I've seen a number of Lego CVTs/automatic transmissions and this is the first time I've seen someone use variable diameter wheels. I suppose there will be some irregularities in the output speed, similar to what you get with a universal joint. I'm definitely interested to see if it works in practice in a vehicle or something. It would be interesting to see if it will go to the highest mechanical advantage on acceleration from a stop of a heavier vehicle.
Something else to consider is the fragility of the Lego chain pieces. I don't have confidence that a high-torque load wouldn't completely annihilate it. Might be a good idea to make the wheels dual chain-drive
I'm amazed at the simplicity of the chain drive automation, but as it stands the gear ratio is currently quite small. I'd be very interested to see a follow-up video that puts two of these stages in series, to see if you can automatically cover a much larger range of reduction speeds.
Thanks! That's a great idea! I actually did a staged automatic transmission like that in the automatic gearbox I made last year to increase the range of gear ratios like you said (though overall that mechanism isn't great compared to this one). I've already got a lot of ideas with this CVT for the next video, but this staged transmission idea could be worth me making a third video on this!
@@techheck3358 I thought about that, but there's two issues with the idea - Both stages won't be at the same gear ratio simultaneously because the force acting on them is not identical, and they may change at different rates. Also, because they are in series, they are not running at the same speed so there would be an interference pattern of the output speeds of the two individual stages as a result.
@@fallen3424 I think (as @techcheck3358 said) they (first stage output gear and second stage input gear) need to be connected out-of-phase. And additionally the shrinking of the first stage's output gear has to be synchronized to the growing of the second stage input gear. Since those gears are usually not at the same size (except for medium gear ratios), their "non-spherical-ness" will usually not fully cancel out, but at least it will not interfere constructively.
If you were willing to sacrifice the cones to live forever for a CVT, you could dip them in liquid rubber. Rubber on rubber is friction city. I think the auto CVT would need to live within a specific RPM range or the centrifugal forces will counter those bands.
@@ArabianShark Yeah it very well could hinder quick CVT shifting. The rotation of the cones should reshape the bands to be in line though. If they aren't rotating you could run into issues shifting.
I seriously expected this channel to be at around a million subscribers like the other technic channels. But this lad is better for giving voiceovers. Keep it up!
I'm shocked that just two rubber bands can replace all that complicated control structure with the differential, and that the other wheel just does what it needs to automatically is really cool! Very ingenious design!
in a realistic application you probably want it to always to default to the shorter ratio, that way you aren't stalling your input instantly, I know the belt drive CVTs for small displacement engines work like that, using centrifugal force to actuate them
Yeah, I'd say that for any high-torque applications, like starting off a car from rest, the automatic transmission will go to the highest gear ratio at first, and then gradually lower the gear ratio as the car builds up speed.
@@BananaGearStudios The biggest factor is that a brushed DC motor makes its maximum torque at 0 RPM which is the opposite of a petrol engine and hence why your design is good as it is
@@evergreatest7644 It's a good question. Imagine an electric motor that has a bar attached to the output shaft and is locked in place. Now ignoring that it won't last long before overheating it will with that setup be exerting the maximum amount of twisting force it can on the bar trying to make it move. You won't be able to visibly see any movement but if you had something like a strain gauge to measure it you would find it's exerting the highest force at that point. You may have experienced this characteristic after when a battery was directly applied to a small motor and see it suddenly jolt from the torque reaction. Note this is for direct current brush type motors only, brushless motors behave differently. Does that make sense? I am not my best with words and it's hard for me to explain.
@@Samqdf Oh, that makes sense, and Internal Combustion Engines can't provide torque at 0 RPMs because they will stall and stop producing torque. Thanks.
I think a lot of the problem is not the lack of friction but the stretch of the rubber bands. If you had a belt that had something to keep it contracted it would work far better as the side load of the torque transfer tightens the belt and makes it tensioned against the variable cones.
