MMU3 filament unload - filament slips back to far and cannot be grabbed anymore
On filament unload, it happens many time that the filament gets unloaded a bit to far. I have identified that if you use a autorewind solution, this may happen more often. But also with stock equipment this may happen. The filament stops behind the pulley axis, so the pulley cannot grab it anymore.
With the upgrade to MMU3 this issues increased, due to less moving resistance.
A solution would be, if there would be a possibility to configure a pause time after the unload before the idler moves to the next position. The filament would then be held tight between idler and pulley and the filament movement (and also some pull momentum from the spools) would be absorbed.
RE: MMU3 filament unload - filament slips back to far and cannot be grabbed anymore
The tubes in the MMU3 went from 2 to a 2.5mm inner diameter so the resistance is less, which is good but really should have even been a 3 internal diameter to reduce friction even more. Friction is to be first to lowered as much as possible, and then controlled in my view. I even use a 3.0 inner diameter small tube inside the MMU in front of the selector and 3.0 tubes throughout. I do not think it is a good idea to rely on the tube diameter alone to keep the filament in place. However, the real issue is the internal twisting and tension that makes the filament want to move in the tubes, and this alone either will pull or push the filament forward or backwards. If forwards it could poke out and interfere with the selector movement. If backwards it slips too far back and the gears cant catch it, which is your issue. The solution is something that is completely missing from the MMU3, and something that an autorewinder will actually worsen, since it adds an extra backwards force that is unwanted and even adds to the existing problem. My solution is to use a slot buffer which adds almost no friction or force, and then to force the filament into a controlled S shaped curve behind the MMU with a set of positioning supports. This S shaped curve puts pressure on the filament and 'locks' it in place after a retraction. This S curve shape is the main design element of my design that I would call your attention to here that solves this slipping problem, but does not add any extra force or undue friction. If you do nothing else, replace your tubes with 3.0 everywhere behind the MMU, and implement an S curve of this shape behind the MMU somehow with brackets or positioners. https://www.printables.com/model/555238-mcs-multi-color-system-for-prusa-mmu