Thanks everyone for 10K subs!!! I hope this is a satisfactory conclusion to this series, I learned so much about CNC and 3D printing over the past two years, if I was able to pass on just a fraction of that in the videos it was well worth it. I'm already designing a much better (and much more conventional) printer that can be printed by this one in true reprap fashion!
*fireworks and upbeat music* "thanks for 10k subs, my parents are getting divorced" is WILD. I'm sorry that you're going through something so stressful and life changing, but that delivery was so outta pocket that it had me hysterical
Lmao, I didn't even realize the irony of that intro until you mentioned it just now! My parent's shit has been going on for longer than this channel exists, I didn't think anything of it. Thanks for the kind words, this comment made my day 💛
This so reminds me of the time when I was building my first printer, a Mendel 90, In the engineering company I worked for I used to go down to the workshops in my lunch hour and run my designs on the printers they had down there, one day one of the chaps showed me his DIY printer, I was amazed, printers back then were around the £20,000 mark for the cheapest models and so having one at home of my very own was a pipe dream, I ordered my kit that same afternoon and printed my plastic parts at work, it took some fernangling and fine tuning to get it to print correctly but my friend had already done the hard work when he built his so mine was relatively straight forward, my frame was built from acrylic but my friend designed some aluminium reinforcements and cut me a set also, he later made all the parts from highly polished CNC'd aluminium and made me some of the fixing block parts also to replace the plastic ones. I often think of reassembling my Mendel 90 as it is currently in bits but stored in an old synthesiser flight case.
Yep, watching this video reminds me of how much the 2012ish era printers were a nonstop faff-fest. Always burning out relatively expensive parts, forums with a million red herrings to chase, outright bad advice sometimes, and prints that were only vaguely okay if you stood 5ft back. The amount of institutional knowledge baked into even the worst cheapest disposable printers of today makes the best FDM hobby printers of the early 2010s look like hot trash.
@@carpespasm I had a good giggle reading your comment, you hit the nail on the head. Luckily my forum was the lunch table with some chaps that had 40 years or so engineering experience and 25 years of industrial 3D printing experience, what they didn't know could be written on the back of a stamp, within a year or so of the first Mendel 90 (ish) appearing In the workshop half the employees of the workshop had one, they were great days!
Most diy 3d printer videos: I built a voron for fast printing Chronic Mechatronic: I built the world's slowest, most unreliable printer And I for sure know which one was more interesting to watch.
I sure know which series I binge watched in entirety over 3 days. I would probably binge the rest of the channel too if I wasn't running out of vacation...
This has been an amazing series displaying perseverance, dedication, and ingenuity. I learned a lot watching you tackle problems from different angles and posting your successes and failures. I see good things in your future. Well done !
I'm so glad everyone ended up enjoying the series, I almost called it off entirely after nobody seemed to care in the beginning (my mental health wasn't the best at the time...) - so glad I didn't!
Hi Ben, for me it's a complete achievement. I've been wondering how that "wood" and cheap 3D printer could output a decent cube or the ubiquitous benchy well. You just proved it can! That series was exciting and full of suspense. Thank you for sharing such a valuable experience Ben. Have a great day despite your familial concern my friend.
Hi Phil, thanks! I'm so glad everyone ended up really liking the series, after all I originally almost killed it off entirely after the first videos seemed to barely interest anyone at all 😂
I love how this printer turned out. It's a huge achievement. Congrats for finishing it ... i still have one in the works but need to make a heated bed for it. I was thinking of making a conductive paint for mine. I look forward to seeing your next project.
Hey man, just wanted to say your project is great inspiration for me. For some time i thought you may fail but now it all comes together and i think your work is really something to be proud of
Thanks, glad to hear! I mean in a way I did fail, it didn't end up being a printer anyone would reasonably want to use to print parts, despite the print quality being okay ;)
You are one hell of an engenier, it's amazing to watch! Yes, there are more effective ways to do this. Yes, for a little more money one could buy a printer and arguably the tools needed for this project cost more. But the learning needed for this, the understanding you have on 3D printing and the craftsmanship. It's amazing.
Congrats on the 10k subs! This project of yours is a reminder of how much better and reliable 3D Printers got by using Aluminum profiles, CNC Machined parts and dedicated injection molded parts. This is a blast from the past - very enjoyable to watch! ❤
Cool seing a printer such as this work at all. That benchy looked very good all things considered. A tip about the leadscrews. Put them in tension instead of compression. That way you don't get leadscrew whip or that bending when compressed too much.
Make the leadscrew end sprung, or consider acme leadscrews. Just running the nut will cause loads of friction heating the threaded rod a lot. You could use very stiff springs, or even some compliant 3D-printed part to do that. You could also use just metal nuts mounted at two opposing places and locked in place against their support using some rotating hex nut fixing assebly part. That way you could adjust friction vs play in their fitting, yet each nut would by default be very durable and rotate with fairly small friction, especially if the threaded rod was well lubricated using MoS2 lube or such.
A little trick for pressing inserts (and nuts) into prints, you can use the printer itself for it. Heat the nozzle up, manually center it over the hole and spin the Z axis lead screw, the nozzle tip will retain the insert and push it in perfectly straight. No need for heat gun shenanigans.
Huge Congrats on the 10K!!!!!! 🎊🎊🎊🎊🎊🎊🎊🎊 Sooooo well deserved man! I remember starting to follow you way back when you were starting out with the Pen Plotter. You have come a long way and have been such an inspiration to me especially with your thriftiness and the way you put to use what others would trash. I rather recently in my sophomore year of school just built myself a 3D printer (Voron) that was not cheap, but prints quite. fast. (I have a couple shorts on my channel) I spend a lot of money and I thought a lot of time, but man you have really done amazing. Again congrats on the 10k
Thank youu!!! And kudos for sticking around so long! I don't know how many folks from back then are still around :p I'm hearing a lot about vorons, my next printer I'm pretty sure will have a stealthburner print head! BTW your FPV flying and editing is amazing :)
ok for such a cheap 3d printer this thing is usable. I mean it is actuaually good enough to make useful parts. I actually hope you keep using it. They are quite handy when you want something made from plastic. Im really impressed. For such a cheap device your prints are actually good quality. I'm not even joking.
I will use it as much as possible, though I need to be careful because the linear rails wear down very quickly, so I won't get nearly as many miles out of it as from any commercial machine. The ultimate goal is to use this printer as a RepStrap to make parts for a better, 3D printed one!
7:48 general design tip to avoid scenarios like this - grind a concave 90 on both pieces and slap a polished ball bearing in between. That way you have mild steel - precision ground contact x 2 which will vastly improve wear resistance and smoothness for like a 2 cent part. Still a bit of a janky solution but I think very much in the spirit of your project!
Good idea! My idea of making a tip on the screw ride in a negative cone was inspired by the balance wheel bearing I saw in a mechanical kitchen timer I took apart as a kid. The ends of the shaft were basically a needle point suspended in a divot on the end of a brass setscrew. Unfortunately I didn't get the indentation in mine to end in a sharp enough point, so I had to round off the tip of the lead screw a bit...
I've known for a long time that for vertical lead screws, the up/down play doesn't matter as much, because gravity preloads it, but it never occurred to me that you could do the same on the X or Y axis with a manual spring based preload. That's smart.
You can also get much higher speeds out of your steppers with a slow and constant acceleration. Direction changes are an issue, but like you can't just jump from one speed to another, you need to slowly increase the speed over time to get a higher speed. Basically you have to take the current momentum of the turning axis into account.
As another next to free upgrade to reduce the axial backlash from the motor you could put another block of wood, with a through hole for the threaded rod, in front of the motor and then put nuts (and counter nuts) before and after the block of wood. Adjusting the nuts for a tight, but no too tight, fit on the block it would take most of the axial backlash out of the rod without adding too much friction and virtually zero load on the motor. Plastic washers or waxing the wood again could help to reduce friction, though thrust bearings (or normal ball bearing that can also take some axial load) would be the better solution. The far end of the threaded rod should not be fixed to the frame for the issues you experienced.
Dude, you have the potential to be an awesome engineer as long as you keep in mind that in relation to everything there is to know in your field you don't know anything!
thank for sharing, I remember that I built from kit 10 years ago, I still have it today. Now, you gonna hate me, I bought a returned printer that I picked up. It is looking brand new never been printing, I only found one problem is the spinning wheels on the base plate, I have to cut one hole through magnetic sheet to get the screw. Everything adjusted and tight and did first print no problem at all. I paid for the printer £110, now what printer am I talking about? It was Creality Ender 3 S1 pro and absolutely beautiful printer at the time of price of new printer is £270. Not bad, as I got 4 of these to repair.
Actually if you just want a cheap 3D printer to use, getting a broken/returned/second hand one and fixing it up definitely seems the way to go. I admittedly wasn't even interested in 3D printing when I came up with this one, journey's the destination and all I guess. But now that I got a taste of what can be done with it, I have to design a better, actually _reliable_ printer that I can use to make parts for videos 😂
Printed m8 nuts(print the thread shape, no cutting) a little undersized and long enough(many threads) are effectively pretty much 0 backlash and out of pla quite slippery. Really the weakness of this is just the crappy steppers.
