Yeah, and this is why we don't sell pellets. Bummer but making good filament with a desktop extruder is not trivial. Maybe some day there will be some good hardware to make Proto-pasta at home. Maybe we'll help make this happen. Let's wait and see...in any case, thanks Joel for sharing this experience, now can make some more filament quickly and easily at our place! ;-)
I look at it like bread. Even if small extruders were to become as easy to use as a bread maker (get ingredients, mix ingredients, put in machine at manual specified temperature and time/speed) - a lot of people have no interest in making their own bread. It makes sense if you want to use a specific type (not a lot of manufacturers state if they're using 4043D, 3D850, 3D870, or what they are using - so if you know you want the properties of a 3D870 - making it yourself is the way to know). It makes sense if you want to make a more off the wall type not generally produced like MABS, ASA, or something like wood filled PETG. It doesn't make sense if all you want to do is grind up some failed prints.
They would not want to run plastic through their machine that random people sent in. Who knows what got in it, on it, or if the person accidentally threw in some PETG or ABS in with their 'here's my box of PLA'. Not to mention the shipping fees both ways, having to mix in raw resin with the recycled, and extra labor in trying to get the settings just right for whatever type of specific resin type (not all PLA is the same) you sent them. Just wouldn't be worth it.
The filament is still molten when it exits the nozzle, so temperature and motor speed aren't the only factors which will affect finished diameter. Any tension as it exits the nozzle will stretch it thinner and any compression will squish it thicker. If it's going to have consistent results, it's going to need some internal cooling system to solidify the filament at exactly the right diameter _before_ it exits the machine. Otherwise, it's basically just a regular extruder spraying irregular spaghetti into the air.
James Tanner good call this is what I was thinking too, it needs a length where it is allowed to run straight and be cooled. If you look at industrial extruded products like glass (sheet) and Aluminium that is how they work. It may be a good idea to use smooth cooled rollers with a gearing (so both sides are driven) to form the meterial. It is a shame this won't do what we need at the moment, recycled material could be used for infill or support really well on a multi material printer or for some interesting colour combinations.
Thats exactly what I was thinking, if the machine had it's own winding spool, and a device to accurately measure the diameter exiting the nozzle it could maintain consistency by increasing or lowering tension applied by the spool. Maybe a double extrusion method might be required, a rough extrusion onto a spool, then hook that spool up to a standard printer heating head/nozzle and then have automatic increase or decrease of filament disposition speed based on measured diameter going into the hot end. But thats more added complication. I think cheap filament recycling is only doable if you just want low accuracy bulky prints with a wide nozzle. Use a machine like this to make pellets, then set up your printer to work straight from those pellets, it won't be terribly consistent but it will work. Perhaps well if you only want .5 mm accuracy.
I wonder if you could just extrude into a teflon tube of the appropriate inner diameter, so that it's cool when it exits the other end. (Note: OP is not an engineer.)
James Tanner I was thinking Teflon tube too but even though it's sliippy it will still cause drag. Moving wheels, would allow precice clamping to form the right diameter and if they were aluminium with a fan blowing through/over then they would contact cool the filament and then run through tefalon tube to a printer/spooler. There is imho a need for a grinder like that used for meat above the hopper on this, then you could break up the waste pla and then grind it into the hopper. From what Joel found though differently sized place chunks melt at different rates so this grinding action needs to produce uniformly sized pla pellets. There are some good cd shredders that produce uniformly sized output. Someone out there must be reading this and have the ability to make one. If you are there is money to be made here, none of us want to be wasting material.
MarcAFK I've been thinking that pellets may be the way to go too. Couple of things came to mind. 1 the weight added to the print head cannot be too much. 2 measuring of the filiment output can be done with motor back pressure/resistance from the extruder (molten plastic stage) and maybe a gear sensor or an optical motion sensor through something like a tefalon tube. 3 retraction may no longer be available without a dual extrusion setup where one feeds into the other (as you suggested).
I'm not a 3D printer user. I design and develop embedded control systems (circuit boards and firmware to drive them, measuring temperatures and speeds, controlling heaters, driving motors (DC and stepper)). I picked up one comment in particular that suggests several system issues weren't adequately resolved. Quoting from memory (likely not an exact quote): "the motor seemed to slow down whenever the heater turned on..." * Why is heater power controlled on-off (seconds on, then seconds off) rather than using PWM ( Pulse Width Modulation - milliseconds on, then msec off, i.e. power cycle is 10Hz-1000Hz). That, and a PID (Proportional, Integral, Differential) control loop would assure a well controlled temperature. * The motor speed varies. This sounds like the motor speed is operating "open loop". They apply a voltage to the motor, and basically hope that the the speed that results is about right. The problem is, the speed is going to change with small mechanical load changes, and minor applied voltage changes. They should be measuring the motor speed, and controlling power to maintain speed. More PID / PWM control loops. * The power supply may not actually be able to supply all the power required. They might have gone a little too cheap. Try a higher current power supply (double the current), or separate supplies for the motor and heater. * Overcurrent trip on the motor? I can't decide if this is to protect the motor from overheating (measure the motor temperature, inadequate motor cooling, or undersized motor), motor driver (needs better motor drivers), the power supply (needs bigger power supply), or the extruder hardware (something breaks). Maybe all of the above... PID loops/PWM outputs on the motor and heater would go a long way towards controlling those parameters. I'm going to venture a guess that they didn't bother with those functions. The machine looks cute.... but it appears to be a starting point for additional work.
Touched all the right issues ... if the price point were closer to $100-150 US, it would be worth buying one and tweaking it in to make it work properly. However, for $700 it should be a turnkey solution that works out of the box. I’m a controls guy, too, and had a plastics professor that had a question on the exam “What are polymers?”. After elaborate responses by all students that got it wrong, the correct answer was ..... “many mers!” .....
20 years ago we automated several of our local pump stations. The number of feedbacks required increased every time you added something for it to control. And all of them had to have a manual override so that if the computer controller took a dump, you could operate it manually. Then just when you think all the bugs are out of the system, along comes a tiny spike at 0200 in the voltage that takes months of chart recordings to finally catch. Because it got past the power supply, the batteries, the opto-isolators, the filter capacitors and everything else that was built into the system controlled by an EMP hardened 386 processor. Addition of another capacitor finally caught the little bugger every time. But hearing an engine running at 1800 going to idle, the gearbox pre-lube pump kick on, the hydraulic clutch disengage, the vacuum valve open dumping the pump vacuum, and the alarms going off finally ended for the most part. Startup failures were even worse as the monitoring of all the parameters added more things that could go wrong. Adding one more required piece of equipment added almost ten more items that had to match to work. Now multiply that time 3 to 5 units at each location, and the requirement that all of them be controllable and monitored remotely over a microwave link. And just for good measure, some of us were also tasked with being able to operate any of the structures ( flood control gates or pump stations ) remotely over an area that stretched from Orlando to the Keys in case there was a headquarters system failure for the gates. Luckily I only had to take care of that one time. Bad thing was, it was on a day when the Area Director decided he needed to visit me. He got rather mad when he saw what I was doing as he was not authorized to even make the changes himself but yet I was ( a total of only about 20 people out of over 2100 employees could ). The whole system was password protected that had a daily change in the password. And none were allowed to give that daily password to someone who was not authorized. Even my supervisor and deputy area director were not allowed to have it so a person 3 levels down had more control they did. Their only way of overriding the system was to actually have someone go to the site and put it in manual control. Without authorization from HQ, that was a firing offense. :-)
So 800$ for a device that ships with no quality control, extrudes correctly about 20% of the time, doesn't put it on a usable spool and takes a month to produce enough filament for a 2 inch print...no thanks
Its not a new product though, sort of. The Filastruder has been around for a long time now, and works ratehr well, especially if a couple of the options like the motorized spooler and tension detector are used. While I'm sure these filament recyclers are much more complicated than they seem, how hard can it really be to heat scraps, and use a low geared auger to force them through a die?
@@zybch > how hard can it really be to heat scraps Probably not hard (I really have no idea). But, to get beyond the "extrudes correctly about 20% of the time, takes a month to produce enough filament" stage? (Presumably that second bit means it out-gases/wastes too much of the input filament in the making process - or it just gunks up the auger slowly that needs to be cleaned once or twice a week). That's probably more difficult (aka expensive). As expensive as Filabot stuff? I don't know. $750 for the EX-6 extruder screw seems awfully pricey for a (albeit very well machined) screw. (www.filabot.com/collections/filabot-core/products/ex6-extruder-screws).
They would be better off making it completely open source etc and people contributing to it. They are basically charging for a prototype that doesn't work at the moment.
Yes ,making your own filament is a journey. I bought an extruderbot when the company was still around and in all honesty it took me about a year to get to a point that I could produce consistent usable filament. You have to hold the filament at a certain angle as it comes out of the machine and have some way of winding it up slowly so that angle does not change . Walking past the thing will put it off, as the breeze you generate will affect it. It has taken me an age to get to the point where I can make about 3 or 4 kilos a day. The biggest problem I saw was the black 'shards', in my experience all stuff being fed into the machine has to be small enough to be readily accepted by the screw thread. My limit is about 4 mm , anything larger in any dimension will start to cause problems. I recommend getting a filawinder and giving it another go.
Come on what a load of rubbish that is all that needed is a truck wiper motor a arcamedeian screw couple of heater elements and a good engineer to know what's needed, that was Chinese crap at its best hoping public will buy ha
You have to set up a rig to wind the filament around a wheel with consistent pressure. You'll also want it to go through water to cool off faster. It's one big piece of the puzzle but there's more to it
That's exactly what I was thinking, there needs to be an extra roller after the plastic comes out of the nozzle to keep the filament moving and consistent.
We have designed a special nozzle that allows to cool down the filament when it comes out. So, as you can see in our video (ua-cam.com/video/JE3up88LEiA/v-deo.html) once found the right temperature, Felfil Evo can work without a spooler or external cooling system, extruding from a table onto the floor.
Yeah I'd say weld some support onto a 2 or 3 foot long section of angle iron so that it sits with grove of the V right at nozzle height and prop the whole thing up so that something like a 15 to 25 degree downward slope, I'd bet you could adjust the size of the filament a bit by doing that. I also think that the way Joel was grabbing the filament and moving it around also led to the wiggly and uneven result. Joel, you need to try this again and put some time into making that work. I mean realy, you don't have 6 hours to spend tinkering around with that thing, what do you do with your time?
I use a Filastruder, but a lot of the principles are the same. Extruding filament is a whole separate field from 3D printing, one that you have to learn the process the same as you did with your 3D printer. Think of getting an assembled extruder like this as if you just got an assembled 3D printer, but you don't know about bed adhesion, you don't know about overhangs, part cooling fans, supports, how to clear jams, what temperatures to use for PLA or ABS, etc. Examples of things you pick up over time: The jagged recycled plastic you put in initially will take longer to extrude than the raw virgin pellets (you can mix recycled WITH virgin plastic, but pure recycled will extrude slow, break easier, and be more out of tolerance). ABS will extrude much easier, quicker, and with a lower tolerance than PLA, so anybody just starting should try the ABS first then move on to PLA. A winder is best (especially with PLA, less needed with ABS), but without one you have to at least extrude hanging over the table enough that the filament starts to coil on it's own on the floor and doesn't hit the table edge on it's way down. Any disturbance in the path of the filament will cause diameter fluctuations. Turning the temperature down will cause motor jams because your higher temperature melted plastic further up the barrel - and now you're cooling it back down while the barrel is filled with plastic. PLA in particular has a very sharp transition - adjusting even in 0.1C increments makes a difference in size.
I use a Filastruder and Filawinder too. Have done for a few years. They work. They are slow and prone to tangles, but they work. Joel's comment that this machine is a start implies that he has not had his eye on the subject. It's not a start. It's been around a long time.
Thanks guys great, you hit the target. Here we are talking about filament extrusion not about 3D printing. Our mistake maybe was to not clarify this HUGE difference. However, according with your experience, if first steps could be hard, then the process become easier and a good chance to learn about filaments and polymer and let you able to customize your prints.
I don't understand how he thought it would have an easier learning curve than a 3d printer. I doubt there was anything wrong with the extruder but with the user.
I have a filastruder too And i made my own winder and puller I use 1.5mm filament(i can tune to make all sizes) Its really consistent(+/-0.05mm) Plus the pellets are cheap 3$/kg Compared to 20$/kg Its not about ABS or PLA its about the MFI value of the plastic you use Thats why some people get bad print with some brand and good print with other brand They need to tune it according to the MFI value of the plastic and not by brand
Great comment Neelmadhav Dookhit, nobody ever talk about MFI ( Melting Flow Index), probably cause is not so easy to explain, but it is also one of the main aspect that influences polymer behavior, actually this parameter is used to define extrusion grade material.
