If you had a flywheel geared to drive that main drive gear and had a small motor running at its most efficient speed, then instead of the neutral space on the sphere gears, have it somehow take that inertia of the wheels back to the fly wheel you could have regenerative breaking and then the whole think in theory could be quite efficient because it would overall be driven by a very small light weight motor?
Interesting, I doubt with the weight required for gyroscopic stabilization the power loss when engaging the drive wheels would have that much of an effect. Seems do'able.
I believe fly wheels must be untouched in use for stabilization. If a heavy weight were to be attached to the output it would give a reaction wheel effect that would be counter intuitive to the robots movement.
I was thinking that as well. But the cvt's would always be spinning even when the wheel was in neutral so changes in fly wheel would really only happen when engaging and dis engagement. Especially if it was inverted and the flywheel was under slung. But then there is the speed of the flywheel to consider. Would need to recalculate reduction to the wheels.
Really clever clutch mechanism, I'd imagined something like the little wheel your omniwheels have so it can idle on the drive wheel, but your way is much simpler and better
There is one downside to its use on this particular testbed style of machine - as there is no braking options that wheel is free to run so you can't 'turn it off' and steer around it 100% predictably - its a very neat execution of the mechanism don't get me wrong, but on that tank steering testbed the clutch freewheeling can lead to steering troubles when you wish to turn around one side or other and the free wheel just rolls so you instead get an uncontrolled veering in the direction you want (as the free wheel still has friction and resistance) - simple enough to avoid in practice if you can accept not being able to just turn the one side against a stopped other i.e the very tight radius but not 'on the spot' turns - as in effect it can only turn 'on the spot' (reversing side B from side A) or at any of the wider radius when side B is just solidly engaged and side A is however much faster you desire to create that radius of turn. Add in a brake for each wheel and that trouble goes away entirely too.
This would be a great setup to use for a huge steam or gas engine powered robot. Something that the power source would not be as responsive as a stepper.
Oh and instead of servos you could use valves and actuators. The sterling engine could also generate the pressure to move them and the power for communication.
Yeah, the engine would only need to maintain constant RPM. The CVT wheels would translate that into useful motion. If the same engine could generate enough surplus power to run its electronics and servos you'd have a pretty long range robot.
The robot arm idea is really interesting! It reminds me of old sawmills, where the equipment would clutch in and out of the main waterwheel which delivered power via a belt and pulley system going through the building. For the arm it means you’d only need one strong drive motor rather than a motor at each joint - that definitely shows promise!
Really really cool--there are few videos that make me exclaim out loud, but this one surprised me with how far it went. The volume and diversity of your output of ideas is inspiring
Yeah, it's quite pointless for actual usage. For robot locomotion, the wheel hub motors seen in hoverboards seem to be the pinnacle in actual design. Proto G recently put for of them together in an omniwheel configuration. But always interesting to see new ideas.
Looks like a modified version of the drive system used by "zero turn" riding mowers? The gas engine turns a flywheel at a constant speed and variable reduction transmissions control each of the two drive wheels, allowing forward and reverse at a range of speeds.
In reality zero turn mowers use two independent hydrostatic drive mechanisms, hydraulic motors powered off of a common pump, the drive levers are connected to valves. Most ordinary lawn mowers with steering wheels have a single hydrostatic drive these days, much easier to use than a conventional transmission when you want to have variable speed and don't want to be messing with the throttle and therefore changing your blade speed.
@@AlRoderick Many zero turns have a pair of independent sealed hydrostatic drive units that are belt driven. A single belt drives both input shafts directly from the motor, and control arms attached to the drive units control their ratios much like the design in the video. The actual reduction mechanism is very different, involving movement of swashplates that change the displacement ratios rather than changing the flow rate with valves. Since they don't have the restriction of valves they are more efficient, and much like a torque converter tend to work great until they don't. I've rebuilt them before for customers and currently own a mower with this system, but my next one will probably be the pump/valve/orbital motor style just because it removes the reliance on proprietary parts (Deere, of course)
This is brilliant!!!! I'm not sure in how far this is patent-able but as robots go this one will go far. The limited amount of moving parts. The load-bearing moving parts.... Man this has a future. I kid you not. As an example I see this riding on mars but also as a more sturdy Roombah. Just imagine this clutch system adapting to different types of flooring by using the clutch. Not just the wheels but also the brushes.
