Well, that was definitely not the expected outcome! This is the nicest machine I built so far, so it's a real shame to see all the effort go to waste... but I learned my lesson! I'll pick this series back up in a few months after I redesigned everything - for now I'll start playing around with this to get at least some basic experience with laser cutting; that way the new version should end up much better still if I can weed out as many of the design flaws in this prototype as possible!
Thank you 🤗 To be really honest I'm not doing great right now, after having had to cancel two planned videos on short notice despite not having any others cued up, and feeling quite burned out in general, but I'm preparing for a new series and will be back with a video as soon as I've got stuff sorted :)
DIY Lathe will be an extreme Project :D I've already considering to get some cheap "mini Lathe" from Vevor for like $200-$300 and then I saw some good TH-camr "Camden Bowen" buying it and fixing everything to make decent lathe jobs. It's 13min well spend entertainment. And now I don't wanna buy the "mini Lathe" :D
There is a quick and honorable fix for this problem with the inconsistency of the screws. You could use only one screw and drive the carriage using the same technic from the parallel ruler of a drawing board. The one I used back on engineering college had two fishing lines and some cleverly put poleies that made the ruler move parallel no mater from witch point it was pushed. I believe you should use some steal cable instead of fishing line, but you can have some for cheap from bicycle parts. Take a look at that before you scrap your machine.
I looked it up and it's definitely cool, but since it requires quite some tension I don't think my machine would withstand it without buckling. The entire frame is designed to grow and shrink with humidity changes, and there just isn't anything structural to take the longitudinal load. Not a big deal though, I'll take the redesign as an opportunity to improve it and change what I don't like about it so far!
It's always soul destroying when a good idea becomes to hard to realise because you hit a road block like that and have to scrap the project. I feel your pain dude, but hopefully you have saved others the pain of running into the same problems, which in a scenario where the person is trying to make the machine for as cheap as possible because they cant afford to spend any money on a machine to improve their quality of life by allowing the to make things for themselves, these insights would ensure they dont get off to a false start and waste what little money they have, your a life saver :). As always a great video dude and i appreciate your commitment to designing machines that anyone can make on a small budget, keep up the good work, much love from England, 8-Bit.
It is! The five days spent grinding those rods are the most painful - I'll try and see if I can reuse them in an actual CNC router one day (with a single screw per axis ofc) But yeah, I really hope this is helpful to other people too - come to think of it, I can't remember ever seeing a video or other documentation online of any CNC projects using threaded rod as lead screws, even though I know it's done rather commonly on budget built machines... Of course it's understandable no one with the means to get something decent even bothers with these cheap options, especially youtubers who get inundated with offers for review machines (heck, I've been offered like 5 laser cutters since starting this project, and my channel isn't exactly large...) Though there really should be more coverage of things like this!
First of all: This video was a pleasure to watch. I also like how you solve some small minor problems! It's amazing! I know, it's crazy idea but since your carriage is running so smooth, you could try to make one lead screw more rigid and discard another one. Then You can make diagonal reinforce and measure the backlash by using only one lead screw. The big thing is: Your laser cable management must be on side, where the lead screw is present! otherwise the cable management will pull on the leadscrewless side and cause the gantry to bend. But its just a crazy idea, you can pull if you not already disassembled the machine :D Greetings and thank you For great content :)
Thanks! I thought about that possibility too, seeing as all the cantilever designs out there do it, but when I took a closer look at it I noticed there's still about half a millimeter of lag between moving one side of the X-carriage and the other one coming with it. I guess it's better than what I have now, but not good enough for a fix. I was actually surprised to see it, given how overkill the entire Y-axis ended up being in rigidness, but it also weighs about half a kilo at the very least so the little play there is in the individual bearings of the POM wheels actually allows some flex. If the rear wheels ran on the rail horizontally it would probably work fine, but as it is the weight introduces too much drag... Also I'm sufficiently fed up with it at this point that I'd rather redesign it and skip the minor inconveniences from having lead screws altogether :)
Wow you have a real 3D printer? And a TT as well? Nice. You can dry PETG with the bed heater. You can use a filament box with one side cut off. You can poke some holes in the top and make sure you have a tiny gap somewhere in the bottom. You can also dress the inside in tinfoil. I don't know why i did it the way i did, but worked for me, takes a day or so to dry. I also suggest getting some double ziploc bags (say IKEA pack of 4.5 and 6 litre bags), and calcium cabronate room dehumidifier packs from the supermarket. You can seal one of those packs with the bag and in a few months when you retrieve PETG it will be super dry. Then i suggest double bagging the whole everything, and never store those dehumidifier packs near anything metal. Wow so much difference in thread? Crazy. I didn't run into this issue since when i used M6, i cut both sides from the same rod. I would have never suspected it! I have never heard it happen to anyone before either!
Yeah, I have a real printer - I mainly took them up on the offer because I figured learning the ropes of 3D printing would be much more difficult on an inherently delicate printer. And having a second one definitely doesn't hurt either ;p I tried the heated bed filament box method dry PLA at some point but it didn't really seem to work - probably cause I half-assed it, I didn't poke any holes in the box, nor did I leave it running for more than 4 hours, since I couldn't stand the constant drone of the cooling fans x) Thanks for the tip with the dehumidifier packs! From my brief googling just now it seems it's the same stuff as road salt, and I happened to find a disused bag of the stuff rotting away in the garage... Just need to figure out how to dehydrate it. Not storing it near anything metal makes sense, having seen the state of the metal scoop that was in there. Yeah the difference in pitch was way more than I expected - never heard of it either, though it doesn't help that I can't remember to ever have seen a video or something where someone tried to make a CNC with it..