2:23 (i think) this is because the formula for the force of friction is F = N(mu), where N is the normal force and mu (the greek letter, not m * u) is the friction coefficient. There's nothing about surface area in the formula; only the normal force and friction coefficient matter. The larger band might have a smaller friction coefficient ("more slippery" essentially, like you said), but what would really help is to have a tighter band. That will apply a higher normal force against the two cylinders, increasing the friction force F.
there is a cvt type like this although usually it uses a rubber belt, one side has fixed diameter and the other uses wheels on the end, it also uses hydraulic/pneumatics for expansion. your design is very cool with counter changing diameters for drive and driven giving a wider gear ratio change, although due to the size of lego pieces, this is slightly limited and is probably very hard to make smaller for lego cars. it will be cool to see that comes of this design in the future
If the chain is also lubricated with oil, then friction losses will be even less than in a manual gearbox! I really want to see how such a variator will accelerate the car and pull it uphill! 👍
3 points: 6:30-ish: You could have used gears to drive the chain, you'd just need to fix them with a ratchet rather than fixing their axle. 0:27 I would have tried wrapping the rubber band around the cones more times. You couldn't wrap it around the same cone more than once, but you could make it wrap around 3 cones per side to complete the 180° turn: 225° around the first cone, 270° around the second one, and then 225° around the last one. The first and last cone can both be geared to the input/output shaft. And the second cone can be used to pinch the rubber band at the contact points of the first/second and second/third cones. 13:01 Route your output into a differential, hook one of the diff's outputs into a device where the drag can increase with speed (like a large fan), and hook the other side into a pointer whose deflection is proportional to the torque (working against a weight or a rubber band). The differential will make sure that both of it's outputs are at the same torque, the pointer will allow you to read off the torque, and the fan will make sure that the output of the CVT can move even though the pointer stays still.
Woah. I've been thinking about a similar chain drive idea for years for a bicycle, but never really actually attempted to make it. Good job for actually implementing this!
Cool concept. I've always had this idea that I could never realise. But what if a transmission would switch gears depending on how much force there's on the chain? In other words, it might be cool if the chain was going over a slack wheel that controls the CVT lever. Could be worth a try.
Frictional force depends on force per unit area. As contact area increase, the force per area decreases proportionally. And so, the width of the rubber band doesn't really effect the available friction. Instead, what matters is the spring force and unsprung diameter of the rubber band. As that will increase the force over the same area.
Because the chain still slips in one direction, this would be a great mechanism to use on a bike as it would allow the bike to coast with out the need for a ratchet mechanism.
I was just having the idea for a variable wheel transmission after seeing the first design and then you pull it out 😂 This is a beautiful concept, i wish you luck on the adventure
The automatic CVT is literally mindblowing I'm not even sure if I understood anything even though you explained it like 3 times, I wonder if you're working as an engineer for some company everytime I watch one of your videos
13:13 measure voltage and current going to the motor That's probably one of the easiest ways to see Cause you're getting more torque out as you're spreading the load over time So power should stay somewhat constant
9:25 I've used a system like this before on the back propeller on a Lego heli I made. It created a bit too much friction though so I went with another design. Still cool to see someone else use it too.
Great concept and execution, the automatic version is especially clever. My only concern is that, since the clips necessarily grab the chain in discrete intervals, this transmission does not appear to truly be continuously variable. Unfortunately a truly continuously variable transmission that does not rely on a friction drive (belt drive, pivoting rollers) is a bit of an unsolved problem in mechanical engineering.
Thanks, but the gear ratio depends on the diameters of the wheels, which can vary continuously, not on where the chain links connect. This is because, when the wheels' diameters change a bit, the chain becomes more/less slack before it slides to the next position, so the gear ratio is continuously variable.
what's the flutter like on the output gear? given the gearing is effectively triangular won't the output speed vary over the course of a rotation, doubly so with the continuously variable design and having inconsistent alignment?