And I know my CVT comment on the previous ball video probably didn't influence the creation of this bot. It would be a natural progression to do this one. But it's nice to think it might have. Well done.
For some reason I was researching this same cvt system earlier this week. What a coincidence you made one as well. Great job. I think it could have a use for a steam power or even gas powered robot. It would really shine if it was 6 wheeled, all direct driven by a single belt and the same 2 servos on a single motor. It could be a great drone vehicle with gas it could run for hours and miles in this configuration.
OH MAN!! Just another 9 more subscribers and you will have 1 million. WOW! Congratulations when you do. You been doing this for a long time. Congrats and enjoy 👏👏👏
I've got a bicycle with a CVT-it's pretty fun. In fact, that sphere-based CVT video you showed is by the company that designed the transmission on my bike. The CVT gearbox is called “Enviolo”, and the bike is called the Priority Continuum Onyx, if you're curious.
That's such an awesome design, can't wait to see it evolve even further. BTW, maybe using small cycloidal drives instead of the servos and arms to change the tilt of the hemispheres I think it could give a more compact and sturdy fit, if placed by the side of the hemisphere's holder, though the current arms design look interesting, looks like there's someone inside manually driving it :D
biggest problem is the operation with high torque. We had this problem in a project for a client, but ended up making a new type of gears which worked well. It is still in the process and will be patented. But for anything with a limited torque, this is grrrreat
I'd imagine that in real world application, debris on the wheel would be pushed between the friction points and would cause quick wear and tear on the half balls driving the wheel. This design probably best to be enclosed in a dust free environment, like perhaps a shoulder for a robot.
@@el-domo Agreed. It becomes a choice in materials used at that point. Metal on rubber, or nylon on some other material. Whichever part being easier and cheaper to replace should wear first. It's a neat design with niche uses... I wouldn't expect it to be used for moving parts that operate 24/7.
@@TheNightstalkerShow They all wear out and can't take high torque anyway. If you find out, it's extremely expensive and limited to the applications anyway
@@el-domo Probably can design a half sphere bevel gear with no spiral these days to handle the torque and get same functionality as the friction only method... but you say you have a patent pending, so I doubt you'd confirm it :D
This is a nice little piece of mechanical engineering. Really awesome how it works. Maybe use a set of clutches on a drive shaft like you suggested to keep the motor weight down and close to the center of mass and then use it just for a hand, like an improvement on your Ultron robot's hand.
Fun stuff as always! You might want to lay out a course or something for your demonstrations to prove that it's really controllable. Seeing it flail around your kitchen doesn't really show that it works, aside from the fact that it doesn't hit the cabinets, but I think it would be a powerful visual to have a track or line or something to really demonstrate controlled motion. Doesn't even have to be fancy, just tape on the floor of a big room or something. Love your work!
This worked way better than I expected! Now I wonder: if you put a big motor on one end with a belt running off it and around a bunch of rollers, could you put these ball drives against the belt (such that it's pinched between the ball and roller)? That could allow for driving a whole lot of individual wheels or whatnot off a single motor with relative ease. Hmm. Could be interesting!
Do the surfaces that touch get hot ? Would that lead to some sort of slippage over time ? Also if the wheel is touching the floor will dirt cause wear between the two surfaces ? Seems like this would be better if the wheel had an inner surface that was drivable rather than an outer one ....
A neat real world application for this sort of transmission might be a gas powered robotic lawn mower. Could use the already steady running gasoline motor for the motive force, and only need very minimal electrical system for the servos. Super neat.
