Printable Filament Spool Hub Skips The Bearings

When you really start fine-tuning your 3D printer, you might start to notice that even the smallest things can have a noticeable impact on your prints. An open window can cause enough of a draft to make your print peel up from the bed, and the slightly askew diameter of that bargain basement filament can mess up your extrusion rate. It can be a deep rabbit hole to fall down if you’re not careful.

One element that’s often overlooked is the filament spool; if it’s not rotating smoothly, the drag it puts on both the extruder and movement of the print head can cause difficult to diagnose issues. For his custom built printer, [Marius Taciuc] developed a very clever printable gadget that helps the filament roll spin using nothing but the properties of the PLA itself. While the design might need a bit of tweaking to work on your own printer, the files he’s shared should get you most of the way there.

All you need to do is print out the hubs which fit your particular filament spools (naturally, they aren’t all a standard size), and snap them on. The four “claws” of the hub lightly contact a piece of 8 mm rod enough to support the spool while limiting the surface area as much as possible. The natural elasticity of PLA helps dampen the moment that would result if you just hung the hub-less spool on the rod.

The STL files [Marius] has provided for his low-friction hubs should work fine for anyone who’s interested in trying out his design, but you’ll need to come up with your own method of mounting the 8 mm rod in a convenient place. The arms he’s included are specifically designed for his customized Prusa Mendel, which is pretty far removed from contemporary desktop 3D printer design. Something to consider might be a piece of 8 mm rod suspended over the printer, with enough space that you could put a couple spools on for quick access to different colors or materials.

Hackers have been trying to solve the spool friction issue for years, and as you might expect we’ve seen some very clever designs in the past. But we especially like how simple [Marius] has made this design, and the fact that you don’t need to source bearings to build it. If you’re thinking of giving this new design a shot, be sure to leave a comment so we know how it worked out for you.

32 thoughts on “Printable Filament Spool Hub Skips The Bearings

  1. I highly recommend against putting bearings in your filament spool holder.
    If it runs too smooth, it can “run ahead” if the printer feeder, due to the tugs of the print head, and then get more and more unspooled. Causing all kinds of issues, which result in print failures.
    (One of the many things we noticed at Ultimaker)

        1. I meant the kind of non-locking “rachet” mechanism, don’t know what to call it, like you had a triangle with a spring against gear teeth.

          But anyway what Jason said about fishing reel, i don’t know how they work, but i would think not exactly like what i mean, so that’s another option.

    1. A reverse bowden tube from a fixed point on the frame running up to the entrance of the print head (left unconnected to allow for retractions) gives enough spooled out filament to allow for free roll without over rolling.

    2. I absolutely agree. I found a source for bearings (dollar store fidget spinners) but then there was too LITTLE pressure, and the spool tended to unwind itself and then it got wound up around the axle

    3. I heavily disagree, bearings are absolutely useful however are only half of the solution. Running a feed tube to decouple filament length from head position is the other half, providing a constant fixed distance from head to filament spool means the only thing pulling on the filament is the extruder stepper, preventing overruns while still allowing for smooth feed and retractions.

  2. Just as an “argument from ignorance” ( i dont have a 3d printer yet) has anyone tried adapting the bearings from an old HDD? And it seems to me that overrun could be conquered with some kind of drag adjustment like on a fishing reel. (There you go… How about using an old Zebco bait cast reel as a hub? They spin smooth and have a decent drag adjustment)

    1. I can’t find any only-bearings in old HDDs, only a combined “hub” of motor, bearings and platter holder/table/shaft.

      But that might be even better if you can use the motor as an electric (auto-)brake…

    2. I had some filiment pulling problems on my ender 3 and I put a hard drive hub motor on it. The fit worked out and i simply pushed the tube that they include on the spindle. Has held up well for 200+ hours of printing

    3. Ok, one last comment, all kidding aside. My mother is a seamstress. She has a Bernina industrial sewing machine. The motor runs at full speed all the time and the speed is controlled, via the footpedal, by an electronic clutch. To my understanding the footpedal engages an electromagnet that controls the slip and therefore the speed and torque of the work. Would it be possible to use such a clutch but have it controlled with feedback from a couple sensors. Maybe one to detect wobble and the other tension or pressure. Again im not familiar with the exact requirements but in my own defense im taking some time to learn about 3d printing before i buy one. I have enough impulsive purchases under my belt…… The mechanism itself seems easy enough to understand but im considering it to be like baking bread, easy to do but hard to master. Unfortunately, there doesnt seem to be a way to eat your mistakes yet…

    1. I never even did that, just shove the rod through with no hubs to center it. It just drags along with no problems I think this may be one of those fixing an issue that doesn’t exist.

      1. Dragging filament can definitely have an impact on print quality. If you don’t have the issue, you have nothing to fix, but that doesn’t mean the issue doesn’t exist elsewhere for others.

