MK4/MK4S with MMU Main-Plate with improovement steps *No more stuck or filament jam in the Extruder*

18.05.2025: OBSOLETE! Successor model IAF Idler (Install and Forget) can be found here:

https://www.printables.com/model/1269218-nextruder-mmu-multi-material-gamechanger-iaf-idler

After many failed prints in which the extruder stops transporting the filament to the nozzle , I spent a lot of time and energy finding the root cause and an improvement for that issue .

Steps to improve the behavior:

1. Replace the grooved idler bearings (groove is out of filament path), with standard 693-RS bearings, I bought this one: https://amzn.eu/d/0in5q3S
    
2. Replace the main plate with the attached one here , the filament is guided as much as possible (without a complete redesign) . I also successfully tested flex filament with Shore 82 inside the enclosure (warmer pulley) .
   
   For printing : Due to currents unavailabile Prusament PC-CF filament, I purchased IEMAI PC-CF filament, and after some adjustments , the print result was like I printed previously with Prusament PC-CF.
   
   Here are my settings for IEMAI PC-CF Filament:
   UPDATE 09.11.2024, due to a small part above the Filamententry, which easly breaks, the Nozzle temp should be higher as described here or you anneal the Plates directly after print to gie the part the chance to reduce internlt forces after print.
   Nozzle: 230°C (yes, really 230°C, maybe without heated chamber it must be hotter)
   Bed 110°C
   Chamber: 60°C
   Extrusion multiplier 1.1
   retraction 4mm
   retraction speed 20mm/s
   deretraction speed 30mm/s
   deretraction extra length 0.1mm
   minimum travel after retraction: 10mm
   
   parameter for printing the main plate:
   0.4mm nozzle
   0.15mm structural preset,
   dynamic layer height
   3 perimenter, 4 bottom and 4 top layer
   50% rectiliner infill
   Support only for the bridge part (enforced)
   snug support
   0.2mm Top and bottom Z distance
   2 layer support interface for top and botto
   shred around support (for snug)
   0.5mm XY seperation of support.
    
3. Flatten the idler screw threads as seen in the picture to realize a smooth return of the idler without strong spring force.
    

4. recommended: replace the existing Idler-tension screws. use the following specefication, then you can more optimize load an unload filament:

0.5mm*5mm*15mm, spring pitch approx. 2,53mm, 8.5pc.number of active coils, approx 8Nm max static force.

I used these springs: https://www.amazon.de/dp/B076LYBZMZ/ref=pe_27091401_487027711_TE_SCE_dp_i1 

UPDATE: 09.11.2024, today i used some cheap ASA filament, where the diameter varies more than it should and this caused some problems.

So I thought again over the idler tension… I already used other springs in the idler.

The original springs have 0.8mm x 5mm x 15mm. The winding slope is here 1.553mm and the spring has 8,5 active colls. A identical spring has here a 35,8 Nm max. static force x2 springs = approx 70Nm max. The spring has max. 4.67mm static spring deflection. - So far nothing impossible, BUT. we did not calculate the Idler lever. 

The to the spring applied lever length is 40mm, the lever from hinge to the middle of both bearings (as average filament pressure point) is 15mm. That means the tever Rationis 4:1,5

In numbers means that, the idle lever has the capability at the bearings to move exactly 1.75mm (4:1,5 Lever Raion / 4,67mm max spring deflection = 1,75mm) 

For the applied force we need to calc: 4:1,5 * 70Nm = 186Nm or 18,6kg force applied to the filament… (of course, this is the worst case and even if there is no gap…)

For me, this spring isnt a good choice here. 
Once you have the extruder design understand, you will see, that the springs arent needed to apply force to the filament, the force the extruder can apply even without any spring is more as enough. The springs are really only needed to move the idler after filament unload back, to disable the sensor… 

Make yourself a picture: 

Very Important Notice:

It is neccesary to cover all three points together!!

If you keep one of them untouched, jamming and stucking will still happens. I tried all combinations vice versa, without success of improoving that issue. For example, yesterday I reinstalled the grooved bearings and after the first print layer, the extruder stucked the filament, after reinstall the bearings without groove, it again works as a charm.

The only one alternative to that improvement is to complete redesign the parts:
idler lever,  Idler-nut-FS, Main-plate, Swivel-A, Swivel-B, Screw-guide together because the bearings needs to be relocated and the screws must be alligned to the holes in the nextruder heatbreak. But Im not sure if that is possible, due to space limitation…

Update: If you also has problems with Layer-shifts, that mod and guide may help to improove:

https://www.printables.com/model/1060870-mk4mk4s-toolhead-ptfe-guide-layer-shift-issue-impr

Further helpful Modification:

Sometimes it happens that a piece of filament blocks the Idlerlever and causes errors on loading filament. 

To remove the idler lever in mid-prints, where it is not good to remove the extruder gear, I came to that idea, that helps here:

From technical side and applying force by PG-Cover to the Gear, it is better to use only 2 diagonal screws instead of 3 unsymetrical mounted screws. (Here the Gear will be pushed to the lower left.

You can also use this solution now, to make it right and fix the PG Cover by 4 flat head-screws now. If all 4 crews are applied by identical force, the Gear is holded correctly.

Issue Explaination:

Important: The issue is a combination of several smaller parts and, like in my case, it will be further worse if you print inside closed housing or even heated chamber.

1. Grooved Idler lever bearings:
   Like seen in the picture, the groove and the filament guide are not in the same axis, which causes the filament to been bent to the front of the mainplate, out of the filament path. Then, the gap where the filament runs is larger then intended which causes in flatend filaments.
   
    
2. Mainplate:
   due to a edge close to the hinge of the idler, the lever could not be pulled close enough to the pulley, after a certain point of thighten the idler screws, the idler will be bent and due to the chamfer at the frontplate the idler will be twisted and pushed towards Mainplate front.
   

Rework of the mainplate, due to design issues of the original Prusa MainPlate, idler-lever and grooved bearings combination that results filament stuck/jammed in Extruder. THese are the most common sympthoms: 
- Filament stops extruding mid - print
- Unloading of filament by the MMU is not possible, manual pull out is necessary .
- Upon filament load, the filament sensor does not enable.
- Unloaded filament has dents, grinds, twists, or bends at the position where the extruder pulley is ( approximately 50 mm behind the filament tip) .