I think the grinding machine you made was a huge success. Even branded bearings usually have specified 6-8 microns of runout, but are press fitted a couple of microns which you didnt achieve. And on top of that you had tape on the threaded rod. So your grinding machine even though low rigidity and accuracy was able to get incredible results
Yeah the results I got are pretty decent, though I had to go really slow to get there. What little play there is in the bearings was one of the biggest sources of trouble, if I tried to take off more than ~0.02mm at a time I would get a lot of chatter
I really enjoy your approach of trying to find solutions for non-ideal designs. So few people have your ability to problem solve. Keep it up! I also appreciate the Matthias Wandel bandsaw in the background. I bought plans for than and need to build it.
Interestingly enough no one on the internet seems to have experienced something quite like this before, based on what I was able to deduce from the comments; so my case here might've been a bit of an outlier in how extreme the pitch difference is. But still good to be aware and check in advance! 👍
@@ChronicMechatronic ironically nobody having experienced this is how i stumbled onto this video lol, was looking up threaded rods on CNC out of curiosity and found this video!
Mistakes happen. That's normal. But you can fix it with Klipper. You can drive the second lead screw separately with another stepper motor. In Klipper, you set a different ratio and it manages it by itself. I know that Klipper is a different coffee and out of the budget, but it's worth a try and you can use the laser cutter with pleasure. Looking forward to your next video and you are very inspiring. Keep going.
I know right? Huge respect for the guys and girls who rapidly type away on their keyboard like some sort of mysterious magical proceeding, and come up with a working piece of code in a matter of minutes! 😂
Thanks! It's not what I had in mind at the time, but I really like this type of potentiometer control for pretty much all testing with stepper motors :D
15:45 Titanium isn't as bad as its reputation. My mini-mill handles Ti6Al4V better than even mild steel. Stainless is the hardest. I can only guess that the problems show up when you have a super rigid machine that can eat steel faster than mine does aluminum. Probably have to slow down to similar speed as my machine and get annoyed that it's slower than you're used to.
Good point, maybe it's an old wife's tale - I don't really know a lot about machining, I heard that it's terrible to machine and verified it in a google search where it ranked second, right under inconel on a list of least machinable metals. Though my guess is that time spent on the process is taken into account as much as hardness since it translates directly into cost?...
Don't beat your self up over that i'm amazed you caught that before installing them, I don't think I would have checked that at all, it never occurred to me that threaded rods could do that, but you can bet I'll be checking that in the future before i even head to the register.
I actually discovered it mostly by chance, I was about to start filming the montage and just fiddling about with them when I noticed they didn't mesh...
Very hard work here! Klipper supports different pitch lead screws on one axis I believe, but not laser cutters (not to mention it's a whole different beast from marlin and GRBL). But coreXY is probably going to serve you better anyway - threaded rods tend to vary in pitch by a noticeable amount over as little as a few millimeters, creating a wavy pattern.
Thanks! I'd never have expected Klipper to support it - though it's no good either way because none of the G-code senders for laser out there are Klipper compatible - but thanks tor the info anyway! Knowledge is power as they say... Yeah, coreXY is definitely more suitable, especially as I have all of the parts at hand already - while I'm at it I'll also purge all the unnecessary weight in the gantry and make it less overkill, it's just a laser after all. Probably a good thing overall for the project. I really didn't expect the thread to be that inaccurate, but it makes sense, threaded rods are meant for screwing on nuts which doesn't need tolerances this tight. As @SianaGearz mentioned, with a constant pitch die in manufacturing the final pitch depends on the diameter of the rod that was fed into the machine to begin with, which hadn't even occurred to me.
You can use the bed from the 3d printer as an "emergency dryer", just put a box around the spool and give it 2 hours each side. I don't know the temps off the top of my head but is around 45C for PLA
I tried that once but didn't feel like it made any difference - maybe I just didn't crank it up enough... But I've since bought a filament dryer, which works great :D
What firmware are you running this with? I know on Klipper you can set the rotation distance for the lead screws. The only way to learn is to do everything wrong at least once. Keep up the good work. BTW Don't forget what Adam Savage says... "failure is always an option".
It'll be running GRBL, there is basically no better option for laser machines since most dedicated Gcode senders don't speak the Marlin or Klipper accents. As far as I'm aware Lightburn and Laser GRBL are the two prime candidates for controlling lasers, so ignoring obscure Inkscape plugins which might result in God knows how many issues, and skipping proprietary Lightburn, I'll pretty much have to use LaserGRBL. Too true, usually failure is the most effective way to learn things!
Well - almost no firmware supports two separate steps/mm settings for one axis, even if it's a dual-motor axis with self-squaring capability. It always applies the same resolution setting to both motors. Apparently Klipper does, (offer individual step/mm settings) as somebody else mentioned here in the comments, but the problem is that no G-code sender for laser cutting supports Klipper. Or even Marlin for that matter. The whole software ecosystem for laser cutting isn't anywhere near as advanced as the one for 3D printing. So I basically HAVE to run this laser on GRBL, which is kinda clunky and doesn't support fancy stuff like screens or SD cards. No reason to fix it if it ain't broke I guess, but it's nowhere near as plug&play as 3D printing has become.