I'm sure there's a bit of variance in the output's speed, but it wasn't all that noticable while I was testing or filming it. Once I put it in a car I'm sure any variations will become more apparent, but I'd say they shouldn't be too bad since the wheels are shaped like very rounded triangles rather than straight triangles.
if you put another of those auto cvts in series it should both multiply low end torque and high end speed and you could probably do this until the friction diminishing returns over take multiplier which should be quite a few if it has as little friction that appears in the video.
I like the version with fixed gears on the wheels, in a bigger version, maybe the gears could spin with the change in wheel size to still be engaged with the belt but allow the gears to roll along the gear whilst the wheels change size if that makes sense
One thing I always wonder when I see a Lego transmission where friction or tension are used to gauge output resistance: is the torque gain from changing ratios greater than the overall power loss from the friction/tension in the resistance measurement? And if so, by how much? Genuinely don’t know. I’m as far from an engineer you can get, but I love Lego technic. Regardless of how efficient this design ultimately is, it’s incredibly clever and undeniably solves the belt slip issue with the classic cone style CVT. Great work on the variable size wheel design, it’s brilliant!
It takes some energy to stretch a rubber band. But once it’s stretched, it doesn’t continuously require energy to keep it that way. It momentarily takes some energy from the input shaft and then the torque gain takes effect. When it shifts the other way, it feeds the energy back into the system (minus what was lost to heat ofc)
So, my love is bicycles, not cars, but we cyclists are stuck with three types of transmission, derailleur, planetary hub and CVT hub. Even people who don't know bicycles probably know the first 2, eg the traditional racing bicycle gears or the sturmey archer 3 speed hub of cruiser bikes. Now, hub gears are both very expensive and very proprietary, while derailleurs can be cheap, are open source, modifiable and easy to work on. It's also easy to understand the torque transfer going on with derailleur gears. The trouble is, you can't belt drive a bicycle with derailleur gears and chains are not the best thing for being exposed to the elements. Belt drives are becoming very popular on hub geared bikes but for what most people are happy to pay for a geared commuter bike will barely buy you a decent hub gear, like a Rohloff, and that's without rim or spokes, or the the rest of the bike. Your automatic, twin variable pinion CVT could revolutionise urban bicycles! Seriously! It's compatible with open gearing (or at least "chain guarded" gearing), compatible with timing belt drive, would feel like riding a single speed (and be just as easy to understand) and matches the rider output to the gradient. It would also be a lot cheaper to make than a planetary or CVT hub gearing system, easier maintenance, only the belt would wear out (and bicycle belts are lasting up to 120000km! Longer than the average life of an urban bicycle, let alone the chain that drives it. Dude! Patent this in the Netherlands or Denmark! You'll be a billionaire! Seriously! This is brilliant for so many exposed drive applications! But really, make it open source, make the world a better place! Then develop this up from lego to timing belt drives, both automotive and for cycles.
I wonder what kind of performance you could get with a pulley based CVT instead. Both the input and the output side have 2 pulleys that get closer or farther apart to change the diameter of the belt. If you do try it out, maybe see if there's some type of V-Belt you can use, it should provide the most friction. A round (instead of a flat) band may also work for a pulley based CVT specifically
Its seems really cool for cars but i think it could be to big to fit in a smaller car. How big is it and could you upload instructions how to build it?
about why the wider rubberband didnt improve as much as hoped: the formula to describe friction does not include the area of contact but instead the friction force f is calculated by multiplying the normal force N with the friction coefficiant μ between the two surfaces: f = N * μ this does not include the area of contact
When a kid with Legos built a CVT that doesn't slip? Damn, someone at Nissan should hire this kid to stop their CVT's from wearing out after 50,000 miles.
I wonder...could you theoretically have four CVTs for four wheels? Basically just a direct drive from the engine to each transmission? I know it wouldnt be too practical in a car, because of the amount of things you have to lubricate and maintain, but could it be possible?
Nice build, I really like the mechanism with the differential. However, It is not a true cvt, you are limited to grabbing onto the individual linkages of the chain, and you cant grab onto half a linkage. Still very impressive work though!
