Here’s a method I’ve developed to splice filaments almost to perfection without any tools. It’s basically the Teflon tube method for cheapskates who don’t want to buy Teflon tubing 🙂

First of all, prepare a 2" x 2" (50mm x 50mm) -ish piece of white paper and a straighten a piece of filament that will serve as a mandrel:

Roll up the paper into a tube around the piece of straightened filament as tight as you can. The hard bit is to start rolling: the paper needs to be really snug against the filament to start with.

Once it’s started right, it’s easy. Roll it up all the way nice and flat. The bit of filament inside should fit inside the tube with quite a lot of friction if you did it right:

You can also wet it a few times with your tongue and it will stay in one piece without holding it. Ex-smoker’s habits die hard 🙂

Cut the ends of the filaments to splice together with a sharp bevel:

Carefully thread the ends into the paper tube so they meet halfway:

They should go in with some force but they should slide smoothly. If you feel any roughness, you’ve snagged the paper inside and it won’t work, so you should start over.

Heat up the center of the tube at 250C to 260C while ramming the filaments into each other firmly, but not so firmly as to collapse the paper tube, until you feel them “go” and melt into one another:

I use a SMD rework station because you can apply heat as much as you want and the paper only browns a bit, even if you overdo it grossly. It takes about 30 seconds for the heat to diffuse through the paper and for the filaments to melt fully. It’s doable with a lighter too while pushing the filaments together with one hand, but it’s less convenient of course.

Then unwrap the paper: some paper should stay stuck to the splice:

Clean up the splice by running a sharp knife along the splice all around. It’s pretty quick, the paper isn’t terminally fused to the filament:

Voila: perfect splice!

And here, seen under a microscope:

  • dual_sport_dork 🐧🗡️English
    10•17 hours ago

    I’m wondering if there’s some obvious reason I’m missing why you couldn’t make a filament splicer by just taking a slug of aluminum billet or copper, drilling a 1.75mm hole through it, and sawing it in half bisecting the hole with, say, a jeweler’s saw so you have a minimal kerf.

    Assemble the halves, stick your two pieces of filament in, cook it, let it cool, split the halves, nirvana achieved. No waste, no consumables.

    Like… Not a single commercial splicer gizmo works that way. Why?

    • @[email protected]OPEnglish
      20•17 hours ago

      I’m wondering if there’s some obvious reason I’m missing why you couldn’t make a filament splicer by just taking a slug of aluminum billet or copper, drilling a 1.75mm hole through it, and sawing it in half bisecting the hole with, say, a jeweler’s saw so you have a minimal kerf.

      I’ll tell you because I tried making one:

      • You need the hole to be at least 30 mm or 40 mm long so that the two filaments are held in place where they’re not melted. Good luck drilling a 1.8 mm hole that deep! Not impossible, but not easy, as any machinist can tell you.

      • You can’t split the block in two through the hole because you always lose material where you saw off. Even EDM machining will lose you 0.2 mm, so your reassembled block would have an oval hole that’s too small. To create two halves that reassemble and create a correct hole, you need to make two identical blocks and mill them to exactly the center plane, which considering the diameter and tolerances of the hole, really isn’t easy to position right. Again, doable but not easy.

      • The block will act as a massive heat-sink. Even if you leave the center part thin so the heat has the best chance to reach the splice, it will still mostly go into heating the masses of metal around it.

      TL;DR: it’s not a trivial bit of machining. And remember: my goal here was to make this ultra-cheap and lazy 🙂

      • FuglyDuckEnglish
        6•16 hours ago

        So, hypothetically, could you use a ball end mill with a 1.75mm diameter? Then mill the channel across two halves?

        Basically, something like a halved heater block. Put a cartridge heater on one side, thermistor on the other, take it up to a melt temp and shut off. (Maybe add some cooling fins beyond the intended melt zone?)

        Also, have regrets because this is more expensive than the boden tube method.

        • @[email protected]English
          7•12 hours ago

          You guys over complicated it. Just cut the block in half first. Then clamp the halves. Drill press down the center.

          No oval. No missing cutouts.

        • @[email protected]OPEnglish
          7•16 hours ago

          So, hypothetically, could you use a ball end mill with a 1.75mm diameter? Then mill the channel across two halves?

          Yeah that would work. Although you’d have to mill at an angle to get a good surface finish, particularly at the bottom of the channel. So ya know, loosen the mill’s head, angle the head, do the thing, then put it back together and realign it nice and parallel to the vise… I.e. a PITA.

          Basically, something like a halved heater block. Put a cartridge heater on one side, thermistor on the other, take it up to a melt temp and shut off. (Maybe add some cooling fins beyond the intended melt zone?)

          Yeah I thought about doing all this too. But for a simple task like that - especially something that you don’t do 10 times a day every day, the over-engineering threshhold is very quickly crossed.

          But you capture the essence of the problem perfectly 🙂

          • FuglyDuckEnglish
            2•12 hours ago

            The only reason practical engineering is more fun than over engineering is that things actually work.

            But, eh, over engineering things is fun too- at least as a thought problem. If OP thinks some leftover Bowden tubing is too expensive, though… it’s only a thought problem…

            If you could get economies of scale working, you might be able to hawk it on Amazon for 20 bucks, though. (Not that I’d ever… that sounds like a desk fire waiting to happen…)

        • @[email protected]English
          1•12 hours ago

          I think the issue still becomes the fact that you hear the metal through it’s entire mass and not focused on the ends, so you might end up with problems with the filament where they go into the contraption

          • FuglyDuckEnglish
            2•12 hours ago

            The way around that is milling in fins on either end, this would basically act like a heat break between the hot and cold ends of a printer.

            Or just load the thing and then turn it on. It heats up, melts the plastic and then immediately shuts off. It would take 30 seconds, maybe.