The manufacturer says to let the filament extrude off the side of the table and allow at least a meter so gravity can help the process along. Seems the motor is under powered so it kept catching on the table. The manufacturer seems to suggest 7 rpms and 175 to 200 heat range depending on plastic... my biggest gripe is a machine like this should not be sold for over 300 bucks when most people buying 3d printers can get a decent low end machine for 200 shipping included!
The home made ones I've seen extrude into a water bath or a fan shroud to cool the filament extremely quickly. Think you could probably extrude vertically down for better results as well
Joel....I used to work in a PVC pipe factory where we made everything form under sink size pipe to 15 inch sch80 PVC. I was a line tender I adjusted certain settings like extruder speed, internal air pressure for controlling inner diameter and one of the ways we controlled wall thikeness was puller speed. Speeding up puller speed reduced wall thickness, slower speeds increased thickness. An inconsistency in pulling the filement from the extruder will explain the inconsistency you were getting.
James Smith, do you think that he would have better results if the machine is placed on the edge of the table and let gravity pull the filament all the way to the ground, so it will always have the same tension in the extruder?
Well...gravity is a constant...but then every 4 ft or so would be an issue.. a puller that would give a constant tension would be the best idea. After the puller a driven wind up spool.
I work in the textile industry and we take nylon 6 and 6.6 and turn it into carpet yarn. I have basically unlimited access to nylon 6 and 6.6 in pellet form, we also make our own colored pellets they are extruded drawn through a water bath then chopped. Afterwards the pellets are mixed with natural color 6 or 6.6 depends on the order and make colored yarn. Recently we've gotten into the 3d printing business and now we make colored filament.
Imo it needs a motorized spool winder at the very least. A rats nest of loose tangly filment isn't that useful. Too bad it doesn't seem to create any useful results in regards to extrusion. There's a good reason why the real deal machines take up an entire room or two.
Hi Pile of carbon, industrial filament extruders works this way but we decided to use a different philosophy: a small tool that could work by itself extruding from the edge of a table to the floor, and we use it for production. (ua-cam.com/video/JE3up88LEiA/v-deo.html) By the way we are developing a spooler that will make easier the extrusion process.
I had a length of it that went over the edge of some stairs testing to see what I would end up with. Pics: twitter.com/Deneteus/status/993734117748609029
You didn't say but I ASSUME you know small changes in temp will take loads of time to stabilize throughout the unit. I agree with Proto Pasta below, it is an involved process and a lot of work goes into set up and QC over time. A home box as you demoed is cool and should get many to start scratching their heads to work out an affordable, yet reliable unit. Maybe for now, store your old wasted PLA for a solution down the road and it does not materialize maybe one of the filament co's will offer a small price for used PLA. Maybe they can shred and re-use it. Perhaps a credit towards new filament. Say $1 per pound credit and they pay the shipping??? Another thing this type of unit is woefully lacking is a take-up system. As the filament is coming out of the extruder it needs to be gently pulled and wound on a take-up reel. Lots of adjustments but once set, the pull at the right speed should stabilize the diameter. HOWEVER, as the take-up bundle grows in diameter the speed will have to be adjusted accordingly to maintain the same pull in Newton's. This could be done, probably, with an arduino and if not, a raspberry pie. One other concern is temperature variations. Try this in a garage whee it's 60F in the AM and 90F in the afternoon and expect headaches. And ONLY GOD and the filament pro's, know how much humidity will play into this. Here in the Sunny South, out humidity can be 90% plus and you folks in the desert areas may have less than 10% H. I grew up in the 60's and remember my sister having an Easy Bake oven. Basically a lightbulb in a box and you could gently cook SMALL amounts of cookie doe or baking mix. Your reviewed item reminds me, in a way, of that toy oven. She did have this oven but it didn't make her a gourmet cook. You do have a neat novelty item and like I mentioned, hopefully, it will encourage some head scratching and innovation...
According to their website the price of a fully assembled version is $860.00 way over priced. For that price you would think that they would at least provide a spool holder.
In any plastic extrusion process you have to make some constant tension in the wire to control the width of it and improve the mechanical properties of the wire. I would recommend you to make a wheel that rotates at constant speed so it takes the wire as it comes out.
LOL I know that is the problem Italian, looks expensive works like crap. Joel knows how to build and setup a 3d printer so this should be much easier just 2 settings.
I'm italian... and honestly i feel ashamed that piece of crap pretend to be italian too. Not even mentioning the disconnected cable. 700€? I can build Lyman Filament Extruder V6 with that money.
@Daniele I've worked a lot with Italian companies and my comment was unfair and meant as tongue in cheek. OTOH Italian manufacturers do spend a lot more time considering the aesthetics of their products than others. When it works it's brilliant (ALESSI) and when it doesn't we like to make fun of it (ALFA ROMEO) 😜
I bought one too and his experience is right on with mine. I do less 3D printing and more experimenting with different materials to be extruded so it’s fine for me. If I had to exclude material to use printer, I’d try a different machine!
If the shaved chips messed it up, then how in the world can someone turn their old prints into "dust" and not run into the same problem? I think this is a good idea, but it needs to be a plug and play. The person shouldn't have to question the parts, dust, etc. I haven't gotten my first printer yet so I don't have a bag of bad prints yet. But I'm keeping an eye our on these things because I know it will be a problem.
i would use something like a bullet blender. can literally turn any plastic into shards, beads, then grains, then dust in a matter of about 30 seconds.
Chrismofer OK but this needs to be plug and play for it to actually sell. Their own material didn't come out right. You shouldn't have to spend so much and then worry about stupid stuff.
We had a very similar filament extruder at the local Hackerspace where I live in Germany, since a few years. It works reasonably if you let the filament fall from a specific height, so gravity keeps the extruded filament under tension etc. (I doubt you'll be able to get any consistent thickness the way you're trying to), we mainly tried extruding ABS though. While it's certainly possible we never really use the filament extruder, since PLA/ABS pellets aren't really cheap enough to justify the hassle (it's obviously not a labor free process and as you already noticed it isn't tremendously fast either) and grinding up used PLA or ABS always comes with the risk of getting impurities into the filament and clogging your printers nozzle. So yeah... It's a pretty neat idea and all, but with the simple "DIY" filament extruders I've seen in the last couple of years, I'd say we're not quite there yet.
I actually picked up a Felfil Evo with their filament winder and so far I'm actually pretty satisfied. Within 2 hours of UPS dropping the package off I was able to figure the machine out, set it up, and extrude some of the included virgin PETG pellets into some very nice filament that I proceeded to print out a test cube on my Prusa MK3S. I haven't played around with it a whole lot yet but so far so good for me, it seems like the filament spooler is crucial for it do have a consistent diameter, which on my first run was 1.75+/-5mm which is honestly kinda impressive to me.
I have done this before and got 3 full spools of good usable ABS out of a similar product called Filastruder V2.0, this product also uses a photo resistor to keep a specific amount of tension in the filament and spools it. the key is to spool the filament as soon as it cools.
Joel! I've been holding this in for so long. Can we please have those holes filled in with Spackle? I don't even care if they are painted. Those holes, are they there on purpose to cause epileptic seizures? For the love of your viewers it really only takes a few moments to fill in those holes. I use an old butter knife and then come back with a damp cloth or sponge or even just a finger to smooth it out.
Oh man, my "whatever it is" will thank you! :D I was a bit overly harsh, but I think I may have snapped. Those holes, those HOLES!!!! LOL You are the best!
keep in mind you also "need" (you can without, but it's a pita) a filament spooler, and if you want to recycle plastic, a shredder. precious plastics have some nice plans for an extruder and a shredder, but not a spooler. to measure the size, the classic way is one fixed bearing, and one bearing on the end of an arm with a spring pressing on a load cell. Though you could probably get away with using an led, a cheap camera, and a pi, measuring the luminosity of the camera gives a fairly accurate size measurement.
They should have added an optical mic and controlled the temp or speed with it via PID loop. You can't ask a user to sit there for hours and constantly measure\adjust. That's not value added - at all.
A well designed machine doesn't need that. There are other filament extruders that run open loop without issues. But they cool the filament, don't have underpowered motors and power supplies and have a motorized winder.
You would need a mini granulator to create consistent size in the recycled plastic (or only use virgin material), in order to give more consistent output. As I work in the plastics industry, I suggest keep the temperature high, and increase the RPM to adjust the size, you always have to have sufficient temperature to give a melt flow. Something you should consider is running the filament through a water bath to cool it as soon as it leaves the extruder, to reduce the stretch on the material don’t allow it to sag or be pulled as it leaves the die.
You'll need to make pellets first. It's not worth it. The machine costs plus the wasted time and energy vs the filament cost...even working in China at their wages it'd be more cost/time effective to just buy the stuff.
I didn't read all the 950+ comments, but I would suggest few modifications to get a decent filament extruder: 1. Instal a larger nozzle 2. Put the outlet on a vertical plan (instead of horizontal, yeah, the hoper will need to be modified as well) 3. Have a guide loop for the filament and a automatic spoiling 4. Get an inline size gauge and link it to the extruder AND the automatic spoiling speed controler Well... 5. Nevermind step 1-2-3-4: Just buy regular filament!
I've never laughed so hard watching a 3D printer review until this one. To bad it doesn't work. There are at least 20 electrical engineering flaws I can see in to design let alone mechanical. Your video production was the best part about this video.
I have started my own filament recycling program to convert waste into usable filament in South Australia. There are at least 8 filament extruders I have found, Felfil is the second cheapest I found, the most expensive is upwards of of $5,000 US for an all in one unit. I agree with your review, they're a good starting point to prove the possibilities, but there are many drawback as you suggest, and I believe most people wouldn't have the time or experience to produce their own filament without significant investment.
I funded their Kickstarted around five years ago, and they shipped late and my kit didn't work. I'm still trying to fix it, and I've spent over $100 trying to get it to work. Not cool! I'd love to get some use out of it!
Thermal runaway - keeps overheating. I've been emailing support for a LONG time about it, and they've sent me two new boards (one I had to pay for) and a handful of other electronics like a new thermistor and heater cartridges. Really disheartening to spend so much time and money on something that, even when fully assembled like yours, doesn't do its job :( Hopefully someday it'll be so easy to recycle 3D prints that it's a no-brainer!
Deneteus Sorry - I wouldn't be able to tell you. I just received a new board from the Felfil guys which was supposedly tested, and it didn't work (even with all new electronics for everything else). They sent me instructions on how to flash new firmware using an Arduino, but my computer wont write sketches to it. If they actually did test it, it should have the newest firmware.....right?
Considering there isn't alot to the FEVO I would expect they sent you a new tested board. If they did send you a board that had the latest firmware I would make sure the XBOX 360 PSU they are using isn't the culprit. If you can't use an arduino then you should try using a USBASP programmer from eBay to do the job. I just verified that there is a port to the right side of the board when looking at it from the rear. Here is the official programmer. All the rest are clones. I use this to program all the 3D printer boards. www.ebay.com/itm/Official-product-USBasp-programmer-red/323209529275?hash=item4b40ca07bb:g:fjoAAOSwQaJXSIGT
In the injection molding industry, which augers the material through a heated tube and tip, we call the “cleaning pellets” purging material. It’s typically clear, I assume they are using recycled scrap with this.
I think that the one major problem is that the filament is gravity pulled, so it stretches unevenly after it exits the nozzle, If it was extruding verticaly down from height it may get more consistent size, or maybe some mechanism that pulls the filament out of the nozzle with constant speed and allow it to cool under constant tention would do the trick.
I get that this machine appears to be a disaster. But, to be fair, it was frustrating that after you initially said “first..." you use the black shards to flush out / clean the new machine. That you then went on to use the output of this cleaning material to assess the performance. Even measuring the width of the cleaning material output! Huh? Well at least near the end of the video you did start to use the proper pellets. :-)
Fair point :) Turns out the black shards of PLA weren't in fact a "cleaning" filament. I read wrong and just the first few meters of filament were considered cleaning. The black pla was suppose to produce good results
Thanks for clarifying. Your findings defiantly suggests this product isn’t really fit for sale, as it doesn’t appear to live up to it’s advertised claims! It would be really interesting to hear from customers who've bought the product in good faith of being able to: "make custom 3D printing filaments at home”, and: "save over 80% of the price of filaments spools”. I guess felfil should at least amend their advertising, to add: “YMMV”. LOL
Well, also to be fair, I would want such a device to re-use grinded stuff back into filament. Exactly like that cleaning stuff. So does that actually work, if it only works nice with perfect pellets than it's not about recycling anymore.
