I will never talk about Z wobbling ever again !
Вставка
- Опубліковано 29 бер 2023
- You can get the WobbleX from here ► s.click.aliexpress.com/e/_Ddk...
WobbleX interface GitHub► github.com/MirageC79/Interfac...
After 2 years of looking for the best Z screw decoupling device, I found the ultimate solution.
This video explains the principle behind the WobbleX and demonstrates their benefits through testing.
Thanks to @EvoMotors for the awesome idea!
Thanks to @3DMELLOWPrinter for their collaboration on materializing this project.
Thanks to the @HevORT community for the support.
WobbleX interface GitHub► github.com/MirageC79/Interfac...
🍺 SUPPORT ME: / miragec
PayPal Donation ► www.paypal.com/donate?hosted_...
You want to learn and contribute to the HevORT project?
FACEBOOK ► / hevort
DISCORD ► / discord
www.HevORT.com
Thank you very much!
MirageC - Наука та технологія
That design is impressive ! I saw your early attempts at fixing the wobble but never saw that solution until now, crazy smooth.
Thank you very much!
This design Is complete nonsense
@@janvostarek4396 Nah, it would work fine for through hole boards. Not so much for high denisty SMD though. Still going to build one and play around with the aperature.
I can show you some very expensive ones that accomplished far less.
It’s kind of amazing how the most minuscule vibration/etc will become apparent in a print.
Though it would actually be cool if it could be taken advantage of since sometimes the effect actually looks nice without being part of the design itself.
Some ceramic printing use this technique.
I think this is a fantastic example of a good mechanical design to solve the issue rather than a software patch to try and fix it.
This Is great example of how not to do it, design Is incredible dumb nonsense.
I just can't believe that it's cheaper/ better than buying straight ball screws or making the frame stiffer. I don't really get this fashion of using proper linear rails and lead screws (all from aliexpress) an mounting the lot to a frame made of jelly. It dosn't seem like good engineering to me
@@michaelwatson7211 Cheaper? Youve never seen the prices for an actual high quality ball screw + ball nut huh? A high quality ball nut itself can easily costs more than the 3d printer itself. $400 for a ball nut is very common, $750 to $800 for a quality ball nut is normal.
Ball screw + linear rails?
@@Sharpless2 Yeah I meant lead nuts, which aren't that expensive even for OK ones. And thats only one of my points, why buy this stuff if you are going to mount it to a frame so weak it's deformed by an 8mm rod? Lazer cut steel is pretty cheap, tapping holes isn't that hard, avoiding that then splurging on custom components, and rails with a >200kg capacity is building your house on the sand.
Mirage my first experiments with the Rat Rig V-Core 3 started with an adaption of your very clever wobble rings. Subsequently I developed the VenterMechs (Simpler and less fiddly than Wobble rings and results were just as good - hundreds of people using them) and then I upgraded the Rat Rig I was using as my test bed to the 1P Z design to also eliminate the 'hula' movement, which other designs don't address very well. This design has proven the best yet. The design is open source for non commercial use and I don't sell any parts for it. I wrote a series of articles on the popular Z designs and the issues and fixes for them including the 1P design.
I really like mellow's parts, they are expensive, but people say something about getting what you pay for... Congrats on succeeding on creating an awesome product!
Thank you!
like straight ball screws?
@@michaelwatson7211 will you sponsor them?
@@MirageC I think I don't really get the problems you are trying to solve, my background is in CNC, the idea of a ball screw deforming the frame is pretty nuts, as is the idea of using a bent screw. Why don't people use a cantilever bed on 4 bearings with a single screw (I've seen this with 2 bearings eg ultimaker but not 4)? Wouldn't it be cheper/ easier to make stiff than using 2 extra ball screws, if you don't care about weight in the frame the whole thing could be welded in thick plate it would be pretty easy to shim it flat once and be done with it. I also don't think you really get any benefit from ball screws over lead screws here, there can't be any backlash becuse the force on the bed is always down, unless there is somthing strange going on with the frame deforming or the 3 screw arangement. Like I said though I don't get the problems, the costs i was using in my head seem to be about an order of magnitude more than you are spending for screws at least, maybe the calculations just all work out different to my intuition.
Instead, how about starting with high quality ball screws and properly machined bearing ends, along with a rigid frame and precision sliders?
I always wondered if you would come back around to the wobble wings you first shared. You're doing some great work in advancing the small things that have big impacts in 3d printing.
I've never once thought about this type of wobble while printing.I just thought the slight variance was due to the unavoidable motion created by moving the print head. Absolutely mind blowing to learn you can get near perfect finishes on printed parts without sanding/finishing.