Those cheap 5 wire stepper motors can move quite fast, you need to add gears to change the extruding ammount. With it currently working, it should be simple to make a sun gear extruder head.
Congratulations with 10K subs! This project is what got me into your channel, it's a little sad that it finished. But great work! I was sceptical about leadscrew design for X and Y, but it worked out better that I expected. Watching video I got a cuple of ideas about backlashes. Firstly the leadscrew itself. When you did that cordless-drill-lathe-setup to grind it to stepper motor axle diameter, you put threaded rod through ball bearing with painters tape. What if instead of pre-tensioning end of threaded rod, just mount a ball bearing on it, fix it in place on rod with tho nuts, and fix outer ring of bearing to Xcarriage or frame (for X or Y axis). That way you wouldn't have your stepper motor axis under stress, and the temperature variation in rod tension would be gone. It needs to be a low backlash ball bearing tho, OR you could use a trick from second idea. So, secondly, a backlash in the leadscrew nut on the. I know about those brass no-backlash nuts with spring, but I really like a more solid solution, that I saw in precision lathe compound. To eliminate backlash there, the used two brass nuts, one mounted permanently on the frame, and second - mounted with a threadede connection. That way, by screwing on unscrewing the second nut it is possible to eliminate backlash without fancy precision brass nuts. And solid mount of both of them eliminates possibility of the spring of no-backlash-nut giving way under high pressure. What if you mouned thi regula nuts in similar way. It is really close to your second plastic no-backlash part, but with metal treads instead of plastic and with both pats aligned and concentric. Returning to threaded rod, with two ball bearings (one mounted solid, second with adjustmnet screw) you could compensate the cheapness of the bearings and tension them in a way to eliminate backlash ant tention in threaded rod. I'll add some video material to illustrate in a reply to this.
First idea - delegating all backlash shenanigans to a ball bearing: 4:13 of this video plus two nuts on both sides of the bearing. Second idea: Using two nuts (one solid and one mounted with a thread) is visible in this video ua-cam.com/video/1l2_uCyBXQ0/v-deo.html, at 2:08 time, and at the end of a video.
Thanks! No worries, the next projects will be very similar (albeit not quite as extreme). Yeah, there's many ways to skin the cat, the single decent ball bearing one is probably what I'll end up doing. I generally don't like to fasten ball bearings to threaded rod with nuts, even with a masking tape shim inside to keep everything concentric, simply because it looks a bit crude. I'd rather grind down the end to a smaller diameter like I did for the motor coupler and have like a 4mm ID ball bearing press-fit onto it. Except I have to order some bearings that size, which is why I didn't try that in the video. Getting rid of axial backlash this way is a little more involved, but definitely doable. Though seeing how little of an improvement in print quality the reduced backlash actually entailed, I think it might not even be worth the hassle. I haven't measured it or anything, but I'd estimate axial backlash in a decent quality (abec-3) bearing to be less than 0.1mm - maybe around 0.05mm?... In any case I'm probably better off building a better extruder and installing a more powerful part cooling fan. The screw-adjusted anti backlash nut is a great design, I have to see if I can print something similar but without threads. The spring being compressed during accelerations was definitely an issue with my spring loaded nut, I couldn't add too much tension so as not to overload the steppers, but very little tension wasn't enough to stand up to the inertia of the bed when zigzaging around for infill patterns...
Wow! I am so jealous! I have those old fashioned wooden printers. And I skipped that era, and jumped into 3d printing from aluminum extrusion models...
I think it's better to have skipped that era of dodgy printers, the modern ones are much less of a pain in the ass! I guess _I didn't_, "for science", but it's also important to know why we moved away from that :D
This is seriously beautiful. Outstanding work! Costs could have been reduced further with a non-heated bed (eg. painter tape, with glue-stick for great adhesion).
Thanks! Yeah I could've made it cheaper, also by reducing the size, but that would've resulted in sort of an easythreed knockoff and I wanted to build a "real" 3D printer at the time. That just included a fully heated bed as a non-negotiable ;p
@@ChronicMechatronic Completely understandable. I'm inching toward building my own, but using 3d printed parts. I think I will incorporate a pellet extruder, as it will offer the benefit of more materials/colors at lower cost, excepting a heavier print head.
@@SeanLumly pellet extruders are the way to go for 3D printing imo, especially for larger machines. That reminds me, I still need to build a shredder and filament extrusion machine to recycle all those failed prints I'm now making...
Regarding your wall's finish, ddi you try doing a pid tune of both the bed and the extruder? A good pid tuning makes a wolrd of a difference in wall finish
Congrats on the 10k subscribers. Wood actually can make an excllent low friction and high stiffnes bearing material. If you boil it in a mixture of wax and oil at around 110 °C/230 °F until all the moisture boils out. But it does make the wood brittle and of course complealty inresponsive to glues, but also so long as it stayes oiled and waxed its pretty much moisture proof and is compatable with high precision set ups with extreme care regarding grain aligment. Hopefully your parents divorce is them getting thier s it together and all will be better for it. Thats rough to deal with (Edit: Wrote the wood guide because you mentioned wood and oil. Then you boiled the wood in wax. Good thinking!)
Thanks, what a coincidence! I totally didn't know wood boiled in wax was even a thing! I even considered using tea light candle wax because it's more oily, but decided against it because it might leach into the clean wood it touches. It makes sense though, safe for the backlash (which I'm willing to bet was more from the wood shrinking in the frying process than actual wear) I found the wooden nut to be lower friction than the PLA I'm now using. Yeah I never imagined it was even possible to mess everything up as badly as my parents did...
I watched the whole series! It seems like anyone trying to repeat this journey would probably want to start with an ewaste printer/scanner combo, since that gives you plenty of motors, gears, a glass bed, and probably other useful hardware for the low,low price of free. Plus, you can brag about having upgraded a 2D printer into a 3D printer. And past that, everything save the motherboards, bearings, and hotend assembly can probably be purchased from any reasonably well-stocked hardware store.
Haha I'm not sure I would recommend anyone to repeat the that journey! At least not if they just want a working printer xD It strongly depends on what Ewaste you have access to, because only large old-ish office printers and stuff actually have any useful components inside. Around here all I can usually get my hands on are new cheap inkjet printers that got thrown out after like 6 months because the print head dried up, and aside form the scanner glass there's basically nothing useful inside (not even a linear rod on Epson models) :(( Bragging rights aside it sounds more promising to keep an eye out for a broken _3D printer_ sold cheaply for parts, to get kind of like a starter pack of components. At least in my neck of the woods...
@ChronicMechatronic I have two perfectly good printers, this is about the journey (and the UA-cam content), not the destination. I found a table with metal legs on the side of the road today that looks like it's a great candidate for the frame :) If most ewaste printers don't have any useful stuff inside, that's a bummer, I'll have to check out other sources of motors and whatnot. I know that floppy drive and CD tray mechanisms have usable motion systems, at the cost of making the print volume quite small. My goal is to learn more about how and why 3D printers work, and just the research for this project has accomplished that before I even started building anything, so I'm happy so far. If I ever get anywhere with making the project/UA-cam video I'll include a disclaimer to the effect of "just buy a used Ender 3" since new parts can't beat used prices, and scrap components, even if free, start to be expensive very quickly if you value your time at even like $10 an hour.
@rosly_yt aah yes, if it serves a second purpose as UA-cam content different rules apply and it makes a lot of sense 😅 I have a "normal" printer too these days, the wooden one would've broken down way too frequently to be a daily driver as I slowly came to realize the more the series progressed. I really should've done a disclaimer like that too, just to dissuade people from unnecessarily going the hard route. I just didn't know it myself when I started out - back then $50 was an investment that I legit didn't know if I would break even on. Metal table legs indeed sound like a great printer frame! The new printer I'm in the process of designing will use square tubing from a trashed pop-up gazebo for the frame x) The ewaste printer situation indeed sucks, I have half a mind to develop a closed loop driver board running FOC or something to make building 3D printers with the DC motors from inkjet printers possible. This would be a long shot though, and there wouldn't be any advantage over traditional stepper motors since getting custom boards manufactured also kind of ruins the "free" aspect of using Ewaste... :(
I was thinking about it today, and the solution might just be to assume a printer is better than the sum of its parts and do only minor retrofitting. Assuming you can root it and run arbitrary code, or write something to do fucked up things to printer code (like partial page retractions) such that a gcode to printer language conversion makes sense, any printer that can do double-sided printing should have mostly the right motors in mostly the right places (minus z motors and extruder motors) to be a 3D printer. If you lean more on software than hardware, and do positron-style upside down printing so you can keep the inkjet gantry where it's supposed to be on the stock printer, you could theoretically make it work. You'd probably need to put in a little extra work to keep the toolhead light, but I can't think of any burning reasons this setup wouldn't work. I've put out some feelers on my local BuyNothing group, if I can get a free printer out of it, I'll tinker with it and post a video if I get anything out of it.