You brought threaded inserts for injection molding, they don't hold well in 3d printed parts due to their geometry Pretty common mistake we all done that))
Honestly, I can't see that it would have been obvious that the threaded rod would have had different pitches. Maybe I've just grown up spoiled but I have come to expect manufactured parts to be reasonably within some form of usable tolerance. So, please allow me to disagree with you. You don't get the "L" for this; your supplier does. Hey, thanks for sticking it out to see how the design would have turned out and being flexible enough to pivot to find a way to improve the project👍
Thing is, in terms of threaded rods these are are well within spec for their intended purposes - I wouldn't have expected tolerances to be this loose, and I haven't verified any numbers here, but it makes sense for manufacturers to not obsess over a 2mm/m difference in pitch if all it's ever meant to do is thread into nuts - which by all means works just fine - I would assume not taking this into consideration and double checking the accuracy before purposely using the rods in a much higher stakes application than what they were intended for, would, objectively, qualify as a failure in the engineering process on my part. But anyway, it's nice of you to see it differently :) One of the reasons for finishing it I didn't mention in the video, was so that I could start experimenting with the machine and get some basic experience with laser cutting. That way the new design should turn out even better!
Hi Ben! Great and awesome post! a candidate for >15K views I'm sure. Great and huge job. Glad to know that you got a commercial 3D printer to save time. The diy one is still useful... Rods aren't born equal -- according to Chinese manufacturers. If you really want high-grade precise rods, you should buy them in... Japan. At a much higher price of course. Selfishly I think that 34-minute video with such a brilliant quality and content, should last... 1 hour. Entertaining, informative, pretty lively and again not boring at all! I'll watch it again - as usual - on my XX-inch TV, just as tribute to your efforts and commitment to post high-quality post. Excuse my English. I should pass my comment through the famous German AI-based language translator Deepl which is really good at improving my poor English. I'll do it next time Ben! Have a great week my friend. (FR) Ton naturel fait partie intégrante de la qualité de tes videos. Elles te reflètent. Reste comme tu es, intéressant et captivant à la fois.
Hi Phil! Thanks! I figured most of my regulars missed the TwoTrees printer video because the thumbnail looked very different - but yeah, it's been good to have that printer to play with around off-camera. The DIY printer is cool, and surprisingly reliable now and all, but the linear rails also wear quickly so I probably shouldn't be printing a lot with it until it's served its main purpose as a RepStrap helping to build a better one that'll eventually replace it. It's pretty annoying how quality concerns become relevant the deeper you dig into a topic - I'm still in the process of designing that new DIY printer, and I really wanted to use the 8mm smooth rods I had salvaged from ewaste printers, except none of the Chinese LM8UU linear bearings I got fit properly on these rods. They're all ~0.03mm oversized and have too much play on the salvaged rods which are actually ~0.05mm UNDERSIZED. The bearings do however fit quite nicely on brand new Chinese smooth rods sold for 3D printers, which are actually 8.03mm in diameter. So a few dozen microns basically render valuable components that could have been reused in a different application almost useless. The problem with a 1-hour long video would be that it takes even longer to make :') I still only manage to produce about ~1min10 of finished video per day on average, less even if I hit a bunch of roadblocks that require extra work to resolve like in this one. I wish I could make them in 4K so the resolution would at least be worthy of a big TV screen, my phone can record 4K no problem after all, but until I splurge on a decent computer later this year or early 2025 that won't be in the cards unfortunately... Your English is just fine to understand by the way :)
@ChronicMechatronic Hi Ben! It's a shame those printers rods couldn't be used in this project. They have great precision and a pretty long life without significant deformation. 70 seconds of finished video per day is pretty decent IMHO. I guess the longest part of your work as a director is setting up lights and cameras to give the viewer an interesting perspective. The Swedish Maker (www.youtube.com/@TheSwedishMaker) once complained that it takes a lot of time -- viewers can't imagine it. By the way, you show how to achieve those acrobatic ;) shots.
pancake motors would be still powerfull enough for rack and pinion drive :P and you could make a rack from a threaded road mounted stationarly, plus it would be supported all the way backlash mannagment is also pretty simple
Not worth it, I get the same minor inconveniences of having lead screws in the first place, but also with the pitch on T8 being a lot steeper I don't get enough granularity with my 48 steps/rev motors. As it is I get a resolution of ~0.01mm in half-step mode, I don't think it would work with less. Overall it'll be cheaper I just stick in a bunch of NEMA 17's like everyone else...
Hey big brother I really really enjoy your content I am 17 year's old mechatronics hobbyist I really struggle building stuff because of inadequate tool and specially money constraints my family doesn't support me that much currently building or I say improving and removing error on my diy drill press. learning onshape and designing work bench and a table saw So so many parts are not available in my country online and offline store AliExpress is ban in here, eBay is not available, even if I use other international website the total import texes are are well over 60% Ok ok That's to much bla bla I just want to ask you that how are you and what about your next video it has been 4 months since your last update Bye
Instead of grinding down the lead screw, could you make the coupler bigger to encompass threads? Hey i developed "gatorCAM for cnc". Grab a copy and I'll give you free license keys. You could use it for laser too i bet...or for future cnc machine.
I use a single lead screw below my cnc machine which is huge. You could use 1 for a laser instead of 2. Mine is centered below with wood to reach the edges.
Threads in the coupler would work too, but result in a lot more runout. I was going for as smooth an operation as possible and in my experience 3D printing would never get below like ~0.1mm runout, which is also why I drilled out a single 3mm hole instead of printing one side of the adapter 3mm and the other 3.5 or 4mm. I'll definitely try to reuse these rods on a CNC machine with single screw axes so the grinding wasn't wasted. Cool! I'll check GatorCAM out!
@@ChronicMechatronic let me know if you have problem downloading. Dropbox makes it look like you need to make an account but you don't. Is that confusing? I may need to make a video helping people.