I'm sure the frictional losses in this thing are quite high. There is essentially 3 places that motor torque must pass through to the wheels. This is an amusing and clever build that is driven from a single motor.
Very cool idea! Looks like some folks beat me to it, but this seems like the perfect setup to use a flywheel to drive the system instead of a motor. If you make a 1 or 2-wheel robot (1-wheel would be awesome!) the flywheel could potentially double as a gyroscope to help you stay upright.
If you rotate the dead zone of the roller spheres over the drive wheel wouldn't that lock the driven wheel in place and disengage the power transfer? Just like when you talked about rotating the dead zone over the driven wheel but would result in free spin
When I was a teen I drew some steampunk scenes, and I would design the machines before drawing them in, I wanted things to look like they actually work. I made spherical gears that would transfer rotation through the joints of a robot modeled on a mountain lion. I didn't think about at the time about the gear ratio changing when it bends and contacts it's limbs, lol. How hard would it be to make a gearing setup that could compensate for that and drive the shafts faster or slower accordingly to effectively use two sphere gears to transfer rotation through a limb's joint? I fear it would be really inefficient.
It`s really nice piece of engineering. In real world application wont there be same issue as with other CVT - excesive wear and tear since actial contact point is really small and will cause warping of material near it?
Very cool. Thank you for sharing this. I think a small improvement would be to make the ball portions a rigid material with surface texture for grip (not flexible), then have replaceable rubber bands on the drive and output disks. *I think that will wear more evenly *be easier to replace without having the original model (later maintenance/rebuild concern) *be less prone for inconsistencies in movement between the 2 sides
I've seen this done with weird ball shaped/wheel configurations ontop of the ball. So simpler wheels (plain ball), with normal wheels driving them on top (one for each direction IIRC).
it works well on a smooth flat surface, but although the CVT allows you to control each wheel's speed independently, doesn't it also trade torque while doing so? That would mean that the wheel spinning faster would push with less force, and in a rough or steep incline turning may behave unexpectedly.
I often wonder if, when James comes home after a day out, several of his robots greet him at the door and say "Home again home again, jiggity jig! Gooood evening, J.B!"
I always wondered about it, but I would assume it lacks power and longevity while using less moving parts than a decently sized gear box. For something bicycle sized it might be feasible though.
The moment you said it'll drive like a tank, I thought about tank treads with a free-rolling capstan and the ball clutches as the pinch-rollers. Not at all sure if it's doable, let alone feasible or even usable. Just a random thought.
I wonder how much torque this setup can handle before the friction breaks between the wheel and the spheres? Perhaps the friction materials molded with some grit could improve the power transfer potential... Great design btw! P.s. perhaps a mechanical lock on the back of the main wheel like a winder gear could provide a park gear for when stopped on an incline?
Dixon Lawn Mowers used this exact system for their zero turn lawn mowers decades ago. They essentially invented the zero turn lawn mower and they used this system for years before switching over to hydraulic zero turn which is the current standard.
The idea is great, the simplicity of it makes it even better. In case of small, light contraptions, it can make sense to develop a prototype and testing it. Although my thoughts can't leave the issue about friction. Imagine a moderately heavy machine using this idea, or a heavier one. I'm not convinced it would be an easy task to solve the possibilities of slipping or grinding. If you can come up with a trusty solution for not "amending" the gear oil with metal shavings or powdered metal, you can make a nice future for yourself. It would be a cool thing to see it working in a long-term use.
You could 3d print some notches that stop the joysticks at different positions so it's easier to drive in a straight line and also easier to choose the turn speed.
is there any advantage of using this over two motors? cost is really the only thing i can think of, and i don't think it would be too expensive to just use a second motor.
Would be interesting to have the angle based on the force the output wheel needs to turn - so you could build a continuous "gear shifter" for bikes that always needs the same input force.