@rosly_yt not gonna lie, the programming for that goes way over my head. Even just trying to jailbreak the proprietary control board would have me totally lost... But it would be great to see someone achieve it, if you end up doing it let me know; it'd definitely be a video I'd watch. In the meantime, I'm happy to reuse the drop-in wifi module they used on one of the printer controllers and maybe even reflash the main MCU after desoldering (if it belongs to one of the prevalent MCU families in the DIY world)...
First time here but I liked the video :) About the X, Y axis how about putting a BALL (single) bearing on the top of the screw? I mean one BALL from bearing between your bolt and the screw. Some grease and should reduce the friction imo. If you are up to the job u can drill a hole in the bolt and put a spring of a pen or smth..
As a rule, you want to run a threaded rod unsupported on the other end. It'll be supported enough on the motor side (ideally you use a slightly flexible mounting there, e.g,. PVC tubing) and on the nut of the moving element. Then the shaft will find its place. If you try to constrain it in 3 points, on the motor end, on the far end, and on the moving part, it WILL necessarily try to bind up, since it necessarily has a little bit of inherent curvature to it. I know that you probably have the problem that the shaft of the motor has a ton of axial backlash itself, and you're trying to take it out on the rod. My motors are... effectively backlash free for some reason. Also realistically, you don't need a fancy zero backlash nut, on a small enough thread, a normal nut is actually fine. Yeah you can feel it sort of being wobbly but in linear terms it's under the printer's tolerance. I have tortured myself a fair bit with these questions, and have ended up actually running M6 as my Z axis. But with a fairly rigid printed coupler. PVC tubing was considered as well. M6 is just flexible enough to not try to bind up too hard in spite of some curvature, i also use the straighter end into the motor and i have ensured that there is ample distance between the motor mount and the nearest position of the Z axis nut to give the shaft some flex. Well at least if you drive it with pretty feisty NEMA17 motors which i happen to use.
Corner bulge is normal unless you have, relative to the movement speed, pretty high acceleration and jerk. Also pressure advance calibration does take it out nicely. In my experience, Marlin's S-Curve doesn't work, it sort of does but it also happens to print worse for some reason, classic jerk is the way to go. You also shouldn't be surprised about leaky hot-end when you have nonprint movement speed which is this low. Fast retraction isn't actually needed, but any retraction can hold back the ooze for a short time anyway. I don't think you need to fix extruder backlash, i don't think it matters really. PLA and PETG do not hold up well to repeat flexing, so i don't know if they are wise for anti backlash nuts. PETG for sure better. But you know what does and can be printed open frame? HIPS. Surprise favourite filament of mine. Poisonous though. 800 is a LOT of acceleration for a printer of dubious construction. Prusa ships 1000 or 1500, something like that. Ender3 ships 500. Also i've heard zero jerk doesn't really work, isn't intended to work, see how far you can push it, and then set a bunch lower i guess. I actually have a lot of unsolved general wobbliness issues on my printer. My attempts to stabilise the X axis have had some success but i'm far from done, can just never quite get around to CAD up the ideas i have. And i think i will try some of your backlash nut designs on my Z. Ideally i want a flexture there to compensate my wobbly screw. I should get rid of the wooden frame someday but it's got so many pieces of threaded rod running across it now that it's fine i guess? I think when i upgrade to alu, it better be a Hypercube derived CoreXY.
I did the unsupported end solution on Z and it worked great, never worried about alignment, but with the screw spinning at 2400RPM it needs to be fully supported otherwise it'll whip around like crazy.. Binding actually ended up less of an issue than I expected, maybe because the M5 is pretty flexible - though I agree with you on M6 being kind of the sweet spot in terms of diameter and flexibility (and pitch, nothing beats nice round numbers haha). Yeah NEMA 17s are beautifully backlash-free, they actually have a spring wave washer preloading the rear ball bearing. No idea how that's not damaging to the bearings in the long run, but it works great.
@SianaGearz I'll try a new extruder anyway, just to see what happens. I also realized part of the messy results I'm getting might be caused by insufficient part cooling, that single 4010 turbine doesn't exactly create the biggest of drafts.. You're probably right about travel speeds being too low to ever entirely get rid of stringing. I didn't give material properties of the filament I used for the nuts any thought really, I only have PLA and PETG, latter still vacuum sealed bc I don't yet have a filament dryer and didn't want to risk it getting wet :} 800 is just max acceleration, the print and travel settings were both 400mm/s, and I've since turned them down to 300. The outdated Repetier Host UI doesn't display it properly. For jerk I never touched Prusaslicer's default 10mm/s, though I have no idea if it's suitable for this machine. Maybe you can make some kind of nut where the screw-adjusted backlash-free thread is suspended by sort of flexible fins all round? I feel I've seen something like that, though it might get difficult for it to resist torsional friction on the lead screw..
I am yet to try this, but i heard a rumour that the gear is actually heat shrunk onto the shaft of the motor. And it can be reversed the same way. Maybe find some way to keep the motor cold while you blast the gear with a jet torch.
Failed prints I have plenty, just made two more haha! I really wanted to do a giveaway with the 10K special, at this point I've collected quite a few PCB projects (like the unipolar stepper driver I recently did) and other stuff that I'd like to actually be of some use rather than sitting on my shelf for a decade before being thrown out. I'll check out how to do giveaways and see if I can organize something :)
I think your original 3dp backlash exerts the force parallel but the newer one tilts the nut (notb parallel). Did you try the 1st with correct acceleration settings? It may work better because of less friction (because of keeping the nut parallel). Maybe 2 screws could be used to keep it parallel ?
Good point, I thought about that too but came to the conclusion it doesn't matter. The thread cut by the tap is actually quite loose on the screw, so if it gets tilted it does cause only one half of the nut to actually engage with the thread, but since there's so much play it doesn't lock up the nut. Over time it'll wear in to have a larger contact area. I didn't try the spring loaded nut with correct accelerations, but the steel nuts were causing much more friction despite being parallel.. Ultimately I might not even switch to brass inserts, plastic seems to work just fine and it makes for significantly quieter operation.
for the leadscrews of the x and y just support them with a normal bearing and let them expand as much as they want at the end no need to constrain them like that also print with a multiplier of your nozzle diameter , for a 0.4 diameter print with 0.14 ,0.18 and 0.2 and 0.24 to eliminate layer lines try using around 1 mm for retraction length and the fastest speed you extruder can go for retraction speed i use 70 mm/s and if you have the option use "wipe while retracting" this will wipe the blob on your part leaving no blob on the part and for god's sake , use prusha slicer it has more options to compensate for faults in printers
I do use Prusaslicer for slicing, Repetier Host is just sending Gcode from the USB drive. My retraction is 1mm, tho I tried 2mm too, and didn't notice any improvement, it just took longer because the 28BYJ-48 can't really go faster than 4mm/s without loosing steps. The normal bearing is what I plan on doing, though it'll increase backlash again - I'd have to use two bearings with a slight preload like on a real CNC. I'll try wipe while retracting, thx for the tip
I know! I unfortunately don't have access to that kind of old high quality ewaste but I was able to acquire some Epson EM-258 steppers for really cheap from a guy who has!
Ben, please ... this is the second video ... don't grind with the cutting disc. Congrats on the 10k, and nice job with the printer :). Impressive job with the wooden nuts.
Buut - I did most of the grinding with the edge of the disc? 🥺 For real tho I'm not sure what this cutting wheel is made of, all the grey grinding stones I have totally disintegrate when used on mild steel, I think they're only for high carbon steel (or simply too old and expired?) but that brown (aluminum oxide judging by the color) cutting disc goes through mild steel like butter. Thanks BTW!
Turn you acceleration from 800 to like 400~ and slice with lower jerk settings...it will cause the print head to accelerate/decelerate slower giving more torque Also...what is the voltage rating of those steppers 24vdc? what are you running it at...and current wise...are the motors getting hot? So awesome that it's running and printing! Keep em coming!!!!
Accelerations were 400mm/s, 800 is just what I set the maximum to, but then Repetier Host messed up and set print accelerations to zero which made Marlin default to the max settings and it failed :D For jerk I never touched the Prusaslicer default 10mm/s. The steppers, IDK, it says 2.8V on the label, but the typewriter I got them from was powered by what I assume to be a 24V transformer. Also I _think_ they were driven in unipolar mode, I'm now driving them in bipolar config with an A4988 set to 150mA if memory serves. Definitely wayyy overclocked, they still reach toasty 60°C
Haha I guess I managed to make a rather trivial fact interesting :p I just collected them over several years, snipping off the cotton after using them. Sounds gross (hence why you shouldn't ask :D), but you don't usually get earwax on the stem (at least I don't)...