It's better to build a Lyman extruder that can extrude almost anything than it is to throw regrind into an extruder with a windshield wiper motor. ua-cam.com/video/H6h1yDzoPMU/v-deo.html
I'm not sure what the "purge" time is on this machine, but on the Filastruder, it takes almost 6 hours to "clean" the machine out... My extrusions looked terrible for almost 2kg of resin.
I heard a lot depends on any pull out speed, or friction showing push out down. Example : more pull out, (has to be perfect all the time), will stretch the hot, just out the nozzle, filament thinner. Another thing with some plastics is moisture
If you really need a 3D-Filament Extruder then i can recomend this guy. He offers a DIY Mini-Extruder which can make 1KG 3D-Filament in 1 hour. Even PA12 and other 3D-Filaments you can do with this thing. I really can recomend it! ua-cam.com/video/44oZ7TSlCx4/v-deo.html
have you ever seen 3devo's machine? It can also extrude up to 1kg/hour and is an all in one desktop extruder. That extruder works very well and is really reliable. You can also make a lot of different filaments with it. Just look it up it's really cool.
Zion: Well, ironically, the machines are actually even more for making new filaments, out of new plastics. It's a great way to experiment with new plastics. The recycling part is definitely something you can do, but it's questionable how interesting it is to actually do that, because, while (for example) spools of PLA filament is pretty expensive, just PLA granulate (little blobs of pla, which you can throw in these extruders) are really cheap. So recycling is probably not worth it. But you should look up the 3devo machine on the internet or youtube, it's really cool! www.3devo.com
CNCkitchen got a good video about recycling scrap prints & pellets. You need a winder unit that starts and stops via photosensor. Let the extruder piss the filament down approx. 1-2 ft to a loop, and wind it up. You need some time to try out until you get the right diameter, but then it works like a charm.
"Tinker" = Fix manufacturer problems. Cheap DC motor with gearbox and no feedback (closed loop). Motor takes strain, drop speed, feed slows down, inconsistent feed. No cooling or support after extrusion. No winding. This really is a POS. Can't believe a company spend money trying to design something cheap and then charge a lot for it. Nice video as always, Joel. I'm sure you had to bite your tongue a few times.
Saran for the making of Saran Wrap is a tube that leaves the extrude nozzle and goes directly into a bath of water and is taken out of the water before it cools down too much. From there it is pinched close and then blown to a tube where it cools fully and makes a solid. The use of water would simplify the process. the tank would be continuously being filled and over flow the heated water at the surface. If sufficient heat remains in the filament the film of water on the surface will evaporate before the take up occurs. So a driven pulley will be needed to remove the filament from the bath a a constant speed. The height of the nozzle to the water will define the diameter of the filament. And the take up mechanism can be very simple. An idler pulley that drives a variable diameter V belt pulley or as we used in the R & D two V belts with the variable pulleys where they transfer motion. Thus as the idler falls, more filament to wind, the take up drum increases in speed and as the idler rises, less filament to wind, the take up drum slows up. You can print this three piece idler V belt fixed side cone on both end and a middle cone on both sides that slides on the shaft. You may print the shaft as a spline that you see on drill presses to have all faces of the cone carry torque.
So the V-belt pulley of the motor and the take up drum are fixed. The three piece V-belt idler moves closer or farther from the take up fixed pulley. the diameter of the two V-belts sum at the three piece V-belt idler is fixed but individually change. Speed changes . The idler measuring the loop of filament moves the three piece V-belt idler assembly.
Maybe the diameter of the filament was varying because you were pulling it out with varying force? It's kind of like a sticky stuff, if you pull it hard it it gets narrow. So maybe it will be a good idea for Felfil to introduce some sort of rolling machine which can put the filaments coming out in roll, and retains proper tension on the filament as it comes out?
Now that it's a year after your initial trial of this machine, I'm curious if you've tried a newer version of this or a similar machine. Are engineers any closer to making printing your own filament a reality?
even when this video came out there were machines available for this, Now there are stl files to print your own parts and build an effective filament maker out of an old 3d printer!
welcome to the world of plastic injection molding. we have a lot of variables that went into plastic injection molding. barrel temp, hopper feed, cooling time, cooling temp, injection speed.... and one problem could have one or multiple sources. my concern with this is it should have an addition nozzle for cooling the plastic down before exiting the machine. so you can get a consistent solid piece that isn't like wet spaghetti. so the heating barrel melts the plastic and then using the screw to feed more plastic in allows the plastic to be pushed through the heated barrel into a cooling barrel that would taper as is exits. that way it wont get bound up from friction but could get a relatively straight piece to form before exiting the machine onto a spool preferably motor driven.
What "proper supporting hardware" was he supposed to use? This machine is supposed to be a stand-alone out of the box working item. The manufacturer himself advises to let the filament fall to the ground. He respected all recommendations from the manufacturer and did corrections as they advised after he emailed them about his issues and still failed to come close to the advertised result. So please tell us what the manufacturer forgot to tell Joel to get it to work.
I think it needs a better/more powerful extruder but most importantly it needs an extrusion take-off system, that's how you could achieve a consistent diameter. It also needs to be able to run for much longer or extrude at 3 or 4 times the rate.
I'm no plastics expert but it would seem a gravity fighting cooling bath, spooling system and a better extruder with graduated larger to smaller internal extrusion dies finally exiting at the 1.75 size. Also graduated slightly lower temperatures each stage would be the trick here, hotter to start out, barely above melting point at the output tip. You also need an absolute constant speed and pressure at the output tip.
It's the power supply. When heater comes on the motor slows because the voltage drops. Very easy problem to solve. Thickness changes happen because the filament is hanging and stretching the material. You can put a fan blowing on the filament to cool it. Have fun.
Thanks for your feedback. We explained that the filament should be falling on the floor, not on a table (at least for now - we are working on improvements). It takes some time to use our product properly.
Isn't the filament meant to be extruded off the edge of a table .... without pulling on it or letting the friction of it pushing against the table surface...that's why it's so squiggly
3D Printing Nerd Thanks for clearing that up... that's such a let down... i really wanted this nice and compact recycler to work at least for future purge blocks and waste etc
i was imagining a messy pile of PLA waste going in and a nice spool of PLA ready to use... this is a great idea that needs some work, I bet in a couple of years it will be amazing !
Filament not 1.75 should be ok. Any directdrive machine should accept 1.5 for example just adjust the slicer. But it must be consistant. If it cabmake 1kg of 1.5 1kg of 1.6 i would be ok witb that as long as it was consistent. Outside my expertise but my hunch tells me they should run dual power supplies so the heater kicking did not effect power to other components and i would gear the grap out of the motor so while slow it would have gobs of torque to eliminate the variances in speed.
Dear Nerys, you are right 3D printers could work also with a filament which is a little tiniest than 1.75mm. However extrude with a goal of 1.60mm (so with +-0.07 optimal tolerance 1.53/1.67 mm filament) is a little bit an hardcore solution, if it totally avoid that the filament get stuck into the printer’s hot end on other side it can influence the print results. What we suggested is to work trying to get the proper diameter 1.75mm and correct the settings only if its measure is over 1.80mm. About our geared motor, it produce a 23Nm torque in full load and, according with our tests it is enough to work with the majority of the 3D print polymers. For who those feel real makers we also provide a basic solution named “Basic Kit” that allow you to full customize the electronics control and alimentation.
Someone designed a filament width detection sensor that you can use with Marlin way back in 2014. A newer version can be found on Tindie. I have not tested this yet but I plan to.
We made some experiment with it! It's a good start but unfortunately it is really susceptible, also making an average of the read values. I'm not telling that is a useless project, just there's still some works to do in order to make it ( or something similar) suitable in our design.
I've built 2 variations of the lyman filament extruders that have been around for about 6 years now and they offer a lot more consistent results compared to this unit and at a fraction of the cost. This just looks to be half a machine without enough real design time put into it, maybe it will get there in the future and at a better price point.
It can also depend on nozzle size and layer height. If it's easier for the melted plastic to flow up and around the cooler filament up in the heat break (into the space that normally would be taken up with filament) rather than be shoved out of the tiny hole in the nozzle then you end up with nasty jams in the heat break from undersized filament flowing upwards.
Filament extrusion consists of much more than just a machine that will chip, melt and push material through a nozzle. The melt exiting the nozzle must be careful controlled in pace while it cools. Allowing it to droop from the tip will induce form and size issues. Typical filament extrusion has carefully controlled cooling baths or sprays depending on the mass of the material being cooled and the rate the filament passes through the cooling zone. They always have a high precision cat or capstan puller drawing the material from the nozzle at a very consistent speed. The extruder is set to a temp and rpm that produces a molten, but still just thicker than liquid melt consistency (tough with PLA) and just larger than the final filament diameter, say 10%. Then the haul off capstan or cat speed is set to stretch the filament to final size (known as Draw Down in the biz) through the cooling system. Any changes to rpm of the extruder or temp of the extruder need time to achieve steady state thermal conditions. A 20 minute trial likely only provided enough time for two adjustments. Any more than that and thermal conditions never reached a steady state. If you try this system again with simple air cooling, try to use gravity to create the precision haul off rate. Tilt the unit on a downward angle and support the filament with a short support plate that will provide steady friction and allow the filament to drop to clear air and then don't touch while running. Should work until the filament hits the floor. Then you need something that keeps that end from influencing the nozzle draw down. A take-up spool set enough distance away would work if synced with the extrudate rate and allowing minimum tension past the support plate. If you don't want to have another go at it, send it to me and I'll give it a shot 😁
I have used filastruder. Requires atleast 45 degree angle, and for PLA a cooling fan. Also every time you touch the filament while extruding it will cause inconsistensies, so once you get it running, do not even touch it. Also in practice a winding machine is also required, and for consistency extrude straight down. You can then get consistency down to around 0.01mm.
@@Felfil the filawinder is open source, just use that with your mentality attached to it, ie. prebuilt, packaging, support, materials etc. Works really good out of the box.
@@Felfil Cool! You checked out the Lyman stuff? That is closed loop industrial style, but open source and desktop form factor. It pulls the filament out and measures it is correct size etc. I should look into that again, might be fun build! But for me, it is more about experimenting, and i quickly found out that to experiment with stuff, i need extra barrels etc. as once it gets gunked up with some weird material, it is very hard to get out. I got a reaally bad batch of Nylon :( Further going back down on temperature range is a lot of work. With closed loop doing higher temp through bigger nozzle one could experiment with alloys and filler materials ... ;) I've been meaning to do some experiments with graphene, got a few kg of big flake (upto 44micron, median somewhere around 20micron) size 80%+ single layer graphene waiting for the right time. PA6 + Graphene 1% + GF50 might be veeeery interesting mix.
@@skaltura Yes we know about Lyman, it's a very nice project! As for materials, we extruded successfully a filament made of nylon PA12 only and it was good for a normal printer, while we never tested graphene.
Having worked on the prototype production of masterbatch for use in plastics manufacturing I'll tell you now that it'll be a while yet before you can make s machine powerful enough to recycle old prints that fits on a desktop and does it's job reliably, plus you need to use a bath of water at the end of the extruder to float the extruded plastic then use an air line to blow off any remaining moisture (Or constantly replace tissue ) Edit: on the subject of consistency, recycling plastics into an extruder is not an easy thing to make consistent, it's not easy at all, and you may need to adjust settings on the fly....If interested in any additional knowledge on plastics extrusion message me, I've got a few years of experience working with pretty much every plastic and polymer that's on the market
Jeffrey Moffitt as cool as that would be the costs to produce such a product would be massive, the issue is that currently the torque and power required plus the knowledge required to operate it consistently means that you'd have a very small target of actual buyers
I drew up a design that takes the torque factor out with a spring system. Imagine a vertical hot glue gun with a cooling tank below it with a hole in it leading to a spool
Well there is already a very reliable consistent desktop machine on the market. It's the 3devo extruder. Works really well, and you can throw all sorts of plastic in it to make filament from. Just look it up on the internet. Only con is that it's a little expensive, but it is a consistent desktop extruder on the market
Most working DIY filament extruders use a spool drive system that is synced to the extruder and that is supposedly where the consistency is derived from because it keeps a constant tension on the extrusion while it cools and winds on to the spool. you can't make filament by just extruding it. you need other parts to make this work well. Also, the best ones uses much longer and bigger heaters and augers.
