Viewer question: the ball bearing trick?

[@NBFL]

Finally found someone who knows what they're talking about!

[@floorpizza8074]

This video just solved my Z axis problem, where 20 other videos couldn’t. Thank you!!

[@floorpizza8074]

I’m sorry for the double comment, but I just had to add yet another huge THANK YOU for this tip. I have a brand new Voxelab Aquila (Ender 3 clone) that was losing a HUGE number of Z axis steps. After following all the advice I could find, I got it improved to the point that I was just losing a few Z steps. I went from not even being able to print a first layer test successfully to being able to print, but with enough lost Z steps to affect print quality. To get to that point, I: Lubed the lead screw Shimmed the motor mount so that my Z axis lead screw was the same distance away from the frame at the bottom as it is from the top Added a flexible coupler After reading your explanation about how a flexible coupler could absorb z steps, I had an “ah ha!” moment, realizing this was my problem. I added the quarter inch ball bearing to the flexible coupler, inserted the lead screw until it rested on the ball, and my troubles were completely solved. NO MORE lost Z steps!! NONE! NADA! ZILCH! ZERO! THE EMPTY SET! After chasing this rabbit for a week and losing the rest of my little remaining hair, I printed the most perfect first layer test I’ve ever seen. Did I mention no lost Z steps?? THANK YOU SO MUCH!!

[@predragbalorda]

smartest thing on youtube in a long time..

[@neilbradley9035]

Alex, I truly appreciate your content! Always making accessible videos, but somehow much more technically in-depth than most other channels! Thanks for the tricks, my friend!

[@PleasantHillRealEstateGroup]

New to 3D printing and this was freaking awesome! I'm on my 3rd real print with a new Ender 3 and having some z axis biding issues. I didn't get everything exactly dialed in out of the box so my first two prints came out supper smooth and then my machine started to come apart....not the machines fault or the design or the slice its the pilot setting up the equipment for sure, since I'm new.....I don't know what I don't know yet. Thank you for imparting your experience and wisdom. This community has been supper supportive and I really think my learning curve has been diminished by learning from other makers experiences. Again Thank you!

[@VikMKW]

Down with the sub 10 minute videos! I'm glad you highlighted your setup, as there's an abundance of wrong information regarding "z-wobble fixes" on forums and Facebook pages. It's becoming difficult to filter knowledgeable information, but you're single handedly debunking these myths. Thank you as always!

[@3dprintandpaint]

This is a must! What an amazing difference this has made using the bearing trick. I have had issues for months with Z binding regardless of numerous adjustments. No more! You sir are a legend! Highly recommended to all Ender 3 owners!

[@matthewframe3340]

Thank you very much sir! I had not run across the bearing trick for load support and this resolved my final print issues in my new corexy machine. I really appreciate the detail you provide.

[@BikerCaf]

I love your videos Alex. To add to the ball bearing trick, I also stretch out the flexible coupling by about 1.5 to 2mm (like pulling a spring apart) and hold it there whilst tightening it onto the shafts. That way it also provides anti-backlash in the other direction (pulling) as well, that's of coarse assuming the linear acceleration is within limits.

[@AlexKenis]

Good call. I do the same thing when using a bearing block and it works well

[@jirij]

Worth adding that this is an old machinist's trick for clamping or measuring 2 (or more) surfaces using a single point. :) Also works in round stock form to cheat dovetail slots. ... Thanks for another useful ... use!

[@AlexKenis]

Old machinists have the best tricks

[@ssjronin3972]

@@AlexKenis can this be effective if 1 motor has a spring coupler and bearing while the 2nd motor has the stock coupler? Or would you only run 2 spring couplers on your machine for it to be efficient?

[@nghin4vacsf]

This is a great tip. I just designed and printed a set of spider couplers to replace the stock ones and noticed they settled down a little over time. This ball bearing trick should solve the problem. Thanks for sharing.

[@cri8tor]

Since getting into the knowledge part, before even putting my Ender 3 Pro together and printing my first Benchy, I've been watching a shit load of tutorials. In my opinion, you have a great balance of detail and "to the point" information on 3D printing, so keep making that epic content. Cheers

[@wangouthangout]

5 years later I hear about this... Brilliant content.

