A compact and reliable filament rewinder based on a linear counterweight mechanism that can be produced in a single 11 hour print.
Provides a complete filament management solution eliminating the need for the space wasting and difficult to load filament buffer.
Offers many PTFE attachment options supporting a wide range of spool, MMU, and printer configurations.
This rewinder supports filament spools with the following dimensions:
- diameter up to 235mm (up to 300mm with this larger stand)
- width up to 90mm
- hub diameter between 50mm and 75mm
The stand included in this design supports three different rewind mechanisms:
My hope is that one stand which supports three rewind mechanisms will encourage more people to print and test a rewind solution as part of their MMU deployment.
The only non-printed parts required are:
- three 608 bearings
- a counterweight (a 0.5" or 12mm bolt and nuts are commonly used)
The linear rewind mechanism now has two additional stand options:
- There is a drybox for moisture sensitive filaments.
- User gcode22 released a very nice wall mounted stand.
Print rewinder using Prusa i3Mk3 gcode:
If you are interested in a filament rewind solution, you most likely own a Prusa i3 with MMU.
In that case, the easiest solution is to print the gcode hosted at the Prusa Community site. All required components, including the stand, fit on the bed at once and produce a rewinder in a single 11 hour print.
Print rewinder with your own gcode:
The .stl files in this thing will work well for most people, especially if you are using PrusaSlicer.
However, significant variance can exist across 608 bearings, slicers, printers, and personal taste for press fit tightness.
Alternative Fit Parts has additional versions of the parts which are affected by the factors listed above. It also contains two printable tools to help you determine which parts are best for you and instruction on their use.
The parts should be oriented on the bed as follows:
- The cap should be oriented with the gear up.
- The bearing pocket on the shaft should be up.
- The wider end of the nut should be down.
- The small gear has a flange on one size which should be facing upwards.
The proper orientation of all other parts (stand, bearing inserts, track, etc.) should be obvious.
You will need to print two instances of the bearing insert.
Use the stand with Vincent’s Universal Auto-Rewind Spool Holder hubs
The instructions can be found in Universal Auto-Rewind hubs.
Use the stand with Martin’s arm
The instructions can be found in Rewind Arm Mechanism.
Additional PTFE mounting options
PTFE Mounting Accessories contains four optional PTFE mounting arms. They are unthreaded but allow for a PC4-M10 fitting to easily cut its own threads. The options are:
- 45° down filament exit placed 50mm in front of the stand
- 90° down exit or extended horizontal exit 44mm in front of the stand
- high with a horizontal filament exit
- high with a 45° upwards filament exit.
The community has graciously generated additional PTFE mounting options:
- 90° down / extended 80mm in front, unthreaded, PC4-M10
- 90° down / extended 80mm in front, threaded, PC4-M6
- 90° down / extended 80mm in front, threaded, PC4-M10
- 90° down / extended 56mm in front, threaded, PC4-M6
- 45° down, 50mm in front, threaded, PC4-M6
Every option listed above works with all three rewind mechanisms.
Other notable community enhancements
Some really creative community improvements have been posted, including:
- The mass cube eliminates the need for nuts and bolts and allows virtually anything to be used as a counterweight.
- Large diameter, coarse thread cap, shaft, and nut replace the M30x3.5 originals and allow for easier loading and unloading of spools.
- Numerical labels track which of the MMU's filament locations the rewinder is feeding.
- This complete drybox solution will keep humidity for affecting your prints.
- The spring track allows for additional rewind simplifying the loading and unloading of filament at the beginning and end of a print.
Insert the track into the stand's channel and remove it 15-20 times to deburr and smooth the parts. If desired, very lightly sand the back and sides of the track.
Press a 608 bearing onto the shaft of the large gear with shaft. Ensure the gear’s flange is flush against the side of the bearing.
Press the small gear onto the shaft with the flange towards the bearing. Be sure to align the small anti-rotation triangles. The gear will go on easily until the last 2-3mm at which time the resistance will increase. Continue pressing the small gear onto the shaft until its flange is against the side of the bearing. The end of the shaft should be approximately flush with the end of the small gear when correctly assembled.