I've suggested it to my local hackspace, but they say its not worth the cost... so I'm slowly getting all the stuff I need to build one and show them. but its an expensive process
theres this one www.filastruder.com/products/filastruder-kit?variant=338126174 Its a kit but it works better than this one and its cheaper, also i have not bought it, i have only seen videos on it
What iv Seen In industrial is a bigger nozzle and spool winder. They get the temp and extrusion set and the spool winder stretches it to size before being cooled at a certain rate.
A few suggestions: 1. For bigger filament, try to increase the capacity of the heating element. 2. If the setting of the temperature changes the motor speed, try to use two independent power supplies for the heater and motor 3. Let the extruded filament pass through running water for uniform cooling. Make a channel of say, one meter long through which the filament is drawn out. 4. Change the size of the nozzle (drill a slightly bigger hole if you use more powerful heater) Hope these modifications can achieve in getting a filament size you are looking for.
To make better filaments you need a larger heating tube and then a water trough and a powered wheel that loads the filament onto a reel and the size is adjusted by the loading wheel speed
Extortionately over priced and unfit to do the job required. you can build yourself a core xy and have enough for 23kg of filament .... this product is a joke.
I don't know much about making filament but I'm pretty sure you need some kind of extraction&cooling system after material exits the nozzle. So like additional, precisely tuned stepper motor pulling it through water which cools it.
I think this extruder is just a portion of a final homemade filament factory. A critical component to be added to the system is a feedback loop to control the filament width and a mechanism to cool the filament down as it extrudes and to collect it on a spool.
I think for a home system without GMPs you would want to extrude the filament under pressure at like 2mm, cool it, them pass it through a cutter than trims it to 1.75mm. That seems a lot less error prone and the trimmings can be recycled again. Relying on the extruder and gravity to size the filament is asking for a home user to mess up.
I have a rule never to correct someone else's prototype which is what you would end up having to do to get it to work. If they are paying then no probs but that unit looks like a Graduates first attempt. Nice idea but it needs to be consistent otherwise why even bother. I think you gave it a fair and honest review and a big thank you for that. :)
I would like a 3D printer that accepts pellets. A complementary grinder, too. Not only failed prints and supports, but also any household plastic (bottles, TV shells (e-waste recycle the innards), broken toys, even plastics not accepted by the local recycle co. can be ground up and used as 3D printer fodder. Pellet 3D printers are a thing, but the tech still needs refining.
How COULD that even work? I mean.. No cooling, no spooling, no measuring, filament just bundles up and tangles in front of the machine, hindering flow of new filament.. Meh. Also 1 meter per minute is an abysmal speed. 4 hours max operation time means 240 meters of filament over that 4 hour period max. Meh.
Yeah it is to cool down the machine, only have to turn it off for a few hours and then back on, could potentially make 3kgs per day or so. Also if you watch the owner use this machine he says it is supposed to be used on the edge of a table, then filament will then spool up in a spiral naturally on the floor or wherever it is landing due to gravity, this will also even out the inconsistencies in the width of the filament as well. The dude in the video got it to 1.74 for instance.
I tried to let it spool on the floor - that's what the big bunch of filament towards the end was from. However, that did not work. There was no consistency, and the machine would shut down every few minutes for protection.
This thing looks insanely awesome, at least in concept, being able to recycle old print material, just got a 3d printer a couple days ago, still not set up properly unfortunately, but I love the idea of near zero waste. Guessing it's not going to work too well, at least not without loads of practice and tweaking. Would be awesome to see the price come down over time though. For a couple hundred bucks it would be awesome. I'm sure things like ambient temperature and how the end is handled play a decent part in how well the filament maintains its shape. edit: man, that was pretty disappointing, and having to run it over 5 and a half hours for the equivalent of a 1kg roll, while only being able to run for 2 hours at a time is pretty rough. Also, if it has trouble with the chips they sent, I can only imagine how bad it would do with the results from a home shredder setup unless it's really keyed in.
After seeing you struggle with sizing , I thought I will share with you the solution, most extruders have downstream equipment ,cooling trough water or air cooled, where filament get cooled as it leaves the extruder. secondly, puller (Haul off) which could simply be two rollers driven to match the speed of extruder output which pulls the filament at a constant speed. Then the spool winder, The puller speed controls the diameter which is usually tiny bit faster that extruder to take up the slack and size the filament to exact size. My experience 1974 to 1993 , producing filament of 1.5 mm to Pipes up to 500mm. all had same set up, except pipes are sized and cooled in vacuum tanks.
This would be really cool... only issues I have is: > the amount of time it takes to actually prepare the unit for use > It takes a damn age to extrude the filament (1 meter per minute!? TOO SLOW!) > *IT'S GOD DAMN EXPENSIVE JUST FOR THE EXTRUDER KIT!! 300 GOD DAMN EUROS!!* > The amount of headache you get trying to actually get consistency with the width of the filament just doesn't seem worth it for the cost. Also, I would like to see a carrier that moves at the same rate as the extruder so that you can wind the filament straight onto a reel ready for use with a 3D printer. This is a cool project and I personally think, as an outsider that this is potentially a really useful tool for saving your excess filament from becoming fodder *IF* they can get it to extrude at a consistent rate.
For it to be good it needs a self spooling mechanism and something to let the strand cool as it comes out, so it's own weight won't pull the filament to a thinner diameter in some areas.
I've worked on big extruders and yes they are way faster but they have cooling tanks to control temp on the output side.. the problem I thought of is there is no way to uniformly grind your scrap..good video though and cool little machine
Check and see what the amps are on this unit and what it can put out. The reason is that consistently the voltage and amps have to work hand and hand to produce product especially when extruding. There theory is true heating it first,then extrude product. But if your amps aren't high enough and stable it's very hard to keep your temperature true. That's why the temperature isn't consistent. It may need a bigger motor also. If you can check the amps and temp.you will see the difference I'm talking about the two. Good luck!
The first thing I noticed is there is no draw from the extruder. the filament needs to be drawn out of the extruder (at a rate consistent with the input speed, temperature, and the size needed). I remember some time back seeing a video from a filament manufacture, where the engineer said that they use 3mm nozzle for 3mm and 1.75, the primary difference is the rate the filament is drawn away from the nozzle. Since this system has no drawing system of pinch rollers it is got no chance of working. Excepting with a very steady human hand. It may be possible to use a gravity based system but that is also going be tricky. My choice would be draw downward with a set of pinch rollers, and have an optical measurement device as close to the extruder as can be managed feeding back so the temperature, auger speed, and draw rate can be adjusted real time.
This, in it's simplest form, is the same principle as plastic injection molding. The cleaning or Purge material looked really rough probably because it was used once. If it was regrind which is used material, then ground up , that throws off the chemical structure of the compound and doesn't melt correctly. Mixing it with virgin material at a 25% regrind and 75% virgin may have produced better results. Also the large chunks of regrind most likely didn't melt as fast as the smaller ones causing a less flowing consistency.
Looks like it needs a speed controller with rpm feedback from the motor. The current draw from the heating elements is messing with the speed. Just slap a vfd driven 30 amp 3 phase motor in there and your problems will be solved. Also needs a good take up winder to keep consistent tension on the extruded plastic. That is affecting the thickness as well.
You need a stepper motor about 30cm away pulling on the filament as it comes out along with cooling fans or water between the two. Tuning the stepper motor speed should make the filament more consistent in size. At the moment it's bunching up while still molten causing it to change diameter
reducing the temp will make the plastic extrude slower, since it needs more pressure from the auger to make it's way out. this increased pressure is what i think increases the diameter. technically slowing or speeding the auger should do that too.
I think that one of the most important part is the roll and cool part once it's extruded. That's not a trivial part and you are just letting it go over your table so the friction is also slowing the filament out of the extruder and creating inconsistence.
Hi, a spool can help but it's not necessary. In this video you can see how Felfil Evo works and maintain the filament constance: ua-cam.com/video/KA3Kf22BsZ4/v-deo.html
The thing needs some kinda motorized take-up/cooling system. Maintain consistent 'draw' on the filament while it is still oozy, cool it before winding onto a reel. Too thick? Speed up the draw a tiny bit to stretch the filament. Too thin? Get a bigger nozzle! You are right: this is a great tinker-toy!
Cool idea, needs a complete redesign though, especially considering the price. Has a brushed motor which would be way better brushless and some way to auto spool the new filament. I don't think another device that requires tinkering on top of the hassle of 3d printing is a winner.
Yeah, and this is why we don't sell pellets. Bummer but making good filament with a desktop extruder is not trivial. Maybe some day there will be some good hardware to make Proto-pasta at home. Maybe we'll help make this happen. Let's wait and see...in any case, thanks Joel for sharing this experience, now can make some more filament quickly and easily at our place! ;-)
Hideki Shinichi shipping the waste back and forth will cost half a spool alone, if not more. Not cost effective. XD
My thought was...is it even cost effective to recycle old filament?
I look at it like bread. Even if small extruders were to become as easy to use as a bread maker (get ingredients, mix ingredients, put in machine at manual specified temperature and time/speed) - a lot of people have no interest in making their own bread.
It makes sense if you want to use a specific type (not a lot of manufacturers state if they're using 4043D, 3D850, 3D870, or what they are using - so if you know you want the properties of a 3D870 - making it yourself is the way to know). It makes sense if you want to make a more off the wall type not generally produced like MABS, ASA, or something like wood filled PETG. It doesn't make sense if all you want to do is grind up some failed prints.
They would not want to run plastic through their machine that random people sent in. Who knows what got in it, on it, or if the person accidentally threw in some PETG or ABS in with their 'here's my box of PLA'. Not to mention the shipping fees both ways, having to mix in raw resin with the recycled, and extra labor in trying to get the settings just right for whatever type of specific resin type (not all PLA is the same) you sent them. Just wouldn't be worth it.
ualdayan exactly, quality control nigtmare
The filament is still molten when it exits the nozzle, so temperature and motor speed aren't the only factors which will affect finished diameter. Any tension as it exits the nozzle will stretch it thinner and any compression will squish it thicker.
If it's going to have consistent results, it's going to need some internal cooling system to solidify the filament at exactly the right diameter _before_ it exits the machine. Otherwise, it's basically just a regular extruder spraying irregular spaghetti into the air.
James Tanner good call this is what I was thinking too, it needs a length where it is allowed to run straight and be cooled. If you look at industrial extruded products like glass (sheet) and Aluminium that is how they work. It may be a good idea to use smooth cooled rollers with a gearing (so both sides are driven) to form the meterial. It is a shame this won't do what we need at the moment, recycled material could be used for infill or support really well on a multi material printer or for some interesting colour combinations.
Thats exactly what I was thinking, if the machine had it's own winding spool, and a device to accurately measure the diameter exiting the nozzle it could maintain consistency by increasing or lowering tension applied by the spool. Maybe a double extrusion method might be required, a rough extrusion onto a spool, then hook that spool up to a standard printer heating head/nozzle and then have automatic increase or decrease of filament disposition speed based on measured diameter going into the hot end. But thats more added complication. I think cheap filament recycling is only doable if you just want low accuracy bulky prints with a wide nozzle. Use a machine like this to make pellets, then set up your printer to work straight from those pellets, it won't be terribly consistent but it will work. Perhaps well if you only want .5 mm accuracy.
I wonder if you could just extrude into a teflon tube of the appropriate inner diameter, so that it's cool when it exits the other end. (Note: OP is not an engineer.)
James Tanner I was thinking Teflon tube too but even though it's sliippy it will still cause drag. Moving wheels, would allow precice clamping to form the right diameter and if they were aluminium with a fan blowing through/over then they would contact cool the filament and then run through tefalon tube to a printer/spooler. There is imho a need for a grinder like that used for meat above the hopper on this, then you could break up the waste pla and then grind it into the hopper. From what Joel found though differently sized place chunks melt at different rates so this grinding action needs to produce uniformly sized pla pellets. There are some good cd shredders that produce uniformly sized output. Someone out there must be reading this and have the ability to make one. If you are there is money to be made here, none of us want to be wasting material.
MarcAFK I've been thinking that pellets may be the way to go too. Couple of things came to mind. 1 the weight added to the print head cannot be too much. 2 measuring of the filiment output can be done with motor back pressure/resistance from the extruder (molten plastic stage) and maybe a gear sensor or an optical motion sensor through something like a tefalon tube. 3 retraction may no longer be available without a dual extrusion setup where one feeds into the other (as you suggested).