[@kremicfein]

LOL, your gesture when you say the word "coupling" ... priceless @ 0:37

[@rentaspoon219]

Yeah I stopped paying attention to him while his hands moved about

[@spock6692]

Thank you very much, Alex! You are analyzing important problems when a solution is not lying on the surface. I agree with most of your points and, particularly, with the final conclusion/solution. However, I have something to add to your and CM 62's words. I have solved this problem on my printer forever. My printer has a lot of other problems, but this Z-axis wobbling and bouncing happen no more. I also use a spiral aluminum coupling, mine just has two setting screws on each end and the screws are not opposite, they are at the straight angle. The tolerance in both holes (on the stepper side and on the acme screw side) is about the same and when I tighten the screws, they make a triangle (the contact spot between the coupling and the screw and the two setting screws). The setting screws are made on the same side (are aligned) on both ends, so they push the acme screw and the stepper shaft the same way inside the coupling. Thus, when the whole thing turns, the coupling is slightly off axis (is wobbling), but the stepper shaft and the acme screw are aligned pretty well. And then about the ball between them. Before assembling I made kind of spherical shape on the end of the acme screw. I turned the screw in a lathe chuck, and used an angle grinder with a sandpaper on it. Lathe is not necessary, you can use a drill and a third hand instead. My spherical end plays the same role as Alex's ball, but is just one part less. And, yes, I slightly pre-load the spiral coupling as CM 62 does before tightening the stepper side screws. Works PERFECTLY for Z axis (in my printers Z is a vertically, and slowly, moving table), absolutely no side effects. I would not recommend this for other axes that move with high accelerations in both ways, but this is another long story. But for the table, slow and under gravity force - perfect. Even the cheap brass (or bronze) nut with a huge play is not a problem, the gravity holds it in place. But don't use such a printer lying on a side :) .

[@AlexKenis]

Sounds like a good way to go about it.

[@spock6692]

@@hoffer_moment Hi Chris! I did not understand what you mean 100%, but it seems to me you got it right. I just can say the same in a little different way. From a factory the stepper's shaft is cylindrical. And the acme screw is cylindrical too in general (if we do not take the thread into account). The point is to make the flat bottom (or lower end) of the acme screw slightly spherical. It does not have to be a half-sphere (with the screw diameter), not at all. It should be just slightly protuberant, a spherical cap of a sphere much bigger in diameter than the screw. And the end of the stepper's shaft should be left untouched, we should not make it spherical. Or actually the opposite way: if it is more convenient for you, you can make the stepper's shaft spherical and leave the acme screw untouched. Then one word about preloading of the couplind. I think, 2 mm of stretching is too much, I would recommend 0,5 mm, maybe up to 1 mm. Then you write "lathe sand the threaded rod to ensure its straightness" - if you mean grinding the side surface (the cylindrical surface) of the acme shaft - NO! You will spoil it by doing that. You should only grind the cylinder bottom at one end of the acme screw to turn it to cpherical cap. If your acme screw came crooked, just streighten it by bending (or UNbending :) ) and a lathe is a good tool for it. I was securing the acme screw in the lathe's chuck, turning the chuck by hand to see the direction of deviation usind a lathe cutting tool as a reference and bent the screw by hand. This process is iterative, but after a few minutes, after bending it in 4-5 points evenly distributed along its length I got an almost perfect screws. By the way, I did the same with polished cylyndrical shafts which the axes move along. And I did not understand your phrase "so it can rotate on the stepper shaft". The stepper's shaft, the coupler and the acme screw turn together, they do not turn against each other, I think, that was just my misunderstanding of you. Well, this is how I see it. If you have more questions, just usk :) . Happy printing!

[@tonylayne2492]

AK, love your videos, thank you for taking the time.

[@TheParachaz]

Ha! Love the subtle innuendos, and your narration.

[@vassoritzen]

I made a small concave recess in the perpendicular face with the dremel micro-rectifier. This ensures that I can place a ball of any size and it will be centered.

[@exotime]

Wowza thanks man! Subscribed for more wisdom. The ball bearing trick is just brilliant, it's a multiaxial joint created with no actual bearings. It's akin to a ball and socket joint like a hip or a shoulder, just brilliant.

[@radry100]

The most effective thing that I found to deal with out of alignment issues is an oldham coupler fixed to the nut.

[@robmckennie4203]

This is a really good idea, I don't have any issues with my z axis but it has always annoyed me that the manufacturer decided to support the gantry directly on the flexible couplings

[@AndrewAHayes]

Overtightening the leadscrew nut can also cause binding of the leadscrew, I have purchased 3 Ender 5 Plus machines and on each one the leadscrew nut screws were fully tightened and caused screeching and binding, one of them was so tight it knocked the part off the bed at exactly the same layer each time, luckily this was the first Ender 5 Plus I bought and this issue was what drew me to the fault, when I built my Ender 3 and CR-10 machines the instructions warned against tightening the leadscrew nut too much, the method I employed was to tighten the screws fully but without torquing them up and then turning them back one half turn.