@@ChronicMechatronic instead of printing an adapter, you can use a little plastic tubing with copper wire to pinch it tight. I don't think run out is an issue. One of my 1st few videos shows a specific way to use the copper wire. It has worked for many years without any problem.
well in my machine I did not in fact check the pitch, they might come from the same rod but who remembers really... I will not check for my sake it does kinda work
I agree, better not check! I won't check the ones on my 3D printer either. From what I read here in the comments it doesn't seem that common for threaded rods to be this different in pitch, maybe I really just got unlucky with two very mismatched batches from the factory...
Pretty sure, unless they manufacture M6 rods with marginally different pitches and get them all mixed up afterward? If they were a different diameter I'd definitely have noticed, and the difference isn't enough to be metric fine vs coarse thread either... Here in the comments we collectively came to the conclusion that it was some unfortunate batch difference since they just aren't manufactured to high standards. No idea how common this is, but it obviously happens ¯\_(ツ)_/¯
Hey, don't think those are manufacturing inconsistencies. Lead screws have different pitch and lead, so my guess is that you just got lead screws with different specs
No no, those were all very much sold as M6 thread, if you look up how threaded rod is manufactured it starts to make sense. The dies used to form the thread just aren't that accurate from one machine to another... I never expected this, but it's obviously the case
@@ChronicMechatronic Yeah it makes sense where the inaccuracy comes from. Let's even assume the thread rolling die is the exact same. The pitch can still depend on the diameter of the rods as they come in! Because the die is basically constant angle not constant pitch, it just drives the rod forward at a rate that happens to emerge from the rod's diameter and that angle. And as you noticed the diameter wasn't quite similar enough either. Though from looking at them at a home improvement store i didn't notice them deviating much at all, the threads of rods that i was picking for straightness and holding against an angle to compare together seemed to slot into each other just fine across 1m length. Perhaps you just got extra unlucky with that batch variance.
@@SianaGearz you're right! I hadn't even thought about how the diameter would affect it.. Measuring the diameter on those rods just now, one is about 5.85mm in diameter while the other one measures ~5.9mm for the most part. Annoyingly the Y-axis screw seems to be about 5.9mm too, which is even more unfortunate :(( So there were probably two much more similar ones among the three I got, and I just happened to pick the most dissimilar ones. It also answers the obvious question whether the pitch is at least consistent over the entire length of a single rod: probably good enough for DIY CNC purposes, but not really...
i think you would have figured out the software stuff in less time than you spent babysitting it. and remember that the knowledge gained will pay dividends for your entire life. imagine how many times you will be able to cash in on not being scared of the code :D
I'm a (minor) Marlin contributor and i wouldn't dare do such a modification. You can't imagine how complex and fragile the movement code in Marlin is. I'm vaguely comfortable with many other areas, i have modified the HAL to suit my system better and enable interrupt features, etc. You will notice how many hundreds of people touched most of the Marlin code and how few have actually contributed to the movement routines, these people are spectacularly sharp, and the code in that area is still very far from ideal in behaviour.
@@SianaGearz i was refering to his arduino based ghetto lathe. you could literally just count steps up to a hard coded value and then reverse direction. doesn't have to be pretty.
@@cho4d Oh yeah sure there's any number of way to do that, easy enough. I meant and thought you meant running different step rates across the same axis two motors. This would require a substantial re-engineering of Marlin, which is complex enough.
Yeah, hard-coding steps and speeds is likely what I would've done, but that still would have taken me a day or so to write and debug, the machine would've taken a day longer to build because this also requires a motorized 3rd axis, and then I'd have to wait another four days till it finished grinding to continue with the video. Sure, I could've done something different in the mean time, but the video wouldn't have been out sooner. Next time I'll do it though! I might also simply run GRBL on it and write some G-code to take the rod down to a preset diameter after probing?...
Maybe.. If I had one on each side, probably yes, the new GRBL even supports dual motor axes for gantry self-squaring. But I'd have to use a single one to drive the entire carriage, which I'd estimate weighs easily over a pound, so accelerations would be very limited. It COULD move it no doubt, but if I do install decent motors I want decent performance to boot ;)
You have reduced possible mechanical errors to their minimum, as here th-cam.com/video/L2Gocdz3vYk/w-d-xo.htmlsi=aYqS8BbsbGGoU9fT&t=1035 , chasing the slightest deformation. You're rigorous as usual. That's a must in mechanical design, and it allows you to play a little with rod clearance. You have the spirit of a professional engineer. I have recommended your videos to Malagasy students in Madagascar. Believe me, most of them would like to have your skills and... I quote "such a great workshop". Have a good weekend my friend.
Well, that was definitely not the expected outcome! This is the nicest machine I built so far, so it's a real shame to see all the effort go to waste... but I learned my lesson! I'll pick this series back up in a few months after I redesigned everything - for now I'll start playing around with this to get at least some basic experience with laser cutting; that way the new version should end up much better still if I can weed out as many of the design flaws in this prototype as possible!
I just hope you are doing well, I really enjoy your videos and ideas. Thanks for all your work and time!
Thank you 🤗
To be really honest I'm not doing great right now, after having had to cancel two planned videos on short notice despite not having any others cued up, and feeling quite burned out in general, but I'm preparing for a new series and will be back with a video as soon as I've got stuff sorted :)
Now make a DIY Lathe using these parts!
DIY Lathe will be an extreme Project :D
I've already considering to get some cheap "mini Lathe" from Vevor for like $200-$300 and then I saw some good TH-camr "Camden Bowen" buying it and fixing everything to make decent lathe jobs. It's 13min well spend entertainment. And now I don't wanna buy the "mini Lathe" :D
There is a quick and honorable fix for this problem with the inconsistency of the screws. You could use only one screw and drive the carriage using the same technic from the parallel ruler of a drawing board. The one I used back on engineering college had two fishing lines and some cleverly put poleies that made the ruler move parallel no mater from witch point it was pushed. I believe you should use some steal cable instead of fishing line, but you can have some for cheap from bicycle parts. Take a look at that before you scrap your machine.