Actual bikes would never use a drive like this because the friction is way too high and it will slip under high load - efficiency is the #1 consideration for road bikes and high torque transfer for off-road bikes.
Thanks to Keeps for sponsoring this video! Head to keeps.com/bruton to get 50% off your first order of Keeps hair loss treatment.
Wait what 1 week ago
I know sponsors pay the bills, but dude... At least get sponsors relevant to the channel!! WTH do hair loss products have to do with robotics?!?!
damn only 7k veiws...
James Bruton
This cardan-gear-based gravity balancer is perfect for your power Efficient robot joints ua-cam.com/video/LNXUASfLbVM/v-deo.html
Lol, new demographic?
This mechanism would be perfect to be powered by a flywheel... *looking at you, Tom Stanton*
Great idea aha
Somehow he would make a trebuchet out of it.
*desire for flywheel monorail grows*
I have no idea what makes you think that, flywheels require efficiency that this mechanism cannot have
If you had a flywheel geared to drive that main drive gear and had a small motor running at its most efficient speed, then instead of the neutral space on the sphere gears, have it somehow take that inertia of the wheels back to the fly wheel you could have regenerative breaking and then the whole think in theory could be quite efficient because it would overall be driven by a very small light weight motor?
Robot arm with CVT. This would be cool. Thank you for inspiration!
hey skyntefic
hi Skyentific , good to see you here,
could you add a gravitational balancing feature, it would be really beneficial for most arm projects.
Excited to see what you do with it, Skyentific! =D
somewhat reminds me of dasa arm
ua-cam.com/video/2-eAZlVFlkM/v-deo.html
a CVT proberbly wont be enogh as it can't put torque on the output shaft while not spinning, you'd need an IVT for that
Given the motor spins continuously I wonder if you could use that as a gyroscopic stabilizing mechanism
Interesting, I doubt with the weight required for gyroscopic stabilization the power loss when engaging the drive wheels would have that much of an effect. Seems do'able.
I believe fly wheels must be untouched in use for stabilization. If a heavy weight were to be attached to the output it would give a reaction wheel effect that would be counter intuitive to the robots movement.
I was thinking that as well. But the cvt's would always be spinning even when the wheel was in neutral so changes in fly wheel would really only happen when engaging and dis engagement. Especially if it was inverted and the flywheel was under slung. But then there is the speed of the flywheel to consider. Would need to recalculate reduction to the wheels.
brilliant
Hi Martin :)
Hello Martin
Hey Martin :-)
Oh, you here? Interresting.
plz upload something
ANYTHING
Remember when this guy was just doing fancy cosplay? This is the first Ive seen from you in years and I cant wait to catch up! Impressive stuff!
Oh, you've missed a lot if you haven't watched since his cosplay builds. You're going to enjoy it. :)
@@StevenIngram Im seeing that haha, ive been binging his new stuff and its great!
Really clever clutch mechanism, I'd imagined something like the little wheel your omniwheels have so it can idle on the drive wheel, but your way is much simpler and better
There is one downside to its use on this particular testbed style of machine - as there is no braking options that wheel is free to run so you can't 'turn it off' and steer around it 100% predictably - its a very neat execution of the mechanism don't get me wrong, but on that tank steering testbed the clutch freewheeling can lead to steering troubles when you wish to turn around one side or other and the free wheel just rolls so you instead get an uncontrolled veering in the direction you want (as the free wheel still has friction and resistance) - simple enough to avoid in practice if you can accept not being able to just turn the one side against a stopped other i.e the very tight radius but not 'on the spot' turns - as in effect it can only turn 'on the spot' (reversing side B from side A) or at any of the wider radius when side B is just solidly engaged and side A is however much faster you desire to create that radius of turn. Add in a brake for each wheel and that trouble goes away entirely too.
I've loved the demos lately, but this is beyond elegant. It is such a simple concept but I've never seen it employed.
Those omni-wheels are mesmerizing as fuck.
you need to work on your vocabulary.