[@Ultrasonictwo]

@alex kenis Now you have forgotten about the spring washer inside of the motor on the bottom that depending on the machining of the housing holding the motor bearing can allow for significant shaft movement

[@Speedlution]

I have spent the last hour trying to get my z axis binding removed. After following every trick in the book, I loosen my z motor and then having everything else tightened I can feel the z motor wobbling as I move it up and down showing that I have a slight bend in my acme screw. I am now looking into the best option for resolving the binding issue and this trick might just do it without having to order a new acme screw which could still be bent. Thank you for the explanation.

[@lacalavera5640]

this was a life saver for mi, thanks, i still having some banding binding ringin kind of problem but this help like no other solution, also simplify 3D is doing some strange things with mi models, never thought that the ringing could be a slizer problem but i printed the same gcode on two of mi printers, exact patern of banding or ringing the two prints, it mean that the problem its in the software for the most part

[@kokodin5895]

fun fact my reprap was oryginally designed to use flexible coupler and it worked fine but as you say it was squishy so every slight problem made a 2mm pattern on the printed part in my case the leadscrew was resting on the motor shaft anyway, so ball bearing wouldn't fix much but swaping the coupling for one from ender 3 fixed all the problems and my printer is using regular m8 rod and brass nuts with backlash spring

[@ryanmacewen511]

Awesome job Alex. I've been trying to preach this to customers forever. Creality printers would often end up with stretched out couplers (non-spring type), so there 'solution' was to go with solid ones. Yikes! I still recommend the spring style couplers, as the gantry on CR-10 style printers is not that heavy, especially if using dual Z threaded rods. Yet other vendors still unwittingly sell couplers that are solid as 'upgrades'. I need a smaller TV, my back isn't what it used to be.

[@TheRainHarvester]

I had the spongy effect on my cnc y axis. I shoved cut-up-credit cards into the coupler's spirals. I have 24 cnc videos with weird fixes/tips like that.

[@dokterzorro]

Came up with that one as well, took some thick paper, cut strips and stuffed pieces of it in the spirals while forcing the axle up so there's more space in the spirals and the paper really bears the load when compressed by weight. It still is flexible in x/y direction so is still functional but the Z-height repeatability dramatically improved. Weird to notice this design flaw was never really adressed by the 3d community, the solution is so simple but the provailing idea seems to be that the z-axis isn't really that important since it doesn't move "fast" I see lots of vids about z-axis wobble, all of them showing the stock spiral couplers and completely ignoring them as a cause. Same goes for upgrading linear bearings, people stuff Igus bearings in the x and y carriages all the time but nobody thinks it necessary to upgrade the z-axis ones as well. The stock ball bearing ones are crap and have terrible slop resulting in ugly prints and rattling. Any loose tolerance parts in a cnc-machine are going to have bad impact on quality. CNC aims for 0% tolerance parts since software assumes a perfect machine without play. The closer to perfect you get, the better the prints will be and these spiral couplers have a serious, underestimated effect on z-axis tolerance. I consider it a flawed design, you really don't want any flexible parts in a cnc drive, luckily it is easily remedied with some paper.

[@adisharr]

One other thing, the dual set screws are not for alignment but for motor shafts that have two flats for additional holding.

[@CDN_Torsten]

Another excellent video - love the detail - thanks Alex!

[@tomu3762]

Your videos are excellent Alex! I'm on a limited budget, but I'll find ways to get you some funds through your links. Please keep teaching us!

[@victorfranco2873]

Works for me, even on a bent lead screw. Thanks!

[@gregc2222]

You showed a flange ball bearing being pressed into that white plastic mount. It should be pointed out that the bearing, in addition to having the flange for mounting purposes, should also be a "thrust bearing" to take the axial load, not just a standard ball bearing meant primarily for radial loads. Now, a radial load bearing *can* take a certain, limited, amount of axial load, but it I think it would be worth using a video like this to point out the difference between the 2 bearing types as I don't think most people are familiar with thrust bearings.

[@kkuenzel56]

I have noticed that since I bought my Cr10 mini, Creality has switched to using a Lovejoy style coupler on the lead screw.

