October 30, 2025
Description
First you need to make sure the one way bearing is installed facing the right way. The one way bearing should only engage when the filament is being pushed back towards the spool when the MMU is extracting the filament from the PTFE.
With one hand lightly hold the large wheel. With the other, take a chunk of filament and feed it in like it was coming from a spool. Now pull the filament back out and compare the resistance you’re getting with the hand holding the wheel. When feeding in, there should be little (ideally no) resistance to the infeed and the bearing should engage in the opposite direction and attempt to move the wheel. If this is the case, you’re good, if not you’ll need to flip the direction of the one way bearing and reassemble the rewinder, but don’t do that until you read through the assembly step by step.
If the one way bearing is engaging and driving the wheel when feeding in from the spool side, you have way too much tension on the adjustment spring. Now on mine, I never used the stock tensioner arm as others had already found several issues with that, so I went directly with MiRe’s fix for that from the get go which I’d strongly advise you using. You can find that here:
https://www.printables.com/model/1343258-collection-of-mods-for-coreboxx-premium
Once that’s in place you shouldn’t have to apply much tension at all, in fact you should only need to turn the tension screw in maybe a turn to a turn & a half (just enough to engage the threads on the heat set screw on the tensioner mount so it doesn’t fall out.
Take a chunk of filament and push it in like it’s coming from the spool side, not the PC4-M6 side. Now repeat from the other direction. Are you getting a fairly loud squeaking metal on metal sound? (It’s normal for there to be a little noise from the reverse direction but from the inlet direction it should be silent or very nearly silent. If not? Read on, as you’ve likely found your culprit.
You may find as I get into the assembly portion that you may end up having to reprint the Tensioner mount and possibly the mid chassis and small idler spacers as well.
If you’ve stripped out any of the holes (particularly on the tensioner mount) with those wood screws that came with the kit, you’ll likely have to reprint. That’s not the right fastener for the job, honestly. I ended up just buying whole mess of countersunk M3’s and redid the tensioner mount and both spacers hole sizes to allow the M3’s to properly act as taps cutting their own threads (still not ideal as I’d far prefer to use Heat Set inserts, but that would have added a lot to the BOM, and provided they get assembled right and not overtorque to the point of damaging the plastic, they should be ok).
If you’re going to have to reprint the tensioner mount (and you’ve been having problems just getting the filament to thread through past the PC4-M6 fitting, you’ll may have a whole lot better luck with my fix for the tensioner mount, which is here:
https://www.printables.com/model/1451821-coreboxx-rewinder-tensioner-arm-mount-fix
I haven’t had the time to post the revised spacers (and a mess of other parts I’ve updated) yet, but I’ll eventually get there. Getting daily radiation treatments for stage 4 cancer while still trying to hold down a the day job to keep the insurance funded takes a big chunk outta my available free time, so please bear with, I’m moving as fast as I can.
The reason you’re likely having lack of torque issues is that that both the one way bearings (and the “main” bearings as well) need to be absolutely square to the axle. If they are skewed at all, it adds a lot of resistance to the entire assembly, and for these rewinders to work correctly they need to be as close to friction free as possible. As a result the alignment is critical both to the operation of the rewinder as a whole, and but also to the efficiency of the one way bearing transferring the linear movement of the filament to rotational torque on the large rim.
Further misalignment of the bearings causes the one way to not engage uniformly across its width, and that squeaking noise you heard? That’s the one way bearings engagement teeth wearing down. The louder that squeak is, the faster you’re acccelerating its failure. It’s sorta beyond the scope of this article to get into the way these things work but they‘re by definition of how they work a wear surface. Leveling out the contact area of engagement has a huge impact on the longevity. Properly installed they should last a good long time, but as more skew is introduced the MTBF decreases exponentially.
The way the original instructions were written, while easy to follow made it way too easy to get the axles out of alignment, which results in most of the issues folks have been seeing.
We‘re going to tackle that here.
So as a result, we’re going to dissemble and reassemble the whole thing. I initially wrote these instructions up for a post on Sanders’ Dischord from the point of view of assembling them for the first time to avoid this whole situation.
The link to that thread is here:
https://discord.com/channels/1331989858520006789/1339712125492527211/1425899946862968902
So what will appear below is written from that “initial assembly“ viewpoint but if you’ll be doing a reassemble the process is the same, albeit with ensuring you‘ve got plastic that hasn’t been torn up by those wood screws.
It’s a lot of steps some of which may be obvious, but I didn’t want to leave anything unstated for folks who haven’t done a lot of this kind of thing before, so bear with if it’s overly detailed.
It seems like a lot of steps (and it is, they‘re just ones I don’t even think about and just do, because this sort of thing is something I’ve done hundreds of times before and they’re just muscle memory best practices sort of approach when assembling stuff with any degree of mechanical complexity).
In practice, it only takes maybe a couple minutes longer per rewinder than the published instructions, and after you’ve done the first one the rest will be a cakewalk.
The core technique demonstrated here (approach assembling anything with a lot of screws going into plastic as though it’s a car tire you‘re mounting on the hub and gradually torque the fasteners down) is an excellent practice to the get in the habit of, and one I’ve used over and over, including both the Coreboxx Proper and Sander’s enXLosure, and most of the mechanisms I’ve designed.
What I ended up doing on assembly (and this isn’t really spelled out in the manual) is before you screw the L&R chassis together you want them lined up exactly as you’re going to be putting each axle between two bearings and if they aren’t square to each other the one way bearing is going to have issues.
1 - Put one of each idler on an axle and lock those down with the axle just flush with the outer holes.
2 - Feed both axles through the left side of the chassis
3 - Add the small idler spacer and the mid chassis spacer (part that clips onto the floor mounts), using the little nubs that stick out to align them to the chassis.
