BoxTurtle filament 'Buffer Ram Sensor' with D2HW switches

Introduction

This is a filament buffer that is suitable to be used in conjunction with a BoxTurtle MMU and the AFC Klipper addon, specifically as described in the 'Buffer Ram Sensor' section. It features 2x D2HW-C201H switches and 2x MR63ZZ bearings, the same ones used as filament sensors in the BoxTurtle extruders. The buffer has a short travel, about 20mm, but this has so far proven sufficient in my tests.

Bill of materials

• 6x M3x4x5 heat inserts ('voron style')

• 4x M3x15 SHCS bolts

• 2x M3x8 SHCS bolts

• 1x 0.6x6x30mm spring (OD 6mm, uncompressed length 30mm)

• 2x ECAS04 collet

• 2x D2HW-C201H microswitches

• 2x MR63ZZ bearings

Printing

The three parts can be printed in the original orientation, and should not require supports. Use the following settings:

• 0.2mm layerheight

• >=3 walls

• >=3 top and bottom layers

• >=15% infill, type doesn't matter much

Filament type shouldn't matter much, I've got two in use where the top and bottom part are in PETG and the carriage is in ASA, but since it is not going to get hot and doesn't have to handle much stress at all, even PLA should be OK (though not tested!).

Assembly

After printing, make sure the carriage travels freely between the top and bottom shells by putting it into it's track. It should go in with a little 'click' as you press it into the bottom. Put the top on (without screwing it into place), squeeze the two together in your hand and shake it. You should hear the carriage slide around and bump against the end stops very easily. If that is the case, you can proceed, otherwise tweak your print settings so the sizing is correct and try again.

If you passed the 'free sliding' test above, pull the carriage out of the bottom (will take a bit of effort but shouldn't be too hard). Insert the ECAS04 sockets into the bottom shell, and the slide. See the pictures for where they need to go.

Now put the heat inserts into their places. Four go at the corners of the bottom shell (see photos), 2 are under the switches, where in the photo the M3 bolts are screwed into. These are a bit tight against the slots the switches are in, but shouldn't be too hard to get them in. Make sure the top is clean and flush with their respective surfaces, or you may not be able to close the shell properly.

Solder wires to the switches and put connectors on as you need for your particular setup. Of special note here is that the switches are in the opposite order from the BoxTurtle TurtleNeck! The switch on the right (on the photos) on the BoxTurtle/MMU side is the one that signals expansion. The switch on the left, closest to the toolhead, signals compression! Insert the switches into their slots, gently screw in the M3x8 bolts. Then put the MR63ZZ bearings into their slots.

Make sure the channel for the spring in the carriage is turned away from the switches (see photos for orientation), drop the spring into its channel. Then make sure the carriage is moved all the way to the right. Place the top shell onto the bottom. If you do this with a bit of a left to right motion, it should catch the spring, and put a very slight bit of tension on it, as intended. Finally put the four M3x15 screws in their spots and screw the two halves together.

Finally, hook up the buffer. The external ECAS04 should point at the BoxTurtle/MMU (also marked with BT on the top shell). Simply cut PTFE tube to length, and slide it into place. Then cut the PTFE tube that will connect the buffer to the toolhead to length, and insert it through the hole on the other side. It may be a little fiddly, but it should go into the ECAS04 on the carriage, inside the buffer. Press until it is seated properly. Tug gently on the PTFE tube to check that it's seated correctly, you should feel the spring compressing.

And.. voila!

Notes

The buffer will automatically, and by design, revert to a compressed state if no filament is in, or if there's exactly enough filament in the path. This is necessary for the 'buffer ram sensor' functionality. This means it is not suitable for buffers that need to be somewhere in between compressed and expanded in those cases.

The buffer is not suitable for very weak/flexible filaments. They'll double over in the space between the entry point and the carriage, as the filament tries to expand and push against the carriage. TPU will definitely not work, but I also had a failure with some very soft matte PETG at some point.