[@LT72884]

Wow, could not get over the fact of how much you remind me of brent off of "corner gas" Oh, great video by the way

[@thiagoennes]

I disagree, man. That is a clean elegant solution! My only concern would be friction between the shaft ends and the bearing in a million years causing some problems. I am a strong advocate of creating error tolerances even when you are doing things right. Aligning properly and using that trick is a good way to go...

[@shiftyjesusfish]

At what point do I stop chasing 0.2 mm misalignment, I've got everything as square as can be with the tools I have (verified with 2 squares, a pair of digital calipers and 2 crappy tape measures). Still binding, probblem is this printer always bound in z and also printed really well (85% of the time with the other 15% being really hard to figgure out bullshit failures leading me to tear down the whole machine and rebuild it).... but for the life of me I can't get it to move smoothly in z. At this point with the learning, tuning and fixing I swear I could make my own printer at this point :), is it possible my vertical z extrusions are twisted (they are not bowed but Idk how to check for twisting on them) . Sorry essay sized comment. Im losing my mind over this one, thanks for the video it did me alot despite the rant :)

[@gushhnet]

I know this is a 9 month old comment but just in case perhaps your acme nut/s are too tight, try to loose the screws one half turn and see if the coupler with a ball bearing also helps align, the binding is most likely caused by the nut being screwed on too tight to the frame and possibly a bent leadscrew.

[@TotoGuy-Original]

i use a flexible coupler with polyurethane insert in the middle and it has support for the head screw and keeps the correct gap between the motor shaft and the lead screw. seems similar to your bearing trick only i dont think its as flexible as the couplers you were using but it works! so thats the main thing

[@bonafidepirate]

I've used this "fix" for a while now on my Anet A8 turned EMT8, with success. I have the more rigid type coupler that came on my Ender 3 and have fought to get the Z axis running smooth since day one... lol

[@seanneal552]

This is NOT a hack. This is a practical solution for a very critical problem. I've sat through endless hours of RCA's. This is the solution mechanics, like myself, would come up with that would annoy the absolute shit out of every engineer. Easy way to find allies (and enemies) in a workplace!

[@ArandomNutter]

Your channel is legit mate well done I liked the video

[@Bednar121]

In Tevo Tarantula Pro there is an o-ring that is meant to do exactly that, just inserted between the motor shaft and the lead screw, although I think that the ball bearing would do a better job :D

[@gushhnet]

Won't the oring crush eventually, merely acting as a wannabe dampener of sorts? I just added the ball bearing to a new flexible coupling on the Z axis, we shall see if wobble is reduced!

[@Bednar121]

@@gushhnet Well, I had the printer for 5 years or so now, printed a literal ton of filament, replaced the lead screw nut twice now, and the o-ring's still looking fine and dandy, so I don't know :v

[@ichigohijkata779]

It would be better if the printer uses flexible couplings to have to mount the motors on top and the screw would be hanging? Would gravity help in this case?

[@3D_Printing]

6:36 Steppers are not built for up/down loads so that is a good idea I guess :)

[@AlexKenis]

Indeed

[@luanmuraro2929]

Can i use 608zz spheres?

[@christopherstaples6758]

thank you this solved my random extrusion issues

[@francescodigiovanni4632]

Hi Alex, could an Oldham coupler between the motor and the Z axis screw solve the problem?

[@thiagokeizo]

what about disk coupling would it work well to mitigate small misalignments ?

[@avejst]

Fantastic video as always 👍 Thanks for sharing 😀👍

[@wernerboden239]

I designed a motor bracket, with an extension to hold the coupler. I printed it with PETG. I should post it on thingyverse or so. When I do, I'll get back to you.

[@AlexKenis]

Sorry man, I have not been deleting anything. If you post frequently or post a link, the comments get held in a filter until I manually approve them, and my computer has been down for a few days with a scrembled boot drive, so I have not been checking. But go ahead and post the link when you get it finished, and I'll keep an eye on the filter if it gets caught up

[@davidprock904]

If I pick a handful of KZbinrs about 3D printing (you being one) and I can get us all all in a conference call with video if possible so “talking with hands to help with visuals” will help. .... And in this live group call we form a “company/Organization” for my truly revolutionary design of a printer to be sold as a product. I’m not asking for money, I am a over the road semi truck driver, I make plenty of money for everything needed, I just don’t have the time to do anything other than talk really. And I live out of my Simi Truck. So I don’t have build space. Most of my design is set in stone, but there are multiple ways to do different factors of my design, can we all discuss it and maybe add tweaks to make it better. But what we all discuss as a collective does not leave the project and stays private until it is ready for sale, can we all agree on this and talk about

[@cfriedalek]

Thank god someone makes videos that are longer than 5 minutes and have some meat! Nice. So how much of the enticingly and incredibly smooth layers you showed in the previous video would you put down to this load supporting "trick"? Of course I'll try it regardless but I'm curious about your sense of the improvement.

