Finally, A Machine That Makes Cheap 3D Printer Filament.

If there’s one problem with the RepRap, it’s the cost of filament. Sure, there’s also the computationally difficult problem of slicing 3D models, but a 5 to 10 times markup on turning plastic pellets into filament is the biggest problem. It’s even a bigger problem than the problems of compatibility and interchangeable parts that comes with everyone forking a ‘standard’ printer design dozens of times. The cost of filament, though, is the biggest problem, right up there with RepRap developers focusing nearly entirely on different printer designs instead of the software, firmware, and electronics that are also vitally important to the RepRap project.

Nearly a year ago, we caught wind of a competition to create a home-based filament manufacturing station that takes cheap plastic pellets available for about $5/kg and turns them in to 3D printer filament that usually sells for $50/kg. A winner for this competion has finally been announced. The winner, [Hugh Lyman] just won $40,000 for his home filament creation station, the Lyman Filament Extruder

The goal of the Desktop Factory Competition was to create a machine that produces filament suitable for 3D printers with a total build cost of under $250 USD. [Lyman] met the goal by using a few motors, 3D printed parts, a PID controller, and off the shelf auger drill bit (that’s the actual model and supplier he used, by the way) that is able to reliably churn out plastic filament.

If you want to build your own Lyman Extruder, all the plans are up on Thingiverse, but LulzBot, the awesome people who gave us a 3D printer, hope to sell a pre-assembled version of this extruder sometime in the future, hopefully with a chain guard around that sprocket.

104 thoughts on “Finally, A Machine That Makes Cheap 3D Printer Filament.

        1. RepRap inflates prices I feel. however now that the community came up with a way to make your own filament, I can see it now, companies will loose business with this, and they will mass buy all pellets up and hoard them for themselves and the General Public will now have to pay a higher supply and demand price. It sucks, and I see it coming.

      1. Hmm. I can’t say I’ve seen that one, but a quick google search didn’t leave me feeling enlightened.
        Was this a clever film reference (link?) or in fact just the usual lack of proofreading we experience when we subject ourselves to this site?
        I, for one, feel sorry for any grade school student who had to read that, reference or otherwise.

    1. I cringed reading that first paragraph too. Something is a problem. But an even bigger problem is the same problem I just mentioned, which compares to another problem. Biggest of all the problems is the first problem I already mentioned is a problem, kind of like a different problem that is similar but not as important as the biggest problem, which is again, the first problem.

      1. I want to defend the flow, as it is rather good, but I am working off a headache due to a 600 level class textbook doing the same thing EVERY DAMN PARAGRAPH. It is one thing to shrug off a bit of inconsistency in a blog, another for a masters level book written by a PhD.

        I guess proof once again that being a public writer on the internet is one of the hardest writing jobs to gain praise for. However, it is not nearly has hard as being a writer whose work is freely available for perusal on a network of networks with a broad base of peers.

        1. I have a telecommunications text that uses “arbitrary” in about every description……………………………………… Ugh! Hmm, my keyboard got stuck on the dots

    2. And this is why everyone should stay away from linguistics majors. It just turns into this big, stupid debate on how sentences should be formed, while adding no usefulness to the world whatsoever.

    3. it also seems to be full of editorial comments that are not really flagged that way. Really, who’s place is it to say where the RepRap project should focus? That’s what is happening – criticism of where they are putting the majority of their development effort.

  1. Seems to use a pretty good “common sense” approach. I haven’t been following the extruder thing, but first thing I thought when I heard there was a “problem” was “Why not use a commercial auger bit???”…

    1. Or we could go right to the source, and make one that processes dinosaurs…

      Could even get paid to remove them if you find someone with a nuisance population of Brontosaurs eating everything in their greenhouse, you could demand $25 a head bounty…

      (If you suspect that I’m kidding and can’t quite put your finger on it, you’re right, the international union of saurian harvesters sets a $35 minimum.)

