I am pleasantly surprised to see the interest in this video! I will put together another video in the coming weeks explaining the design and showing how to print and assemble it. I'd love to see people incorporating this style of mechanism into projects in the future!
First off, brilliant work on recreating this design! It seems at first glance like quite a daunting task building this, but yours turned out amazing! We would love to incorporate this type of joint in the humanoid robot we've been working on. Pairing this with a brushless motor would make for quite a fast and powerful joint! Cheers!
you could probably make a 15 minute video talking about the parts, how you made it, how much force reduction or multiplication there is between A and B and C... and you'd get even more views and shares.. cuz people love this stuff and want to know more.. even just what you're thinking, or what you're building. are all the pullies required? are the 2 supports on the far side basically an elbow? why not have that on both sides? or.. wither side, since there's already a bolt holding things in place.. soooo many questions. :-)
The dozen pulleys multiplies the force by at most 12x the tension on the pulling rod but there are frictional losses. More pulleys means less tension is required, but it also leads to more weight and more friction, as well as possibly slower motion. The far side of the elbow has two cams with strings running from bottom left to top right, and top left to bottom right, and that constrains the system so that the elbow lever bisects the system, and the whole system doesn't awkwardly collapse like the snakey tetris piece.
The Skyentific channel has already done an explainer. watch?v=utDagouxM5U He also links to earlier work done using this principle. Love that guy's journey btw.
I call it! This joint is the future... I've been looking for a good robotic joint since forever ago, and this is the one I need. Thank you for posting.
What you mean, is that you have been looking for a cheap robotic joint forever. If you were looking for a good one you would simply go with strain wave gear like all the industrial solutions do.
@@mattmurphy7030 I agree that precision also belongs on that list and the best you can get is probably 5 out of 3. But this principle done right is so much cheaper than a harmonic drive that you could have several to extend the range of motion as well as speed, so you get precise (steel cable duh), strong, fast (doubling up), high range (360 deg with 6 cheapo joints?), but kind of bulky, heavy and although cheap, not super cheap if you need several of them to get the desired range and speed.
I've been looking at ways to clamp and vice grip things. There are three great ways: Hydraulics Pneumatics Wire/Rope/Pulleys Cost wise using mountain bike brake line (bowden cable) and pulleys is absolutely 100% the way to go. This proves my point.
@@mattmurphy7030 Definition of sick (Entry 1 of 2) 1a(1) : affected with disease or ill health : AILING (2) : of, relating to, or intended for use in sickness took five sick days this month a sick ward b : QUEASY, NAUSEATED sick to one's stomach was sick in the car c : undergoing menstruation 2 : spiritually or morally unsound or corrupt
biggest draw back for this drive mechanism is the cable can slip on the shaft over time and loose accuracy, however for angular precision and duty cycle of a robot arm elbow it is good enough
TBH, I don't see the cable slipping off the shaft being a thing, but I am worried about the difficulty of keeping tension in the system after a while of use - Though that could be overcome as well with the use of a spring tensioned ( during maintenance ) and lockable ( when operating ) return roller before attaching the ends to the drum.
@@A3Kr0n You'd really want the springiness out of the system while in operation - Hence the need to lock down the return roller ( actually more like an idler now that I think about it... ) - If that dial system allows for that, why not.
I don't think the cable will slip on the shaft, it's locked to it by being fed through a hole on the shaft. as long as the cable is pinned to that hole, it won't slip anywhere. similar systems are used for fwd/back movement from a single reversible motor driving a shaft, which then moves "object' back and forth. can be used on doors, 3d printer gantries (instead of rubber belts, etc). it's pretty solid. Durahl is right about the tension though.. others have mentioned springs, which I can't talk to at all.. (they never listen to me).. ;-)
Would it be possible to feed cable to other joints further down the arm? Just thinking that would be a good way to centralize all the motors and not add weight to the arm.
Put a motor to that thing your spinning, and you got some Transformers type joint...cool! This combined with those new hydraulic mechanical muscles. Careful government might take your idea!!👍👍
Really nicely done! I'm wondering why did you choose not to make it symetric? Even if you have the cbles only on one side, I assume it would provde more stability? Are you using Kevlar as a cable? thx
Wow amazing. Are you planning to go on details about this? I think if you do you will have lot of people interested in this. Please do it. Like many other I am curious about the capabilities of this model.
Hello, I hope to simulate a rope-driven robot through software. Later I found that mujoco does not seem to be suitable. Can you provide some suggestions?
There is literally a picture of a mechanism that looks almost exactly like this in the paper on page 3 on the left side? Not sure what you are talking about