Thanks for the comment. However, the gear ratio depends on the diameters of the wheels, not on the chain links that they connect to - the chain is able to become more or less slack to accommodate the wheels changing size slightly before the chain linkages move over. Since the wheels' diameters can vary continuously, I would say that it functions as a CVT.
@@BananaGearStudios well, in practice it functions like a CVT, but it is not truly continues. It can only grap an integer amount of chain links per rotation, and so can the other side. When you increase the size of the wheel, at some point it goes from let's say 20 to 21 chain links per "teeth" of the wheel. it can't grab 20,37582 chain links. This results in a fixed number of ratios. Although there are a lot of ratios, there are not infinite like in a true CVT. A true CVT which is not based on friction is yet to be invented. Non the less, the contraption you made comes close, so keep thinkering, maybe you are the person to find a solution some day! (And probably get rich xD)
make it 4 or more (preferably 8) gear wheels per size-changing wheel and make only one of the gear wheels fixed so that the chain stays unfixed to rotating gears for the most amount of time (when the one fixed wheel is not in contact with the chain
I've seen a number of Lego CVTs/automatic transmissions and this is the first time I've seen someone use variable diameter wheels. I suppose there will be some irregularities in the output speed, similar to what you get with a universal joint. I'm definitely interested to see if it works in practice in a vehicle or something. It would be interesting to see if it will go to the highest mechanical advantage on acceleration from a stop of a heavier vehicle.
Thank you! Yep, putting it in a car and using it to accelerate some heavy vehicles is exactly what I've got planned for the next video.
That will bee amazing
@@BananaGearStudios you could try using 2 of these side by side but offset 180⁰ to help smooth the output
@@BananaGearStudiosawesome can’t wait to see that
Something else to consider is the fragility of the Lego chain pieces. I don't have confidence that a high-torque load wouldn't completely annihilate it. Might be a good idea to make the wheels dual chain-drive
The automatic CVT blew my mind
Haha, thank you!
Those are called variators. They have been used in snowmobiles forever and are much simpler than this.
I'm amazed at the simplicity of the chain drive automation, but as it stands the gear ratio is currently quite small. I'd be very interested to see a follow-up video that puts two of these stages in series, to see if you can automatically cover a much larger range of reduction speeds.
Thanks! That's a great idea! I actually did a staged automatic transmission like that in the automatic gearbox I made last year to increase the range of gear ratios like you said (though overall that mechanism isn't great compared to this one). I've already got a lot of ideas with this CVT for the next video, but this staged transmission idea could be worth me making a third video on this!
I think having two stages in series would exaggerate the irregularities in the output speed more, so that might be rough.
@@fallen3424i think it could actually help if they’re connected out-of-phase!
@@techheck3358 I thought about that, but there's two issues with the idea - Both stages won't be at the same gear ratio simultaneously because the force acting on them is not identical, and they may change at different rates.
Also, because they are in series, they are not running at the same speed so there would be an interference pattern of the output speeds of the two individual stages as a result.
@@fallen3424 I think (as @techcheck3358 said) they (first stage output gear and second stage input gear) need to be connected out-of-phase. And additionally the shrinking of the first stage's output gear has to be synchronized to the growing of the second stage input gear.
Since those gears are usually not at the same size (except for medium gear ratios), their "non-spherical-ness" will usually not fully cancel out, but at least it will not interfere constructively.
If you were willing to sacrifice the cones to live forever for a CVT, you could dip them in liquid rubber. Rubber on rubber is friction city. I think the auto CVT would need to live within a specific RPM range or the centrifugal forces will counter those bands.
That's a cool idea!
Wouldn't that hinder the movement of the rubber band along the height of the cones as well?
@@ArabianShark Yeah it very well could hinder quick CVT shifting. The rotation of the cones should reshape the bands to be in line though. If they aren't rotating you could run into issues shifting.
You should work for tesla
Very interesting concept! If you want a smoother output, you could run another pair in parallel but offset by 45°.
This is a solid build, probably the most impressive lego transmission I've seen!