4 - Slide the tensioner mount with attached tensioner onto the large wheel axle. No need for the one way bearing piece yet.
5 - Line the nubs up to the tensioner mount ends up with the PC4-M6 passthrough facing towards the chassis mid spacer pointing downwards.
6 - Holding that mess with the spacers and tensioner facing up, slide the right side chassis over the axles and bring it flush with the spacers.
7 - Hold that mess together as a sandwich and grab the non wheeled end of the axles and spin them. If they’re lined up correctly and not skewed they will continue to rotate (freewheel) after you’ve let go of the axle.
8 - Very carefully add the screws to the chassis right side. Note this is backwards order shown in the manual, but there’s a good reason for it.
Those really should be countersunk M3’s with a smaller diameter holes to give them something to bite into - I’ve corrected this in my rework of the wheels you can find in the mods and remixes, and the Tensioner Mount. I haven’t got the mid and small Idler spaces posted yet.
The Blurolls supplied screws are wood screws, not machine bolts. As such they won’t have squat for holding power, and very easily will strip out the plastic when torquing them down.
Further they won’t survive but maybe a couple disassemblies at best. SO if you are going to use them treat it as a one shot deal, as you’ll need to reprint the spacers and tensioner mount if the plastic gives way from over torquing.
That being said, be very painstaking in your initial threading of the screws.
Make sure the tip is centered in the hold and be as close to perpendicular all the way around as you can, as if they go in cockeyed, the countersinks will force the spacers and tensioner mount off kilter with the chassis resulting in binding.
So DON’T Torque them all the way down yet (stop just before the countersink flange on the screw engages the chamfer on the chassis part.
Test the axle spins again. If they’re still freewheeling you’re good. If not find which of the screws is off center, and back that one out.
9 - Pull the wheels and axles out and reinsert them from them other side (where you have the partially seated screws. Check the axle spins and make sure they are still freewheeling.
10 - Begin to seat the corresponding screws on the left chassis side, using the same technique as above and once you get near the countersink flange hitting the chamfer, check the spin.
11 - do a second pass on all 6 screws, bringing them close but not fully tight (stop when you begin to feel resistace.
Check the spins again.
12 - pull the axles and wheels out and insert on the right side, and do another pass on the screws, again stopping when you start to hit resistance.
13- Swap sides again, this time fully seating the left side screws (once you hit resistance maybe go 1/2 -3/4 of a turn, if you go too much further you’ll strip them out.
14 - swap sides again, and fully seat the right side.
15 - If you had to back any of the screws out due to going in cockeyed, you can now put those in as the assembly will be well enough held together to not get skewed by one screw, and it should pull into perpendicular when you put it in.
Now insert the one way bearing piece. Screw the tensioner adjust screw in with maybe a couple turns just so the threads mate. Stuff filament through the “slot” until it protrudes out the PC4-M6.
Hold the wheel loosely and continue to push filament in.
The wheel should not move (might feel a little resistance) and you shouldn’t hear any squeaking out of the one way bearing.
Now push the filament from the PC4-M6 side and the one way bearing should engage and the wheel should turn.
Hold it loosely and make sure it turns. If the one way bearing part is oriented correctly you should have nearly no resistance to feeding filament toward the PC4-M6 and the wheel if held should not turn as filament goes in, and in the opposite direction the wheel SHOULD turn.
If this is not the case, take out the adjuster screw, open it up and flip the oneway bearing piece over and try again.
If you’re using MiRe’s fix for the Tensioner (which I highly recommend), you should barely need any turns at all to make the one way bearing do its’ thing.
I can’t speak to the “stock” version as I skipped that right out of the gate knowing it was probematic.
Once you’ve got the direction of the one way verified, add the wheels to the other side and proceed on.
Note on the ”set screws” for the wheels -
Don’t overtorque these as you’ll mar the surface of the axles (they’re just extremely lightweight aluminum and deform easily). You want them just firm enough to lock the wheel rotationally to the axle so the rewinder can impart its motion to the spool. That’s all thats needed. More than that will tear up the axles making subsequent disassembly difficult. I’ve you’ve already chewed them up, hitting the axle with some 80-100 grit emory cloth will smooth out the deformation.
Grab ahold of both large wheels, hold the whole thing far enough away from your chest that when it rotates around the axles it’s not going to clobber you, and move both hands in a circular motion to get the entire assembly in the middle to spin around the wheels.
Speed that up as much as possible and feel for any vibrations or loud squeaks out of the one way bearing (there should be none as it should NOT be engages with no filament inserted).
Then reverse directions and repeat.
If it’s squeaking a lot the one way bearing is not square to the mating surface on the axle which is exactly what you’re trying to prevent with all the above.
Repeat with the small idlers (just to make sure they’re offering as little resistance as possible thus ensuring those bearings are properly aligned to their axle.
The reason for all this is you’re asking the MMU to not just pull the filament out of the PTFE’s during changes (and its’ expecting next to zero resistance for that) but you’re now using the MMU stepper as a linear motor via the filament itself, and you want it to take as little of a force to do so as you can, as the MMU’s firmware for the stepper is written expecting nearly zero resistance current wise.
Further, I think you’ll find the one way bearings will last a LOT longer than if you just slammed the stuff together off center.
Also, using my wheel remix with actual M3 countersunk screws makes the above a LOT easier with a lot less margin for error.
Hopefully that’ll help folks that have had reliability issues with the rewinders. I’ve printed a ton of prints on them with literally thousands of tool changes and have had ZERO rewinder related issues.
License:
Creative Commons — Attribution — Noncommercial
7
LewnWorx