[@Tomas-ir8xl]

Are you working for "Ruthex"? They are selling exactly the same type of steel ball 3d printer couplings you describe in this video.

[@Levisgil47]

Do you know a trick to compensate bending on a lead screw driven by a motor placed on top instead? I believe the ball placed in the coupler won't work in that case. I got an old tevo tarantula. Thanks

[@mertcapkin7263]

I bought a Nema 17 which has the lead screw integrated. That is better, right?

[@AlexKenis]

I have not done any long-term testing, but an integrated leadscrew should be good in terms of alignment and preventing eccentric motion as long as it is straignt. My only long-term concerns are that the weight of the gantry is resting on the motor (As Werner mentioned) which is one of the reasons I use a bearing block, BUT companies like Prusa seems to be having good luck with it, so it may be fine. I'll be keeping an eye on Prusa user feedback to see how it does long term, and if it seems to be an issue, I'll address it in a vid.

[@nightsaber2272]

Stepper has spring washer at one od it's bearings so this ball trick don't eliminate all axial movements.

[@alfrecletero]

If you have severe axial missalignment as mentioned in the last part of the video, and you can't fix it for some reason, search for "oldham couplers". For this application maybe you can even print them yourself 'cause they are a bit pricey (and hence even combine them with the bearing trick with some ingenuity ;) ).

[@deathcube2006]

I just found this video and its very knowledgeable. I thank you very much for all the info, I had a small binding wich I fixed with cheking the lead screw all the way and adjusting things. Also I installed a second Z leadscrew and the z movement is very consistent. I think Im going to upgrade to POM t8 nuts and Im not sure yet of going for the flexible couplings, I am very interested tho of taking the weight of the z axis from the steppers to the frame. Have you seen the ptfe t8 lead screws? How do they compare to regular ones?

[@timmturner]

I use "anti-backlash" POM T-nuts which are so much better than brass.

[@3dfxvoodoo]

Wow thanks my friend, cheers from Argentina

[@alinioanmoroi1370]

Awsome video as always!

[@123bookra]

why not use spider cupler?

[@Icrahukuku]

Great explanation. thanks man.

[@hobbyrob313]

yeah that's good information! friendly greetings from The Netherlands! Rob.

[@wernerboden239]

I am from the netherlands too ;) And I built 3 3D printers. What he says is not good at all, but he keeps removing my recommendations.

[@hobbyrob313]

@@wernerboden239 How okay? but what is wrong with a ball in the coupling?Wat is er mis mee dan? Een kogeltje in de koppeling? VrGr Rob.

[@wernerboden239]

@@hobbyrob313 The coupler is also designed to detach the motor from axial forces. But now, the load of the printer acts upon the bearing of the motor, which usually deal with radial forces. I built 3 printers; I should put some stuff online to demonstrate how it is supposed to be done (or at least, how I did it). But, if you look through the specs of sfu2005, you will see how the BK15 bolts the lead screw into place with 2 bearings and a nut, to protect the motor from such impact. I could upload a motor mount design I used for my printer onto thingyverse. When I do, I will get back to you. I am tempted to do a video on some of his topics, that can destroy a 3d printer. Like the floating lead screws, without bearing on one end; good way to destroy you tr8x8 lead screw. Well .. the longer ones. If I did that with my 700mm lead screw...ouch.