      1. Can’t we just use that fancy RWGS cycle apparatus that people have thought up to create HDPE from the very air on mars? It takes CO2 and Hydrogen as inputs, throws it into an endothermic(?) reaction to create propane and similar gasses, then goes through a exothermic reaction to create plastics. It is even Eco-friendly.

        …okay that was from memory.. I found a link that is significantly better:

  2. Why is there still a 5 to 10 times markup for off the shelf filament? Is there still insufficient demand for cost lowering mass production? Or is something else hindering competitive mechanisms from pushing prices down toward marginal production costs?

    1. A couple reasons: The high startup costs of a large extrusion setup. The expense of the pellets themselves. Yes, they are super cheap in small quantities but good luck buying them that way at that price. The cynical view of this situation is that if home extruders take off, the cost will just shift to the pellets. To get the $5/kg price from China, you have to buy them by the ton.

      1. I don’t follow your logic. If pellets are 5USD/Kg how could they cost more than that when bought by the ton?
        There might eventually be a price increase in pellets but not because the demand is outstripping production. There’s too much of an industry built around these pellets for personal 3D printers to have in impact in the near future. The only reason for a price increase is due to retailers knowing people will pay it.

        No, QC tolerances on weed trimmer line is significantly looser than the +/- 0.05mm required for 3D printers to feed reliably. Also They are they wrong plastic.

          1. You listed cost of pellets as a reason for mass production markup being so high, but presumably the companies mass producing filament are buying it at the discounted bulk prices. So he didn’t misread, you just presented an illogical argument.

          2. I get what you are saying T.M. – however, if you take a look at the article linked to above ( click the first green link in this post ) you will see the following paragraph:

            This home-made filament dramatically improves the economics of 3D printing. For instance, producing 392 chess pieces in a particular color requires one kilogram (2.2 pounds) of plastic. Buy one spool of mass-produced filament, and that will cost you about $50. Buy a kilogram of pellets and make your own filament, and the cost goes down to $10. Buy 25 kilograms of pellets in bulk, and you can print the chess pieces for just $5.

            So the savings come sooner than by the ton, however, shipping costs have to be factored in as well.

          3. What have you been smoking? In China plastic pellets are not only free, they are so abundant that they have become a nuisance!

            to wit:

            Tongue in cheek of course. Some cargo container(s) got wiped by a storm and now there are plastic pellets everywhere. Estimates say it will take some years before they are no longer a problem fro the wildlife.

  3. It’s a very clever scale model of a regular extrusion press. The only thing missing, and on the scale Lyman’s working at it’s probably not necessary, is a pressure transducer at the die. On the commercial ones, pressure at the extrusion head is phenomenal, and a clog can have catastrophic results.

    1. What’s a pressure transducer?
      The problems with these machines is that you need to closely control the feed rate, pressure, and heating. A small change in filament diameter has a significant effect on print quality. I don’t know exactly what techniques commercial filament extruders use, but i know they are rather complex and big machines.

      1. … which is something a pressure transducer would undoubtedly help with. Monitor temp and pressure, and you get an idea of the ‘ideal’ viscosity of the material to produce the filament size you want… (or something like that). Use this information to control the extrusion speed, since I’m guessing that it’s open loop here.

        If you were really concerned, you’d use some sort of diffraction setup to very accurately image the width of the filament (same as with fiber drawing towers for glass optical fiber) and adjust parameters accordingly.

        But why bother? You’re not building rockets here.

    2. commercial extruders normally have a variable pitch auger, to compress the pellets and better squeeze out any entrained air. I imagine the filament produced by this thing may have porosity issues which will cause a loss of print quality.

      IMHO producing quality filament is best suited to high volume commercial production; the real issue here is lack of competition in the filament market.

      Not to take away frmo this device though, which is impressive.

