Ikea Skadis mounted Spool holder with filament buffer

Online “recipe-style” backstory:

With the addition of the MMU to my printer I found myself in need of a buffer. I had tried the Multiple Mechanism Auto-Rewind Spool Holder but wasn't a fan of the constant “thud” the geared rack made every time it “skipped” a tooth (I also had issues with the gear binding and never actually rewinding) Knowing other geared versions would be the same I started looking at the various buffer options. 

I built the RMU inspired buffer box and started to set it up and immediately found that dealing with all filaments in the same box with the same lid was a pain. Before I even got it setup and working I decided I wanted something modular, so if I had to change one filament I didn't have to fight with the rest.

I was a big fan of how the spool axle mounted in the Multiple Mechanism Auto-Rewind Spool holder so I reused the bearing caps and retainers from that, as well as the bracket and socket idea from my Ikea Skadis mounts for it. I wanted to keep things as narrow as possible so I could get 5 spools in a row on the Skadis pegboard I have mounted above/behind my printer. 

I liked the simplicity of the Improved MMU Buffer actual buffer mechanism so I sliced it down to be an individual setup. I added a M6 PTFE coupler (Thanks again @MisterD) directly to the mechanism and modified the path of the PTFE tube a bit to be more centered. Because there was now only one buffer I had to make an axle for just it. Through my various iterations I found the axle to be a weak point so I ultimately made it “snap” in place. This fixed width axle also helps hold the sides of the buffer where they should be and allowed me to modify the wheel to keep tolerances tighter as I found the filament would sometimes get bound between the idler wheel and the buffer cage. 

The buffer cage I also found to be a weak point as its published design. I ended up making a honeycomb design in the sides and that improved strength greatly. 

I spent a fair bit of time, numerous revisions, and had a couple of downright failures of ideas before I came up with this as my (current) solution. I hope someone else is able to make use of the design.

Printing:

I tried to optimize as much “laying down flat” printing as I could with the various components. You should not need any any supports for anything and draft settings have worked great for me. I've tried to include photos of the various components, how to assemble them, and have included one model which is a simplified view of how I printed them. If you find a better way to do something please feel free to share. 

As per a request I have added a Buffer Cage with a M10 PTFE fitting. I have yet to test this though so I have no idea if it works (I'd appreciate some feedback)

Material:

For the most of my development of this buffer/roller I used PETG, some of the components/variations are made from ASA, and one made entirely from recycled PLA. Honestly I haven't had any issues with any of the materials other than the ASA wanted to warp a bit when it printed. For strength you likely want to use PETG, but again, my PLA one seems to work just fine. 

Assembly:

The buffer wheel requires the sides of the buffer cage to be spread a bit to fit it in place. Since I moved to the honeycomb sides I've not had any issues with breaking. It's a bit of a squeeze but it will fit. Once in and assembled I suggest you “squeeze” the sides together and spin the wheel some. This helps smooth out the surfaces and after doing this I find the wheel will freely spin inside the cage (you may have to adjust where it sits on the axle). 

The axle only fits in the buffer cage one way, look closely and you should see the different diameter holes. The axle is a tight fit! Intentionally! Personally I use something hard to press down against my table-top to seat the small end then just use my thumb against the axle and fingers on the pulley to seat the large end. The cage will deflect JUST enough that the large end generally doesn't seat when the small end does. 

Assembly of the rest should be fairly straight-forward. Various parts “snap” together and rely on friction fit vs fasteners. In theory you could glue pieces together, or use screws, but I haven't needed to as of yet. While the fasteners that hold the main base to the arms could be tighter I think I'd rather knock it apart there than transfer too much to the Skadis pegboard that is now supporting a bunch of weight.  

I have added a “jig” for cutting the PTFE tube. You will have to insert two pieces, one that acts as a guide for the filament entering the buffer and another as the filament exits the buffer, before it gets to the PTFE tube fitting. The piece at the back of the buffer passes through the base mechanism and helps lock the components in place. The jig includes a couple of small holes at the opposite end so you can use an old piece of filament to remove the tube from the jig.

Also included is a set of “drop” arms for mounting the entire setup to the Skadis pegboard. They allow you to place one roll in the immediate next set of holes on the pegboard and allows you to easily get 5 spools all in a row at the same height.

Feeding the Filament in to the buffer:

You'll want to cut your filament in a chisel or tapered tip in order to get it to feed through the PTFE tubes. In early revisions of this I had issues with the filament binding in the PTFE fitting however, since I optimized a few things, I've not had any issues. If you do encounter binding at this fitting just put your PTFE tube in until it feeds through. 

I find it easier to feed the filament through in steps, once through the buffer, then in to the PTFE tube to the MMU, then once in the MMU I pull the slack back in to the buffer enough to tighten everything up. 
 

Multiple Mechanism Auto-Rewind Spool Holder 
Used the bearing caps and cap retainer concepts.

Improved MMU Buffer 
Sliced the model down to separate one buffer to be integrated in to the spool holder, ended up changing orientation of how filament enters and exits as well as modifying the pulley (thickness), axle and cage - but I'll give credit where credit is due. 

M6 PTFE Coupler Joiner
Used your design yet again ;) 

Bowden Coupler M10
Used half of your coupler to model the buffer cage with a M10 fitting as per a request.