DIY mechatronics always has some unique challenges when relying on simple tools. 3D printing enables some great abilities but high precision gearboxes are still a difficult problem for many. Answering this problem, [Sergei Mishin] has developed a very interesting gearbox solution based on a research paper looking into simple rollers instead of traditional gears. The unique attributes of the design come from the ability to have a compact angled gearbox similar to a bevel gearbox.
Multiple rollers rest on a simple shaft allowing each roller to have independent rotation. This is important because having a circular crown gear for angled transmission creates different rotation speeds. In [Sergei]’s testing, he found that his example gearbox could withstand 9 Nm with the actual adapter breaking before the gearbox showing decent strength.
Of course, how does this differ from a normal bevel gear setup or other 3D printed gearboxes? While 3D printed gears have great flexibility in their simplicity to make, having plastic on plastic is generally very difficult to get precise and long lasting. [Sergei]’s design allows for a highly complex crown gear to take advantage of 3D printing while allowing for simple rollers for improved strength and precision.
While claims of “zero backlash” may be a bit far-fetched, this design still shows great potential in helping make some cool projects. Unique gearboxes are somewhat common here at Hackaday such as this wobbly pericyclic gearbox, but they almost always have a fun spin!
Thanks to [M] for the tip!
pick one
Everyone’s design is “zero backlash” at first.
for some definition of “zero”
Plastic gearboxes are easy to make zero backlash at zero torque. Just design for a slight interference and let it bend.
Depending on the material is a decent but complicated way of welding a joint, even with no torque via the slight friction alone.
Super squishy TPE gears to enable zero backlash (but some elastic play)… surely it would be useful for something, but probably not much
Guy probably spent more money on bushings and filament than it would cost to buy a mini lathe to turn some conical rollers and have an actually useful gearbox. Friction will tear this thing apart quickly.
To my understanding plastic-on-plastic contact ought to be fairly long lasting, so long as it is a rolling contact (as achieved by involute teeth or within a cycloidal gear) and not a sliding contact (as achieved in worm gears and compound “perpetual wedge” dual layered planetaries). A rolling contact usually shouldn’t wear very much at all, particularly if you apply a lubricant grease to it, a sliding contact is always suffering friction so will wear quickly.
Why so much negativity in this thread?
It starts with Hackaday doubting the “zero backlash” claim.
Myself, I like this idea very much. It combines a lot of things to make a very good compromise between 3D printing and standard (relatively cheap) hardware.
First, the hackaday doubt of “zero backlash”. These things are commercially available. Just a simple search:
https://duckduckgo.com/?t=h_&q=roller+pinion+gear&ia=web
finds several manufacturers, and they also claim “zero backlash”. Quite understandable, as these things can be made with a preload, and that eliminates backlash. How much of that remains in a 3D printed design (and how a 3D printed design holds up over time) may be a different question, but because there is only rolling friction between the 3D printed gear and the rollers, I expect these may hold out quite well. This is likely the best best design of a 3D printed gear transmission I’ve seen, and achievable by anyone with a 3D printer (and drawing skills?) The most wear will be in locations with sliding friction, and that is (probably hardened) standard pins, and (also standard, oil impregnated) bronze bushings. And that is a well proven combination.
Now, about the actual design:
@03:50 he mentions two solutions, conical rollers, or multiple bearings on each roller. But there is a third option in between. You can make the pinion itself conical, but still use cylindrical rollers. Normally the hart lines of both gears both converge to the same point (intersection of both axles). I don’t know if you can go that far with these rollers, but you can go in that direction.
Second:
As steel is much stronger then plastic, you want a relative large amount of plastic in the gears. With this size, reducing both the steel pins and bronze bushings radius helps. Although bronze bushings with a big wall thickness improve efficiency (bigger lever action).
And last, about the patent. I don’t know what that’s about. It’s a Taiwanese (chinese?) patent.
https://patents.google.com/patent/TW202100895A/en?oq=TW108143739A
I do like the idea of making notches into the inner support ring, so there is no interference with the big gear.
A general search for “roller pinion gear” finds US1289508A and that is a patent from 1918.
https://patents.google.com/patent/US1289508A/en
Details on patents also vary. Best I know, I can make a copy of any patented thing for my personal use here in the Netherlands, I think it’s even legal to give a drawing of a patented thing to someone, and pay them for making such a thing for me. In other jurisdictions it’s apparently already questionable if you make a copy of a patented thing and then place it in your lawn so the neighbors can see it.
This design looks very promising, and there are lots of design parameters to tweak, and it needs more experimentation to find the best geometry for long term use and stability. The only downside of this video is the annoying background music.
Don’t get the negativity either, except readers may expect polished solutions when the presentation appeared to be more about how to approach a geometrically non-trivial concept.
Between the conpromises like making do with relative co- and counter-rotating regions and bruteforcing a surface that nore likely.than not won’t even have a closed-form representation, it’s surprising (and therefore delightful) to even see anything practical come of it.
People who don’t actually build things and push paper assume everything must be perfect or its garbage. Sadly the world isn’t so black and white. You see this with a lot of “enthusiasts” in tech who don’t understand how much goes into building anything
Sign of immaturity and lack of pragmatism, combined with ignorance/sheer lack of experience understanding larger scale projects.
Everything in engineering is a trade off, and an important one at that here with this gear design. If anyone has an opinion that’s negative, just I just know, most of them would be dead in the woods since they can’t even make something basic, or understand how much goes into making a simple screw. Even better lets be pragmatic, a nail! How many people know how nails are made? Many don’t
Time, Cost, Maintainability, Ease of use, Repeatability, and Longevity are axis how things are measured.
Using a 3d printer is always going to sacrifice some longevity for the trade off of Time, Cost and Repeatability, ease of use
Sure someone can “get a lathe and make XYZ” or “use some expensive thing” but how many people have the skills for a lathe, let alone how long does it take to fabricate out of metal, gears with precision? Metal isn’t 10 bucks a roll, and standing over a lathe has a time and ease of us cost, along with repeatability between batches. Then why not get a high end Tormach CNC and machine everything out of titanium with the top end grease? Well that’s cost, a lot of cost.
Sure its not going to last as long as your grandpa’s hip implants, but heck, you can churn out 10 in a day while going about your life I’d call this a massive win. Seriously you can spend the time designing something that leverages the “downsides” (again a time cost) but get something that will work/last longer (longevity)
Everything is a trade. Nothing is Free.
Useless gadgets have negative value and should be called out as such. Being able to “churn out” 10 plastic gearboxes a day is just making the world worse.
I disagree. Not everything requires a metal gearbox. Just this evening I finished assembling another dozen phone stands that use a 3D printed gearbox. Because they are under minimal load and don’t get a ton of wear and tear, a 3D printed gearbox works just fine.
Ho ho ho, hold your horses! You mean that people without any real knowledge could not critisise something they barely understand ? Go back to the 20 century!!
Criticism is more useful than praise.
Criticism without a constructive feedback is just being contrarian for the sake of being cynical. Skepticism is incredibly important, and some of the claims in the video seem a bit naive, but “zero” is an incredibly relative term, ESPECIALLY in the world of gearboxes.
Criticism without a constructive call to action or advice is what deters people from innovating. It creates a needless barrier that makes people either hide and/or doubt their unique ways of thinking.
So maybe instead we stop the “ur dumb” responses and instead approach with health skepticism and feedback? Like, this ain’t a snowflake this, this is just about being antisocial.
This looks great! Can we get access to print the parts ourselves? This is what open source is about. Collective refinement. Keep going with your work, I love it. :)