I looked it up and it's definitely cool, but since it requires quite some tension I don't think my machine would withstand it without buckling. The entire frame is designed to grow and shrink with humidity changes, and there just isn't anything structural to take the longitudinal load.
Not a big deal though, I'll take the redesign as an opportunity to improve it and change what I don't like about it so far!
It's always soul destroying when a good idea becomes to hard to realise because you hit a road block like that and have to scrap the project. I feel your pain dude, but hopefully you have saved others the pain of running into the same problems, which in a scenario where the person is trying to make the machine for as cheap as possible because they cant afford to spend any money on a machine to improve their quality of life by allowing the to make things for themselves, these insights would ensure they dont get off to a false start and waste what little money they have, your a life saver :). As always a great video dude and i appreciate your commitment to designing machines that anyone can make on a small budget, keep up the good work, much love from England, 8-Bit.
It is! The five days spent grinding those rods are the most painful - I'll try and see if I can reuse them in an actual CNC router one day (with a single screw per axis ofc)
But yeah, I really hope this is helpful to other people too - come to think of it, I can't remember ever seeing a video or other documentation online of any CNC projects using threaded rod as lead screws, even though I know it's done rather commonly on budget built machines...
Of course it's understandable no one with the means to get something decent even bothers with these cheap options, especially youtubers who get inundated with offers for review machines (heck, I've been offered like 5 laser cutters since starting this project, and my channel isn't exactly large...)
Though there really should be more coverage of things like this!
First of all: This video was a pleasure to watch.
I also like how you solve some small minor problems! It's amazing!
I know, it's crazy idea but since your carriage is running so smooth, you could try to make one lead screw more rigid and discard another one. Then You can make diagonal reinforce and measure the backlash by using only one lead screw. The big thing is: Your laser cable management must be on side, where the lead screw is present! otherwise the cable management will pull on the leadscrewless side and cause the gantry to bend.
But its just a crazy idea, you can pull if you not already disassembled the machine :D
Greetings and thank you For great content :)
Thanks! I thought about that possibility too, seeing as all the cantilever designs out there do it, but when I took a closer look at it I noticed there's still about half a millimeter of lag between moving one side of the X-carriage and the other one coming with it. I guess it's better than what I have now, but not good enough for a fix. I was actually surprised to see it, given how overkill the entire Y-axis ended up being in rigidness, but it also weighs about half a kilo at the very least so the little play there is in the individual bearings of the POM wheels actually allows some flex. If the rear wheels ran on the rail horizontally it would probably work fine, but as it is the weight introduces too much drag...
Also I'm sufficiently fed up with it at this point that I'd rather redesign it and skip the minor inconveniences from having lead screws altogether :)
Wow you have a real 3D printer? And a TT as well? Nice.
You can dry PETG with the bed heater. You can use a filament box with one side cut off. You can poke some holes in the top and make sure you have a tiny gap somewhere in the bottom. You can also dress the inside in tinfoil. I don't know why i did it the way i did, but worked for me, takes a day or so to dry.
I also suggest getting some double ziploc bags (say IKEA pack of 4.5 and 6 litre bags), and calcium cabronate room dehumidifier packs from the supermarket. You can seal one of those packs with the bag and in a few months when you retrieve PETG it will be super dry. Then i suggest double bagging the whole everything, and never store those dehumidifier packs near anything metal.
Wow so much difference in thread? Crazy. I didn't run into this issue since when i used M6, i cut both sides from the same rod. I would have never suspected it! I have never heard it happen to anyone before either!
Yeah, I have a real printer - I mainly took them up on the offer because I figured learning the ropes of 3D printing would be much more difficult on an inherently delicate printer. And having a second one definitely doesn't hurt either ;p
I tried the heated bed filament box method dry PLA at some point but it didn't really seem to work - probably cause I half-assed it, I didn't poke any holes in the box, nor did I leave it running for more than 4 hours, since I couldn't stand the constant drone of the cooling fans x)
Thanks for the tip with the dehumidifier packs! From my brief googling just now it seems it's the same stuff as road salt, and I happened to find a disused bag of the stuff rotting away in the garage... Just need to figure out how to dehydrate it. Not storing it near anything metal makes sense, having seen the state of the metal scoop that was in there.
Yeah the difference in pitch was way more than I expected - never heard of it either, though it doesn't help that I can't remember to ever have seen a video or something where someone tried to make a CNC with it..
I think the grinding machine you made was a huge success. Even branded bearings usually have specified 6-8 microns of runout, but are press fitted a couple of microns which you didnt achieve. And on top of that you had tape on the threaded rod. So your grinding machine even though low rigidity and accuracy was able to get incredible results
Yeah the results I got are pretty decent, though I had to go really slow to get there. What little play there is in the bearings was one of the biggest sources of trouble, if I tried to take off more than ~0.02mm at a time I would get a lot of chatter
Love your videos! Mistakes happen to everyone, so don't be down.
I'm not down, just annoyed I wasted so much time on it.. But glad you all still enjoyed it!
I really enjoy your approach of trying to find solutions for non-ideal designs. So few people have your ability to problem solve. Keep it up! I also appreciate the Matthias Wandel bandsaw in the background. I bought plans for than and need to build it.
Thank you for the featured message 😅
A good idea is a good idea :)
SUPER helpful look at cheap threaded rod. saved me from finding this out myself...