  3. if your getting artifacts from pulling filament off the spool its not the spool that the problem.The lack of a guide tube to transfer the pulling forces away from the head and onto the frame that is the issue. The pulling force can jostle the print head without this. Setup correctly and with a good extruder one can pull from a spool sitting on the floor or as often in my case out of a ziplock bag for moisture control without any issue.

    1. Kinda depends how you look at it. The whole point of 3D printing is putting the right amount of material at the right spot.
      And the right amount of material is where it gets difficult. As material comming out of the nozzle depends on the pressure in the nozzle and the diameter of the nozzle hole. The pressure in the nozzle depends on how much force you put on it with the feeder, and the amount of force the feeder gives depends on the grip, which depends on a lot of factors. Including the amount of force required to pull the material from the spool, so you would want to keep that force constant for better flow control.
      Now, this is all in the range where you usually cannot really see it visually anymore, and thus you need a machine with a lot more sensors to actually measure these effects. Or, if you can see this effect, you have simpler fixes to do to get good results, like a stronger feeder.

        1. Small reminder, I’ve been in this field for 7 years now. And one thing I’ve learned the last few years, that everything I know is wrong.

          So, on this topic, let’s get our termonology aligned. We call “slip” when the requested material does not match up with the actual material. And we call “grind” when the material flow stop to a near 0%. While the bondtech feeders are great, they experience slip, just like any feeder. Compared to moving without a hotend attached, this is generally in the 20% range. If your back pressure is constant and the force required to get the material from the spool is constant then you can calibrate this away (which is what everyone does)

          Why does this all matter if you cannot see it? Well, there are multiple aspects of a 3D print. Depending on the usecase. Aesthetics is one of them, strength, dimensional accuracy, reproducability are others. And these all can suffer without visual impact.

          (FYI, I work at Ultimaker, might know me for starting Cura)

      1. Or in other words, its more art than science. Im going to put my vote on the aluminum bar, in a good setup with just a drop of vaseline for the mech. and a cold beer to sit back and watch..
        I like the way you said that, it reminds me of something out of “The Art of Electronics”.

  4. Isn’t it easier to source 608s than it is to source rods and cut them though? And then figuring out how to mount all this business?

    I’m a big fan of the TUSH. There’s nothing to mount, spool changes are super quick and friction is right where it needs to be. Low enough not to cause extrusion issues and high enough that tugs won’t cause the whole spool to unroll.

    1. The Tush was the first spool holder I tried, and I never had to try another one. In my opinion, it is perfect. 608 bearings are super cheap and you can use them for many other things. I ordered a pack of 50 and have only 20 left…It’s worth it having these in stock.

  5. All that plastic and threaded rod, so many wires… That is quite a build, compared to my $200 Ender 3, not quite 3 weeks old. Not sure you could tell if it’s the filament tugging, or something needs snugging. I’m just getting started, only two parts printed to be proud of, but learning, and getting better. I went straight to ABS, and it’s been fun figuring out all the problems and settings. Anyway, seems to me that what the spool is doing, is pretty well dampened, by the time the filament gets to the hotend. The extruder holds the filament pretty steady, and the resistance of the feed tube should smooth out the rest. Guess sometimes simpler is better… Besides, if it was a common issue, wouldn’t have been addressed by now? Maybe these fancy machines run faster, I’m finding that a slower print speed is improving my prints, seem stronger less brittle, but still not as expected. I bought cheap ABS, since I figured on many failures, but kind of suspecting it might be old stock, or something wrong. Next spool, plan on finding out who makes it first, and getting it direct, rather than save a couple of bucks, someone who might have dumpster dived to stock his online store.

    1. On my Ender 2 it was not the pressure or other effects of the filament, but how the spool would shake the entire machine when unspooling. I printed a thicker and longer pipe adapter, now it unspools smoothly. No bearings, just a better match for the spool.

  6. The proper way to eliminate variations in filament tension leading into the extruder is to employ an auxiliary mechanism that pulls filament from the spool independently. This would eliminate tension variations due to spool friction, inertial, and filament cross wrapping differences. Such a mechanism should be able to detect filament unspooling issues (ex. tangles) that would normally cause a print to fail by “pausing” the print.

  7. Brilliant idea. Bearings are too smooth and the filament unrolls itself by spring force alone. Wish i had thought of this.

    I wonder how long till the slight friction wears ruts down and it springs out ? Maybe rev two will resolve this potential .

    Nice work on this.

  8. Personally, I’ve *never* had a problem with ‘run ahead’. I just printed a TUSH set that clipped to the frame of my printer, then designed a foam-filled filter/guide that also clips there. Done. The bearings make sure that there can’t be too much friction, and the foam rubbing on the filament makes sure that there can’t be too little. (I’m pretty sure that I could stick a 10kg roll on and it would act just the same, though the frame wouldn’t enjoy it.)

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.