[@hobbyrob313]

@@wernerboden239 I had to read twice before I understood! But yes I understand now and to be honest you are right! But I also have a but! ( LOL ;-) ) and that is that I also understand why they want to do the trick with the ball. And to be honest, I would also like to do this, despite the fact that you: Werner Boden are right! Because what would become the problem in the long run is that the bearings of the stepper motors could break (and I would take that for granted) As a check, did I understand you correctly? And Yessss, I would make a small video right now to show what you mean. (I think and am afraid that not everyone understands it what you mean!) Kind regards from Rotterdam! Rob. Ik moest twee keer lezen voor ik het begreep! Maar ja ik begrijp het nu en eerlijk is eerlijk je hebt gelijk! Maar ik heb ook een maar! en dat is dat ik ook wel begrijp waarvoor ze het truukje met de kogel WEL willen doen. En eerlijk gezegd zou ik dit ook toch gerust doen, Ondanks dat jij: Werner Boden gelijk hebt! Want wat dus op lange duur het probleem zou worden is dat de lagers van de stappen motoren stuk zouden kunnen gaan.( en ik zou dat dan voor lief nemen ) Ter controle heb ik je goed begrepen? ik denk van wel! En nog even dit, ik zou juist nu een kleine video maken om te laten zien wat jij bedoelt. (ik denk en ben bang dat niet iedereen het snapt!) Vriendelijke groeten uit Rotterdam! Rob.

[@NicksStuff]

That's a great trick, Alex!

[@markivankarlsancho9152]

use POM nuts instead of brass nuts. - less maintenance in lubrication because its self lubricating material.

[@jimgreene3863]

Thank you I am converting an R1+ 3D printer into a CNC router and it had the flew couple's like you just put the berg into So this poses a different problem that I had not thought about so instead of support I have the problem of thrust when the spindle moves down to cut the thrust in the downward move (could) stretch the coupler as the Z axis moves down I will have to think about this would like to hear your thoughts on this please

[@AlexKenis]

Sorry about the delay. I would not depend on flex couplers for support on a CNC without using a bearing block if that is what you are saying, but I don't know much about that printer, so I can't really speak to it

[@3dpfeeney]

Looks good and STILL practical.

[@Willy_LGM]

M4 acorn nut works really well! Thanks! :)

[@johnmoser3594]

Would you recommend 3D printing the nut out of ABS instead of using a shitty bronze nut machined with a broken piece of glass? Also it occurs to me that the method with a thrust bearing isolates the ball screw from the motor shaft, decoupling it from any vibration in the Z axis. Abutting it to a bearing allows it to pivot and rotate, but doesn't this create a solid connection in the Z direction?

[@SianaGearz]

ABS is usually not abrasion resistant, choose another material. Indeed some people use 3D printed nuts (PETG works fine, Nylon works better); whether it is something to recommend... puh i don't know always worth a try isn't it :D you can also just buy POM nuts.

[@buzzerco]

great idea for a simple load bearing. great video

[@peterberendsen9236]

awesome straight to the point . thanks Alex

[@stevenwydler3641]

Like the idea. Good work. I would us a thrust bearing instead though.

[@guatagel2454]

It is 2019. Are we still talking about z wooble?

[@AM-pi7jy]

Thanks, this is a great idea and well explained!

[@tonypinzarrone7993]

awesome content and subscribed!

[@3dtitan379]

you sir... are a genius!

[@GregTheroneggythetech]

Any tips on prepping lm8luu for install I need to replace mine and thought I might lube them correctly from the start.

[@laharl2k]

makes sure theres no debris inside the channels and then clean them with DW40, wash with water and detergent, rinse and onces dry (make sure theres no detergent left) use motor oil. Most of the chinese crap doesnt even come lubricated, the come coated in kerosene or some shit like that. Generalpurpose oil works nice but dries too quickly and doesnt remove the particled that wear out, motor oil is thicker but sticks well to the metal and has good properties for this application. Make sure you do your maintenance every 1 month if you run it 24/7 or 3-4 month if you use it every so often to avoid oil drying and thinks just ceasing due to lack of proper lubrication. Also, if you can design the parts, just use brass bushings. Chinese LM8uu are crap and bushings run well even on cheap rods

[@GregTheroneggythetech]

@@laharl2k I was thinking of following the similar idea that Alex did in his linear rail video. Yours sounds very close to what he did. What about the super lube that he used instead of motor oil? I kind of like lithium grease but wasn't sure if the super lube lithium is too thick.

[@GregTheroneggythetech]

@@laharl2k I'm also not too keen on putting water and detergent it's through bearings.