Interestingly enough no one on the internet seems to have experienced something quite like this before, based on what I was able to deduce from the comments; so my case here might've been a bit of an outlier in how extreme the pitch difference is. But still good to be aware and check in advance! 👍
@@ChronicMechatronic ironically nobody having experienced this is how i stumbled onto this video lol, was looking up threaded rods on CNC out of curiosity and found this video!
Mistakes happen. That's normal. But you can fix it with Klipper. You can drive the second lead screw separately with another stepper motor. In Klipper, you set a different ratio and it manages it by itself. I know that Klipper is a different coffee and out of the budget, but it's worth a try and you can use the laser cutter with pleasure.
Looking forward to your next video and you are very inspiring. Keep going.
The "Hardwares easy but software is absolutely not" is something i relate too all too much😂
I know right? Huge respect for the guys and girls who rapidly type away on their keyboard like some sort of mysterious magical proceeding, and come up with a working piece of code in a matter of minutes! 😂
Great video!
now i know what you needed the code for :)
Thanks! It's not what I had in mind at the time, but I really like this type of potentiometer control for pretty much all testing with stepper motors :D
Hello, Very good video as always.
It is a great machine, I can't wait to see it work.
thanks for sharing
Thanks!
15:45 Titanium isn't as bad as its reputation. My mini-mill handles Ti6Al4V better than even mild steel. Stainless is the hardest. I can only guess that the problems show up when you have a super rigid machine that can eat steel faster than mine does aluminum. Probably have to slow down to similar speed as my machine and get annoyed that it's slower than you're used to.
Good point, maybe it's an old wife's tale - I don't really know a lot about machining, I heard that it's terrible to machine and verified it in a google search where it ranked second, right under inconel on a list of least machinable metals. Though my guess is that time spent on the process is taken into account as much as hardness since it translates directly into cost?...
Don't beat your self up over that i'm amazed you caught that before installing them, I don't think I would have checked that at all, it never occurred to me that threaded rods could do that, but you can bet I'll be checking that in the future before i even head to the register.
I actually discovered it mostly by chance, I was about to start filming the montage and just fiddling about with them when I noticed they didn't mesh...
Bravo......thread.......ball screw.......better......cheers
Very hard work here! Klipper supports different pitch lead screws on one axis I believe, but not laser cutters (not to mention it's a whole different beast from marlin and GRBL). But coreXY is probably going to serve you better anyway - threaded rods tend to vary in pitch by a noticeable amount over as little as a few millimeters, creating a wavy pattern.
Thanks! I'd never have expected Klipper to support it - though it's no good either way because none of the G-code senders for laser out there are Klipper compatible - but thanks tor the info anyway! Knowledge is power as they say...
Yeah, coreXY is definitely more suitable, especially as I have all of the parts at hand already - while I'm at it I'll also purge all the unnecessary weight in the gantry and make it less overkill, it's just a laser after all. Probably a good thing overall for the project.
I really didn't expect the thread to be that inaccurate, but it makes sense, threaded rods are meant for screwing on nuts which doesn't need tolerances this tight. As @SianaGearz mentioned, with a constant pitch die in manufacturing the final pitch depends on the diameter of the rod that was fed into the machine to begin with, which hadn't even occurred to me.
You can use the bed from the 3d printer as an "emergency dryer", just put a box around the spool and give it 2 hours each side. I don't know the temps off the top of my head but is around 45C for PLA
I tried that once but didn't feel like it made any difference - maybe I just didn't crank it up enough...
But I've since bought a filament dryer, which works great :D
@@ChronicMechatronic Awesome!
What firmware are you running this with? I know on Klipper you can set the rotation distance for the lead screws. The only way to learn is to do everything wrong at least once. Keep up the good work. BTW Don't forget what Adam Savage says... "failure is always an option".
It'll be running GRBL, there is basically no better option for laser machines since most dedicated Gcode senders don't speak the Marlin or Klipper accents.
As far as I'm aware Lightburn and Laser GRBL are the two prime candidates for controlling lasers, so ignoring obscure Inkscape plugins which might result in God knows how many issues, and skipping proprietary Lightburn, I'll pretty much have to use LaserGRBL.
Too true, usually failure is the most effective way to learn things!
19:21 appreciated 😊
I was setting up the shot and thought I had to take a 3rd person view just in case because of how absurd it would look 😅
Is it not possible to use dual motor Y and adjust steps per mm? You'd get double the power too.
Well - almost no firmware supports two separate steps/mm settings for one axis, even if it's a dual-motor axis with self-squaring capability. It always applies the same resolution setting to both motors. Apparently Klipper does, (offer individual step/mm settings) as somebody else mentioned here in the comments, but the problem is that no G-code sender for laser cutting supports Klipper. Or even Marlin for that matter. The whole software ecosystem for laser cutting isn't anywhere near as advanced as the one for 3D printing. So I basically HAVE to run this laser on GRBL, which is kinda clunky and doesn't support fancy stuff like screens or SD cards. No reason to fix it if it ain't broke I guess, but it's nowhere near as plug&play as 3D printing has become.
You brought threaded inserts for injection molding, they don't hold well in 3d printed parts due to their geometry
Pretty common mistake
we all done that))
Oh, I didn't realize that's why they looked different! Thanks for the info!
Honestly, I can't see that it would have been obvious that the threaded rod would have had different pitches. Maybe I've just grown up spoiled but I have come to expect manufactured parts to be reasonably within some form of usable tolerance. So, please allow me to disagree with you. You don't get the "L" for this; your supplier does.