Thank you!
I seriously expected this channel to be at around a million subscribers like the other technic channels. But this lad is better for giving voiceovers. Keep it up!
Thank you! I'm really glad you like the style of the videos.
@@BananaGearStudios ofcourse! I went and watched almost every other video you posted 😅😅
I'm shocked that just two rubber bands can replace all that complicated control structure with the differential, and that the other wheel just does what it needs to automatically is really cool! Very ingenious design!
in a realistic application you probably want it to always to default to the shorter ratio, that way you aren't stalling your input instantly, I know the belt drive CVTs for small displacement engines work like that, using centrifugal force to actuate them
Yeah, I'd say that for any high-torque applications, like starting off a car from rest, the automatic transmission will go to the highest gear ratio at first, and then gradually lower the gear ratio as the car builds up speed.
@@BananaGearStudios The biggest factor is that a brushed DC motor makes its maximum torque at 0 RPM which is the opposite of a petrol engine and hence why your design is good as it is
@@Samqdf Forgive my ignorance but how can there be any torque at 0 RPM?
@@evergreatest7644 It's a good question. Imagine an electric motor that has a bar attached to the output shaft and is locked in place. Now ignoring that it won't last long before overheating it will with that setup be exerting the maximum amount of twisting force it can on the bar trying to make it move. You won't be able to visibly see any movement but if you had something like a strain gauge to measure it you would find it's exerting the highest force at that point. You may have experienced this characteristic after when a battery was directly applied to a small motor and see it suddenly jolt from the torque reaction. Note this is for direct current brush type motors only, brushless motors behave differently. Does that make sense? I am not my best with words and it's hard for me to explain.
@@Samqdf Oh, that makes sense, and Internal Combustion Engines can't provide torque at 0 RPMs because they will stall and stop producing torque. Thanks.
Very elegant device
Thanks!
THIS IS FIRST FULLY AUTOMATIC "GEARBOX" OUT OF LEGO!!! AMAZING BRO!
possibly the first automatic CVT, but definitely not the first automatic gearbox
idc about the legos his voices is so calming
Thanks, that's really kind of you.
@@BananaGearStudios tbh i almost fell asleep, your voice is so calm i love it
underrated af!
Thanks!
Watching these makes me feel the same as watching a Ben Eater video
thats so cool!! you make the mechanics seem so approachable and you explain stuff really well. great builds :)
Thank you! I'm glad you liked it!
Love the videos man!
Thank you, I really appreciate that!
That design is awesome!
I really like how you walk us through the whole thinking process.
This was one of the coolest designs I have ever seen!!!!
I think a lot of the problem is not the lack of friction but the stretch of the rubber bands. If you had a belt that had something to keep it contracted it would work far better as the side load of the torque transfer tightens the belt and makes it tensioned against the variable cones.
That's a good point, though making the belt tighter might also increase the amount of friction in the CVT.
A massive improvement on your previous CVT design, and I also can't wait to see this in a car.
2:23 (i think) this is because the formula for the force of friction is F = N(mu), where N is the normal force and mu (the greek letter, not m * u) is the friction coefficient. There's nothing about surface area in the formula; only the normal force and friction coefficient matter. The larger band might have a smaller friction coefficient ("more slippery" essentially, like you said), but what would really help is to have a tighter band. That will apply a higher normal force against the two cylinders, increasing the friction force F.
This is really cool! I love the design of the wheels with the variable diameter : )
Thanks!
You are so clever. Indeed the best lego transmisión i have ever seen
Thank you, I'm glad you liked it!
there is a cvt type like this although usually it uses a rubber belt, one side has fixed diameter and the other uses wheels on the end, it also uses hydraulic/pneumatics for expansion. your design is very cool with counter changing diameters for drive and driven giving a wider gear ratio change, although due to the size of lego pieces, this is slightly limited and is probably very hard to make smaller for lego cars. it will be cool to see that comes of this design in the future
If the chain is also lubricated with oil, then friction losses will be even less than in a manual gearbox! I really want to see how such a variator will accelerate the car and pull it uphill! 👍
I remember seeing another similar lego cvt that also used a chain and size changing wheel but it used one size changing wheel instead of two.