I'm not a 3D printer user. I design and develop embedded control systems (circuit boards and firmware to drive them, measuring temperatures and speeds, controlling heaters, driving motors (DC and stepper)). I picked up one comment in particular that suggests several system issues weren't adequately resolved. Quoting from memory (likely not an exact quote):
"the motor seemed to slow down whenever the heater turned on..."
* Why is heater power controlled on-off (seconds on, then seconds off) rather than using PWM ( Pulse Width Modulation - milliseconds on, then msec off, i.e. power cycle is 10Hz-1000Hz). That, and a PID (Proportional, Integral, Differential) control loop would assure a well controlled temperature.
* The motor speed varies. This sounds like the motor speed is operating "open loop". They apply a voltage to the motor, and basically hope that the the speed that results is about right. The problem is, the speed is going to change with small mechanical load changes, and minor applied voltage changes. They should be measuring the motor speed, and controlling power to maintain speed. More PID / PWM control loops.
* The power supply may not actually be able to supply all the power required. They might have gone a little too cheap. Try a higher current power supply (double the current), or separate supplies for the motor and heater.
* Overcurrent trip on the motor? I can't decide if this is to protect the motor from overheating (measure the motor temperature, inadequate motor cooling, or undersized motor), motor driver (needs better motor drivers), the power supply (needs bigger power supply), or the extruder hardware (something breaks). Maybe all of the above...
PID loops/PWM outputs on the motor and heater would go a long way towards controlling those parameters. I'm going to venture a guess that they didn't bother with those functions.
The machine looks cute.... but it appears to be a starting point for additional work.
Your view is actually that felfel developers must have.
I agree with you on all points!
You are absolutely right, and what it all boils down to is thoughtfulness, or rather the lack of, on the part of the product developer.
Touched all the right issues ... if the price point were closer to $100-150 US, it would be worth buying one and tweaking it in to make it work properly. However, for $700 it should be a turnkey solution that works out of the box. I’m a controls guy, too, and had a plastics professor that had a question on the exam “What are polymers?”.
After elaborate responses by all students that got it wrong, the correct answer was ..... “many mers!” .....
20 years ago we automated several of our local pump stations. The number of feedbacks required increased every time you added something for it to control. And all of them had to have a manual override so that if the computer controller took a dump, you could operate it manually. Then just when you think all the bugs are out of the system, along comes a tiny spike at 0200 in the voltage that takes months of chart recordings to finally catch. Because it got past the power supply, the batteries, the opto-isolators, the filter capacitors and everything else that was built into the system controlled by an EMP hardened 386 processor. Addition of another capacitor finally caught the little bugger every time. But hearing an engine running at 1800 going to idle, the gearbox pre-lube pump kick on, the hydraulic clutch disengage, the vacuum valve open dumping the pump vacuum, and the alarms going off finally ended for the most part. Startup failures were even worse as the monitoring of all the parameters added more things that could go wrong. Adding one more required piece of equipment added almost ten more items that had to match to work. Now multiply that time 3 to 5 units at each location, and the requirement that all of them be controllable and monitored remotely over a microwave link. And just for good measure, some of us were also tasked with being able to operate any of the structures ( flood control gates or pump stations ) remotely over an area that stretched from Orlando to the Keys in case there was a headquarters system failure for the gates. Luckily I only had to take care of that one time. Bad thing was, it was on a day when the Area Director decided he needed to visit me. He got rather mad when he saw what I was doing as he was not authorized to even make the changes himself but yet I was ( a total of only about 20 people out of over 2100 employees could ). The whole system was password protected that had a daily change in the password. And none were allowed to give that daily password to someone who was not authorized. Even my supervisor and deputy area director were not allowed to have it so a person 3 levels down had more control they did. Their only way of overriding the system was to actually have someone go to the site and put it in manual control. Without authorization from HQ, that was a firing offense. :-)
So 800$ for a device that ships with no quality control, extrudes correctly about 20% of the time, doesn't put it on a usable spool and takes a month to produce enough filament for a 2 inch print...no thanks
bdub z it's a new product a pioneer later in the future I'll start better ones more reliable one
Its not a new product though, sort of. The Filastruder has been around for a long time now, and works ratehr well, especially if a couple of the options like the motorized spooler and tension detector are used.
While I'm sure these filament recyclers are much more complicated than they seem, how hard can it really be to heat scraps, and use a low geared auger to force them through a die?
@@zybch > how hard can it really be to heat scraps
Probably not hard (I really have no idea). But, to get beyond the "extrudes correctly about 20% of the time, takes a month to produce enough filament" stage? (Presumably that second bit means it out-gases/wastes too much of the input filament in the making process - or it just gunks up the auger slowly that needs to be cleaned once or twice a week). That's probably more difficult (aka expensive).
As expensive as Filabot stuff? I don't know. $750 for the EX-6 extruder screw seems awfully pricey for a (albeit very well machined) screw. (www.filabot.com/collections/filabot-core/products/ex6-extruder-screws).
They would be better off making it completely open source etc and people contributing to it. They are basically charging for a prototype that doesn't work at the moment.
Cnc kitchen has a better unit.
Yes ,making your own filament is a journey. I bought an extruderbot when the company was still around and in all honesty it took me about a year to get to a point that I could produce consistent usable filament. You have to hold the filament at a certain angle as it comes out of the machine and have some way of winding it up slowly so that angle does not change . Walking past the thing will put it off, as the breeze you generate will affect it. It has taken me an age to get to the point where I can make about 3 or 4 kilos a day. The biggest problem I saw was the black 'shards', in my experience all stuff being fed into the machine has to be small enough to be readily accepted by the screw thread. My limit is about 4 mm , anything larger in any dimension will start to cause problems. I recommend getting a filawinder and giving it another go.
I have a huge trash bag of failed pla plastic I am waiting for a reliable filament maker that works easily.
3DSage filastruder?
Come on what a load of rubbish that is all that needed is a truck wiper motor a arcamedeian screw couple of heater elements and a good engineer to know what's needed, that was Chinese crap at its best hoping public will buy ha
Not sure I'm getting this Joel endorcing rubbish my hero telling me Chinese goods are good mmmm
Search for 3devo next 1.0 on Google
How did I endorse this, as at the end I said, "I do not recommend this product"?
You have to set up a rig to wind the filament around a wheel with consistent pressure. You'll also want it to go through water to cool off faster. It's one big piece of the puzzle but there's more to it
You need a way to support the hot filament as it comes out. Id bet the filament is stretching and that is why it is too thin.
That's exactly what I was thinking, there needs to be an extra roller after the plastic comes out of the nozzle to keep the filament moving and consistent.
I had thought that too, but, it's what the manufacturer recommends. Their recommended use is to let it extrude from a table onto the floor.
We have designed a special nozzle that allows to cool down the filament when it comes out. So, as you can see in our video (ua-cam.com/video/JE3up88LEiA/v-deo.html) once found the right temperature, Felfil Evo can work without a spooler or external cooling system, extruding from a table onto the floor.
also needs at least a fan to cool the filament quickly after it leaves the extruder, both things the Filabot and Filastruder have had for years.....
Yeah I'd say weld some support onto a 2 or 3 foot long section of angle iron so that it sits with grove of the V right at nozzle height and prop the whole thing up so that something like a 15 to 25 degree downward slope, I'd bet you could adjust the size of the filament a bit by doing that. I also think that the way Joel was grabbing the filament and moving it around also led to the wiggly and uneven result. Joel, you need to try this again and put some time into making that work. I mean realy, you don't have 6 hours to spend tinkering around with that thing, what do you do with your time?
Gotta say, I wouldn't buy filament that says "Dimensional Accuracy: idk lol"
I use a Filastruder, but a lot of the principles are the same. Extruding filament is a whole separate field from 3D printing, one that you have to learn the process the same as you did with your 3D printer. Think of getting an assembled extruder like this as if you just got an assembled 3D printer, but you don't know about bed adhesion, you don't know about overhangs, part cooling fans, supports, how to clear jams, what temperatures to use for PLA or ABS, etc. Examples of things you pick up over time:
The jagged recycled plastic you put in initially will take longer to extrude than the raw virgin pellets (you can mix recycled WITH virgin plastic, but pure recycled will extrude slow, break easier, and be more out of tolerance). ABS will extrude much easier, quicker, and with a lower tolerance than PLA, so anybody just starting should try the ABS first then move on to PLA. A winder is best (especially with PLA, less needed with ABS), but without one you have to at least extrude hanging over the table enough that the filament starts to coil on it's own on the floor and doesn't hit the table edge on it's way down. Any disturbance in the path of the filament will cause diameter fluctuations. Turning the temperature down will cause motor jams because your higher temperature melted plastic further up the barrel - and now you're cooling it back down while the barrel is filled with plastic. PLA in particular has a very sharp transition - adjusting even in 0.1C increments makes a difference in size.
I use a Filastruder and Filawinder too. Have done for a few years. They work. They are slow and prone to tangles, but they work. Joel's comment that this machine is a start implies that he has not had his eye on the subject. It's not a start. It's been around a long time.
Thanks guys great, you hit the target. Here we are talking about filament extrusion not about 3D printing. Our mistake maybe was to not clarify this HUGE difference.
However, according with your experience, if first steps could be hard, then the process become easier and a good chance to learn about filaments and polymer and let you able to customize your prints.
I don't understand how he thought it would have an easier learning curve than a 3d printer. I doubt there was anything wrong with the extruder but with the user.
I have a filastruder too
And i made my own winder and puller
I use 1.5mm filament(i can tune to make all sizes)
Its really consistent(+/-0.05mm)
Plus the pellets are cheap 3$/kg
Compared to 20$/kg
Its not about ABS or PLA its about the MFI value of the plastic you use
Thats why some people get bad print with some brand and good print with other brand
They need to tune it according to the MFI value of the plastic and not by brand
Great comment Neelmadhav Dookhit, nobody ever talk about MFI ( Melting Flow Index), probably cause is not so easy to explain, but it is also one of the main aspect that influences polymer behavior, actually this parameter is used to define extrusion grade material.
The manufacturer says to let the filament extrude off the side of the table and allow at least a meter so gravity can help the process along. Seems the motor is under powered so it kept catching on the table. The manufacturer seems to suggest 7 rpms and 175 to 200 heat range depending on plastic... my biggest gripe is a machine like this should not be sold for over 300 bucks when most people buying 3d printers can get a decent low end machine for 200 shipping included!
The home made ones I've seen extrude into a water bath or a fan shroud to cool the filament extremely quickly. Think you could probably extrude vertically down for better results as well
Joel....I used to work in a PVC pipe factory where we made everything form under sink size pipe to 15 inch sch80 PVC. I was a line tender I adjusted certain settings like extruder speed, internal air pressure for controlling inner diameter and one of the ways we controlled wall thikeness was puller speed. Speeding up puller speed reduced wall thickness, slower speeds increased thickness. An inconsistency in pulling the filement from the extruder will explain the inconsistency you were getting.
James Smith, do you think that he would have better results if the machine is placed on the edge of the table and let gravity pull the filament all the way to the ground, so it will always have the same tension in the extruder?
Well...gravity is a constant...but then every 4 ft or so would be an issue.. a puller that would give a constant tension would be the best idea. After the puller a driven wind up spool.
Exactly this.
Without controlling the output, every time you move it, you’ll change the diameter and the angle.
I work in the textile industry and we take nylon 6 and 6.6 and turn it into carpet yarn. I have basically unlimited access to nylon 6 and 6.6 in pellet form, we also make our own colored pellets they are extruded drawn through a water bath then chopped. Afterwards the pellets are mixed with natural color 6 or 6.6 depends on the order and make colored yarn. Recently we've gotten into the 3d printing business and now we make colored filament.
Imo it needs a motorized spool winder at the very least. A rats nest of loose tangly filment isn't that useful. Too bad it doesn't seem to create any useful results in regards to extrusion. There's a good reason why the real deal machines take up an entire room or two.
Yeah, like a lego-kind motorized spool winder?
Hi Pile of carbon, industrial filament extruders works this way but we decided to use a different philosophy: a small tool that could work by itself extruding from the edge of a table to the floor, and we use it for production. (ua-cam.com/video/JE3up88LEiA/v-deo.html)
By the way we are developing a spooler that will make easier the extrusion process.