"IVE GOT LOTS OF HAIR!" the best intro for an ad for hair loss prevention 😁
This would be a great setup to use for a huge steam or gas engine powered robot. Something that the power source would not be as responsive as a stepper.
My thought exactly. Hook this up to like a lawn mower engine, which isn't normally conductive to electronic control. Make a nimble robotic lawnmower!
Like a sterling engine!
Nice, made all in copper. The cvt's could be copper toilet floats.
Oh and instead of servos you could use valves and actuators. The sterling engine could also generate the pressure to move them and the power for communication.
Yeah, the engine would only need to maintain constant RPM. The CVT wheels would translate that into useful motion. If the same engine could generate enough surplus power to run its electronics and servos you'd have a pretty long range robot.
The robot arm idea is really interesting! It reminds me of old sawmills, where the equipment would clutch in and out of the main waterwheel which delivered power via a belt and pulley system going through the building. For the arm it means you’d only need one strong drive motor rather than a motor at each joint - that definitely shows promise!
These video are literally highlight of my week. I look forward to them every single week and I cant explain enough how much I enjoy them.
Do you have lots of hair ?
@@travisash8180 wth
That makes for a great zero-point turning system. Lots of maneuverability - but I would be worried about slipping on those transmission spheres alot.
That's super clever! You are definitely one of my favorite and most inspirational UA-camrs!
The servo arms waving around makes it look like there's some little creature up there pulling in the reigns of its mechanical beast.
Now that sound awesome to make.
The final robot is super cool! and seeing the servo arms swinging back and forth to control the robot is very cool and a little mesmerizing.
Really really cool--there are few videos that make me exclaim out loud, but this one surprised me with how far it went. The volume and diversity of your output of ideas is inspiring
If civilization collapses,guys like you,tom stanton,clickspring,peterspiol,integza and other diy geniuses will reign supreme.
Is this the most practical way of making wheels turn at different speeds? No.
But is it the most FUN way? Definitely!
And quite impressive too!
Yeah, it's quite pointless for actual usage. For robot locomotion, the wheel hub motors seen in hoverboards seem to be the pinnacle in actual design. Proto G recently put for of them together in an omniwheel configuration. But always interesting to see new ideas.
These sort of mechanical solutions to common ways to do things make your builds really fun to watch in action.
yea, i've got one of these CVTs on my bycycle...'nuvinci harmony' now enviolo... fully automatic...awesome.
Looks like a modified version of the drive system used by "zero turn" riding mowers? The gas engine turns a flywheel at a constant speed and variable reduction transmissions control each of the two drive wheels, allowing forward and reverse at a range of speeds.
sounds like that also has the advantage of being able to keep the engine at the ideal rev count for power/efficiency
In reality zero turn mowers use two independent hydrostatic drive mechanisms, hydraulic motors powered off of a common pump, the drive levers are connected to valves. Most ordinary lawn mowers with steering wheels have a single hydrostatic drive these days, much easier to use than a conventional transmission when you want to have variable speed and don't want to be messing with the throttle and therefore changing your blade speed.
@@AlRoderick Many zero turns have a pair of independent sealed hydrostatic drive units that are belt driven. A single belt drives both input shafts directly from the motor, and control arms attached to the drive units control their ratios much like the design in the video. The actual reduction mechanism is very different, involving movement of swashplates that change the displacement ratios rather than changing the flow rate with valves.
Since they don't have the restriction of valves they are more efficient, and much like a torque converter tend to work great until they don't. I've rebuilt them before for customers and currently own a mower with this system, but my next one will probably be the pump/valve/orbital motor style just because it removes the reliance on proprietary parts (Deere, of course)
Wow, I love the broad visual movement in the servo connected arms as it maneuvers. Makes it really nice to watch
I would connect the joystick on disgonals. That way, pushing forward is forward on both, and pulling back is back on both, as well as turning.