      1. If it was really that bad, you could integrate the lower and clip into a single disposable unit. Instead of popping in a new clip, pop in a new LR/clip. :-)

        But it is really not that bad, and even plastic will suffice when SHTF! :-o

          1. Yeah, never gonna happen. If society breaks down (ha!) then suddenly everything that was once easy will become labor intensive. Things like farming and transport. Any loss of life would be compounded on to the local community. Just one person’s effort being lost could mean the ones that did the killing will starve, along with the rest of their family.

            Real value in a SHTF world would be knowing how to care for heirloom plants, the raising of sheep, lamb, and fowl (for insulation), creation of soap, and the knowledge to purify mass sums of water without fuel waste.

            This has already happened. During the black death the armed classes of knights and nobles had to tend their own fields or starve (not enough lower classes to support a labor-less lifestyle) while the peasants suddenly had the land to create luxury crops like vegetables, fruit, and dairy. It really was one of the primary forces that sparked the renaissance by improving the quality-of-life of the common people who survived.

          2. ^^^ That reminds me to have a serious crack at running my mower off grass clippings this year…. basically a solar still cum pyrolitic converter to terpene and wood alcohol liquid fuel.. my apocalypse has tractors in it TYVM :-P

          3. Roadwarrior, mount your still / pyrolitic converter directly where the grass bag should go. Then point it toward some grass and set it off across the plains! See how far it gets, a self-powered grass-eating machine, cutting a 2-foot wide swath across the country!

          4. @Greenaum <3 That was great. Reminds me of a (probably common) idea to make tiny, solar robots that move toward light. So long as they have light, they live and keep moving toward brighter sources of light. Kinda entertaining in a silly way.

          5. @ Smee – Great post. I’ve long felt that in emergency situations and ‘village’ based societies, the Jack of All Trades is a much more valuable asset than we get credit for in the ‘modern’ world. I’d much rather be the guy who can build shelter, make a weapon and stay alive long enough to build himself a machine shop, versus an average ‘sheeple’ (forgive the slur, but I’m sure you know what I mean.) Even in less dire circumstances… the person with the skills to open, repair, and recharge a phone from what might be on hand, without a WalMart around, would be valuable.

          6. @ Greenaum, it’s not really intended to be mounted on the mower at this stage, or produce fuel at the rate at which the mower consumes it, more like you mow your lawn one weekend, and in a couple of days you have fuel to mow it again…. though you probably have to leave it grow 2 weeks even in peak season for enough biomass to fuel a run… but you can cheat and steal your neighbours grass clippings :-D

  4. You can buy those auger bits at Harbor Freight pretty cheap when they have the coupons — they come in a 3 pack/sizes (I think 1/2″ 5/8″ and 3/4″?) for about the price of that ebay bit

  5. Why hasn’t some one made a print head that takes in pellets instead of filament. I understand that filament allows for a certain level of speed/material output control but if 1Kg of pellets is 1/6 the cost of filament surely a more expensive pellet extruderwould save money after 2-3Kg.

    1. Because print heads require controllable flow but with consistent start/stop times and consistent rates. There’s no way to predict when pellets will melt, so there’s no way to make the plastic flow consistent.

      Also, it’s difficult to generate a consistent pressure at the nozzle without having lots of complicated piping. The nozzle has to be lightweight and easily movable, so you can’t attach an auger mechanism to it.

      (among other reasons)

      1. I am curious why you say, “The nozzle has to be lightweight”. Why is that the case? Spindle CNC machines seem to be able to accelerate large masses to speeds well beyond what one could hope to deposit plastic at.

        1. Most consumer 3D printers are not built like CNC machines and cannot handle heavyweight parts. Other than that it could probably be done. Of course someone would have to design the proper system and all that junk. Many of the 3D printers are ripoffs of the Reprap and are not designed to be “Industrial” machines.

