains improvement in the situation when the printer recognizes a failed cut. After the filament is cut, printer tries to load the filament and trigger the F.I.N.D.A. sensor in the MMU2S unit, in case this operation fails, all movable parts of the MMU2S unit are rehomed to ensure proper alignment.
The filament cut feature is disabled for now, we are still working on it. We plan to enable it in a future firmware release.
Disabled unwanted selector movement (during fail state)
When the MMU2S experiences a filament load/unload fail, it asks the user to solve the issue. After resolving the problem, the user has to confirm it by pushing the right button on the unit. However, if the right button was pressed for a longer period, the unit also moved the selector to the next filament. The current release is changing this behavior and selector can be moved only after the right button is pressed again.
Automatic idler disengagement
During the filament load, the MMU2S unit is pushing the filament using the idler until getting information from the extruder (IR-sensor). This information was sometimes lost and the idler was still pushing even when no longer needed. In the current release as soon as the preset distance is loaded through the MMU2S unit, the idler automatically disengages and returns only when prompted by the extruder (IR-sensor).
Thanks to Brian Smith for contributing to the code, more information in Make sure idler disengages