It might be necessary to lightly sand the stand's bearing pocket, especially at the top and bottom. The extrusions in these areas may imperfect which will make the gear bearing difficult to insert.
Install the gear/bearing assembly into the stand from the inside, small gear first. Ensure the bearing is seated firmly against the wall on the outside of the stand. The picture below shows a full inserted gear assembly. Note that the large gear does not rub on the stand and that the small gear is approximately centered in the opening of the track's channel.
Insert a 608 bearing into the gear side of the cap and another into the pocket on the end of the shaft.
Screw the cap onto the end of the shaft without the bearing. The cap is not removed from the shaft in use, so try to get this reasonably tight without breaking the printed threads. You should now have a bearing on both ends of the assembly.
Press a bearing insert into the cap’s bearing.
Repeat with the second insert and the bearing on the shaft.
Install your counterweight (likely nuts and bolts) into the mounting hole at the end of the track.
The track's hole accepts a 0.5" bolt. I am using a standard 1/2-13 x 1" bolt (one inch long) and two 1/2-13 nuts. The total weight of the bolt and both nuts is 2.5 ounces. I've also used a 1.25" inch bolt and three nuts weighing to a total of 3.3 ounces which produces a more rapid rewind. If you need more than 3.5 ounces, you likely have some source of friction which you will need to investigate and repair.
If metric hardware is more readily available, use a 12mm bolt and nut(s) weighing between 70 and 90 grams total. If you need more than 100 grams of weight, you probably have a source of friction.
Nuts and bolts are convenient as they are heavy and easy to attach but anything which achieves the target weight will work. The mass cube from behn66 makes it easy to use arbitrary objects as a counterweight if nut and bolts are not available.
I've also successfully used rubber bands and elastic cord to store rewind energy. One end of the band/cord should be attached to the hole in the track and the other to the stand.
Install the track in the stand from the top with the teeth facing the small gear.
The teeth on the track should be roughly aligned with those on the small gear. If they are not, inspect the large gear, bearing, and small gear for proper assembly and fit within the stand.
Screw a pass-through style festo fitting into the hole on the bottom rail of the stand (or optional arm if used). These holes are not threaded but are sized to allow the fitting to cut its own threads with little resistance.
Each side of the stand is labeled with the type(s) of optional PTFE fitting arm that can be used on that side.
If desired, a filament number label can be installed to identify which filament location the spool is feeding.
Install a filament spool on the shaft. The cap includes arrows showing the direction filament must feed, so please follow them when deciding which side of the spool to insert the shaft into.
Install and tighten the nut to hold the spool firmly against the cap.
Place the spool onto the stand so that the cap’s gears mesh with the large gear previously installed into the stand.
History and Inspiration
This started with the parametric auto-rewind spool holder from Vincent Groenhuis. I printed a set of these and found they worked very well. Unfortunately, my overall experience with the initial MMU2 release was poor. I decided to shelve it until improvements were made and bought a Palette 2 Pro to work with instead.
My MMU2S/MK3S upgrade has arrived and I am now having a =MUCH= better experience. I shifted back to the MMU and decided to evaluate and upgrade my entire setup as part of the process.
I saw that Vincent had released a new "Universal" design, so I printed one and found success with it as well. The use of a large diameter threaded shaft and nut versus multiple inserts was a great innovation. The new design is also much easier to print, assemble, and adjust. He somehow even managed to maintain compatibility with the large ecosystem of 3rd party stand designs which spawned to support his earlier version. Unfortunately, I still had one concern with its operation.
In any rewind solution, the filament spool must be able to slip during continuous long extrusions of a single filament. Vincent's earlier version allows for this through the fingers on the hub. His newer version does so with an adjustable clutch mechanism. Both work well but have one unpredictable side effect. In my experience, the size of each slip is non-deterministic. The slip is usually very small. But, on rare occasions, the spool will slip significantly, releasing most of the spring's tension. When this occurs, it usually prevents full rewind during a filament ejection.