  6. Nice. The folks at Nerd Nights came up with a variant of this which uses an outer and inner layer of Polymorph, ABS etc and a centre layer of low melt 75C (BiInSn) alloy which is the lowest possible quantity of indium possible.
    It would work best if wound as very fine wires at manufacture, so as the filament is extruded it liquifies enough that upon solidifying it then sets into a contiguous conductor.
    LEDs and other through hole and possibly some SMD parts could then be poked into it, and the wire “pinched” when set to break the conductor at the pinched site.

    Thought about applying for a patent but what would be the point, much rather “donate”! it to the H-a-D crowd.

    have fun with this :-)

    1. You’re wrong.

      According to the link in the main story, the auger is $12. Total cost is supposed to be about $250, including the motor, motor controller, sprockets, temperature controller, etc.

    1. Hah, yeah if it was made by regular printer makers you’d get a fancy looking little box with about 5 linear feet curled up in it for $80.

      (I run a laserjet on remanned carts here… )

      1. More likely it would be a big box with your 5-feet curled up in it. Along with that there would be a clutch to stop the filament from coming out and a chip that engages the clutch after an expiration date. There would also be a little bag of sand to make it feel like there is much more stuff inside.

        But it’s ok because for half the price of their filament box they would sell you a whole new printer that comes with a new filament box. Of course… that one would only have about 1 foot of filament in it but the box would look the same and we would all believe it came with 5 feet.

        And if replacing the whole printer seems wasteful.. well, that’s ok to because in this alternate universe we would all be running windows and have to replace our printer anyway since the latest version of windows wouldn’t have drivers for our old printers.

  7. Now I totally want to make one and see if I can’t make it possible to recycle the mountain of plastic that accumulates on my curb every week. even with crushing milk jugs and plastic bottles my house (of 8 people) produces enough recyclables to fill the can twice, our neighbors are kind enough to let us top theirs off instead of putting it in with the trash.

    I have 3 of those wal-mart personal pedicure machines that give your feet an ultra sonic bubble bath and would be willing to part with them to make a sanitizing tub if I could just mulch my plastics into filament. way cheaper than the current $50 a spool, cheaper still than the $10 recycled spool as all the plastics that entered my house could be processed in the shop and I’m sure at the end of the month I would have enough spools of graded recycled filament to sell on Ebay (with enough excess to print off the spools themselves to roll it all up on lol)

    1. Harvesting plastic from dead all-in-one printers is also a good idea, as it tends to be high grade ABS.
      I looked into this but the problem is the WEEE legislation, setting up as a “licenced WEEE handler” is fairly complex.
      However if you buy the dead printer off whoever for say £0.20 then there is nothing they can do as long as you have a signed certificate of sale :-)

      1. Lol, I wasn’t even thinking of the enclosures when I wrote that, more the wadges of waste plastic bottles but that’s a very interesting point about waste electronics such as those that end up in landfill. As for the law you mention it sounds like a strange one on the books in the isles sir. I’m not saying go out and break it but it seems as likely to get arrested for recycling that plastic as it douse being caught smoking cannabis in Colorado.

        On the note of electronic enclosures, I wonder how much filament one could get out of the average sized desk cube? Another comment farther up the page linked to a Kickstarter page that was along the same lines as my comment; that yielded around 2.5 meters of filament to a milk jug and a detergent bottle. As sad as it is to say I have a few VCR/DVRs in storage I could probably use to make a fair length of industrial grade filament with. It’d be an interesting project for YouTube to see how much filament one could reclaim from used electronics enclosures. Spinning back around to your comment for my closing; I’m sure there is a similar law here in the States as well so I hope in doing so I don’t become the martyr for “pirated” plastic recycling =P not to mention the wealth of recyclable components one could strip from them like the linux server someone found in an all-in-one printer []. Though I would be more interested in harvesting the threw hole components like LEDs Diodes and memory if I could work it without cooking the components =3

  8. Is this the first success with an auger drill bit? I suspect it would work even better to cut the tang end off, chuck it in a lathe and turn a new driving shaft on the formerly cutting end. These bits are made to push wood shavings in the opposite direction from which they’re being used to push plastic pellets.