Thanks for the tip! I'll have to look online, the hardware stores around me have basically nothing in terms of brass hardware that isn't ridiculously overpriced
So far I didn't find the Geeetech stuff to be all that bad honestly. Maybe I have low expectations, but aside from the occasional tiny bubble it prints pretty good I think?
definitely think people underestimate how useful wood is. the fact you dont need a 3d printer to build your 3d printer is great. if it really costs $50 to put together, thats genius
Yeah well, I cannot deny that the extra zero on the price tag helps with performance... I was definitely under $50 for the printer itself, (there was a little spreadsheet towards the end, don't know if you saw) but only because I had all the wood sitting around already. There's a lot that can be done with wood, it just needs to be designed right to allow for humidity changes - I don't think this printer is anywhere close to ideal in that regard, but it works, so it's close enough I guess
@@ChronicMechatronicthough to be fair, there's some changes that could be done with this printer for fairly cheap that don't line up with the "Do it from scraps from scratch" to significantly increase its usability for not a lot. An extra 100-150 dollars I am fairly certain will go a long way with like another mainboard and some nema-17 ender 3 motors or something. And I imagine those kinds of upgrades you will probably do if you continue to find use in the printer and deem it worthwhile. That kind of thing is something I imagine a second series could be about. Value upgrades for the printer that make it print well
@@donovan6320 yeah, even just a slightly higher initial budget of $100 would've gone a long way toward making it a usable printer. NEMA 17s would definitely be at the top of the list in terms of components worth splurging for. Different controller board - I don't know, Ramps works just fine I think. More important would've been a screen so I didn't need a host for sending the G-code, as well as having the aluminum rails anodized so they don't wear out as fast. And a pack of real POM wheels I guess. It's funny you should say that, I definitely plan to do all these upgrades, as well as doing a new series about them! However I'm not much of an upgrade afficionado - I usually deem old designs not worth upgrading since it's all just superficial fixes on a base that ideally would be fully redesigned - so I'm actually designing a whole new printer to build using this one to print the parts :D
Would you be interested in trying 3d printed ball bearings with copper bbs for your threaded rods? Link to video from designer ua-cam.com/video/Pj2N08Fq3bA/v-deo.htmlsi=srlaQ2Egy8171buY
Cool test, though I consider 3D printed ball bearings more of a gimmick for anything other than large, low load applications. For tiny like 4mm ID bearings the tolerances on printed parts will always be too low, it's either too stiff to rotate easily, has a lot of backlash and/or runout. Keep in mind 3D printers never actually print circles, it's always polygons with x amount of facets, so it'll never be smooth enough.
It was simple enough of a shape to machine... It did run smoother and with lower friction than the printed PLA nut though, maybe I should try to modify it to add the springy tab to take out backlash
i built a printer almost as bad, its a corexy from years back (6?) cost me 1400$ over the years working on it. and it still performs horrible. it would be a perfect match for you and your farm.
Please make a new video in which to show how to build it the right way from the first try, for people who want to DIY themselves with you as a guide. Even if it ends up being 3 hours long, people will gladly watch the video at least once, and many of them will even make their own version. And maybe make another series in which you try to improve it while going as little over $50 as possible, in which to start with ranking each improvement by the importance in one table, by the price in another table, and by your choice if which to improve in another table.
I'm designing a new, much better printer that should be possible for more people to make, and it'll be fully open source. I'm aiming for a budget of no more than $100 if possible, because $50 just doesn't allow for the right components to get a machine that lasts any reasonable length of time. I had to spend a significant amount of time to keep even this one below the limit, it's just not worth it for people to replicate, since all that time would have to be accounted for too, and off-the-shelf printers these days are just too damn cheap for how good they are, it would be more economical to just work a couple of hours overtime to have money to buy an Ender 3 or an A1 mini from. Filament is fairly expensive too, and that's a consumable - at this point I've spent 3 times as much on filament as on this printer, so yeah... There will be a new printer series, it'll be cheap and made from a lot of trash too, but not as bare bones as this one. It's also how 3D printing goes, as soon as you get used to one printer and its features you'll immediately want to upgrade for more speed/blinky lights/cameras, etc
KiCad! It's incredibly easy to pick up, even with no prior CAD skills, just watch a tutorial or two and you're set, at least as far as core functionality goes. Plus it's open source 👍
@@ChronicMechatronic I have actually tried KiCAD a few days ago, and I couldn't for the life of me get it to connect a trace to multiple pins. It just refused to make the connection! I tried for 3 hours before I gave up. All I need is 4 pins with GND for some fans, dammit!
What is the z axis steps u used for Marlin framework I am using 8mm thread rod from hardware store please calculate for me and give me the number and I'm using nema 17 moters
It's very simple math, I explained it in a different episode, but you can also use a calculator like this one: www.circuitist.com/how-to-calculate-steps-per-mm-lead-screw-and-gt2-timing-belt/ Just look up M8 thread pitch and plug in the values
Is there a technical reason you didn't use cheap drawer Slides with ball bearings? Or did you not think of it? Also, I would have use vegetable oil and Silicone based lubricant instead, Because Vaseline is a petroleum product which can break down wax. I would have also used used super glued nuts instead of theading the wood, Would take just a few minutes to cut out a hole for the nuts to lock into with a small Chisel, If you used a Softer wood, you could press the nuts then, super glue the nut in and then flood the wood to fill the pores would be probably been a lot better. In the US where people put a lot of trash products along the street for people to pick up so everything i've mentioned could have been gotten for free.
@@ChronicMechatronic As a matter of fact colour does affect some things. Light, opaque colours can contain minerals such as zinc oxide and titanium dioxide that can cause the filament to have a sharper more unwieldy transition from not melty enough to too melty, more difficult to work with, since they absorb some heat until they get to the temperature. Though realistically there shouldn't be a lot of that in yellow. Unless there's also way too much yellow pigment. Another fun fact, the mentioned minerals are white pigments, but they aren't actually white, they are almost glass clear. The white comes about due to structure and dispersion, and their increased refraction index compared to carrier polymer.
The worst idea! 😅 3D printing is so easy, unless you do it like me! For a first printer (if it needs to be cheap) I'd say your best bet is a used and/or broken one someone is trying to get rid of on like Facebook marketplace or something. That way you just need to fix it up a little or have most of the parts for a DIY, and it'll be way better than this, haha!
don't worry my fully 3d printed 3d printer didn't work either so i converted it to be 90% 3d printed prusa i3 mark 3,5 and it prints worse than your current model i make my gear pullers from m14 nuts filed on one side and drilled and thereaded from the other side, they can't handle long puleys though
Yeah that is the unfortunate nature of experimental machines I guess... Thanks a bunch for the tip, I'll pick up some big nuts next time I'm at the hardware store! I've seen welded versions using square tubing but nuts sound better because they're sturdier.
Thanks everyone for 10K subs!!! I hope this is a satisfactory conclusion to this series, I learned so much about CNC and 3D printing over the past two years, if I was able to pass on just a fraction of that in the videos it was well worth it. I'm already designing a much better (and much more conventional) printer that can be printed by this one in true reprap fashion!
ur amazing
Can't wait.
Keep up the great work man
Congrats
Loved the video. Huge reprap fan. Cant wait to see whats next!
Don't forget to dry your printer before printing.
Yeah pretty much!
@@ChronicMechatronic Awesome work BTW!
@@mvaar thx!
*fireworks and upbeat music* "thanks for 10k subs, my parents are getting divorced" is WILD. I'm sorry that you're going through something so stressful and life changing, but that delivery was so outta pocket that it had me hysterical
Lmao, I didn't even realize the irony of that intro until you mentioned it just now! My parent's shit has been going on for longer than this channel exists, I didn't think anything of it. Thanks for the kind words, this comment made my day 💛
This so reminds me of the time when I was building my first printer, a Mendel 90, In the engineering company I worked for I used to go down to the workshops in my lunch hour and run my designs on the printers they had down there, one day one of the chaps showed me his DIY printer, I was amazed, printers back then were around the £20,000 mark for the cheapest models and so having one at home of my very own was a pipe dream, I ordered my kit that same afternoon and printed my plastic parts at work, it took some fernangling and fine tuning to get it to print correctly but my friend had already done the hard work when he built his so mine was relatively straight forward, my frame was built from acrylic but my friend designed some aluminium reinforcements and cut me a set also, he later made all the parts from highly polished CNC'd aluminium and made me some of the fixing block parts also to replace the plastic ones.
I often think of reassembling my Mendel 90 as it is currently in bits but stored in an old synthesiser flight case.
I love hearing these stories from way back when! Meanwhile I was still taking apart toy boats and my dad's broken shavers to get the motors :D
Yep, watching this video reminds me of how much the 2012ish era printers were a nonstop faff-fest. Always burning out relatively expensive parts, forums with a million red herrings to chase, outright bad advice sometimes, and prints that were only vaguely okay if you stood 5ft back. The amount of institutional knowledge baked into even the worst cheapest disposable printers of today makes the best FDM hobby printers of the early 2010s look like hot trash.
@@carpespasm I had a good giggle reading your comment, you hit the nail on the head.
Luckily my forum was the lunch table with some chaps that had 40 years or so engineering experience and 25 years of industrial 3D printing experience, what they didn't know could be written on the back of a stamp, within a year or so of the first Mendel 90 (ish) appearing In the workshop half the employees of the workshop had one, they were great days!
Most diy 3d printer videos: I built a voron for fast printing
Chronic Mechatronic: I built the world's slowest, most unreliable printer
And I for sure know which one was more interesting to watch.
