MMU2S not ready for public use
Personally in retrospective I find frustrating that most MMU2s problems are quite obvious and could have been adressed by PR with no extra cost. Basically, my with current knowledge I would never assemble the unit as delivered. There are some mods everyone seems to install at some point to get it working. PR could have included those in the design at little or no extra cost:
- some selector with improved path (optionally with magnet)
- wider diameter bowden tubes
- better quality idler door
- filament inlet with 5x pushfit connector
- ditch the buffer
- some sort of auto rewind spool holder
- several users reported issues with PTFE in hotend, why not include 50cm of material and a cutting/chamfering aid?
A lot of these things are obviously self-printable, or could replace printed parts already at the factory. I have not seen a post by a happy user who had not at least implemented one or two of those.
While I agree with the sentiment,Prusa has been doing that. The filament wallet you poo-poo is proof of that. The people complained and Prusa agreed. It does its job while still being cost effective.
I do think Prusa could adopt the most commonly used mods as new part revs, not sure why they are not.
I'm a new owner of both printer and MMU and I must say reading this topic has me rather annoyed. I now feel conned out of £269! Was thsat cheap for a MMU, I don't know, Prusa has a reputation for producing an excellent printer, and it is, I tested it and ran several prints before attempting the MMU upgrade.
Once on I printed 2 small squres 15cm x 15cm x 3mm in 2 colours and it printed them perfectly. I was over the moon, it was a joy to watch.
I moved on to a 4 colour print, it failed twice on the first layer after 2 colours. I've tried everything I can think of to get this thing working, I thought loading to extruder would prove that at least it could load each colour. I selected filament 1 and the selector moved to the right place but thr idler selected and subsequently fed filament 4 into freash air! Noiw that's nothing to do with the PTFE tubing or shape of the filament end or the buffer or anything else, in my opinion thats firmware!
My bottom line, I can't get any filament to load no matter what I try, I now have a £1000 paper weight and to be honest that ain't funny!
This morning I would have said "I'm with you!" Hours of futzing with this thing to try and get it to load. Then I read this thread. One fellow said something about 2.5mm ID tubing ( I could use a link/source please). Following that train of though I took mine apart this afternoon. I removed those short little tubes to inspect the bevel. Looking at the little tubes closely I noticed it was misshapen/ oval inside. Using a ejected filament end I found way to much resistance in all the tubes. Now I have none and it hasn't missed a beat since!
I drilled the tubes with a #45 drill bit and beveled the other end. Solved. Resistance in the path is the problem.
I also notice that the tube coming out of the selector in the air hose connector had a little too much resistance. I drilled that end of the hose for good measure. I believe the hose isn't quite long enough and bends a little too sharp almost causing a kink where it goes into the connector on the selector.
A nice touch on the selector bearing/gear assembly would be a quick release for cleaning... and a dust cover.
I got mine from Ebay...
this seller has four colours
Note I have not bought from this particular seller, the seller I got mine from is currently not advertising, or at least I can't find them!
there are revised models for the filament selector that slides along the front of the mmu2, on thingiverse, some revisions address the feed path through the selector, making transition from the main MMU body, across the gap into the selector, more successful, others add a magnet to ensure the Finda Ball drops more effectively when the filament is withdrawn,
there are also models that allow the use of 10mm screw thread, 'Pass Through' pneumatic fittings on the rear of the MMU, to make the filament feed PTFE tubes more secure without squashing them.
JoanTabb: "there are also models that allow the use of 10mm screw thread, 'Pass Through' pneumatic fittings on the rear of the MMU, to make the filament feed PTFE tubes more secure without squashing them."
BTW that PTFE link won't send to the US.
I was just going to ask for that, because guess what happened... one of the feed tubes is pushing out of place during eject.
My Die maker friend owes me favors for inventing the control process for their Titanium injection mold machine. I'm going to have the short tubes turned in aluminum or brass or some such. Titanium is too expensive 🙁
I still think the tube between the extruder and selector is too short, the bend out of the selector is too sharp = a bit more friction which seems to be the enemy of the MMU2. I'm thinking that a clear dust cover with integrated tube support and a longer tube for the MMU2 is in order.
Whenever a cable is used in a push - pull control mechanism, the cable manufacturer always insists the first x inches on each end are straight and have a limit on the radius of the bend.
Good news is after the rework I did I had one eject error in the 10 hour trial run + the tube feed being pushed out sometime while I slept.
I'm about over this printer.. No matter how may tweaks i make... no matter how many forums.. i spend more time working on the MMU than printing.. seriously ready to sell this thing. I bought it to print not to troubleshoot ant tweak.. numerous tings that are problematic and i'm fairly technically capable... is this still worth the cost?
Life should be fun. Not frustrating. You might consider just using one channel to print in one color and revisiting it when other people come up with mods that make it work better.
This product is rough but it is pretty cool when it works. I printed a project after I though I had all the bugs worked out. My partner works for a non profit and she hands the gear bearings out to graduating kids as fidget spinners.
It printed flawlessly for the first 20 hours and then started misbehaving with load problems in this 28 hour project. I nursed it along for 2 more days to finish it after coming home from work or waking up to find it stalled yet again.
I will be ripping this thing apart and generating a wish list and fix list.
I see it as more of a challenge. I am determined to make it work more reliably. Maybe the more tenacious among us here will come up with ways to make this thing work better. It's not the most elegant bit of engineering IMO. But I'll address that inn another post.