It doesn't turn into a rats nest and I told them the same thing when I got mine and the last I heard they haven't released the kit to the US.
Demeteus - agree, mine didn't turn into a rats nest on the floor.
I had a length of it that went over the edge of some stairs testing to see what I would end up with. Pics: twitter.com/Deneteus/status/993734117748609029
You didn't say but I ASSUME you know small changes in temp will take loads of time to stabilize throughout the unit.
I agree with Proto Pasta below, it is an involved process and a lot of work goes into set up and QC over time. A home box as you demoed is cool and should get many to start scratching their heads to work out an affordable, yet reliable unit.
Maybe for now, store your old wasted PLA for a solution down the road and it does not materialize maybe one of the filament co's will offer a small price for used PLA. Maybe they can shred and re-use it. Perhaps a credit towards new filament. Say $1 per pound credit and they pay the shipping???
Another thing this type of unit is woefully lacking is a take-up system. As the filament is coming out of the extruder it needs to be gently pulled and wound on a take-up reel. Lots of adjustments but once set, the pull at the right speed should stabilize the diameter. HOWEVER, as the take-up bundle grows in diameter the speed will have to be adjusted accordingly to maintain the same pull in Newton's. This could be done, probably, with an arduino and if not, a raspberry pie.
One other concern is temperature variations. Try this in a garage whee it's 60F in the AM and 90F in the afternoon and expect headaches.
And ONLY GOD and the filament pro's, know how much humidity will play into this. Here in the Sunny South, out humidity can be 90% plus and you folks in the desert areas may have less than 10% H.
I grew up in the 60's and remember my sister having an Easy Bake oven. Basically a lightbulb in a box and you could gently cook SMALL amounts of cookie doe or baking mix. Your reviewed item reminds me, in a way, of that toy oven. She did have this oven but it didn't make her a gourmet cook.
You do have a neat novelty item and like I mentioned, hopefully, it will encourage some head scratching and innovation...
According to their website the price of a fully assembled version is $860.00 way over priced. For that price you would think that they would at least provide a spool holder.
Ya you can make your own for much cheaper
In any plastic extrusion process you have to make some constant tension in the wire to control the width of it and improve the mechanical properties of the wire. I would recommend you to make a wheel that rotates at constant speed so it takes the wire as it comes out.
So 700 Euros for something that had a wire disconnected and it doesn't work now lets complain about Chinese workmanship.
To be fair it's Italian, so it's pretty first, functional second...
LOL I know that is the problem Italian, looks expensive works like crap. Joel knows how to build and setup a 3d printer so this should be much easier just 2 settings.
I'm italian... and honestly i feel ashamed that piece of crap pretend to be italian too. Not even mentioning the disconnected cable.
700€? I can build Lyman Filament Extruder V6 with that money.
@Daniele
I've worked a lot with Italian companies and my comment was unfair and meant as tongue in cheek. OTOH Italian manufacturers do spend a lot more time considering the aesthetics of their products than others. When it works it's brilliant (ALESSI) and when it doesn't we like to make fun of it (ALFA ROMEO) 😜
dont be so fast with your judgement with italians, they had build Lamborghini and Ferrari ;D
I bought one too and his experience is right on with mine. I do less 3D printing and more experimenting with different materials to be extruded so it’s fine for me. If I had to exclude material to use printer, I’d try a different machine!
If the shaved chips messed it up, then how in the world can someone turn their old prints into "dust" and not run into the same problem? I think this is a good idea, but it needs to be a plug and play. The person shouldn't have to question the parts, dust, etc.
I haven't gotten my first printer yet so I don't have a bag of bad prints yet. But I'm keeping an eye our on these things because I know it will be a problem.
www.3dhubs.com/talk/thread/which-shredding-and-extruding-machines-are-available-part-3
i would use something like a bullet blender. can literally turn any plastic into shards, beads, then grains, then dust in a matter of about 30 seconds.
Chrismofer OK but this needs to be plug and play for it to actually sell. Their own material didn't come out right.
You shouldn't have to spend so much and then worry about stupid stuff.
what do you mean plug and play? do you expect to be able to put an entire footlong failed print in there sideways?
In reality the different sizes melts at different temps and pressures which is why you can run into problems.
We had a very similar filament extruder at the local Hackerspace where I live in Germany, since a few years. It works reasonably if you let the filament fall from a specific height, so gravity keeps the extruded filament under tension etc. (I doubt you'll be able to get any consistent thickness the way you're trying to), we mainly tried extruding ABS though. While it's certainly possible we never really use the filament extruder, since PLA/ABS pellets aren't really cheap enough to justify the hassle (it's obviously not a labor free process and as you already noticed it isn't tremendously fast either) and grinding up used PLA or ABS always comes with the risk of getting impurities into the filament and clogging your printers nozzle.
So yeah... It's a pretty neat idea and all, but with the simple "DIY" filament extruders I've seen in the last couple of years, I'd say we're not quite there yet.
Thanks for your honesty on this and all your reviews.
I actually picked up a Felfil Evo with their filament winder and so far I'm actually pretty satisfied. Within 2 hours of UPS dropping the package off I was able to figure the machine out, set it up, and extrude some of the included virgin PETG pellets into some very nice filament that I proceeded to print out a test cube on my Prusa MK3S. I haven't played around with it a whole lot yet but so far so good for me, it seems like the filament spooler is crucial for it do have a consistent diameter, which on my first run was 1.75+/-5mm which is honestly kinda impressive to me.
How much did they pay you to say this
With a QC like this in assembly, I hope the rest of the casing is fireproof.
I have done this before and got 3 full spools of good usable ABS out of a similar product called Filastruder V2.0, this product also uses a photo resistor to keep a specific amount of tension in the filament and spools it. the key is to spool the filament as soon as it cools.
Joel! I've been holding this in for so long. Can we please have those holes filled in with Spackle? I don't even care if they are painted. Those holes, are they there on purpose to cause epileptic seizures? For the love of your viewers it really only takes a few moments to fill in those holes. I use an old butter knife and then come back with a damp cloth or sponge or even just a finger to smooth it out.
Willie - for you, my friend, I will make it happen this week.
Thanks Joel, was pretty much going to 'ask' the same. :)
Oh man, my "whatever it is" will thank you! :D I was a bit overly harsh, but I think I may have snapped. Those holes, those HOLES!!!! LOL You are the best!
I highly recommend Wrigleys spearmint for hole filling - chew until flavor is gone & fill hole. Sands nice once it hardens....
Love it :)
keep in mind you also "need" (you can without, but it's a pita) a filament spooler, and if you want to recycle plastic, a shredder.
precious plastics have some nice plans for an extruder and a shredder, but not a spooler.
to measure the size, the classic way is one fixed bearing, and one bearing on the end of an arm with a spring pressing on a load cell. Though you could probably get away with using an led, a cheap camera, and a pi, measuring the luminosity of the camera gives a fairly accurate size measurement.
They should have added an optical mic and controlled the temp or speed with it via PID loop. You can't ask a user to sit there for hours and constantly measure\adjust. That's not value added - at all.
A well designed machine doesn't need that. There are other filament extruders that run open loop without issues. But they cool the filament, don't have underpowered motors and power supplies and have a motorized winder.
You would need a mini granulator to create consistent size in the recycled plastic (or only use virgin material), in order to give more consistent output.
As I work in the plastics industry, I suggest keep the temperature high, and increase the RPM to adjust the size, you always have to have sufficient temperature to give a melt flow. Something you should consider is running the filament through a water bath to cool it as soon as it leaves the extruder, to reduce the stretch on the material don’t allow it to sag or be pulled as it leaves the die.
Holy moly! How bad can your Quality Control be that a tight screwed wire can get loose? Booo Felfil!
IgnoreMyChan It even has a ferrule on it! That shouldn't fall out.
And it took for ever to extrude I like how he said it's well built then realized he was wrong
Ask most of the 3D printer manufacturers that same question. There isn't one out there that hasn't had QC issues.
Especially for 700+ bucks...
Wait this is 700+ USD??
it works, you need to have it sit on the edge of a table and a slight tilt. for it to slowly cool on its own as it comes out.
I’ve been saving all the scraps since I started so this would be great
me too
Lukas if it did work right. It’s a real good idea and definitely goin in the right direction
You’d think but it’s 800$
Sideways Gamer get a filastruder instead cheaper and works
You'll need to make pellets first. It's not worth it. The machine costs plus the wasted time and energy vs the filament cost...even working in China at their wages it'd be more cost/time effective to just buy the stuff.
I didn't read all the 950+ comments, but I would suggest few modifications to get a decent filament extruder:
1. Instal a larger nozzle
2. Put the outlet on a vertical plan (instead of horizontal, yeah, the hoper will need to be modified as well)
3. Have a guide loop for the filament and a automatic spoiling
4. Get an inline size gauge and link it to the extruder AND the automatic spoiling speed controler
Well...
5. Nevermind step 1-2-3-4: Just buy regular filament!
I've never laughed so hard watching a 3D printer review until this one.
To bad it doesn't work. There are at least 20 electrical engineering flaws I can see in to design let alone mechanical.
Your video production was the best part about this video.
I have started my own filament recycling program to convert waste into usable filament in South Australia. There are at least 8 filament extruders I have found, Felfil is the second cheapest I found, the most expensive is upwards of of $5,000 US for an all in one unit. I agree with your review, they're a good starting point to prove the possibilities, but there are many drawback as you suggest, and I believe most people wouldn't have the time or experience to produce their own filament without significant investment.
I funded their Kickstarted around five years ago, and they shipped late and my kit didn't work. I'm still trying to fix it, and I've spent over $100 trying to get it to work. Not cool! I'd love to get some use out of it!
Oh no! What isn't working on your kit?
Thermal runaway - keeps overheating. I've been emailing support for a LONG time about it, and they've sent me two new boards (one I had to pay for) and a handful of other electronics like a new thermistor and heater cartridges. Really disheartening to spend so much time and money on something that, even when fully assembled like yours, doesn't do its job :( Hopefully someday it'll be so easy to recycle 3D prints that it's a no-brainer!
What firmware version are you using?
Deneteus Sorry - I wouldn't be able to tell you. I just received a new board from the Felfil guys which was supposedly tested, and it didn't work (even with all new electronics for everything else). They sent me instructions on how to flash new firmware using an Arduino, but my computer wont write sketches to it. If they actually did test it, it should have the newest firmware.....right?
Considering there isn't alot to the FEVO I would expect they sent you a new tested board. If they did send you a board that had the latest firmware I would make sure the XBOX 360 PSU they are using isn't the culprit. If you can't use an arduino then you should try using a USBASP programmer from eBay to do the job. I just verified that there is a port to the right side of the board when looking at it from the rear. Here is the official programmer. All the rest are clones. I use this to program all the 3D printer boards. www.ebay.com/itm/Official-product-USBasp-programmer-red/323209529275?hash=item4b40ca07bb:g:fjoAAOSwQaJXSIGT
In the injection molding industry, which augers the material through a heated tube and tip, we call the “cleaning pellets” purging material. It’s typically clear, I assume they are using recycled scrap with this.
It is pronounced "acrylonitrile butadiene styrene."
Yeah, words and I aren't good friends sometimes :)
We can be friends Joel!
could you please repeat that?
Maybe next time we can send to you “polyethylene terephthalate glycol-modified” aka PETG 😉
Felfil, maybe next time you can send a machine that works and is not a 700$ scam.
I think that the one major problem is that the filament is gravity pulled, so it stretches unevenly after it exits the nozzle, If it was extruding verticaly down from height it may get more consistent size, or maybe some mechanism that pulls the filament out of the nozzle with constant speed and allow it to cool under constant tention would do the trick.
I get that this machine appears to be a disaster. But, to be fair, it was frustrating that after you initially said “first..." you use the black shards to flush out / clean the new machine. That you then went on to use the output of this cleaning material to assess the performance. Even measuring the width of the cleaning material output! Huh?
Well at least near the end of the video you did start to use the proper pellets. :-)
Fair point :)
Turns out the black shards of PLA weren't in fact a "cleaning" filament. I read wrong and just the first few meters of filament were considered cleaning. The black pla was suppose to produce good results
Thanks for clarifying. Your findings defiantly suggests this product isn’t really fit for sale, as it doesn’t appear to live up to it’s advertised claims!
It would be really interesting to hear from customers who've bought the product in good faith of being able to: "make custom 3D printing filaments at home”, and: "save over 80% of the price of filaments spools”.