Amazing concept, the variable size wheels are genius by themselves.
Definetely worth stacking a couple of these together to get a wider range of ratios
3 points:
6:30-ish: You could have used gears to drive the chain, you'd just need to fix them with a ratchet rather than fixing their axle.
0:27 I would have tried wrapping the rubber band around the cones more times. You couldn't wrap it around the same cone more than once, but you could make it wrap around 3 cones per side to complete the 180° turn: 225° around the first cone, 270° around the second one, and then 225° around the last one. The first and last cone can both be geared to the input/output shaft. And the second cone can be used to pinch the rubber band at the contact points of the first/second and second/third cones.
13:01 Route your output into a differential, hook one of the diff's outputs into a device where the drag can increase with speed (like a large fan), and hook the other side into a pointer whose deflection is proportional to the torque (working against a weight or a rubber band).
The differential will make sure that both of it's outputs are at the same torque, the pointer will allow you to read off the torque, and the fan will make sure that the output of the CVT can move even though the pointer stays still.
Really like the torque measurement idea.
This is such a beautiful mechanism
Thank you!
Woah. I've been thinking about a similar chain drive idea for years for a bicycle, but never really actually attempted to make it. Good job for actually implementing this!
Your automatic CVT is incredible 🤯
Now I just wanna see a complete car with all Lego mechanics
This is crazy. You sir are a smart man
nice simple mechanism to achieve a complex result. well done :)
So awesome !
Please, if it didn't exist before, name it!
I'm sure this mechanism will be really useful in more than one contraption
Brilliant! Variable-radius wheels!
Cool concept.
I've always had this idea that I could never realise. But what if a transmission would switch gears depending on how much force there's on the chain? In other words, it might be cool if the chain was going over a slack wheel that controls the CVT lever. Could be worth a try.
Frictional force depends on force per unit area. As contact area increase, the force per area decreases proportionally. And so, the width of the rubber band doesn't really effect the available friction. Instead, what matters is the spring force and unsprung diameter of the rubber band. As that will increase the force over the same area.
Because the chain still slips in one direction, this would be a great mechanism to use on a bike as it would allow the bike to coast with out the need for a ratchet mechanism.
I was just having the idea for a variable wheel transmission after seeing the first design and then you pull it out 😂
This is a beautiful concept, i wish you luck on the adventure
this deserves more views this is incredible
The automatic CVT is literally mindblowing I'm not even sure if I understood anything even though you explained it like 3 times, I wonder if you're working as an engineer for some company everytime I watch one of your videos
Very very cool I’m intrigued to see new videos
13:13 measure voltage and current going to the motor
That's probably one of the easiest ways to see
Cause you're getting more torque out as you're spreading the load over time
So power should stay somewhat constant
One idea for a car it to try to put 2 of these in series giving more torque and greater spead ranges! Excellent work!
Cool concept. Of course a differential eCVT like Toyota is using in their Hybrid cars would be better but this one shows how its working :)
9:25 I've used a system like this before on the back propeller on a Lego heli I made.
It created a bit too much friction though so I went with another design.
Still cool to see someone else use it too.
Great concept and execution, the automatic version is especially clever. My only concern is that, since the clips necessarily grab the chain in discrete intervals, this transmission does not appear to truly be continuously variable. Unfortunately a truly continuously variable transmission that does not rely on a friction drive (belt drive, pivoting rollers) is a bit of an unsolved problem in mechanical engineering.
Thanks, but the gear ratio depends on the diameters of the wheels, which can vary continuously, not on where the chain links connect. This is because, when the wheels' diameters change a bit, the chain becomes more/less slack before it slides to the next position, so the gear ratio is continuously variable.