You never thought about it because it's NON-EXISTENT!
@@Theexplorographerif you use belts
...
or a delta.
amazing!
I love how much the community help each other out and come up with amazing ideas then share them!
Thanks for taking the time to create and put it all into a video :)
I know you are beating a dead horse, HOWEVER i super appreciate the systematic approach of testing and validating mods. Too many it seems are done in the name of performance, but end up causing more issues that they fixed. I'm currently trying to print as fast as possible with the best quality on ender 3. Lack of time does not allow for the cool stuff like voron.
this kind of problem was already solved for over a century
Reminds me of a documentary I've seen many years ago about how some bridges in quake-prone areas decouple their structures from the ground using large bearings in concave cavities to make it self-centering.
Wow, great video. I built a ~2000mm tall bed slinger, and wondered about my artifacts. This is really great. Thanks for making these available to purchase.
So glad I found your channel.
This hobby desperately needs more evidence based examination of "known" things.
I especially like that you're not afraid to show when one of your hypothesis turns out not to work (as in the extruder gear issue for example).
This is a gift to the printing community. Thanks!
Glad you like it!
A great way to add precision without making everything from cast iron and concrete. Will be checking out those 8mm ones.
You and the community are amazing for such solution!!!
I'm especially happy that's also available for common bed slinger's lead screw!
This is so awesome man! Been following this journey for a few years now, and I'm glad to see it concluded
Glad you enjoyed it!
Given how not-straight Ender 3 style lead screws tend to be, this is going to be a must have upgrade for anyone running dual Z’s.
Cause 3 d printers print up down, not in layers?...one whole layer the Z axes don't move .then move what, 0.2mm to get the next layer hight , then the whole layer print z stays still, not moving a 1000of a mm...so this "gimmick" doesnt solve anything, it's a false narrative, nicely presented, but nothing of truth to it..a bit like Ukraine war and NATO excuses...
Thanks wobble man, instantly ordered 3 WobbleX. Been fighting with the Z wobble for 3 years on my custom printer. I will soon add a custom mount for the Sparkcube XL.
Glad I could help Keep me posted on the results!
This came to me on my recommendations, and definitely not disappointed as the design more than exceeds its intended purpose.
Definitely be worth its pricing, and also good Segway to the shop and pricing. Subscribed especially if you’re working on other projects, I like seeing how they progress.
the sound of the stepper motor slowly moving the screw and it being affected by the screw wobble is satisfying
I seriously love following this not only as a Mechatronics Engineer but now I’m torn between a design like this or Voron so hard to decide now
It depends on your needs, really.
A Voron's quad gantry belt assembly hits its limits in size quicker than this design, and I would say that the pin mod (replaces screws as axles with hardened pins) and GE5C gantry tilt blocks instead of the stock plastic on plastic tilters are almost a requirement for high performance ones. It is, however, much simpler to isolate motion problems and there's fewer of them in the first place as belts don't tend to introduce Z-axis artifacts to nearly this extent.
I'm not sure the fixed bed VS tri-screw bed is really making much difference at this scale of design.
Looking forward to Prusa support
Eureka! You solved a huge "mystery" problem! The world of 3D printing owe you a great debt!
(BTW; VERY cool camera zoom, out from the reflection of the chrome bearing ball; nice stylistic choice!)
Very iimpressive & clever design! Also love how you edit your videos!
Hi guy! It's fantastic design. Congratulations for that and for all the development... have you thought about or already have a plan/project for prusas?
one of the best 3d printer channels out ther, you and vez3d rly are leading the industry RN!!!!
keep it goin!!!!
Thank you! I am glad you enjoy what we do 😊
❤❤
I haven´t followed the progress after your first video in this topic (refered to at 0:40), well, that must be years ago... But I always know this video here would one day come and find me... I never doubted you are gonna solve the problem, even in a mass-production-scale. I even thought the probel was solved years ago, little did I know about engineering... But that here is on another level! Longevity and good prices, not just a DIY-solution.
Currently building a Trident and will be installing the WS8 when Mellow makes them available. Awesome work.
Very great video! I have made the somewhat exact same experiences you talk about in this video. I have an Ender 3 with a single Z-axis screw. I used a anti-Z-wobble constriction from Tekti from Thingiverse as inspiration, designed it completely from scratch from CAD and use it in my printer right now. It works, but as you said at 7:30 you need to fixate both ends of the screw, so the "wobble" will be transferred to the part that is made to take on the wobble. It works very great, even thoug it is 3D-printed and uses magnets instead of little rods. So far, so good.