    1. I imagine that’s the beauty of the design. simply running it in reverse pushes material into the cavity like a low pressure injection mold. that said its not the most efficient design since its made of tool steel for the wear and tear of masonry drilling. I’m sure the only reason, mathematically speaking, that a larger volume of filament can’t be produced is the efficiency of the auger in moving media into the nozzle/heating-chamber (as well as the issue of heating that material to temp fast enough) as you said, by design its meant to excavate material for a relatively small hole. if someone could find out a way to preheat the plastic to near melting point as it was churned into the nozzle I’m sure someone could just use a paint auger, a larger hotter nozzle, and likely produce more filament faster than it could be used.

      That said great care would be needed so one didn’t burn down their shop if it spontaneously decided to piss out a volcanic stream of plastic instead of filament =P

  9. There are some practical issues that haven’t been raise here, for those thinking about bulk production of filament.

    How long does it take to start up from cold – how much scrap do you make, and how much attention is required to get it up and running stable without burning the initial pellets (from being too hot) or snapping the auger (from being too cold) ?
    What happens when you want to stop ? Big cleanout ready for next time ?
    What happens when you want to change colour, or material ?

    I guess these would be good for a hackerspace, where several could be run pumping out filament for shared use, but I suspect the expense and hassle for a single ‘home’ user will take forever to pay back (you’re a hacker, you already have a hobby to pass your spare hours, so you don’t need a new hobby namely ‘running a plastic extruder’)

  10. Anybody know how the plastic pelets are made and care to comment?

    All the ones I have looked at appear to have started out as filament and been chopped up! They are tiny cylinders with ends that look a little pinched like they have been sheared off a longer cylinder (or filament) in some kind of machine.

    It seems like quite a shame if we are taking plastic that was originally formed as a filament, cut up and now we have to apply energy to make it back into filament again! How inefficient! But… it’s still a fact of life that the pellets are being sold much cheaper than the filament so even if this is so, great work and carry on!

    1. Well that’s just how plastics are processed, extruding filament is just a byproduct of recycling. They are indeed “filaments” for a few split seconds before they are cleaved into pellets. The reason for this is that pellets are more efficient to transport. While some plants do purchase plastics by the spool these are industrial customers who will further process the plastics, likely in the same way that they would be in the first factory if they where processed into pellets.

      The key difference to recycling and the shop solution we see here is the industrial background work. Not only is there extrusion but sorting, grading, cleaning as well as any number of other formulations of plastics for specific product grades. While I made mention of simply mulching my recyclables in a previous post I’m sure others are considering finding that perfect plastic formula for the nearly indestructible arduiuno case =P

      bottom line is the bottom line.

      If you chose the shop solution over buying pellets you cut out the middle man in your recycling process, first you have to source your plastics (hopefully you have enough consumables in your house) then sort and process and so on. While it would be cheaper its another novelty that we pay other people to do. For a $10 spool of filament you are paying maybe 30-50¢ in labor, at least half in shipping and the rest is a gray split between materials and mark up.

  11. I wonder what mixing colors would look like. Would they really mix or would it create a cool speckled effect? How about letting the pellets almost run out and then filling the hopper with a different color. Maybe one could make striped filament that way, could be some interesting prints.

  12. Nice idea. I came across a material called “heat-it-up” which is a granular coloured plastic sold for making artworks.
    It mixes with PCL (aka Polymorph ™ ) and the bright colours such as red seem to work best, and obviously Glowymorph aka strontium aluminate doped PCL also works.
    I’d give that a try, can even send you some if that would help. Mailto testing_h at yahoo dot OMGWTFcom

    demunge email to reply..

  13. Well actually while making own filament is cheap the quality can be of different levels.

    ON the other hand 3Dfilamenta as I run are going to offer cheap filaments by making it fit for sending cheap

  14. Is there an update on the preassembled station? Also, besides plastic pellets, is it possible to mix metal powder and make customized metal-filled filaments?

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

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