[@laharl2k]

@@GregTheroneggythetech Theres nothing wrong with water on bearings, if you dry them and lubrecate them they wont rust, only if you leave them wet or washed clean of any oil will they rust. Grease forget it, too thick, it would only jam the balls inside and wont provide any benefit over thnner oil. Unless you find a very soft grease, then i wouldnt recomend it. The rule is simple, oil for fast light things, grease for slow heavy load stuff. Always use oil if possible. You usually need good sealing to use grease, otherwise the bearing will push it aside over time, oil being more flowy would crawl back onto the race for when the next ball passes over it. Lubrication is quite a good reading if you are interested. Ideally you'd use good quality brearings on good quality rods, all in specific materials with specific characteristics, with specific lubricant thats appropiate for said material, speed, and workload. Thats why you cant use any oil in your car. Sure you could use vegetable oil and IT WOULD WORK, but your engine wouldnt last as long and the oil would probably become rancid in a shirt time, it works but its not optimal for long term use.

[@leofortey7561]

igus bushings FTW.

[@wernerboden239]

I really would not suggest to attach the axial forces from a machine onto the radial bearing inside the motor.

[@AlexKenis]

It depends on the motor. For cheapo motors, I agree completely, you can't expect a long life with a high axial load. I posted this a a quickie hack, not a recommendation, and I personally use a bearing block to keep the load off the motor. But if you want to look into it further, bearing life expectancy is listed as it's L10, and other ratings are tied to it. I'll add what I can, but my bearing and motor experience before printers is from building industrial automation control systems, which have more of a tendency to list all the relevant parameters in the data sheets since most gear has to pass certification. Many stepper data sheets will have the max axial (sometimes called thrust) load listed as max permissible to meet the L10 rating. The better nema 17 sheets I have here generally list the axial load anywhere from 10 Newtons to several hundred, so it can vary drastically. operating conditions are also important, so as an example, one 42mm nema 17 sheet had 28N max axial load at continuous operation at 300RPM to be able to meet the quoted 10,000 hour spec, but obviously we don't run continuously at RPM that high, so I would consider that motor okay for a gantry under 10 lbs, BUT many cheap motors don't give a life estimate, and I have found that they arbitrarily just put "10 Newtons" as the axial rating, so whenever I see that, I am suspicious and would not trust the motor bearings to last very long with the gantry weight on them. They are probably using lightl duty deep groove bearings (which are only rated to handle 10-20% of their static load as an axial force... at least from what I remember), so you could do the math to figure out what the bearings can handle. Light duty motors also have a wave washer for compression, which can flatten under axial load, so that's another consideration with "10N" or unrated motors.

[@ianide2480]

If the motor is at the top of the machine this gets easier. This is how I designed my printer. The flexible coupler rests on a washer which rests on a skate bearing. No need for a separate collar but your motor ends up being upside down, so all the magnetism might fall out......

[@AlexKenis]

Ohnoes! Better 3d print a bucket to catch the magnetism. Sounds like it should work well though. I have not tried top mounted motors in a while. Years back that was done because the regular threaded rod worked better that way, but it was flipped when leadscrews got cheaper since they could take the weight. THe rationale was to lower the center of mass and prevent motor vibrations from causing print artifacts, but I don't see vibrations being an issue with solid design and the smoother stepper drivers we have these days.

[@ianide2480]

@@AlexKenis I built my own printer and my X gantry (actually all axis') is completely captivated by 2 linear rails in a doubly solid frame (I overbuilt the entire thing on purpose), there would need to be a serious amount of force to introduce "Z-banding". As it sets my acme-rods pretty much float at the lower end, there was just no reason to capture the bottom of the rods. I just installed TMC2208's in stand alone mode and holy crap, I don't know why I waited. I also lightened up the moving bed by removing some aluminum from the bed: original design (sandwich) was 1/4" alu plate 8"x8", 1/8" Alu plate 12"x"12 stood off from the thick plate, a 1/8" silicone mat with a channel for the heatpad wires, heatpad stuck to the bottom of another 1/8"x12"x12" Alu plate, all topped with a piece of glass. So I redesigned the drive motor brackets, lowered things a bit and removed the 1/4" plate and the first 1/8" alu plate, replaced those with a 1/8 x 12 x 12 piece of garolite and cut the silicone mat into strips to reduce even more weight. So now it's just garolite with silicone strips glued on (with silicone caulking heh), then the alu heat spreader and heat mat with the glass atop. Reducing the bed weight by nearly half and the new TMC drivers has nearly gotten rid of vibration effects in my prints. Then again I don't really go beyond 40mm/s for print speed. I just don't need my prints that quickly. Hell I didn't even care about the ringing effects to begin with, but tinkering with the printer is the larger part of the hobby for me, so I like to try new things out here and there. BTW, I'm using a Ramps 1.6 and its working great with zero issues. I cannot comment on the mosfet for the heatbed because mine is plugged into the wall and the mosfet is controlling an SSR. The mosfet for the extruder is working like a champ, I might head to digikey and order a known name brand mosfet of the same type or better to avoid the possibility of this one being a clone.