Very cool video, I like it.
Thank you! Glad you liked it. Yes, many systems will accomplish the same role. Many ways to skin a cat as they say... My grandfather did anyways ... lol
@@MirageC Yes, I had to do some iterations, but since the print is actually 3 piece and quite small, it doesnt take long to print. I sureley had some space trouble because of the extruder stepper motor. But it works pretty good, even though not engineered/desgined to every standard. I hate to go for some shortcuts to safe a bit of space ... no threaded insert but screws directly into the plastic. Works fine if you dont tighten it too much and dont unscrew it and rescrew it every half an hour.
@@heavyweight6440 Have you observed the hardened ball bearings chewing up the PLA layers buffering between the magnet pulling from the other side...in the same way MirageC noticed bearings directly on magnets chewed up the magnet surface?
@@patrickboyd8368 Not yet. Right now it all looks kinda good, but I can imagine that after enough print hours the magnets will probably suffer the same fate as they did in MirageC's video. I also am pretty sure that Tekti has developed a different version that doesnt use magnets anymore but 2 dowl pins as guidance for the ball bearing. They have much greater HRC then the neodym magnets and will probably not suffer such a fate. But as of now, it all looks ok. I think they weight of the tray is not great enough to damage it very fast. Otherwise you could upgrade to double Z-Axis on the ender 3 and split the load up on two anti-wobble systems. This doubles the area where force is applied and the surfaces of the magnets will live much longer.
The PLA looks all fine as of now. Prints also come out good.
Oliver, what are your thoughts on printers using belts instead of screws?
I don't have experience with Belted Z, Only seen user mods of them and no official design around them. I have many questions... Will the bed jump up and down as the belt vibrates from XY vibration sent through the frame? How many parts will be involved to lift an 8mm 500*500 bed and keep it up there?.... Somehow... I don't feel the need to investigate these question anymore 😁
@@MirageC I have a custom Ender 3 IDEX, well the only thing Ender is the frame I used. I use dual side mounted Z lifts but it's not like the ones you commonly see. It runs on an 8mm hardened axle that passed under the frame. Most use flimsly shafts and mounts which can be an issue. I've seen no reaction in Z from X/Y moves on mine, was worried about it early on.
It would be a pain on a 500mm cast bed, take your pick, a braking system or insanely low gear reduction. I run gear and pulley reduction at 40:1 to keep the heavier IDEX X carriage from dropping when the steppers powered off. DId the brake route first using a servo, it worked but a PIA to sync the release.
Belts are nightmare at high speed, unless you can afford to go very wide and tight. The heavier the XY movements, the more vibration will be sent down.
@@Soravia so how is it that vorons use them apparantly without z wobble? better figure this out before i order a voron :)
@@sanderveenman599 how wide is the belt vs print size for XY and the assembly weight?
A very elegant solution. I was actually designing something like this for my Ender 3. I used ball bearings from motorcycle parts and the pins from an old DVD writer I had around. I made some prototypes but stopped once I learned that Oldham-type nut exists.
Yep, got them on my CNC X1 otherwise all the stepper motor torque goes into bending the screw.
Amazing! Love to see some hardcore mechanical design grace the printing space :D
Biggest issue I have with my Sovol SV06 (and my Ender-3 V2 for that matter) is all z axis related. Not sure how that would fit to either of them but if the 8mm one is compatible, definitely something I'd be interested in when I have a bit of spare money.
Wow guys, i've been working on designing my own printer. I had long since decided to go with ballscrews and I never considered something like this on a cheap set of screws. Im really glad I haven't ordered anything for the Z axis yet! This is an ingenious solution.
Superb work!
making it available for everyone takes it to the next level 👍
Cool! Design reminds me of kinematic mounts we use in optics systems.
excellent work! 👏😎
Thank you very much! I am glad you liked it.
@@MirageC I don't see in the archive the file in between Wobble X_TR8 and ZCarriage_WS8 for ender. can you share it?
@@MirageC When can we expect a version for the Voron 0.2 with Kirigami Bed? That would be an incredible asset for the many 0.2 users
Z wobble isn't always the sole reason for Z artifacts or Z banding - No matter how much we compensate! There's always others factors coming into play, too.
well of course! but like, that was a HUGE reason. It's like seeing your house has mold everywhere, no electricity and also half of it is under the water bc of the latest hurricane or whatever: Sure, fungi and lack of energy are very important, but solving the issue of the flood would greatly improve your house, and also would help to solve the other two things 😆
Yup... On my custom Delta printer I had the problem where the bed is static, but still experienced z-banding. Turns out it was the PSU I was using, it had a very very loud fan, that would only turn on when "needed". The fan drew enough power that the voltage dropped, causing the bed to cool a bit faster than the PID tune could kick it back up, and the expansion was enough to cause visible artifacting. Took me months to figure out.