😅
I sure know which series I binge watched in entirety over 3 days. I would probably binge the rest of the channel too if I wasn't running out of vacation...
@@monophonic_og I binged half the series in one night. Not good.
Waxed wood t-nuts make this almost a steampunk project! Love it!
This has been an amazing series displaying perseverance, dedication, and ingenuity. I learned a lot watching you tackle problems from different angles and posting your successes and failures. I see good things in your future. Well done !
I'm so glad everyone ended up enjoying the series, I almost called it off entirely after nobody seemed to care in the beginning (my mental health wasn't the best at the time...) - so glad I didn't!
Hi Ben, for me it's a complete achievement. I've been wondering how that "wood" and cheap 3D printer could output a decent cube or the ubiquitous benchy well. You just proved it can! That series was exciting and full of suspense. Thank you for sharing such a valuable experience Ben. Have a great day despite your familial concern my friend.
Hi Phil, thanks! I'm so glad everyone ended up really liking the series, after all I originally almost killed it off entirely after the first videos seemed to barely interest anyone at all 😂
I love how this printer turned out. It's a huge achievement. Congrats for finishing it ... i still have one in the works but need to make a heated bed for it. I was thinking of making a conductive paint for mine. I look forward to seeing your next project.
Hi Enkidu! Thanks!
I lost it when you deep fried wood in paraffin wax. That's a MacGyver solution I would never have come up with!
Congratulations on 10k man, really enjoy your content.
Thanks, glad to hear!
Hey man, just wanted to say your project is great inspiration for me. For some time i thought you may fail but now it all comes together and i think your work is really something to be proud of
Thanks, glad to hear! I mean in a way I did fail, it didn't end up being a printer anyone would reasonably want to use to print parts, despite the print quality being okay ;)
You are one hell of an engenier, it's amazing to watch! Yes, there are more effective ways to do this. Yes, for a little more money one could buy a printer and arguably the tools needed for this project cost more. But the learning needed for this, the understanding you have on 3D printing and the craftsmanship. It's amazing.
Congrats on the 10k subs! This project of yours is a reminder of how much better and reliable 3D Printers got by using Aluminum profiles, CNC Machined parts and dedicated injection molded parts. This is a blast from the past - very enjoyable to watch! ❤
Thank you! Yeah it's kinda surprising just how much of a difference these construction methods make!
Cool seing a printer such as this work at all. That benchy looked very good all things considered.
A tip about the leadscrews. Put them in tension instead of compression. That way you don't get leadscrew whip or that bending when compressed too much.
Yeah it's kinda surprising..
Thanks for the tip! I'll incorporate that idea into my laser engraver project since I've got a very similar issue there
Make the leadscrew end sprung, or consider acme leadscrews. Just running the nut will cause loads of friction heating the threaded rod a lot. You could use very stiff springs, or even some compliant 3D-printed part to do that. You could also use just metal nuts mounted at two opposing places and locked in place against their support using some rotating hex nut fixing assebly part. That way you could adjust friction vs play in their fitting, yet each nut would by default be very durable and rotate with fairly small friction, especially if the threaded rod was well lubricated using MoS2 lube or such.
Dude, this was massively entertaining to watch and also a great flashback to my Anet A8 (before the mods)
Glad to hear!
I absolutely love the use of cheap materials here!
A little trick for pressing inserts (and nuts) into prints, you can use the printer itself for it. Heat the nozzle up, manually center it over the hole and spin the Z axis lead screw, the nozzle tip will retain the insert and push it in perfectly straight. No need for heat gun shenanigans.
Great tip, never thought of it, thanks!
Even quicker, use a solder iron
It's working beautifully! I never expected it to work so well!
Wow, what a fancy "powered" turntable! Great work.
Thanks, haha yeah I need to build a proper motorized rotating platform for videoing
those drivers bring back old memories of my cr-10s before silent driver upgrade on a 32bit board...
31:22 i actually am really impressed. good job!
Huge Congrats on the 10K!!!!!! 🎊🎊🎊🎊🎊🎊🎊🎊 Sooooo well deserved man! I remember starting to follow you way back when you were starting out with the Pen Plotter. You have come a long way and have been such an inspiration to me especially with your thriftiness and the way you put to use what others would trash. I rather recently in my sophomore year of school just built myself a 3D printer (Voron) that was not cheap, but prints quite. fast. (I have a couple shorts on my channel) I spend a lot of money and I thought a lot of time, but man you have really done amazing.
Again congrats on the 10k
Thank youu!!! And kudos for sticking around so long! I don't know how many folks from back then are still around :p
I'm hearing a lot about vorons, my next printer I'm pretty sure will have a stealthburner print head!
BTW your FPV flying and editing is amazing :)
ok for such a cheap 3d printer this thing is usable. I mean it is actuaually good enough to make useful parts. I actually hope you keep using it. They are quite handy when you want something made from plastic. Im really impressed. For such a cheap device your prints are actually good quality. I'm not even joking.
I will use it as much as possible, though I need to be careful because the linear rails wear down very quickly, so I won't get nearly as many miles out of it as from any commercial machine. The ultimate goal is to use this printer as a RepStrap to make parts for a better, 3D printed one!
7:48 general design tip to avoid scenarios like this - grind a concave 90 on both pieces and slap a polished ball bearing in between. That way you have mild steel - precision ground contact x 2 which will vastly improve wear resistance and smoothness for like a 2 cent part. Still a bit of a janky solution but I think very much in the spirit of your project!
Good idea! My idea of making a tip on the screw ride in a negative cone was inspired by the balance wheel bearing I saw in a mechanical kitchen timer I took apart as a kid. The ends of the shaft were basically a needle point suspended in a divot on the end of a brass setscrew. Unfortunately I didn't get the indentation in mine to end in a sharp enough point, so I had to round off the tip of the lead screw a bit...
I've known for a long time that for vertical lead screws, the up/down play doesn't matter as much, because gravity preloads it, but it never occurred to me that you could do the same on the X or Y axis with a manual spring based preload. That's smart.
You can also get much higher speeds out of your steppers with a slow and constant acceleration. Direction changes are an issue, but like you can't just jump from one speed to another, you need to slowly increase the speed over time to get a higher speed. Basically you have to take the current momentum of the turning axis into account.
As another next to free upgrade to reduce the axial backlash from the motor you could put another block of wood, with a through hole for the threaded rod, in front of the motor and then put nuts (and counter nuts) before and after the block of wood.
Adjusting the nuts for a tight, but no too tight, fit on the block it would take most of the axial backlash out of the rod without adding too much friction and virtually zero load on the motor. Plastic washers or waxing the wood again could help to reduce friction, though thrust bearings (or normal ball bearing that can also take some axial load) would be the better solution.
The far end of the threaded rod should not be fixed to the frame for the issues you experienced.
Congratulations on both the project and also reaching this 10K milestone.
Thanks!
Dude, you have the potential to be an awesome engineer as long as you keep in mind that in relation to everything there is to know in your field you don't know anything!
Some improvements, Lets gooooo!
Congrats on the 10k btw.
Thaanks!
thank for sharing, I remember that I built from kit 10 years ago, I still have it today. Now, you gonna hate me, I bought a returned printer that I picked up. It is looking brand new never been printing, I only found one problem is the spinning wheels on the base plate, I have to cut one hole through magnetic sheet to get the screw. Everything adjusted and tight and did first print no problem at all. I paid for the printer £110, now what printer am I talking about? It was Creality Ender 3 S1 pro and absolutely beautiful printer at the time of price of new printer is £270. Not bad, as I got 4 of these to repair.
Actually if you just want a cheap 3D printer to use, getting a broken/returned/second hand one and fixing it up definitely seems the way to go.
I admittedly wasn't even interested in 3D printing when I came up with this one, journey's the destination and all I guess. But now that I got a taste of what can be done with it, I have to design a better, actually _reliable_ printer that I can use to make parts for videos 😂
Most beautiful thing i've ever witnessed
Printed m8 nuts(print the thread shape, no cutting) a little undersized and long enough(many threads) are effectively pretty much 0 backlash and out of pla quite slippery.
Really the weakness of this is just the crappy steppers.
Bro you are light in my life please keep your head strong and tour heart stronger. Great stuff!🎉🎉🎉
Wow. Great work man. Pretty mind blowing what you were able to get done
thx!
Those cheap 5 wire stepper motors can move quite fast, you need to add gears to change the extruding ammount. With it currently working, it should be simple to make a sun gear extruder head.
Strippers? o.o
@@Atlessa spuul chuck struck again!
Congrats for 10k!
You can't rely on PLA springiness. PLA will deform under constant load.
