When this works, it works perfectly
I just upgraded my MK3S+ to use the MMU2S and did the "who follows the manual?" thing. I did invent a few new swear words. Prior to my "upgrade" I was as happy as a clam. This thing would do what ever I wanted. Then I added the MMU2S and everything went to hell in a handbag. I rebuilt everything twice. I had hotend's clog up and fail (cheap nozzles will break if you breath on them wrong and once broken, you can't tap them out). All of this frustration, after I sat back and analyzed, had nothing to do with the MMU2S upgrade! Just me not following instructions. Once I got the first layer calibrated (several adjustments of the PINDA) and now all 5 of the filaments work perfectly. Most are PLA and one is PTEG, but the Slicer knows and everything works. I swear I thought this thing would never work. Dialed in and it just works. It is a pleasure to use.
RE: When this works, it works perfectly
Congratulations! You and I obviously have two of the 'good' ones.
I would like to offer you a few hints, learned by experience (baptism by fire) while taming a MMU2S a couple of years ago.
1. Install a remixed manifold with bowden tube fittings. This will save a lot of hassle when one tube needs to be removed for any reason.
2. Install those 3d printed thumbscrew extensions so you can release the idler quickly and conveniently without tools. (Barely visible in top photo.)
3. Install a power switch that removes the power to the MMU2S so that it can be used as a plain old MK3* when you don't need multi color.
A few notes on #3. The housing used to be in the old Prints section but it is apparently difficult to find. PM me and I will send you the .stl. Also, use a DPST switch and break both positive and negative power. Some people have switched the data lines instead. There are two threaded holes in the MK3* frame that almost appear to be made for something like this.
There's also a lot of chatter in another thread WRT the MMU2S malfunctioning due to marginal voltage. Mine does not seem to be affected by this issue.
RE: When this works, it works perfectly
LOL! Those were the idea's I had in between the swearing sessions. I was thinking bowden at every end of a tube because of the ease of access. The buffer could use a few as I find feeding new filament is easier when you can disconnect the tube between the buffer and the manifold. I can see why it's not done with the production machine as the cost goes up considerably for something that is not used during normal use.