I'm amazed at the consistent high quality of your videos
This is brilliant!!!! I'm not sure in how far this is patent-able but as robots go this one will go far. The limited amount of moving parts. The load-bearing moving parts.... Man this has a future.
I kid you not. As an example I see this riding on mars but also as a more sturdy Roombah. Just imagine this clutch system adapting to different types of flooring by using the clutch. Not just the wheels but also the brushes.
I really enjoy these proof-of-concept videos!
Always coming up with something new and I am here for it! Really interesting system, visually very cool and mechanically fascinating.
Wow, you make it look so easy. You're a very busy person, and I really appreciate everything you share here.
And I know my CVT comment on the previous ball video probably didn't influence the creation of this bot. It would be a natural progression to do this one. But it's nice to think it might have. Well done.
For some reason I was researching this same cvt system earlier this week. What a coincidence you made one as well. Great job. I think it could have a use for a steam power or even gas powered robot. It would really shine if it was 6 wheeled, all direct driven by a single belt and the same 2 servos on a single motor. It could be a great drone vehicle with gas it could run for hours and miles in this configuration.
OH MAN!! Just another 9 more subscribers and you will have 1 million. WOW! Congratulations when you do. You been doing this for a long time. Congrats and enjoy 👏👏👏
This clutch is nothing short of brilliant! Love the idea of a mini robot with these fellas on!
This setup would be great for an internal combustion motor based robot since that's usually constantly running.
This would probably be more important for motors who like specific RPMs, like ICEs.
Wow!!...Crazy concept. You are truly a genius. Your thought process and overall implementation is so inspiring. Keep up the good work.
Deeply neat. Thanks for exploring these new wheel and transmission designs.
James its 12am in the usa go to sleep we understand you want to keep putting out the amazing content but seriously get some sleep D:
Luckily he released the vid at 7 am (GMT + 1) which I think is his timezone.
He’s from britain so right now its pretty early
I think it's a joke friends
I work swing in Oregon so it was kinda nice to come home and watch a new vid from James. Love the content 👌🏾
Sleep loss can be a major contribution to hair loss.
I am most impressed with the demo of this, well done. Also, excellent music used in the demo too.
I've got a bicycle with a CVT-it's pretty fun. In fact, that sphere-based CVT video you showed is by the company that designed the transmission on my bike.
The CVT gearbox is called “Enviolo”, and the bike is called the Priority Continuum Onyx, if you're curious.
That's such an awesome design, can't wait to see it evolve even further.
BTW, maybe using small cycloidal drives instead of the servos and arms to change the tilt of the hemispheres
I think it could give a more compact and sturdy fit, if placed by the side of the hemisphere's holder, though the current arms design look interesting, looks like there's someone inside manually driving it :D
Wonderful video....excellent content for around a decade.......
biggest problem is the operation with high torque. We had this problem in a project for a client, but ended up making a new type of gears which worked well. It is still in the process and will be patented. But for anything with a limited torque, this is grrrreat
I'd imagine that in real world application, debris on the wheel would be pushed between the friction points and would cause quick wear and tear on the half balls driving the wheel.
This design probably best to be enclosed in a dust free environment, like perhaps a shoulder for a robot.
@@TheNightstalkerShow yes. But even then you will have the problem, just less. Imagine when a lot of torque applies.
@@el-domo Agreed. It becomes a choice in materials used at that point. Metal on rubber, or nylon on some other material. Whichever part being easier and cheaper to replace should wear first. It's a neat design with niche uses... I wouldn't expect it to be used for moving parts that operate 24/7.
@@TheNightstalkerShow They all wear out and can't take high torque anyway. If you find out, it's extremely expensive and limited to the applications anyway
@@el-domo Probably can design a half sphere bevel gear with no spiral these days to handle the torque and get same functionality as the friction only method... but you say you have a patent pending, so I doubt you'd confirm it :D
Very impressive!
Lots of friction, but very good!
So you have a spinning disk that is almost at constant speed. Does this mean you can turn this into a flywheel for some sort of self balancing?