While playing with Vincent's "Universal" design, I found Martin Hoffmann's implementation. His use of a counterweight rather than a printed component made his design easy to print and assemble. He also had a creative solution for slipping the spool during long extrusions. The arm Martin designed makes a "cogging" noise when it reaches its travel limit and slips. But it retains full rewind energy in all cases. I printed a unit and it worked perfectly.
My two concerns with the design were size and rewind turns. Martin’s rewinder is necessarily large to give the counterweight enough moment arm to function reliably. For me, it also only rewinds the spool approximately 1.5 revolutions. This becomes marginal near the end of a spool as the circumference is smaller near the hub.
My design attempts to take the best elements from both Vincent and Martin, allow for an additional half turn of rewind, and minimize overall size. I also want to support a wide array of spool sizes and PTFE tube mounting options.
- Modified the bearing insert to include a flange. This is intended to keep the stand sides from spreading outwards over time.
- Updated to accept the flanged bearing insert.
- Added PrusaSlic3r specific .stl files for the gears.
- Clarified counterweight information.
- Posted an optional arm to allow filament to pull from the top of the spool (on the opposite side) and go into a downward facing festo fitting.
- Added marking to the stand and cap showing the required filament direction.
- Added lettering to stand showing which optional inserts can be used in which locations.
- Posted revised gears and inserts that are a compromise between what is optimal for Simplify3D and for PrusaSlicer.
- Made changes to the cap to eliminate stay filaments at the top of the threads created by unsupported extrusions.
- Threaded shaft shortened slightly in order to maintain the correct overall length when assembled with the modified cap.
- Added a wall to the stand for the gear bearing to seat against.
- Adjusted the teeth width, flange width, and flange diameter of both the small and large gears. These changes should eliminate most problems with assembling the gears and installing them in the stand.
- Modified the final tooth on the track to reduce the "cogging" noise when a slip occurs.
- Tightened up the bearing pockets on the cap and shaft.
- Changed optional arm mounting locators to triangles from semi-circles to eliminate some bridging perimeters..
- Modified the stand around the channel gear opening to eliminate an area that printed well on 0.2mm, but was often poor quality at 0.3mm.
- Extended the track to have a section without teeth at the bottom. This part remains in the channel below the gear even when the last tooth is slipping on the gear. This aids alignment and will prevent the track from becoming too skewed in the channel.
- Optimized the nut and cap to consume less space, but also support a larger range of hub diameters.
- Changed the profile of the track to have less contact on the sides and back.
- Tightened up the channel to further eliminate skew.
- Version 1.0 released with the following new features:
- Support for tighter or looser fits of all bearings and shafts
- Enablement of both the universal spring rewind hubs from Vincent Groenhuis and Martin Hoffman's rewind arm design.
- v1.1 released
- Revised gears on cap, both gears, and track. Slight modification of stand to create optimal meshing. All revised parts will work in the old stand and provide much smoother operation.
- Significant smoothing of cap and nut to eliminate sharp edges.
- Shaft beveled so threads are easier to start.
- Moved the lettering to the face of the stand's bottom rail as "detect thin walls" was causing a problem at the top of the channel.
- Fixed an error where the fitting opening in the stand was partially blocked.
- Updated STEP file to v1.2 to include the stand revisions.
- Added teeth to the track which now provides slightly more than two turns of rewind. This track is backwards compatible with all previous versions of the stand and gears.
- Chamfered large and small gears for easier assembly. There is no need to update existing rewinders with these parts.
- Posted v1.2.1 STEP file including the new gears and track.
- Shot myself in the foot again with the Fusion 360 timeline. Failed to roll it to the end before exporting a stl and released flawed gears.
- Corrected gears now posed as v1.2.1
- STEP archive updated to 1.2.2
- Made another correction to the gears.
- Place a slight chamfer on the nut gear's flange.
- optimized the stand to improve print time and quality
- added a slot for gbzh01's filament location numbers
- added holes to the bottom of the stand allowing it to be mounted to any horizontal surface
- updated many (but not all) of the pictures
- revised the instructions for clarity
- added a link to the new stand and track supporting 300mm spools