I guess felfil should at least amend their advertising, to add: “YMMV”. LOL
Well, also to be fair, I would want such a device to re-use grinded stuff back into filament. Exactly like that cleaning stuff. So does that actually work, if it only works nice with perfect pellets than it's not about recycling anymore.
It's better to build a Lyman extruder that can extrude almost anything than it is to throw regrind into an extruder with a windshield wiper motor. ua-cam.com/video/H6h1yDzoPMU/v-deo.html
I'm not sure what the "purge" time is on this machine, but on the Filastruder, it takes almost 6 hours to "clean" the machine out... My extrusions looked terrible for almost 2kg of resin.
I heard a lot depends on any pull out speed, or friction showing push out down. Example : more pull out, (has to be perfect all the time), will stretch the hot, just out the nozzle, filament thinner. Another thing with some plastics is moisture
Great production quality on this one Joel 👌 Such a shame about the machine.. I can't wait until the day we can actually reuse that waste!
If you really need a 3D-Filament Extruder then i can recomend this guy. He offers a DIY Mini-Extruder which can make 1KG 3D-Filament in 1 hour. Even PA12 and other 3D-Filaments you can do with this thing. I really can recomend it! ua-cam.com/video/44oZ7TSlCx4/v-deo.html
Hello dude
have you ever seen 3devo's machine? It can also extrude up to 1kg/hour and is an all in one desktop extruder. That extruder works very well and is really reliable. You can also make a lot of different filaments with it. Just look it up it's really cool.
Maybe you could make a video where you try to make one of these machines . It’s like the space race just with a filament recycler!
(Btw in a huge fan)
Zion: Well, ironically, the machines are actually even more for making new filaments, out of new plastics. It's a great way to experiment with new plastics. The recycling part is definitely something you can do, but it's questionable how interesting it is to actually do that, because, while (for example) spools of PLA filament is pretty expensive, just PLA granulate (little blobs of pla, which you can throw in these extruders) are really cheap. So recycling is probably not worth it. But you should look up the 3devo machine on the internet or youtube, it's really cool! www.3devo.com
CNCkitchen got a good video about recycling scrap prints & pellets. You need a winder unit that starts and stops via photosensor. Let the extruder piss the filament down approx. 1-2 ft to a loop, and wind it up. You need some time to try out until you get the right diameter, but then it works like a charm.
"Tinker" = Fix manufacturer problems.
Cheap DC motor with gearbox and no feedback (closed loop). Motor takes strain, drop speed, feed slows down, inconsistent feed.
No cooling or support after extrusion.
No winding.
This really is a POS.
Can't believe a company spend money trying to design something cheap and then charge a lot for it.
Nice video as always, Joel. I'm sure you had to bite your tongue a few times.
Saran for the making of Saran Wrap is a tube that leaves the extrude nozzle and goes directly into a bath of water and is taken out of the water before it cools down too much. From there it is pinched close and then blown to a tube where it cools fully and makes a solid.
The use of water would simplify the process. the tank would be continuously being filled and over flow the heated water at the surface. If sufficient heat remains in the filament the film of water on the surface will evaporate before the take up occurs. So a driven pulley will be needed to remove the filament from the bath a a constant speed. The height of the nozzle to the water will define the diameter of the filament.
And the take up mechanism can be very simple. An idler pulley that drives a variable diameter V belt pulley or as we used in the R & D two V belts with the variable pulleys where they transfer motion. Thus as the idler falls, more filament to wind, the take up drum increases in speed and as the idler rises, less filament to wind, the take up drum slows up.
You can print this three piece idler V belt fixed side cone on both end and a middle cone on both sides that slides on the shaft. You may print the shaft as a spline that you see on drill presses to have all faces of the cone carry torque.
So the V-belt pulley of the motor and the take up drum are fixed. The three piece V-belt idler moves closer or farther from the take up fixed pulley. the diameter of the two V-belts sum at the three piece V-belt idler is fixed but individually change. Speed changes . The idler measuring the loop of filament moves the three piece V-belt idler assembly.
A thought. The take up device above could eliminate the fixed speed pulley removing the filament from the water bath.
I’ve been saving my failed prints ever since I started 3D printing just Incase I ever get one of those.
Maybe the diameter of the filament was varying because you were pulling it out with varying force? It's kind of like a sticky stuff, if you pull it hard it it gets narrow. So maybe it will be a good idea for Felfil to introduce some sort of rolling machine which can put the filaments coming out in roll, and retains proper tension on the filament as it comes out?
Now that it's a year after your initial trial of this machine, I'm curious if you've tried a newer version of this or a similar machine. Are engineers any closer to making printing your own filament a reality?
even when this video came out there were machines available for this, Now there are stl files to print your own parts and build an effective filament maker out of an old 3d printer!
welcome to the world of plastic injection molding. we have a lot of variables that went into plastic injection molding. barrel temp, hopper feed, cooling time, cooling temp, injection speed.... and one problem could have one or multiple sources. my concern with this is it should have an addition nozzle for cooling the plastic down before exiting the machine. so you can get a consistent solid piece that isn't like wet spaghetti. so the heating barrel melts the plastic and then using the screw to feed more plastic in allows the plastic to be pushed through the heated barrel into a cooling barrel that would taper as is exits. that way it wont get bound up from friction but could get a relatively straight piece to form before exiting the machine onto a spool preferably motor driven.
well, that was disappointing
What "proper supporting hardware" was he supposed to use?
This machine is supposed to be a stand-alone out of the box working item.
The manufacturer himself advises to let the filament fall to the ground. He respected all recommendations from the manufacturer and did corrections as they advised after he emailed them about his issues and still failed to come close to the advertised result.
So please tell us what the manufacturer forgot to tell Joel to get it to work.
So you admit the manufacturer is lying about its own product abilities but you put the blame on the reviewer. That's an interesting way of thinking...
I think it needs a better/more powerful extruder but most importantly it needs an extrusion take-off system, that's how you could achieve a consistent diameter. It also needs to be able to run for much longer or extrude at 3 or 4 times the rate.
Him: Polylactic acid
me: *GASP*
him: aka PLA
me: Oh...
Hydric Acid.... waiting for GASP..... aka Water ;o)
@@gavincrouch Gotta watch out for that dihydrogen monoxide. I hear it's corrosive to most metals and deadly if inhaled.
@@qumefox LOL. Classic. ...Que California, demanding one more unenforceable ban...
I'm no plastics expert but it would seem a gravity fighting cooling bath, spooling system and a better extruder with graduated larger to smaller internal extrusion dies finally exiting at the 1.75 size. Also graduated slightly lower temperatures each stage would be the trick here, hotter to start out, barely above melting point at the output tip. You also need an absolute constant speed and pressure at the output tip.
Oh no what a let down that it's not quite up to speed. (puns intended) Hopefully they can improve it from your review
I really hope so. I want an out-of-box experience like what they expect it to be. I think that would be cool.
It's the power supply. When heater comes on the motor slows because the voltage drops. Very easy problem to solve. Thickness changes happen because the filament is hanging and stretching the material. You can put a fan blowing on the filament to cool it. Have fun.
Thanks for your feedback. We explained that the filament should be falling on the floor, not on a table (at least for now - we are working on improvements). It takes some time to use our product properly.
And thankfully I did have the filament attempt to fall to the floor and my results were subpar.
Isn't the filament meant to be extruded off the edge of a table .... without pulling on it or letting the friction of it pushing against the table surface...that's why it's so squiggly
Yes, that's why it started so squiggly. However, at the end when you see the long strands of filament, it was extruded over a table edge.
3D Printing Nerd Thanks for clearing that up... that's such a let down... i really wanted this nice and compact recycler to work at least for future purge blocks and waste etc
reduce the height and add a fan.
Hi Steven, yes Felfil Evo is designed to extrude from a table edge. In this video you can see how it works (ua-cam.com/video/JE3up88LEiA/v-deo.html)
i was imagining a messy pile of PLA waste going in and a nice spool of PLA ready to use... this is a great idea that needs some work, I bet in a couple of years it will be amazing !
Filament not 1.75 should be ok. Any directdrive machine should accept 1.5 for example just adjust the slicer. But it must be consistant. If it cabmake 1kg of 1.5 1kg of 1.6 i would be ok witb that as long as it was consistent.
Outside my expertise but my hunch tells me they should run dual power supplies so the heater kicking did not effect power to other components and i would gear the grap out of the motor so while slow it would have gobs of torque to eliminate the variances in speed.
Dear Nerys, you are right 3D printers could work also with a filament which is a little tiniest than 1.75mm. However extrude with a goal of 1.60mm (so with +-0.07 optimal tolerance 1.53/1.67 mm filament) is a little bit an hardcore solution, if it totally avoid that the filament get stuck into the printer’s hot end on other side it can influence the print results. What we suggested is to work trying to get the proper diameter 1.75mm and correct the settings only if its measure is over 1.80mm.
About our geared motor, it produce a 23Nm torque in full load and, according with our tests it is enough to work with the majority of the 3D print polymers. For who those feel real makers we also provide a basic solution named “Basic Kit” that allow you to full customize the electronics control and alimentation.
Someone designed a filament width detection sensor that you can use with Marlin way back in 2014. A newer version can be found on Tindie. I have not tested this yet but I plan to.
We made some experiment with it! It's a good start but unfortunately it is really susceptible, also making an average of the read values. I'm not telling that is a useless project, just there's still some works to do in order to make it ( or something similar) suitable in our design.
I've built 2 variations of the lyman filament extruders that have been around for about 6 years now and they offer a lot more consistent results compared to this unit and at a fraction of the cost. This just looks to be half a machine without enough real design time put into it, maybe it will get there in the future and at a better price point.
It can also depend on nozzle size and layer height. If it's easier for the melted plastic to flow up and around the cooler filament up in the heat break (into the space that normally would be taken up with filament) rather than be shoved out of the tiny hole in the nozzle then you end up with nasty jams in the heat break from undersized filament flowing upwards.
Filament extrusion consists of much more than just a machine that will chip, melt and push material through a nozzle. The melt exiting the nozzle must be careful controlled in pace while it cools. Allowing it to droop from the tip will induce form and size issues. Typical filament extrusion has carefully controlled cooling baths or sprays depending on the mass of the material being cooled and the rate the filament passes through the cooling zone. They always have a high precision cat or capstan puller drawing the material from the nozzle at a very consistent speed. The extruder is set to a temp and rpm that produces a molten, but still just thicker than liquid melt consistency (tough with PLA) and just larger than the final filament diameter, say 10%. Then the haul off capstan or cat speed is set to stretch the filament to final size (known as Draw Down in the biz) through the cooling system.
Any changes to rpm of the extruder or temp of the extruder need time to achieve steady state thermal conditions. A 20 minute trial likely only provided enough time for two adjustments. Any more than that and thermal conditions never reached a steady state.
If you try this system again with simple air cooling, try to use gravity to create the precision haul off rate. Tilt the unit on a downward angle and support the filament with a short support plate that will provide steady friction and allow the filament to drop to clear air and then don't touch while running. Should work until the filament hits the floor. Then you need something that keeps that end from influencing the nozzle draw down. A take-up spool set enough distance away would work if synced with the extrudate rate and allowing minimum tension past the support plate.
If you don't want to have another go at it, send it to me and I'll give it a shot 😁
For 719,00€ I will buy Proto-pasta instead. One of you best videos yet. Never laughed so much.
Indeed, btw the components don't look 719 Euro's worth to me in the slightest. I guess because they will not sell a lot it has to be this expensive.
bruh I expected it to cost like 150 euro max but 700... it's definitely not worth that price
I have used filastruder. Requires atleast 45 degree angle, and for PLA a cooling fan. Also every time you touch the filament while extruding it will cause inconsistensies, so once you get it running, do not even touch it. Also in practice a winding machine is also required, and for consistency extrude straight down. You can then get consistency down to around 0.01mm.
Hi, good point - we are developing a winder to achieve easier extrusion and more consistency. Hopefully will be ready soon
@@Felfil the filawinder is open source, just use that with your mentality attached to it, ie. prebuilt, packaging, support, materials etc.
Works really good out of the box.
@@skaltura we are trying to use a different solution, which allows to calibrate the filament diameter automatically. With our mentality, of course:)
@@Felfil Cool! You checked out the Lyman stuff? That is closed loop industrial style, but open source and desktop form factor. It pulls the filament out and measures it is correct size etc.
I should look into that again, might be fun build!
But for me, it is more about experimenting, and i quickly found out that to experiment with stuff, i need extra barrels etc. as once it gets gunked up with some weird material, it is very hard to get out. I got a reaally bad batch of Nylon :( Further going back down on temperature range is a lot of work.