This is a nice little piece of mechanical engineering. Really awesome how it works. Maybe use a set of clutches on a drive shaft like you suggested to keep the motor weight down and close to the center of mass and then use it just for a hand, like an improvement on your Ultron robot's hand.
i have never seen any thing like this and it is awesome
Fun stuff as always! You might want to lay out a course or something for your demonstrations to prove that it's really controllable. Seeing it flail around your kitchen doesn't really show that it works, aside from the fact that it doesn't hit the cabinets, but I think it would be a powerful visual to have a track or line or something to really demonstrate controlled motion. Doesn't even have to be fancy, just tape on the floor of a big room or something. Love your work!
It's like a regular two wheeled robot, but with extra steps! Super cool.
This worked way better than I expected! Now I wonder: if you put a big motor on one end with a belt running off it and around a bunch of rollers, could you put these ball drives against the belt (such that it's pinched between the ball and roller)? That could allow for driving a whole lot of individual wheels or whatnot off a single motor with relative ease. Hmm. Could be interesting!
Do the surfaces that touch get hot ? Would that lead to some sort of slippage over time ? Also if the wheel is touching the floor will dirt cause wear between the two surfaces ? Seems like this would be better if the wheel had an inner surface that was drivable rather than an outer one ....
The clutch rods points in the same side direction when it travels straight frustrates my need for symmetry so much, lol
A neat real world application for this sort of transmission might be a gas powered robotic lawn mower. Could use the already steady running gasoline motor for the motive force, and only need very minimal electrical system for the servos. Super neat.
I'm sure the frictional losses in this thing are quite high. There is essentially 3 places that motor torque must pass through to the wheels. This is an amusing and clever build that is driven from a single motor.
Channels like yours are the ones I love subscribing to. ✌️👏😎
Awesome mechanism! I can't wait for an arm driven by this
Very cool idea! Looks like some folks beat me to it, but this seems like the perfect setup to use a flywheel to drive the system instead of a motor. If you make a 1 or 2-wheel robot (1-wheel would be awesome!) the flywheel could potentially double as a gyroscope to help you stay upright.
1:45 pufferfish having fun
If your recent projects have looked like hacks I don't know what mine look like. I have really been enjoying the recent content.
If his are messy then mine might just summon Cthulhu for the sole purpose of him looking at me with whatever passes for a raised eyebrow
I am extremely impressed you are an extremely smart man i never would have thought of that, and that's why I love your videos
Thank you
If you rotate the dead zone of the roller spheres over the drive wheel
wouldn't that lock the driven wheel in place and disengage the power transfer?
Just like when you talked about rotating the dead zone over the driven wheel but would result in free spin
When I was a teen I drew some steampunk scenes, and I would design the machines before drawing them in, I wanted things to look like they actually work. I made spherical gears that would transfer rotation through the joints of a robot modeled on a mountain lion. I didn't think about at the time about the gear ratio changing when it bends and contacts it's limbs, lol.
How hard would it be to make a gearing setup that could compensate for that and drive the shafts faster or slower accordingly to effectively use two sphere gears to transfer rotation through a limb's joint? I fear it would be really inefficient.
It`s really nice piece of engineering. In real world application wont there be same issue as with other CVT - excesive wear and tear since actial contact point is really small and will cause warping of material near it?
Very cool. Thank you for sharing this.
I think a small improvement would be to make the ball portions a rigid material with surface texture for grip (not flexible), then have replaceable rubber bands on the drive and output disks.
*I think that will wear more evenly
*be easier to replace without having the original model (later maintenance/rebuild concern)
*be less prone for inconsistencies in movement between the 2 sides
This is so satisfying to watch it work idk why but I love the arms moving the balls
I've seen this done with weird ball shaped/wheel configurations ontop of the ball. So simpler wheels (plain ball), with normal wheels driving them on top (one for each direction IIRC).
it works well on a smooth flat surface, but although the CVT allows you to control each wheel's speed independently, doesn't it also trade torque while doing so? That would mean that the wheel spinning faster would push with less force, and in a rough or steep incline turning may behave unexpectedly.