@@NeoAcheron daaaaaaaamnnn! Awesome that you found out! Sounds like something you might not find easy at all untill everything else is correct
Right, i was just thinking about this, you dont really print while the z is moving up, so that extrusion line widh shouldnt be all uneven because of Z wobble. Well not all of it, thats more likely to due bed temp fluctuations, that will cause the z to change ever so slightly while is shirnking and contracting, something more probable since its happening constantly while on the same layer, whereas a Zrod problem would only show up in layer changes. And in this example here it does seem to be Z wobble since the pattern reflected the rods. But the uneven "bumpy" random lines he was showing arent z wobble, but more likely bed temp fluctuations. Same effect, different reason. Why I think its important to PID tune your bed while the fan is blowing as well so it has that constant cooling to fight against, should make the temperature more consistant.
This channel has a lot of videos on the topic of Z artifacts and even things that look like Z artifacts.
So Awesome. On my cart now for my Ender 3 Pro. Thank you
Just bought a set after the restock. Can't wait to get them installed!!! I knew as soon as I saw this video how good of an idea this was. Simple yet elegant and solves the problem once and for all.
WHERE did you buy these? I have been looking every 5 minutes since this video came out and they are never in stock. :)
@Jim Drew I had to wake up at 9am today to get them as I knew they would sell out fast. If you message the seller and ask for the next restock date this will give you a better time to check.
@@drumid1881 Thanks... I have left several messages via Aliexpress as well as their Discord channel with no response. :(
They are now! Qty100 have been restocked today.
@@MirageC WHOOP WHOOP. grab them while you can
This was a nice presentation but I can make system using belts 5 times cheaper and 5 times simpler and yes 5 times better.🤔😉🤘
Go for it! cant wait to see your system! 😉
Prove it.
Amazing. Your scientific approach really makes me believe everything you say
Yep. This looks amazing! Totally gonna machine my own at my local school
Vielen Dank 😊 für das Video und für deine Mühe. Tolle Arbeit 😊
This looks interesting wonder if this could be incorporated in the Bambu lab printers? As they are tri z axis like your example
It's amazing to see how one person's perseverance to create the absolutely best print results can improve the whole future of a technology. I'm looking forward to see Wobble X adapted by the Voron community soon. ;-)
Our most useful tools have came from someone solving a 'minor problem'!
man ... great work, the level of ur content is increadible ...
Congratulation, excelent work, thank you for your dedication! 👍
I'm planning on buying a bambulab X1 and leave my ender 5 plus as a backup big volume printer, but maybe with this I can extend its life a bit longer!
Edit: the ws8 is out of stock, any idea when they will be back?
Edit2: I installed the wobbleX, and under the correct amount of light I can see a tiny amount of movement. However, I still have massive Z issues. I gave up on my ender 5 plus, I have a bambulab X1C now and it prints flawless.
I got so exasperated at looking for the cause of my wobbling and banding that, when I discovered that the second lead screw was going through a malaligned flange attached to the backplate, I lost it.
Got unlucky with Creality's quality control (or lack of it).
I stripped out the lead screws and modded the z movement to use belts instead, driven by a stepping motor on top of the assembly. It rotates a horizontsl bar via a transmission, which provides the vertical movement for the belts.
Worked a treat.
great contribution to the world of 3d printing, thank you
Nice! I can't wait to try this on my V-Minion!
I did much the same on a precision X-Y table I designed in 1993. The precision-ground NSK ball screw I used was very good geometrically but under high (1G) acceleration against an offset load, the ballnut would skew under deflection of the bracket and subsequently jam on the screw. What I implemented was basically a U-joint around the ballnut. Rolled-thread lead screws are pretty bad for precision but at low acceleration I suspect you could get away with a rubber mount.
Me sitting here with a Sermoon D1 (nothing special, just a very sturdy "bed dropper" machine) and no Z wobble, shift, banding from the top of a 310mm print. This is with one of the lead screws having a slight bend. I did the math on deflection, and less than 1mm deflection translates to almost nothing. I also tightened the brass nut screws, then loosened it 1/4 turn to allow lateral shifting but no Z slop. The bed rides on two vertical 20x40mm v-slot extrusions
Two things i did, i put the bed all the way down until it bottomed out and loosened the z-rods then activated the stepper motors as to jump into phase. Then i carefully tightened the couplings as to not move them in the slightest, the stepper motor will give a little while being energized. That made the z-banding nearly invisible and you had to really really look for it. Then i did the same thing again but this time i installed a timing belt, making sure they where in perfect phase i energized the motors on and off listening and feeling for any shift. That completely removed any z-banding visible.