With closed loop doing higher temp through bigger nozzle one could experiment with alloys and filler materials ... ;)
I've been meaning to do some experiments with graphene, got a few kg of big flake (upto 44micron, median somewhere around 20micron) size 80%+ single layer graphene waiting for the right time. PA6 + Graphene 1% + GF50 might be veeeery interesting mix.
@@skaltura Yes we know about Lyman, it's a very nice project! As for materials, we extruded successfully a filament made of nylon PA12 only and it was good for a normal printer, while we never tested graphene.
You referenced Austin Powers, take my subscription sir.
Thank YOU, good sir.
you gotta love this guy. He makes a joke about a zamboni and then laughs himself to closed eyes. I enjoy this kind of company early in the morning.
You’re easily my favorite person right now.
Having worked on the prototype production of masterbatch for use in plastics manufacturing I'll tell you now that it'll be a while yet before you can make s machine powerful enough to recycle old prints that fits on a desktop and does it's job reliably, plus you need to use a bath of water at the end of the extruder to float the extruded plastic then use an air line to blow off any remaining moisture (Or constantly replace tissue )
Edit: on the subject of consistency, recycling plastics into an extruder is not an easy thing to make consistent, it's not easy at all, and you may need to adjust settings on the fly....If interested in any additional knowledge on plastics extrusion message me, I've got a few years of experience working with pretty much every plastic and polymer that's on the market
peter geary let's try to develop one that works, patent it, and make a Kickstarter for it!
Jeffrey Moffitt as cool as that would be the costs to produce such a product would be massive, the issue is that currently the torque and power required plus the knowledge required to operate it consistently means that you'd have a very small target of actual buyers
I drew up a design that takes the torque factor out with a spring system. Imagine a vertical hot glue gun with a cooling tank below it with a hole in it leading to a spool
Well there is already a very reliable consistent desktop machine on the market. It's the 3devo extruder. Works really well, and you can throw all sorts of plastic in it to make filament from. Just look it up on the internet. Only con is that it's a little expensive, but it is a consistent desktop extruder on the market
Do you happen to know what the strongest formulation of PLA that I can order? I switched to PETG but I always wondered if there was an alternative.
Most working DIY filament extruders use a spool drive system that is synced to the extruder and that is supposedly where the consistency is derived from because it keeps a constant tension on the extrusion while it cools and winds on to the spool. you can't make filament by just extruding it. you need other parts to make this work well. Also, the best ones uses much longer and bigger heaters and augers.
Id love a filament exturder... if they wont so bloomin expensive lol
They would be the kind of thing that would be nice to have as say, a 3d printing group, than an individual.
I've suggested it to my local hackspace, but they say its not worth the cost... so I'm slowly getting all the stuff I need to build one and show them. but its an expensive process
theres this one www.filastruder.com/products/filastruder-kit?variant=338126174 Its a kit but it works better than this one and its cheaper, also i have not bought it, i have only seen videos on it
Filastruder is the only affordable one that actually works
I'd bet this one would work fine as well with the correct settings, that Filastruder looks like a good deal though.
What iv Seen In industrial is a bigger nozzle and spool winder. They get the temp and extrusion set and the spool winder stretches it to size before being cooled at a certain rate.
Infinity War was awesome, but to quote a friend of mine who saw it too, "I am not ok!"
A few suggestions:
1. For bigger filament, try to increase the capacity of the heating element.
2. If the setting of the temperature changes the motor speed, try to use two independent power supplies for the heater and motor
3. Let the extruded filament pass through running water for uniform cooling. Make a channel of say, one meter long through which the filament is drawn out.
4. Change the size of the nozzle (drill a slightly bigger hole if you use more powerful heater)
Hope these modifications can achieve in getting a filament size you are looking for.
Hi, thanks for your feedback. Our engineers are already working to improve the extrusion;)
A shame this is the quality you get for the money and premise.
To make better filaments you need a larger heating tube and then a water trough and a powered wheel that loads the filament onto a reel and the size is adjusted by the loading wheel speed
Extortionately over priced and unfit to do the job required. you can build yourself a core xy and have enough for 23kg of filament .... this product is a joke.
I don't know much about making filament but I'm pretty sure you need some kind of extraction&cooling system after material exits the nozzle.
So like additional, precisely tuned stepper motor pulling it through water which cools it.
Felfil Evo doesn't need a cooler system : ua-cam.com/video/KA3Kf22BsZ4/v-deo.html watch this video
You look like markiplier mixed with tony stark
I think this extruder is just a portion of a final homemade filament factory. A critical component to be added to the system is a feedback loop to control the filament width and a mechanism to cool the filament down as it extrudes and to collect it on a spool.
Hi Rooholla, this is exactly what we did, we developed a winder. Anyway the extruder can work alone, if used in the right way.
Oh my god. "Dust in the wind" was freaking hilarious
What you need is a cooling and spooling system. Without consistent cooling and spooling the diameter varies by irregular stretching.
.... so I can spend like $700 USD, for something that might work? ...Lol
just like the 3d printers really.
I think for a home system without GMPs you would want to extrude the filament under pressure at like 2mm, cool it, them pass it through a cutter than trims it to 1.75mm. That seems a lot less error prone and the trimmings can be recycled again. Relying on the extruder and gravity to size the filament is asking for a home user to mess up.
"across the pond" = anywhere else :D
I have a rule never to correct someone else's prototype which is what you would end up having to do to get it to work. If they are paying then no probs but that unit looks like a Graduates first attempt. Nice idea but it needs to be consistent otherwise why even bother. I think you gave it a fair and honest review and a big thank you for that. :)
8:40 Did.... did you just mix Austin Powers and Deadpool?????
EDIT, LOL, nevermind, you caught it post.
I would like a 3D printer that accepts pellets. A complementary grinder, too. Not only failed prints and supports, but also any household plastic (bottles, TV shells (e-waste recycle the innards), broken toys, even plastics not accepted by the local recycle co. can be ground up and used as 3D printer fodder.
Pellet 3D printers are a thing, but the tech still needs refining.
How COULD that even work? I mean.. No cooling, no spooling, no measuring, filament just bundles up and tangles in front of the machine, hindering flow of new filament.. Meh.
Also 1 meter per minute is an abysmal speed. 4 hours max operation time means 240 meters of filament over that 4 hour period max. Meh.
240 meters is quite a bit, 1kg is 330m.
rdizzy1 yeah it's not nothing but that means you can't even produce a normal 1kg roll of filament before you HAVE TO turn off the machine. :/
Yeah it is to cool down the machine, only have to turn it off for a few hours and then back on, could potentially make 3kgs per day or so. Also if you watch the owner use this machine he says it is supposed to be used on the edge of a table, then filament will then spool up in a spiral naturally on the floor or wherever it is landing due to gravity, this will also even out the inconsistencies in the width of the filament as well. The dude in the video got it to 1.74 for instance.
I tried to let it spool on the floor - that's what the big bunch of filament towards the end was from. However, that did not work. There was no consistency, and the machine would shut down every few minutes for protection.
I think you need a buffer ahead of the machine to preheat the plastic to close to the melt zone to allow a consistent supply of melted plastic
This thing looks insanely awesome, at least in concept, being able to recycle old print material, just got a 3d printer a couple days ago, still not set up properly unfortunately, but I love the idea of near zero waste.
Guessing it's not going to work too well, at least not without loads of practice and tweaking. Would be awesome to see the price come down over time though. For a couple hundred bucks it would be awesome.
I'm sure things like ambient temperature and how the end is handled play a decent part in how well the filament maintains its shape.
edit: man, that was pretty disappointing, and having to run it over 5 and a half hours for the equivalent of a 1kg roll, while only being able to run for 2 hours at a time is pretty rough.
Also, if it has trouble with the chips they sent, I can only imagine how bad it would do with the results from a home shredder setup unless it's really keyed in.
Hello Jesse, if you have troubles while extruding please contact us, we can find a solution together!
After seeing you struggle with sizing , I thought I will share with you the solution, most extruders have downstream equipment ,cooling trough water or air cooled, where filament get cooled as it leaves the extruder. secondly, puller (Haul off) which could simply be two rollers driven to match the speed of extruder output which pulls the filament at a constant speed. Then the spool winder, The puller speed controls the diameter which is usually tiny bit faster that extruder to take up the slack and size the filament to exact size.
My experience 1974 to 1993 , producing filament of 1.5 mm to Pipes up to 500mm. all had same set up, except pipes are sized and cooled in vacuum tanks.
This would be really cool... only issues I have is:
> the amount of time it takes to actually prepare the unit for use
> It takes a damn age to extrude the filament (1 meter per minute!? TOO SLOW!)
> *IT'S GOD DAMN EXPENSIVE JUST FOR THE EXTRUDER KIT!! 300 GOD DAMN EUROS!!*
> The amount of headache you get trying to actually get consistency with the width of the filament just doesn't seem worth it for the cost.
Also, I would like to see a carrier that moves at the same rate as the extruder so that you can wind the filament straight onto a reel ready for use with a 3D printer. This is a cool project and I personally think, as an outsider that this is potentially a really useful tool for saving your excess filament from becoming fodder *IF* they can get it to extrude at a consistent rate.
For it to be good it needs a self spooling mechanism and something to let the strand cool as it comes out, so it's own weight won't pull the filament to a thinner diameter in some areas.
I've worked on big extruders and yes they are way faster but they have cooling tanks to control temp on the output side.. the problem I thought of is there is no way to uniformly grind your scrap..good video though and cool little machine
Check and see what the amps are on this unit and what it can put out. The reason is that consistently the voltage and amps have to work hand and hand to produce product especially when extruding. There theory is true heating it first,then extrude product. But if your amps aren't high enough and stable it's very hard to keep your temperature true. That's why the temperature isn't consistent. It may need a bigger motor also. If you can check the amps and temp.you will see the difference I'm talking about the two. Good luck!
The first thing I noticed is there is no draw from the extruder. the filament needs to be drawn out of the extruder (at a rate consistent with the input speed, temperature, and the size needed).
I remember some time back seeing a video from a filament manufacture, where the engineer said that they use 3mm nozzle for 3mm and 1.75, the primary difference is the rate the filament is drawn away from the nozzle.
Since this system has no drawing system of pinch rollers it is got no chance of working. Excepting with a very steady human hand. It may be possible to use a gravity based system but that is also going be tricky.
My choice would be draw downward with a set of pinch rollers, and have an optical measurement device as close to the extruder as can be managed feeding back so the temperature, auger speed, and draw rate can be adjusted real time.
This, in it's simplest form, is the same principle as plastic injection molding. The cleaning or Purge material looked really rough probably because it was used once. If it was regrind which is used material, then ground up , that throws off the chemical structure of the compound and doesn't melt correctly. Mixing it with virgin material at a 25% regrind and 75% virgin may have produced better results. Also the large chunks of regrind most likely didn't melt as fast as the smaller ones causing a less flowing consistency.
Looks like it needs a speed controller with rpm feedback from the motor. The current draw from the heating elements is messing with the speed. Just slap a vfd driven 30 amp 3 phase motor in there and your problems will be solved. Also needs a good take up winder to keep consistent tension on the extruded plastic. That is affecting the thickness as well.
You need a stepper motor about 30cm away pulling on the filament as it comes out along with cooling fans or water between the two. Tuning the stepper motor speed should make the filament more consistent in size. At the moment it's bunching up while still molten causing it to change diameter
reducing the temp will make the plastic extrude slower, since it needs more pressure from the auger to make it's way out. this increased pressure is what i think increases the diameter. technically slowing or speeding the auger should do that too.
I think that one of the most important part is the roll and cool part once it's extruded. That's not a trivial part and you are just letting it go over your table so the friction is also slowing the filament out of the extruder and creating inconsistence.
Hi, a spool can help but it's not necessary. In this video you can see how Felfil Evo works and maintain the filament constance: ua-cam.com/video/KA3Kf22BsZ4/v-deo.html
The thing needs some kinda motorized take-up/cooling system.
Maintain consistent 'draw' on the filament while it is still oozy, cool it before winding onto a reel.
Too thick? Speed up the draw a tiny bit to stretch the filament.
Too thin? Get a bigger nozzle!
You are right: this is a great tinker-toy!
Cool idea, needs a complete redesign though, especially considering the price. Has a brushed motor which would be way better brushless and some way to auto spool the new filament. I don't think another device that requires tinkering on top of the hassle of 3d printing is a winner.