Thanks! Yeah I kind of skipped material properties on this one
Congratulations with 10K subs! This project is what got me into your channel, it's a little sad that it finished. But great work! I was sceptical about leadscrew design for X and Y, but it worked out better that I expected. Watching video I got a cuple of ideas about backlashes. Firstly the leadscrew itself. When you did that cordless-drill-lathe-setup to grind it to stepper motor axle diameter, you put threaded rod through ball bearing with painters tape. What if instead of pre-tensioning end of threaded rod, just mount a ball bearing on it, fix it in place on rod with tho nuts, and fix outer ring of bearing to Xcarriage or frame (for X or Y axis). That way you wouldn't have your stepper motor axis under stress, and the temperature variation in rod tension would be gone. It needs to be a low backlash ball bearing tho, OR you could use a trick from second idea. So, secondly, a backlash in the leadscrew nut on the. I know about those brass no-backlash nuts with spring, but I really like a more solid solution, that I saw in precision lathe compound. To eliminate backlash there, the used two brass nuts, one mounted permanently on the frame, and second - mounted with a threadede connection. That way, by screwing on unscrewing the second nut it is possible to eliminate backlash without fancy precision brass nuts. And solid mount of both of them eliminates possibility of the spring of no-backlash-nut giving way under high pressure. What if you mouned thi regula nuts in similar way. It is really close to your second plastic no-backlash part, but with metal treads instead of plastic and with both pats aligned and concentric. Returning to threaded rod, with two ball bearings (one mounted solid, second with adjustmnet screw) you could compensate the cheapness of the bearings and tension them in a way to eliminate backlash ant tention in threaded rod. I'll add some video material to illustrate in a reply to this.
First idea - delegating all backlash shenanigans to a ball bearing: 4:13 of this video plus two nuts on both sides of the bearing.
Second idea: Using two nuts (one solid and one mounted with a thread) is visible in this video ua-cam.com/video/1l2_uCyBXQ0/v-deo.html, at 2:08 time, and at the end of a video.
Thanks! No worries, the next projects will be very similar (albeit not quite as extreme).
Yeah, there's many ways to skin the cat, the single decent ball bearing one is probably what I'll end up doing. I generally don't like to fasten ball bearings to threaded rod with nuts, even with a masking tape shim inside to keep everything concentric, simply because it looks a bit crude. I'd rather grind down the end to a smaller diameter like I did for the motor coupler and have like a 4mm ID ball bearing press-fit onto it. Except I have to order some bearings that size, which is why I didn't try that in the video. Getting rid of axial backlash this way is a little more involved, but definitely doable. Though seeing how little of an improvement in print quality the reduced backlash actually entailed, I think it might not even be worth the hassle. I haven't measured it or anything, but I'd estimate axial backlash in a decent quality (abec-3) bearing to be less than 0.1mm - maybe around 0.05mm?... In any case I'm probably better off building a better extruder and installing a more powerful part cooling fan.
The screw-adjusted anti backlash nut is a great design, I have to see if I can print something similar but without threads. The spring being compressed during accelerations was definitely an issue with my spring loaded nut, I couldn't add too much tension so as not to overload the steppers, but very little tension wasn't enough to stand up to the inertia of the bed when zigzaging around for infill patterns...
@@ChronicMechatronic can't wait for next project!
My sincere respect for the work you r doing and the achievement keep up 😊
Man, u just made a printer from scrap better than easythreed k1
Haha easythreed machines aren't hard to beat
works better than my old anet A8 clone straight out of the box
Very nice project!
Wow! I am so jealous! I have those old fashioned wooden printers. And I skipped that era, and jumped into 3d printing from aluminum extrusion models...
I think it's better to have skipped that era of dodgy printers, the modern ones are much less of a pain in the ass!
I guess _I didn't_, "for science", but it's also important to know why we moved away from that :D
it sound like an inkjet printer with a stutter! I love it
congrats on 10k bro!
love your content, keep it up
Thanks! Will do!
This is seriously beautiful. Outstanding work!
Costs could have been reduced further with a non-heated bed (eg. painter tape, with glue-stick for great adhesion).
Thanks!
Yeah I could've made it cheaper, also by reducing the size, but that would've resulted in sort of an easythreed knockoff and I wanted to build a "real" 3D printer at the time. That just included a fully heated bed as a non-negotiable ;p
@@ChronicMechatronic Completely understandable. I'm inching toward building my own, but using 3d printed parts.
I think I will incorporate a pellet extruder, as it will offer the benefit of more materials/colors at lower cost, excepting a heavier print head.
@@SeanLumly pellet extruders are the way to go for 3D printing imo, especially for larger machines. That reminds me, I still need to build a shredder and filament extrusion machine to recycle all those failed prints I'm now making...
Regarding your wall's finish, ddi you try doing a pid tune of both the bed and the extruder? A good pid tuning makes a wolrd of a difference in wall finish
First thing I did! :D
Congrats on the 10k subscribers.
Wood actually can make an excllent low friction and high stiffnes bearing material. If you boil it in a mixture of wax and oil at around 110 °C/230 °F until all the moisture boils out.
But it does make the wood brittle and of course complealty inresponsive to glues, but also so long as it stayes oiled and waxed its pretty much moisture proof and is compatable with high precision set ups with extreme care regarding grain aligment.
Hopefully your parents divorce is them getting thier s it together and all will be better for it. Thats rough to deal with
(Edit: Wrote the wood guide because you mentioned wood and oil. Then you boiled the wood in wax. Good thinking!)
Thanks, what a coincidence! I totally didn't know wood boiled in wax was even a thing! I even considered using tea light candle wax because it's more oily, but decided against it because it might leach into the clean wood it touches. It makes sense though, safe for the backlash (which I'm willing to bet was more from the wood shrinking in the frying process than actual wear) I found the wooden nut to be lower friction than the PLA I'm now using.
Yeah I never imagined it was even possible to mess everything up as badly as my parents did...
20:20 friction -> heat -> expansion
Congrats on 10K subs!!
I dub the the "WoodM8" [AM8..but made outta wood] (ya know woodmate)
Very awesome!
Keep em coming!!!!
Thanks, haha!
I watched the whole series! It seems like anyone trying to repeat this journey would probably want to start with an ewaste printer/scanner combo, since that gives you plenty of motors, gears, a glass bed, and probably other useful hardware for the low,low price of free. Plus, you can brag about having upgraded a 2D printer into a 3D printer. And past that, everything save the motherboards, bearings, and hotend assembly can probably be purchased from any reasonably well-stocked hardware store.
Haha I'm not sure I would recommend anyone to repeat the that journey! At least not if they just want a working printer xD
It strongly depends on what Ewaste you have access to, because only large old-ish office printers and stuff actually have any useful components inside. Around here all I can usually get my hands on are new cheap inkjet printers that got thrown out after like 6 months because the print head dried up, and aside form the scanner glass there's basically nothing useful inside (not even a linear rod on Epson models) :((
Bragging rights aside it sounds more promising to keep an eye out for a broken _3D printer_ sold cheaply for parts, to get kind of like a starter pack of components. At least in my neck of the woods...
@ChronicMechatronic I have two perfectly good printers, this is about the journey (and the UA-cam content), not the destination. I found a table with metal legs on the side of the road today that looks like it's a great candidate for the frame :)
If most ewaste printers don't have any useful stuff inside, that's a bummer, I'll have to check out other sources of motors and whatnot. I know that floppy drive and CD tray mechanisms have usable motion systems, at the cost of making the print volume quite small.
My goal is to learn more about how and why 3D printers work, and just the research for this project has accomplished that before I even started building anything, so I'm happy so far.
If I ever get anywhere with making the project/UA-cam video I'll include a disclaimer to the effect of "just buy a used Ender 3" since new parts can't beat used prices, and scrap components, even if free, start to be expensive very quickly if you value your time at even like $10 an hour.
@rosly_yt aah yes, if it serves a second purpose as UA-cam content different rules apply and it makes a lot of sense 😅
I have a "normal" printer too these days, the wooden one would've broken down way too frequently to be a daily driver as I slowly came to realize the more the series progressed.
I really should've done a disclaimer like that too, just to dissuade people from unnecessarily going the hard route. I just didn't know it myself when I started out - back then $50 was an investment that I legit didn't know if I would break even on.
Metal table legs indeed sound like a great printer frame! The new printer I'm in the process of designing will use square tubing from a trashed pop-up gazebo for the frame x)
The ewaste printer situation indeed sucks, I have half a mind to develop a closed loop driver board running FOC or something to make building 3D printers with the DC motors from inkjet printers possible. This would be a long shot though, and there wouldn't be any advantage over traditional stepper motors since getting custom boards manufactured also kind of ruins the "free" aspect of using Ewaste... :(
I was thinking about it today, and the solution might just be to assume a printer is better than the sum of its parts and do only minor retrofitting. Assuming you can root it and run arbitrary code, or write something to do fucked up things to printer code (like partial page retractions) such that a gcode to printer language conversion makes sense, any printer that can do double-sided printing should have mostly the right motors in mostly the right places (minus z motors and extruder motors) to be a 3D printer. If you lean more on software than hardware, and do positron-style upside down printing so you can keep the inkjet gantry where it's supposed to be on the stock printer, you could theoretically make it work. You'd probably need to put in a little extra work to keep the toolhead light, but I can't think of any burning reasons this setup wouldn't work.
I've put out some feelers on my local BuyNothing group, if I can get a free printer out of it, I'll tinker with it and post a video if I get anything out of it.
@rosly_yt not gonna lie, the programming for that goes way over my head. Even just trying to jailbreak the proprietary control board would have me totally lost...