Hey, thanks for sticking it out to see how the design would have turned out and being flexible enough to pivot to find a way to improve the project👍
Thing is, in terms of threaded rods these are are well within spec for their intended purposes - I wouldn't have expected tolerances to be this loose, and I haven't verified any numbers here, but it makes sense for manufacturers to not obsess over a 2mm/m difference in pitch if all it's ever meant to do is thread into nuts - which by all means works just fine -
I would assume not taking this into consideration and double checking the accuracy before purposely using the rods in a much higher stakes application than what they were intended for, would, objectively, qualify as a failure in the engineering process on my part.
But anyway, it's nice of you to see it differently :)
One of the reasons for finishing it I didn't mention in the video, was so that I could start experimenting with the machine and get some basic experience with laser cutting. That way the new design should turn out even better!
Love the laser, somehow it reminds me the dog Peggy from the last Dead Pool movie
😅
What if you stretched the shorter rod to undergo plastic uniform elongation? You could build a new jig!!
Definitely one of the crazier ways of fixing it, but also one of the coolest!
thank you, youtube algorithm, for bringing me this video.
Glad it did!
Thanks for watching!
Hi Ben! Great and awesome post! a candidate for >15K views I'm sure. Great and huge job. Glad to know that you got a commercial 3D printer to save time. The diy one is still useful... Rods aren't born equal -- according to Chinese manufacturers. If you really want high-grade precise rods, you should buy them in... Japan. At a much higher price of course.
Selfishly I think that 34-minute video with such a brilliant quality and content, should last... 1 hour. Entertaining, informative, pretty lively and again not boring at all! I'll watch it again - as usual - on my XX-inch TV, just as tribute to your efforts and commitment to post high-quality post.
Excuse my English. I should pass my comment through the famous German AI-based language translator Deepl which is really good at improving my poor English. I'll do it next time Ben! Have a great week my friend.
(FR) Ton naturel fait partie intégrante de la qualité de tes videos. Elles te reflètent. Reste comme tu es, intéressant et captivant à la fois.
Hi Phil! Thanks!
I figured most of my regulars missed the TwoTrees printer video because the thumbnail looked very different - but yeah, it's been good to have that printer to play with around off-camera. The DIY printer is cool, and surprisingly reliable now and all, but the linear rails also wear quickly so I probably shouldn't be printing a lot with it until it's served its main purpose as a RepStrap helping to build a better one that'll eventually replace it.
It's pretty annoying how quality concerns become relevant the deeper you dig into a topic - I'm still in the process of designing that new DIY printer, and I really wanted to use the 8mm smooth rods I had salvaged from ewaste printers, except none of the Chinese LM8UU linear bearings I got fit properly on these rods. They're all ~0.03mm oversized and have too much play on the salvaged rods which are actually ~0.05mm UNDERSIZED. The bearings do however fit quite nicely on brand new Chinese smooth rods sold for 3D printers, which are actually 8.03mm in diameter. So a few dozen microns basically render valuable components that could have been reused in a different application almost useless.
The problem with a 1-hour long video would be that it takes even longer to make :') I still only manage to produce about ~1min10 of finished video per day on average, less even if I hit a bunch of roadblocks that require extra work to resolve like in this one. I wish I could make them in 4K so the resolution would at least be worthy of a big TV screen, my phone can record 4K no problem after all, but until I splurge on a decent computer later this year or early 2025 that won't be in the cards unfortunately...
Your English is just fine to understand by the way :)
@ChronicMechatronic Hi Ben! It's a shame those printers rods couldn't be used in this project. They have great precision and a pretty long life without significant deformation.
70 seconds of finished video per day is pretty decent IMHO. I guess the longest part of your work as a director is setting up lights and cameras to give the viewer an interesting perspective. The Swedish Maker (www.youtube.com/@TheSwedishMaker) once complained that it takes a lot of time -- viewers can't imagine it. By the way, you show how to achieve those acrobatic ;) shots.
pancake motors would be still powerfull enough for rack and pinion drive :P and you could make a rack from a threaded road mounted stationarly, plus it would be supported all the way
backlash mannagment is also pretty simple
First! Love your vids
'grats! thanks!
Would switching to T8x2 3DP rods be worth it? They should be easier to lubricate and with brass nuts work well. But i‘am not really sure about that.
Not worth it, I get the same minor inconveniences of having lead screws in the first place, but also with the pitch on T8 being a lot steeper I don't get enough granularity with my 48 steps/rev motors.
As it is I get a resolution of ~0.01mm in half-step mode, I don't think it would work with less. Overall it'll be cheaper I just stick in a bunch of NEMA 17's like everyone else...
Hey big brother I really really enjoy your content
I am 17 year's old mechatronics hobbyist I really struggle building stuff because of inadequate tool and specially money constraints my family doesn't support me that much currently building or I say improving and removing error on my diy drill press. learning onshape and designing work bench and a table saw
So so many parts are not available in my country online and offline store AliExpress is ban in here, eBay is not available, even if I use other international website the total import texes are are well over 60%
Ok ok
That's to much bla bla
I just want to ask you that how are you and what about your next video it has been 4 months since your last update
Bye
Instead of grinding down the lead screw, could you make the coupler bigger to encompass threads?
Hey i developed "gatorCAM for cnc". Grab a copy and I'll give you free license keys. You could use it for laser too i bet...or for future cnc machine.
I use a single lead screw below my cnc machine which is huge. You could use 1 for a laser instead of 2. Mine is centered below with wood to reach the edges.
Its a little slow for a laser , but that would make a good cnc....with 1 lead screw on the bottom.