[@wernerboden239]

I put a design on thingyverse: www.thingiverse.com/thing:3579695 You can see a working model on my channel. Of course, it is still a prototype and we have imporved a lot, since this one. Currently working on a larger printer.

[@alexscarbro796]

As @Spock alludes to, a poor finish on the end of the stepper motor shaft and lead screw could have a detrimental effect. Perhaps exaggerating Z-banding in a worse case scenario!

[@AlexKenis]

That's possible, but I'd have to test to confirm. But you're right that there is no harm in taking a couple passes with a metal file to make sure the end is clean. I'd also be more worried about the quality of the threads if the end machining is that bad, and there is no way of really knowing if the threads themselves are rolled, machined, or ground until you have it in your hands.

[@ba7338]

This flexible coupling kills me!

[@adisharr]

The problem is most of the 'couplings' they use for these printers are garbage. No one uses these in industry. Also, direct clamping to a threaded shaft is not the correct way to do it and it's done everywhere in the printer market. It's all about saving money.

[@originaltrilogy1]

What is the correct way?

[@danieldimitri6133]

Meh, I just install the screw in th coupler such that the screw and motor shaft touch. Basically install the coupler as far down the 5mm motor shaft as possible and then put the 8mm in so it hits the motor shaft before it hits the shoulder. Then I stretch the coupler so it acts as a spring pulling the shafts together and tighten it up. It can never float this way.

[@robscanlon6568]

and why not just delete the ball bearing and run the screw on the motor shaft? full disclosure I've been running my screw on the motor shaft for over 2 years for the same benefit and as far as i can tell no issues.

[@originaltrilogy1]

It can flex more freely on the ball-baring than on the flat shaft.

[@Wezix1]

Buttt i have a heavy gantry riding on two z-axis motors using these springy couplers. No problems at all. Never had z banding or whatever people usually complain from. and that's on a cheapo DIY wooden i3 clone. I just have the 8mm lead screw meeting the 5mm motor shaft inside the coupler with virtually no space between them

[@Bob_Dub]

8:59 My nuts are typically off-kilter and I've ended up accidentally sitting on them a few times now. :0

[@chinaminiyumako1415]

Great works alex!

[@TD3DMakes]

Hey, awesome video. Thanks for sharing.

[@lidarman2]

Seems good to me and more than a hack.

[@GregTheroneggythetech]

awesome, thanks

[@franksemi_modular]

thats a good hack.. nice

[@injektorofficial]

AWESOME GOT A SUBSCRIBE, good work ; )

[@hex1101]

Great vid earned a sub

[@-Gothicgirl-]

Or you let instead of a Ball the rod on the rod of the motor. Same effect

[@wernerboden239]

Also, I wanted to say: You know your stuff. But you can't allways be right. I posted some comments on this video, but they got deleted. I don't know what happened there, exactly. But somehow, my comments were removed. If you do not agree with what I say, just reply, instead of delete. If that is what you did. Becaue I would like to learn, if I was wrong. As far as what I know: I do not particularly agree with this hack.

[@AlexKenis]

I just checked and they got caught in the auto-filter. I didn't read them yet, but I'll pull them out and make sure they are posted. I don't delete anything unless someone is attacking another poster or being inappropriate, but people can disagree with me all they want, that's half the point

[@roybrown3018]

Put a sleeve on motor shaft

[@davidprock904]

I just realized my design for a 3D printer does not use a lead screw nut... oh wait never mind it’s also the sun gear.... duh, my screws don’t spin

[@AlexKenis]

ooOOooo, you tease. possibly a strain-wave or planetary non-captive design, or something like that? That would be interesting to see

[@davidprock904]

Alex Kenis : what is a Strain-Wave? And what do you mean by planetary non-captive design? I am making my own motors that will give high speeds with less that micron precision but will just round it up to a micron. They are planetary setup. Butt my gears are not exactly gears in the traditional sense. Key point here, no slipping. And my encoder and controller are placed inside the stationery planet gears. By the way my total design gets rid of all forms of “lash” 100% rigidity. Any my style has never been done before, very close to a delta but still totally different. Side note, don’t wanna Reveal too much information but I will be doing SLA done like FDM. But also can be 100% FDM, also can do CNC.