I'm getting a used MK4s with a MMU3 in addition to my MK3S+ and I would like to use the MMU3 on both machines. I already have the information that the Rambo Borad needs a different cable - so I wonder if i could customize an "switch" on the MMU3 where both cables could connect to the controlboard. Are the pins/cables the same? is there a diagram which is which?

I’ve developed an open-source script that helps to solve the common issues with wipe towers in multi-material printing:

● no more “elephant’s foot” at the tower base

● fewer blobs, strings, and clumps

● reduced risk of nozzle collisions and crashes

● less delamination of layers when printing with different materials

● stronger towers thanks to improved geometry (also supports spiral towers with inner hole)

● configurable parameters via a simple GUI

👉 The script works as a post-processing tool for PrusaSlicer / OrcaSlicer:

it modifies the generated G-code and automatically rearranges the wipe towers.

GitHub (download & documentation):

https://github.com/nic1972/SmartWiper

On GitHub you will also find the link to the explanatory video.

If you like to try it out, feedback is very welcome – especially from other MMU users struggling with unstable towers.

Best regards,

Nicola

TL;DR: First rewinder that actually works reliably for me.

Over the years my experience with multi-material units has been disappointing. The MMU2 was a hot mess, and the Mosaic Palette arguably even worse. So when the MMU3 came along, I ordered it with much trepidation and little hope. Yet so far, it's been working flawlessly. (I should say that I also have an XL-5TH, which I use for true multi-material printing; the MMU3 is mostly used for multiple colors). The main downside (for me) is the buffer system. It's one of many printers in my workshop, and I just don't have the space to set up the original buffer as suggested by Prusa. Plus, it's a pain to load (and yes, I understand that folks have come up with improved loading options).

So I looked into various auto-rewinders, including Groenhuis's https://www.thingiverse.com/thing:3338467 , Jerrari's https://www.printables.com/model/292276 , and Broomé's https://www.printables.com/model/961047 , and spent way too much money, filament, and time on it.

I want to be very clear–if those systems didn't work for me, it's probably my fault. Plenty of people have written good things about them, so they're clearly not "broken". In fact, in each case I was very impressed with the ingenuity of the designers. In my hands though, I had too many reliability issues to put them in production.

For all of those designs, to make sure I didn't have printer-to-printer variability, I printed them on one machine, with the same gcode, and the same filament. Printed parts always looked excellent. I usually ended up with one or two units that worked fine, and three or four that didn't. Most of the time I saw binding during the rewind at some point. Cleaned parts again, used super lube, etc., but I never got to the point that I had confidence in any of those systems. To reiterate, given that others have been able to use these systems successfully I suspect "it's me, not you", but I just don't have time nor inclination to work out issues with solutions that have limited tolerance to inevitable variations in FDM printing. My solution eventually was to just not use a buffer at all, and I actually never experienced any issues.

So when I saw the post about the Filamentalist, like the MMU3, I went into it with much trepidation. But, hey, I'm a sucker for pain, so I printed the parts from https://github.com/Enraged-Rabbit-Community/ERCF_v2/tree/master/Recommended_Options/Filamentalist_Rewinder and ordered the hardware as a kit from TriangleLab https://trianglelab.net/products/filamentalist-passive-rewinder-full-kit?VariantsId=11834

The plastic parts printed easily. I followed the recommendations to use ASA, 40% cubic infill, and 4 perimeters. I suspect they will print just fine with PETG and less infill.

The Trianglelab hardware kit was excellent. All parts (and some extras). Not cheap but I didn't want to introduce another complication by buying hardware piecemeal that then didn't work together. 

The documentation on the GitHub site was excellent, with nice pictures, but a couple of (obvious) steps were missing in the assembly instructions. Assembly is very straightforward, and once you've built the first, the next ones are easy to do without referring to the instructions. 

There are a few variations, including a rear load option, which is the one I chose. I also designed and printed a base to put the Filamentalist units on my RepRack. So with the rear load option, the PTFE tubes run down along the workshop wall into the backside of the Mk4S LACK enclosure to the MMU3 (until—heaven forbid—they ship the Core ONE upgrade kit and the Core ONE MMU3). Rear-load really means the PTFE tube is routed to the rear, but it actually loads the filament from the front, which makes it IMHO much easier to feed the filament than with the standard option. Plus, the tensioner screw is easily accessible from the front.

Every unit I printed worked right away, with zero fuss. I think the "secret" is the one-way bearing, a machined part that just works, unlike 3d printed clutches and similar solutions.

I've only had a week or so of use but so far it's been flawless. Sure it's early days, maybe I'll run into issues later, but so far I'm pretty giddy about it, and as others have asked me to post my experiences with the Filamentalist, I thought I should share this "early look".

Here is the original design from Prusa.

Here my design

I'm aware that when the MMU3 appeared the only nozzle that could be used with it was the 0.4.

But now, I wonder if it is already approved by Prusa that we can use other nozzles, specially the 0.25, for more detailed text in prints?

Thanks!

I wanted to show you my pre-design for a buffer setup. I kind of fell into a creativity low, maybe you have ideas how to make it better and discuss it.

For the system I attached a Check-Valve (Filament Check Valve von Yannik) to the spoolholder. Therefore the filament can't get pushed back into the spool. I then attached about 80 cm PTFE-tubing (2,5 mm inner diameter/4 mm outer) to the check valve which reaches all the way to the MMU3. But I didn't attache the tube to the MMU3. I attached another short piece of PTFE-Tube to the MMU3 and then a fitting to attache a bigger PEN Tube (4,5/6mm). The PEN-Tube is 60cm long. I was then able to thread the long PTFE-Tube into the PEN tube.

Now it is easy to load filament because the two PTFE Tubes are almost touching. I just need to make sure, that the long PTFE is pushed all the way inside the PEN, which isn't difficult.
For unloading I just need to release the Check-Valve and pull the Filament. 

When the MMU3 does a filament change, it trys to push back the filament. But because of the Check-Valve, it pushes the PTFE Tube out of the PEN tube, to extend the path from the check-Valve to the MMU3. 
When the MMU3 loads Filament, it first pulls the PTFE Tube back into the PEN tube. When the two PTFE-tubes are touching again, it pulls filament from the spool. 

Now my problems are:

• The pushed out PTFE tubing hangs loosely. I am afraid, that it gets tangled with the other tubes. Is there a way to somehow guide the tube which gets pushed out of the other tube?
  I only printed short prints yet, I will update as soon as I printed a long print with lots of filament changes.
• It works that way, but when I switch the attachement of the tubes (big 4,5/6 tube attached to the check valve) it does not work. When the mmu3 loads Filament, it immediately takes filament from the spool, without pushing the ptfe tube into the other tube. I am not quite sure why that is...maybe because of friction? Is there a way to force the behaviour I need without relying on magic =P

Here are a video and some pictures of the setup:

Btw: This is my first forum topic ever made. I hope I described everything well.