Threads in the coupler would work too, but result in a lot more runout. I was going for as smooth an operation as possible and in my experience 3D printing would never get below like ~0.1mm runout, which is also why I drilled out a single 3mm hole instead of printing one side of the adapter 3mm and the other 3.5 or 4mm.
I'll definitely try to reuse these rods on a CNC machine with single screw axes so the grinding wasn't wasted.
Cool! I'll check GatorCAM out!
@@ChronicMechatronic let me know if you have problem downloading. Dropbox makes it look like you need to make an account but you don't. Is that confusing? I may need to make a video helping people.
@@ChronicMechatronic instead of printing an adapter, you can use a little plastic tubing with copper wire to pinch it tight. I don't think run out is an issue. One of my 1st few videos shows a specific way to use the copper wire. It has worked for many years without any problem.
well in my machine I did not in fact check the pitch, they might come from the same rod but who remembers really... I will not check for my sake it does kinda work
I agree, better not check! I won't check the ones on my 3D printer either. From what I read here in the comments it doesn't seem that common for threaded rods to be this different in pitch, maybe I really just got unlucky with two very mismatched batches from the factory...
Are you sure you actually had the same thread pitch? And not just two different sticks mislabeled?
I use alot of all thread and have never seen that
Pretty sure, unless they manufacture M6 rods with marginally different pitches and get them all mixed up afterward? If they were a different diameter I'd definitely have noticed, and the difference isn't enough to be metric fine vs coarse thread either...
Here in the comments we collectively came to the conclusion that it was some unfortunate batch difference since they just aren't manufactured to high standards. No idea how common this is, but it obviously happens ¯\_(ツ)_/¯
Hey, don't think those are manufacturing inconsistencies. Lead screws have different pitch and lead, so my guess is that you just got lead screws with different specs
No no, those were all very much sold as M6 thread, if you look up how threaded rod is manufactured it starts to make sense. The dies used to form the thread just aren't that accurate from one machine to another... I never expected this, but it's obviously the case
@@ChronicMechatronic Yeah it makes sense where the inaccuracy comes from. Let's even assume the thread rolling die is the exact same. The pitch can still depend on the diameter of the rods as they come in! Because the die is basically constant angle not constant pitch, it just drives the rod forward at a rate that happens to emerge from the rod's diameter and that angle. And as you noticed the diameter wasn't quite similar enough either.
Though from looking at them at a home improvement store i didn't notice them deviating much at all, the threads of rods that i was picking for straightness and holding against an angle to compare together seemed to slot into each other just fine across 1m length. Perhaps you just got extra unlucky with that batch variance.
@@SianaGearz you're right! I hadn't even thought about how the diameter would affect it.. Measuring the diameter on those rods just now, one is about 5.85mm in diameter while the other one measures ~5.9mm for the most part. Annoyingly the Y-axis screw seems to be about 5.9mm too, which is even more unfortunate :((
So there were probably two much more similar ones among the three I got, and I just happened to pick the most dissimilar ones.
It also answers the obvious question whether the pitch is at least consistent over the entire length of a single rod: probably good enough for DIY CNC purposes, but not really...
Get a hammer and squish the longer wavelength screw until it's in tolerance
😂
i think you would have figured out the software stuff in less time than you spent babysitting it. and remember that the knowledge gained will pay dividends for your entire life. imagine how many times you will be able to cash in on not being scared of the code :D
I'm a (minor) Marlin contributor and i wouldn't dare do such a modification. You can't imagine how complex and fragile the movement code in Marlin is. I'm vaguely comfortable with many other areas, i have modified the HAL to suit my system better and enable interrupt features, etc.
You will notice how many hundreds of people touched most of the Marlin code and how few have actually contributed to the movement routines, these people are spectacularly sharp, and the code in that area is still very far from ideal in behaviour.
@@SianaGearz i was refering to his arduino based ghetto lathe. you could literally just count steps up to a hard coded value and then reverse direction. doesn't have to be pretty.
@@cho4d Oh yeah sure there's any number of way to do that, easy enough. I meant and thought you meant running different step rates across the same axis two motors. This would require a substantial re-engineering of Marlin, which is complex enough.
Yeah, hard-coding steps and speeds is likely what I would've done, but that still would have taken me a day or so to write and debug, the machine would've taken a day longer to build because this also requires a motorized 3rd axis, and then I'd have to wait another four days till it finished grinding to continue with the video. Sure, I could've done something different in the mean time, but the video wouldn't have been out sooner.
Next time I'll do it though!
I might also simply run GRBL on it and write some G-code to take the rod down to a preset diameter after probing?...
nema 17 pancakes on a carriage that smooth i think would work still.
Maybe.. If I had one on each side, probably yes, the new GRBL even supports dual motor axes for gantry self-squaring. But I'd have to use a single one to drive the entire carriage, which I'd estimate weighs easily over a pound, so accelerations would be very limited. It COULD move it no doubt, but if I do install decent motors I want decent performance to boot ;)
A millimeter per foot! That's a weird error in the screws.
Well, a millimeter every 35cm approximately, I just used imperial because 1 foot was a nice single unit to visualize things
You have reduced possible mechanical errors to their minimum, as here th-cam.com/video/L2Gocdz3vYk/w-d-xo.htmlsi=aYqS8BbsbGGoU9fT&t=1035 , chasing the slightest deformation. You're rigorous as usual. That's a must in mechanical design, and it allows you to play a little with rod clearance. You have the spirit of a professional engineer. I have recommended your videos to Malagasy students in Madagascar. Believe me, most of them would like to have your skills and... I quote "such a great workshop". Have a good weekend my friend.
Video at kanka kuruduk aq