But it would be great to see someone achieve it, if you end up doing it let me know; it'd definitely be a video I'd watch.
In the meantime, I'm happy to reuse the drop-in wifi module they used on one of the printer controllers and maybe even reflash the main MCU after desoldering (if it belongs to one of the prevalent MCU families in the DIY world)...
good job you just figured out how to tighten belts on a leadscrew
this is so cool! definitely didnt think a wooden 3d printer "wood" work so well
bro you need to use a heatsink on the stepper motors, the heating modifies the steps or sometimes it fails
I’ve really liked this video, I hope you’re going to make more and more !
That's at least the plan!
First time here but I liked the video :)
About the X, Y axis how about putting a BALL (single) bearing on the top of the screw? I mean one BALL from bearing between your bolt and the screw. Some grease and should reduce the friction imo. If you are up to the job u can drill a hole in the bolt and put a spring of a pen or smth..
As a rule, you want to run a threaded rod unsupported on the other end. It'll be supported enough on the motor side (ideally you use a slightly flexible mounting there, e.g,. PVC tubing) and on the nut of the moving element. Then the shaft will find its place.
If you try to constrain it in 3 points, on the motor end, on the far end, and on the moving part, it WILL necessarily try to bind up, since it necessarily has a little bit of inherent curvature to it.
I know that you probably have the problem that the shaft of the motor has a ton of axial backlash itself, and you're trying to take it out on the rod. My motors are... effectively backlash free for some reason.
Also realistically, you don't need a fancy zero backlash nut, on a small enough thread, a normal nut is actually fine. Yeah you can feel it sort of being wobbly but in linear terms it's under the printer's tolerance.
I have tortured myself a fair bit with these questions, and have ended up actually running M6 as my Z axis. But with a fairly rigid printed coupler. PVC tubing was considered as well. M6 is just flexible enough to not try to bind up too hard in spite of some curvature, i also use the straighter end into the motor and i have ensured that there is ample distance between the motor mount and the nearest position of the Z axis nut to give the shaft some flex. Well at least if you drive it with pretty feisty NEMA17 motors which i happen to use.
Corner bulge is normal unless you have, relative to the movement speed, pretty high acceleration and jerk. Also pressure advance calibration does take it out nicely. In my experience, Marlin's S-Curve doesn't work, it sort of does but it also happens to print worse for some reason, classic jerk is the way to go. You also shouldn't be surprised about leaky hot-end when you have nonprint movement speed which is this low. Fast retraction isn't actually needed, but any retraction can hold back the ooze for a short time anyway. I don't think you need to fix extruder backlash, i don't think it matters really.
PLA and PETG do not hold up well to repeat flexing, so i don't know if they are wise for anti backlash nuts. PETG for sure better. But you know what does and can be printed open frame? HIPS. Surprise favourite filament of mine. Poisonous though.
800 is a LOT of acceleration for a printer of dubious construction. Prusa ships 1000 or 1500, something like that. Ender3 ships 500. Also i've heard zero jerk doesn't really work, isn't intended to work, see how far you can push it, and then set a bunch lower i guess.
I actually have a lot of unsolved general wobbliness issues on my printer. My attempts to stabilise the X axis have had some success but i'm far from done, can just never quite get around to CAD up the ideas i have. And i think i will try some of your backlash nut designs on my Z. Ideally i want a flexture there to compensate my wobbly screw. I should get rid of the wooden frame someday but it's got so many pieces of threaded rod running across it now that it's fine i guess? I think when i upgrade to alu, it better be a Hypercube derived CoreXY.
I did the unsupported end solution on Z and it worked great, never worried about alignment, but with the screw spinning at 2400RPM it needs to be fully supported otherwise it'll whip around like crazy.. Binding actually ended up less of an issue than I expected, maybe because the M5 is pretty flexible - though I agree with you on M6 being kind of the sweet spot in terms of diameter and flexibility (and pitch, nothing beats nice round numbers haha).
Yeah NEMA 17s are beautifully backlash-free, they actually have a spring wave washer preloading the rear ball bearing. No idea how that's not damaging to the bearings in the long run, but it works great.
@SianaGearz I'll try a new extruder anyway, just to see what happens. I also realized part of the messy results I'm getting might be caused by insufficient part cooling, that single 4010 turbine doesn't exactly create the biggest of drafts.. You're probably right about travel speeds being too low to ever entirely get rid of stringing.
I didn't give material properties of the filament I used for the nuts any thought really, I only have PLA and PETG, latter still vacuum sealed bc I don't yet have a filament dryer and didn't want to risk it getting wet :}
800 is just max acceleration, the print and travel settings were both 400mm/s, and I've since turned them down to 300. The outdated Repetier Host UI doesn't display it properly. For jerk I never touched Prusaslicer's default 10mm/s, though I have no idea if it's suitable for this machine.
Maybe you can make some kind of nut where the screw-adjusted backlash-free thread is suspended by sort of flexible fins all round? I feel I've seen something like that, though it might get difficult for it to resist torsional friction on the lead screw..
This is a true reprap a printer that can replicate itself
I am yet to try this, but i heard a rumour that the gear is actually heat shrunk onto the shaft of the motor. And it can be reversed the same way. Maybe find some way to keep the motor cold while you blast the gear with a jet torch.
Well for 10k subs you can organize giveaway of failed prints! Those might be valuable collectibles one day!
Failed prints I have plenty, just made two more haha! I really wanted to do a giveaway with the 10K special, at this point I've collected quite a few PCB projects (like the unipolar stepper driver I recently did) and other stuff that I'd like to actually be of some use rather than sitting on my shelf for a decade before being thrown out. I'll check out how to do giveaways and see if I can organize something :)
I think your original 3dp backlash exerts the force parallel but the newer one tilts the nut (notb parallel).
Did you try the 1st with correct acceleration settings? It may work better because of less friction (because of keeping the nut parallel).
Maybe 2 screws could be used to keep it parallel ?
Good point, I thought about that too but came to the conclusion it doesn't matter. The thread cut by the tap is actually quite loose on the screw, so if it gets tilted it does cause only one half of the nut to actually engage with the thread, but since there's so much play it doesn't lock up the nut. Over time it'll wear in to have a larger contact area.
I didn't try the spring loaded nut with correct accelerations, but the steel nuts were causing much more friction despite being parallel.. Ultimately I might not even switch to brass inserts, plastic seems to work just fine and it makes for significantly quieter operation.
a torch lighter is amazing for removing stringing, you just has to move it fast so you don't melt the print itself
That's what I did before talking a picture of the benchy for the thumbnail :p
very clever build
for the leadscrews of the x and y just support them with a normal bearing and let them expand as much as they want at the end no need to constrain them like that
also print with a multiplier of your nozzle diameter , for a 0.4 diameter print with 0.14 ,0.18 and 0.2 and 0.24 to eliminate layer lines
try using around 1 mm for retraction length and the fastest speed you extruder can go for retraction speed i use 70 mm/s and if you have the option use "wipe while retracting" this will wipe the blob on your part leaving no blob on the part
and for god's sake , use prusha slicer it has more options to compensate for faults in printers
Hey, I think u mean 0.28, 0.24, 0.2, 0.16 and 0.12?
I do use Prusaslicer for slicing, Repetier Host is just sending Gcode from the USB drive. My retraction is 1mm, tho I tried 2mm too, and didn't notice any improvement, it just took longer because the 28BYJ-48 can't really go faster than 4mm/s without loosing steps.
The normal bearing is what I plan on doing, though it'll increase backlash again - I'd have to use two bearings with a slight preload like on a real CNC. I'll try wipe while retracting, thx for the tip
U can get normal (nema 17) stepper motors from old normal printers
I know! I unfortunately don't have access to that kind of old high quality ewaste but I was able to acquire some Epson EM-258 steppers for really cheap from a guy who has!
7:41 "mild steel on mild steel bush" (derogatory)
Lesbian AIs: "what's wrong with that?"
Nothing wrong with lesbian (or gay) mild steel, gust too much grinding action 😜
@@ChronicMechatronic heresy
Ben, please ... this is the second video ... don't grind with the cutting disc.
Congrats on the 10k, and nice job with the printer :). Impressive job with the wooden nuts.
Buut - I did most of the grinding with the edge of the disc? 🥺
For real tho I'm not sure what this cutting wheel is made of, all the grey grinding stones I have totally disintegrate when used on mild steel, I think they're only for high carbon steel (or simply too old and expired?) but that brown (aluminum oxide judging by the color) cutting disc goes through mild steel like butter.
Thanks BTW!
Well done
Turn you acceleration from 800 to like 400~ and slice with lower jerk settings...it will cause the print head to accelerate/decelerate slower giving more torque
Also...what is the voltage rating of those steppers 24vdc? what are you running it at...and current wise...are the motors getting hot?
So awesome that it's running and printing!
Keep em coming!!!!
Accelerations were 400mm/s, 800 is just what I set the maximum to, but then Repetier Host messed up and set print accelerations to zero which made Marlin default to the max settings and it failed :D
For jerk I never touched the Prusaslicer default 10mm/s.