Awesome! Been curious about how effective this type of cvt would be for a while, a similar ivt has been on my list of to do for a while.
How much does it take for the wheels to slip, and what is the efficiency?
Really cool! Rather than add complexity to stop, why not have a rubber cap on the transmission wheel so vertical is a brake instead of freewheeling?
10:07 I‘m having a flashback to Ross screaming „PIVOT!!!“
Again, excellent proof of concept development and display .
I often wonder if, when James comes home after a day out, several of his robots greet him at the door and say "Home again home again, jiggity jig! Gooood evening, J.B!"
James you are an absolute genius
So much awesomeness going on it's hard to keep up in a first watch through but this is a great video!
That is such a genius design, WOW!!!!
The back little wheel just standing there... *M E N A C I N G L Y*
I always wondered about it, but I would assume it lacks power and longevity while using less moving parts than a decently sized gear box. For something bicycle sized it might be feasible though.
The moment you said it'll drive like a tank, I thought about tank treads with a free-rolling capstan and the ball clutches as the pinch-rollers. Not at all sure if it's doable, let alone feasible or even usable. Just a random thought.
It works great! You can also use a flywheel instead of the electric motor and still have excellent control
With those two levers on the top, it looks like a strange version of a steam locomotive. Really cool! :)
someone please give James a proper TV show so he doesn't have to plug these companies lol. The guy should be on TV - he's so awesome
I wonder how much torque this setup can handle before the friction breaks between the wheel and the spheres? Perhaps the friction materials molded with some grit could improve the power transfer potential...
Great design btw!
P.s. perhaps a mechanical lock on the back of the main wheel like a winder gear could provide a park gear for when stopped on an incline?
Dixon Lawn Mowers used this exact system for their zero turn lawn mowers decades ago. They essentially invented the zero turn lawn mower and they used this system for years before switching over to hydraulic zero turn which is the current standard.
really awesome inspirational video. Thanks for the great work you are doing. what is the TPU material you have used for the balls and even wheels?
Dude you’re an absolute genius
The idea is great, the simplicity of it makes it even better. In case of small, light contraptions, it can make sense to develop a prototype and testing it. Although my thoughts can't leave the issue about friction. Imagine a moderately heavy machine using this idea, or a heavier one. I'm not convinced it would be an easy task to solve the possibilities of slipping or grinding. If you can come up with a trusty solution for not "amending" the gear oil with metal shavings or powdered metal, you can make a nice future for yourself. It would be a cool thing to see it working in a long-term use.
This is so cool! I imagine this is how steam powered car engines control their speed.
It's not, but you could make one that uses it.
These robots are getting more and more interesting
Dude watching your videos are the best part of my day ❤
Nice Toroidal Cycloidal Transmission Drive.
this is genius and would make a cute robot
You could 3d print some notches that stop the joysticks at different positions so it's easier to drive in a straight line and also easier to choose the turn speed.
is there any advantage of using this over two motors? cost is really the only thing i can think of, and i don't think it would be too expensive to just use a second motor.
If James Burton manages to build this concept into a self balanced robot, i will grant him the title god of the mechanical world and beyond!
Would be interesting to have the angle based on the force the output wheel needs to turn - so you could build a continuous "gear shifter" for bikes that always needs the same input force.
Actual bikes would never use a drive like this because the friction is way too high and it will slip under high load - efficiency is the #1 consideration for road bikes and high torque transfer for off-road bikes.
@@PKMartin but that's not fun.
Amazing Video 👍 love your work 😁🙌
Wow, great stuff. This could be applied to a stirling engine to make a drivetrain, as the output of a stirling engine is quite constant...
can you imagine someone like Leonardo DeVinci with a 3D printer?. That's the ingenuity I see here on this channel. Awesome.