Definitely wanna see this in a lego RC car climbing hills and pulling loads
Love the videos
So much better with you talking
Thank you! That's great to hear, I enjoy making them this way a lot more as well.
what's the flutter like on the output gear? given the gearing is effectively triangular won't the output speed vary over the course of a rotation, doubly so with the continuously variable design and having inconsistent alignment?
I'm sure there's a bit of variance in the output's speed, but it wasn't all that noticable while I was testing or filming it. Once I put it in a car I'm sure any variations will become more apparent, but I'd say they shouldn't be too bad since the wheels are shaped like very rounded triangles rather than straight triangles.
Dud, you are so smart it is scary.
Haha thanks, you're too kind.
genius man, incredible!
That is such a cool invention! I think if this was made out of metal with many more circle segments it could actually viable for a car
Thank you!
if you put another of those auto cvts in series it should both multiply low end torque and high end speed and you could probably do this until the friction diminishing returns over take multiplier which should be quite a few if it has as little friction that appears in the video.
I like the version with fixed gears on the wheels, in a bigger version, maybe the gears could spin with the change in wheel size to still be engaged with the belt but allow the gears to roll along the gear whilst the wheels change size if that makes sense
One thing I always wonder when I see a Lego transmission where friction or tension are used to gauge output resistance: is the torque gain from changing ratios greater than the overall power loss from the friction/tension in the resistance measurement? And if so, by how much? Genuinely don’t know. I’m as far from an engineer you can get, but I love Lego technic. Regardless of how efficient this design ultimately is, it’s incredibly clever and undeniably solves the belt slip issue with the classic cone style CVT. Great work on the variable size wheel design, it’s brilliant!
It takes some energy to stretch a rubber band. But once it’s stretched, it doesn’t continuously require energy to keep it that way. It momentarily takes some energy from the input shaft and then the torque gain takes effect. When it shifts the other way, it feeds the energy back into the system (minus what was lost to heat ofc)
Whoa. Expandable dorito.
Braaaaap!
Rotary owners be like…..
Very cool. Good job
Thanks!
i wonder if a design like this is feasible in real cars (assuming it hasn't been tested already).
cool video, very creative.
Honestly i dont know why lego hasnt tried making pieces for a CVT. Maybe because it is such a specific thing to want but still.
So, my love is bicycles, not cars, but we cyclists are stuck with three types of transmission, derailleur, planetary hub and CVT hub. Even people who don't know bicycles probably know the first 2, eg the traditional racing bicycle gears or the sturmey archer 3 speed hub of cruiser bikes. Now, hub gears are both very expensive and very proprietary, while derailleurs can be cheap, are open source, modifiable and easy to work on. It's also easy to understand the torque transfer going on with derailleur gears. The trouble is, you can't belt drive a bicycle with derailleur gears and chains are not the best thing for being exposed to the elements.
Belt drives are becoming very popular on hub geared bikes but for what most people are happy to pay for a geared commuter bike will barely buy you a decent hub gear, like a Rohloff, and that's without rim or spokes, or the the rest of the bike.
Your automatic, twin variable pinion CVT could revolutionise urban bicycles! Seriously! It's compatible with open gearing (or at least "chain guarded" gearing), compatible with timing belt drive, would feel like riding a single speed (and be just as easy to understand) and matches the rider output to the gradient. It would also be a lot cheaper to make than a planetary or CVT hub gearing system, easier maintenance, only the belt would wear out (and bicycle belts are lasting up to 120000km! Longer than the average life of an urban bicycle, let alone the chain that drives it.
Dude! Patent this in the Netherlands or Denmark! You'll be a billionaire! Seriously! This is brilliant for so many exposed drive applications! But really, make it open source, make the world a better place! Then develop this up from lego to timing belt drives, both automotive and for cycles.
As soon as I saw the new system I immediately remembered Veer Shift Drive. It's very similar)
You are a genius.
This is so cool!