The steppers, IDK, it says 2.8V on the label, but the typewriter I got them from was powered by what I assume to be a 24V transformer. Also I _think_ they were driven in unipolar mode, I'm now driving them in bipolar config with an A4988 set to 150mA if memory serves. Definitely wayyy overclocked, they still reach toasty 60°C
But.... why do you have a whole tub of them? I didn't think anything of it, but now I need to know!
Haha I guess I managed to make a rather trivial fact interesting :p
I just collected them over several years, snipping off the cotton after using them. Sounds gross (hence why you shouldn't ask :D), but you don't usually get earwax on the stem (at least I don't)...
28:50 your nozzle look very ... flat ! what nozzle are you using ?
It's an E3D V6 clone, I actually don't like those flat nozzles at all...
Parafin is not grease resistent, use just parafine in both pieces or the adorable messy parafin with graphyte
Congrats on hitting 10k🎉🎉
Thank you ✨
why not copy the titan extruder and add mechanical advantage to your extruder stepper?
With gear reducers the retractions will get even slower
Gratz dude!
Thanks!
try using brass hex rivnuts! Cheap and reliable. i have them on my first 3d printer's z axis to this day. That printer worked like a rented mule.
Thanks for the tip! I'll have to look online, the hardware stores around me have basically nothing in terms of brass hardware that isn't ridiculously overpriced
It can print even shitty Geeetech filament?? that printer is amazing
So far I didn't find the Geeetech stuff to be all that bad honestly. Maybe I have low expectations, but aside from the occasional tiny bubble it prints pretty good I think?
definitely think people underestimate how useful wood is. the fact you dont need a 3d printer to build your 3d printer is great. if it really costs $50 to put together, thats genius
Yeah well, I cannot deny that the extra zero on the price tag helps with performance... I was definitely under $50 for the printer itself, (there was a little spreadsheet towards the end, don't know if you saw) but only because I had all the wood sitting around already. There's a lot that can be done with wood, it just needs to be designed right to allow for humidity changes - I don't think this printer is anywhere close to ideal in that regard, but it works, so it's close enough I guess
@@ChronicMechatronicthough to be fair, there's some changes that could be done with this printer for fairly cheap that don't line up with the "Do it from scraps from scratch" to significantly increase its usability for not a lot. An extra 100-150 dollars I am fairly certain will go a long way with like another mainboard and some nema-17 ender 3 motors or something. And I imagine those kinds of upgrades you will probably do if you continue to find use in the printer and deem it worthwhile.
That kind of thing is something I imagine a second series could be about. Value upgrades for the printer that make it print well
@@donovan6320 yeah, even just a slightly higher initial budget of $100 would've gone a long way toward making it a usable printer. NEMA 17s would definitely be at the top of the list in terms of components worth splurging for. Different controller board - I don't know, Ramps works just fine I think. More important would've been a screen so I didn't need a host for sending the G-code, as well as having the aluminum rails anodized so they don't wear out as fast. And a pack of real POM wheels I guess.
It's funny you should say that, I definitely plan to do all these upgrades, as well as doing a new series about them! However I'm not much of an upgrade afficionado - I usually deem old designs not worth upgrading since it's all just superficial fixes on a base that ideally would be fully redesigned - so I'm actually designing a whole new printer to build using this one to print the parts :D
@@ChronicMechatronic good luck can't wait to see what comes out.
Would you be interested in trying 3d printed ball bearings with copper bbs for your threaded rods?
Link to video from designer
ua-cam.com/video/Pj2N08Fq3bA/v-deo.htmlsi=srlaQ2Egy8171buY
Cool test, though I consider 3D printed ball bearings more of a gimmick for anything other than large, low load applications. For tiny like 4mm ID bearings the tolerances on printed parts will always be too low, it's either too stiff to rotate easily, has a lot of backlash and/or runout. Keep in mind 3D printers never actually print circles, it's always polygons with x amount of facets, so it'll never be smooth enough.
This is amazing and all but we can we talk about how you hand crafted an artisanal nut out of exotic wood?
It was simple enough of a shape to machine... It did run smoother and with lower friction than the printed PLA nut though, maybe I should try to modify it to add the springy tab to take out backlash
How about Brushed DC motor instead of these stepper motor?
That's what I plan to try next! (at least after I built an actually decent printer using steppers haha)
i built a printer almost as bad, its a corexy from years back (6?) cost me 1400$ over the years working on it. and it still performs horrible. it would be a perfect match for you and your farm.
$50 in materials, $1000 in skill and time.
Pretty much 😂😂
Please make a new video in which to show how to build it the right way from the first try, for people who want to DIY themselves with you as a guide. Even if it ends up being 3 hours long, people will gladly watch the video at least once, and many of them will even make their own version. And maybe make another series in which you try to improve it while going as little over $50 as possible, in which to start with ranking each improvement by the importance in one table, by the price in another table, and by your choice if which to improve in another table.
I'm designing a new, much better printer that should be possible for more people to make, and it'll be fully open source. I'm aiming for a budget of no more than $100 if possible, because $50 just doesn't allow for the right components to get a machine that lasts any reasonable length of time. I had to spend a significant amount of time to keep even this one below the limit, it's just not worth it for people to replicate, since all that time would have to be accounted for too, and off-the-shelf printers these days are just too damn cheap for how good they are, it would be more economical to just work a couple of hours overtime to have money to buy an Ender 3 or an A1 mini from. Filament is fairly expensive too, and that's a consumable - at this point I've spent 3 times as much on filament as on this printer, so yeah...
There will be a new printer series, it'll be cheap and made from a lot of trash too, but not as bare bones as this one. It's also how 3D printing goes, as soon as you get used to one printer and its features you'll immediately want to upgrade for more speed/blinky lights/cameras, etc
What program do you use to design your PCBs in the sponsor segment?
KiCad! It's incredibly easy to pick up, even with no prior CAD skills, just watch a tutorial or two and you're set, at least as far as core functionality goes. Plus it's open source 👍
@@ChronicMechatronic I have actually tried KiCAD a few days ago, and I couldn't for the life of me get it to connect a trace to multiple pins. It just refused to make the connection! I tried for 3 hours before I gave up.
All I need is 4 pins with GND for some fans, dammit!
What is the z axis steps u used for Marlin framework I am using 8mm thread rod from hardware store please calculate for me and give me the number and I'm using nema 17 moters
It's very simple math, I explained it in a different episode, but you can also use a calculator like this one: www.circuitist.com/how-to-calculate-steps-per-mm-lead-screw-and-gt2-timing-belt/
Just look up M8 thread pitch and plug in the values
Is there a technical reason you didn't use cheap drawer Slides with ball bearings?
Or did you not think of it? Also, I would have use vegetable oil and Silicone based lubricant instead, Because Vaseline is a petroleum product which can break down wax.
I would have also used used super glued nuts instead of theading the wood, Would take just a few minutes to cut out a hole for the nuts to lock into with a small Chisel, If you used a Softer wood, you could press the nuts then, super glue the nut in and then flood the wood to fill the pores would be probably been a lot better.
In the US where people put a lot of trash products along the street for people to pick up so everything i've mentioned could have been gotten for free.
need more videos of upgrades HAHAHA
Bro im again. I just made ur attiny driver module using my own at89c2051 with same functionality.
Cool!
congratulations from north Africa
Thx!
hey please provide me sketch up file.
ofcourse the problem was color of the filament :P
OMG I never thought of that! That's probably the reason 🤔
@@ChronicMechatronic As a matter of fact colour does affect some things. Light, opaque colours can contain minerals such as zinc oxide and titanium dioxide that can cause the filament to have a sharper more unwieldy transition from not melty enough to too melty, more difficult to work with, since they absorb some heat until they get to the temperature. Though realistically there shouldn't be a lot of that in yellow. Unless there's also way too much yellow pigment.
Another fun fact, the mentioned minerals are white pigments, but they aren't actually white, they are almost glass clear. The white comes about due to structure and dispersion, and their increased refraction index compared to carrier polymer.
I’m assuming this would be a horrible idea for my first 3d printer?
The worst idea! 😅 3D printing is so easy, unless you do it like me!
For a first printer (if it needs to be cheap) I'd say your best bet is a used and/or broken one someone is trying to get rid of on like Facebook marketplace or something. That way you just need to fix it up a little or have most of the parts for a DIY, and it'll be way better than this, haha!
don't worry my fully 3d printed 3d printer didn't work either so i converted it to be 90% 3d printed prusa i3 mark 3,5 and it prints worse than your current model
i make my gear pullers from m14 nuts filed on one side and drilled and thereaded from the other side, they can't handle long puleys though
Yeah that is the unfortunate nature of experimental machines I guess...
Thanks a bunch for the tip, I'll pick up some big nuts next time I'm at the hardware store! I've seen welded versions using square tubing but nuts sound better because they're sturdier.
PUT A PEN SPRING IN THJE LEAD ASSEMBLY TO MAKE IT WORK
Bros using Windows 7 and still doing better than me
Actually a Linux distro called Q4OS, it's made to look like windows
Congratulations 🎉😊
Thank you 🤗
Wood work!