Sunday, November 19, 2023

Prusa MK3s+ & MMU3 Upgrade Review

 MMU2s --> MMU3 Upgrade Review

I had an early MMU2, which I later upgraded to a MMU2S. This meant I was using an old 2018 version of the MMU PCB and stepper motor set. The MMU3 upgrade kit doesn't include any of these - it is basically just a hardware kit (as in nuts and bolts), a new power supply board (which is probably the actual critical item of the MMU3), and a full set of redesigned and optimized printed parts (which you print yourself from supplied filament). All of this is well documented elsewhere, so I won't say anything here about the actual build other than it was fairly straightforward and easy to follow. 

MMU control board
Ancient MMU control board- still being used!


Prusa MK3S+ with MMU3
Assembled MMU3 in action

MMU3
Looks familiar but many improvements under the hood

Clown fish printed with MMU3

Clownfish initial test print in eSUN PLA+. Model source HERE

The initial test prints were 'out of the box' better than the old MMU2s - but still not perfect. Interventions were still needed - but a large improvement from the MMU2s. The errors were also much easier to deal with as there is now some trouble-shooting text displayed on the printer's LCD. 

Errors typically were of two types:

  • The FINDA or print head sensor wouldn't recognize the filament change had been successful and reached the head sensor
  • During filament retraction, the filament tip would back too far out of the selector, so the selector couldn't grab and load it during the next exchange

FINDA Tuning

To address the FINDA error described above, I adjusted FINDA sensor. The installation manual recommends setting this to to barely turn on when a filament is inserted - however didn't always recognize when a filament was inserted during changes. I adjusted this to barely turn off when a filament was removed - this seemed to increase reliability of filament changes, in particular with flexible filaments. The new method of tuning this sensor with a screw is a big improvement over the MMU2. 

After adjusting the FINDA- the only failures seen was from filament retracting too far out of the MMU selector- probably due to the auto-rewind spool holder. I'm also printing TPU using the MMU, which seems to work better than expected as this isn't a supported filament type from PRUSA. 

Too much auto-rewind action

I'm using gravity auto-rewind filament spools. These seem to cause intermittent problems by pulling too much filament out of the selector during filament change. I'm planning on switching to a simpler spool holder with no auto-rewind, and just a method to manage the slack created during filament unload. I haven't found an existing design yet that I like, I might just design and build one. 



3D Printing FDM Multi-material adhesion experiment

Background

Most multi-material prints I've seen on Thingiverse or Printables involve using multiple colors of the same material for cosmetic effects - printing characters, making art, etc. 

While this is interesting, being able to print multiple types of filament in the same 3D print also enables using different materials for mechanical design reasons. Examples include adding a soft touch surface to a rigid mechanical part or flexible hinges to a rigid part. 

For this to work, the different materials must adhere together during printing. If there is no bonding during printing, mechanical interlocks could also be used. 

I'm planning on designing and making a part which good adhesion during printing is desired, so I conducted a quick experiment to test adhesion between various materials.

Method

I designed a small block- 25mm x 25mm with two solid bodies. Test block available on thingiverse here

I had 4 material combinations I wanted to test:

  • PLA to TPU
  • PETG to TPU
  • Nylon to TPU
  • PC to TPU
For this experiment, for all experiment trials conducted, the experiment material was the base, printed directly on the build plate, and the TPU was the top material. 

Equipment

  • Prusa Research MK3S+ with MMU3
  • Prusa Research smooth PEI build plate

Prusa Slicer really didn't like using TPU as a material option to print with the MMU, so you need to manually select this material profile in the filament tab.

PLA / TPU test block

All Test blocks


Results Table

Base MaterialTop Material Adhesion
PLA
eSUN PLA+, Blue
TPU
ANYCUBIC A11795 TPU, red
Very Good
PETG
Prusament, Orange
TPU
ANYCUBIC A11795 TPU, red
Good
Nylon
https://www.lay3rs.nl/ Nylon, Black
TPU
ANYCUBIC A11795 TPU, red
Very Poor
PC Blend
Prusament PC Blend, Black
TPU
ANYCUBIC A11795 TPU, red
Poor

Conclusions
This TPU appears to adhere well to both PLA and PETG. Qualitatively, PLA may be slightly better, but both seemed firmly adhered with this test piece.