Brilliant video
You should try building a lego car that uses this CVT to really show off the capabilities of this machine
I wonder what kind of performance you could get with a pulley based CVT instead. Both the input and the output side have 2 pulleys that get closer or farther apart to change the diameter of the belt. If you do try it out, maybe see if there's some type of V-Belt you can use, it should provide the most friction. A round (instead of a flat) band may also work for a pulley based CVT specifically
This is so cool.
This is better than what's in my hyundai
Haha!
Has this been built for real life applications already? Seems like a very practical design for real cars if done right.
Its seems really cool for cars but i think it could be to big to fit in a smaller car. How big is it and could you upload instructions how to build it?
about why the wider rubberband didnt improve as much as hoped: the formula to describe friction does not include the area of contact but instead the friction force f is calculated by multiplying the normal force N with the friction coefficiant μ between the two surfaces: f = N * μ this does not include the area of contact
Yeah, that's a good point. I know that formula well but I kind of just intuitively assumed that a wider belt would mean more friction :)
The formula is true, but to get μ you don't just take the materials into account, the surface of contact is baked into it.
@@nicodiaz2662 oh ok thx
Brillant 🤩
When a kid with Legos built a CVT that doesn't slip? Damn, someone at Nissan should hire this kid to stop their CVT's from wearing out after 50,000 miles.
Very good design. Now, can you make an e-CVT (Toyota Prius transmission)?
wonderful, i love it
I wonder...could you theoretically have four CVTs for four wheels? Basically just a direct drive from the engine to each transmission? I know it wouldnt be too practical in a car, because of the amount of things you have to lubricate and maintain, but could it be possible?
the shape of the wheels at 4:14 is uh, interesting..
It’s the flag of the Isle of Man 🇮🇲, or at least the same shape
That would be four arms, not three
love it, but very scared the adjustable wheels would hit midway though
Does the triangular design causes an oscillation in the speed of the output?
So uh... toroidal lego gears to reduce drag on the chain due to chaning geometry?
Wouldn't this work like... very well on a bicycle? Even more so with a couple of these in series?
Pleaseee make a video where you build a little test car using this transmission
hallelujah, the rial CVT works EXACTLY like that ( 10:25 )
I am amazed
Better then a Nissan CVT
Man's literally recreated the Subaru ECVT in Lego lol
Make the battle:
Car without transmission
VS
Car with manual transmission
VS
Car with this CVT
isn't CVT what's used in snowmobiles?
Didn’t u technically also integrated a slipper clutch into Uhr design?
Nice build, I really like the mechanism with the differential.
However, It is not a true cvt, you are limited to grabbing onto the individual linkages of the chain, and you cant grab onto half a linkage. Still very impressive work though!
Thanks for the comment. However, the gear ratio depends on the diameters of the wheels, not on the chain links that they connect to - the chain is able to become more or less slack to accommodate the wheels changing size slightly before the chain linkages move over. Since the wheels' diameters can vary continuously, I would say that it functions as a CVT.
@@BananaGearStudios well, in practice it functions like a CVT, but it is not truly continues. It can only grap an integer amount of chain links per rotation, and so can the other side. When you increase the size of the wheel, at some point it goes from let's say 20 to 21 chain links per "teeth" of the wheel. it can't grab 20,37582 chain links. This results in a fixed number of ratios. Although there are a lot of ratios, there are not infinite like in a true CVT.
A true CVT which is not based on friction is yet to be invented. Non the less, the contraption you made comes close, so keep thinkering, maybe you are the person to find a solution some day! (And probably get rich xD)
next: use CVT & differential. make the output spin 0rpm in gear & the engine is runnning
nissan should take notes!
Wait i need stl files for somrthing like this id 100% LOVE to 3d print something like this.
What about putting the rubber band in a figure 8.
excellent
I really nedd to get myself some Lego technic.
6:40 why didn't the gear wheels move apart spinning freely along the chain?
make it 4 or more (preferably 8) gear wheels per size-changing wheel and make only one of the gear wheels fixed so that the chain stays unfixed to rotating gears for the most amount of time (when the one fixed wheel is not in contact with the chain
I wanna see this made out of actual parts and tested in a combustion engine