Note; I'm using a 0.5mm nozzle and regularly print at 0.12mm to 0.32mm layer heights. I always use PTFE grease on lead screws and PFTE oil/grease mix on LM8UU bearings and meticulously square my machine. Which annoyingly exposes any slop in the bearings and makes them slightly noisy.
This is amazing! Excellent work!
Major props on this design
Really great product! Well engineered.
What a great advancement for 3d printing
thank u man
you are the best!
Thanks a lot!
Seriously considering tweaking my ender 3 pro... Thanks for all the hard work...
Stunning vertical print quality
Dude I just saw these on aliexpress and was very curious, great video
I am a welder and I use 1 solid stainless rod to hold and stabilize and 1 threaded for moving up and down nor left and right movement and it fixed the issue, you can also find in most industrial how they already fixed your problem specially with CNC
Wow this is incredibly cool. Is this a fully new invention or was there already a dual bar and ball style wobble table like this used in industry? Awesome video demonstrating the issue and fix.
0:04 close. The actual best is to remove the screws all together and use linear motors ( though it can be a bit pricey for off the shelf or hard to figure out diy )
Nicely done! Will definitely try to tame my wobbly Ender 3 S1. Then, I'll come back for more!
Try just to reorient the rod in the coupler. Try to center the rod. If it's bent, replace it, not worth the hassle. Basically I never saw z wobble in real life.
simple, yet awesome engineering.
Another great video! I am wondering how you do all the animations? Seems very convenient to model it and analyze the problem in 3D.
Very beautiful your work my congratulations
That's some sick engineering!
Amazing outcome
So, you recreated oldham couplers. This is why we used 5mm threaded rods in the RepRap community instead of 8mm. This way the rod could deflect and not overpower the smooth rods. Everything old is new again!
Nope. OldHam coupler deal with wobbling only. These will manage the nut tilting as well as teh wobbling: 5:18.
this is really cool! you have a brilliant mind.
years ago we started with M5 threaded Z Rods .. they bend, mostly eliminating effects of wobble seen with current massive lead screw rods being used today.
You're making this world a better place!
Can't to wait for maglev technology implementation in 3d printers ☺️
This is next level! Well done 👏😄
I am happy with the belted z lift on the Valkyrie triple z. It eliminate any wobble due to screws.
Sweet! glad to hear that. Did you have to install a brake system to secure the bed when motors are not energized?
WOW 🤩!!!
Practical example of "The Power of Open Source Innovation at it's best" 😀!
Are there any use-of-open-source-in-the-industry "Industry Awards Contests" looking for contestants ? -this sure is a Winner 😀 !
Very informative! 👍🏻👍🏻👍🏻
ngl, its cool to see original idea straight to Aliexpress or Chinese volume manufacture, relieves some of the guilt having to cheap out to knockoffs that may've gone too far, either being off spec or actual cheapness for profit's sake
Amazing job as always!
Great to see man!
Funny how I got to precisely the same solution after being inspired by your wobble wings...😀😀😀
Never ceasing to amaze us :) Well done.
Glad you enjoyed it!
ADD A THIN STEEL COATING OVER THE WEAR SPOT ON THE MAGNETS. BEARING WILL STILL BE HELD, BUT THE WEAR WILL OCCUR ON THE STEEL SHIM.
I solved this by just printing a stiff TPU coupler between the stepper and the threaded rod. It doesn't really compress, but it gives just enough laterally to remove all z artifacts.
Great Video! Earned my sub!
I went with belted Z, no more Z-wobble what so ever and very easy to make.
AMAZING!!! Would that support the PrusaMK3 or Mini printers at all? thanks
Great work!
I can't wait for the next episode on Z wobble!
hahah! I hope next olympics video is before that episode! lol
Your CAD and animation skills are legend. If you are employed, I hope you are working in the R$D department of a 3D printer company. BTW, what kind of printer did you use in this video?
absolutely incredible!
this is amazing work.
I would have preferred that you kept the old mounts in order to change only one variable, but this design deserves some consideration.
Great work. What's the CAD software you are using to demonstrate your work ? thanks
Cool animations and use of 3D mouse ;^)
An idea to help the magnets last longer: maybe a small hardened steel disc there to stick right on the megnet to act as a bearing race rather than bearing strsight to magnet
Geniunely learnt something there.
This is amazing and beautiful.