Half a croissant, on a plate, with a sign in front of it saying '50c'
h a l f b a k e r y
Not so much a thought experiment as a single neuron misfire.

idea: add, search, annotate, link, view, overview, recent, by name, random

meta: news, help, about, links, report a problem

account: browse anonymously, or get an account and write.



Continuously variable bike transmission

No more switching gears or kicking chains
  (+5, -3)
(+5, -3)
  [vote for,

So the basic principle behind continuously variable transmissions are nothing new, and have been used successfully in snowmobiles and various cars over the years. So why not simplify it and apply it to a bike, which doesn't have nearly the torque requirements?

The basic variator mechanism consists of two cones (which act as the gears) connected by a belt. In order to change the input:output ratio you just need to slide the belt laterally along the cones.

In order to make sure that the change in gear ratio corresponds with the change in the bike's speed, have a spring-tensioned centrifugal-governor style system so that as the rear wheel increases its rpm, flyout weights linked to the two cones automatically adjust their position.

Ideally, you'd be able to pedal at a consistant rate and the variable gearing would automatically do all the work for you.

Turbofrog, Jun 11 2006

Continuously variable transmission technology http://auto.howstuffworks.com/cvt.htm
[Turbofrog, Jun 11 2006]

req.bicycles FAQ 8i.3 http://www.faqs.org...rt4/section-30.html
Rather bluntly denies its usefulness and possibility. I don't understand the details; but maybe some of the discussions in rec.bicycles go into more depth? [jutta, Jun 11 2006]

Strida folding bike http://www.strida.com/
Uses a single-speed toothed belt. It's not a friction belt, as this would need to be, but it's not a chain, either. [Turbofrog, Jun 11 2006]

Variator animated gif http://www.obvio.in...ator%20animated.gif
Mechanism in isolation, without howstuffworks explanation [Turbofrog, Jun 12 2006]

Centrifugal governor style linkage http://www.math.buf...d/306/s00/cgrot.gif
Flyout weights linked to drive speed used to slide the variator cones [Turbofrog, Jun 12 2006]

Baked! the NuVinci CVT http://www.fallbroo....com/03_Bicycle.asp
CVT planetary drive used in the Ellsworth Ride bicycle (US$3000) [jcomeau_ictx, Jun 30 2007]

Description of a bike found in trash http://oldroads.com...2_311_43_59_PM.html
As mentioned in an annotation. [Vernon, Aug 21 2007]

Another description of the "Yankee" bike http://www.freethou...thread.php?p=381538
Again search for "yankee". It occurs to me that by now the patent on that transmission must have expired. I still want one! [Vernon, Aug 21 2007]


       I know these already exist, I've seen them on an infomercial.   

       But it automatically changes the speeds, so you don't have to.
BJS, Jun 11 2006

       The only advertisement I've seen for something like this was the "Autobike" and that is not a CVT. It automatically shifts between discrete gears just as you do with a standard bike.   

       Regarding the bicycles-faq: The working CVT designs I have seen do rely on friction, but I fail to see why that is such a bad thing. There is friction in the standard bike chain drives. Also the objection to weight makes little sense since the CVT designs I've seen were not much heavier than a chain and sprocket. He does have a point about the human motor being very adaptable and CVT not being all that necessary in bicycling.   

       The main objection I can see is the CVT is expensive; requiring extreme forces and strong materials to work effectively.
James Newton, Jun 11 2006

       The friction in a good chain drive is so low that it is in excess of 98% efficient. Even a toothed belt drive has higher friction. That's why I've never seen one on a bicycle, I suppose.
I think a cone type CVT just couldn't compete with a standard derailer system.
Ling, Jun 11 2006

       I don't think the poster in the bicycles FAQ makes any especially strong arguments. CVTs actually are being adopted in cars (Audi, Nissan, and others), offering both better fuel economy and improved acceleration to boot.   

       The human motor is highly adaptable, but why not abstract that argument even further and saying we're flexible enough that we'll never need 18 speeds on a bike?   

       I will agree that this CVT will likely be less mechanically efficient than a standard derailleur/sprocket system, but reduced maintenance and improved ease of use are definite benefits. One of my motivations for suggesting this was that my current bike's derailleur is quite flaky, so I'm stuck on one of 3 main cogs, and I can't be bothered to fix it.
Turbofrog, Jun 11 2006

       This is not a new idea, and the posting doesn't give enough infomation as to how to accomplish it. It certainly isn't even halfbaked. [-]   

       Buy a good-quality bicycle, and learn how to shift. I have no doubt that the first clunky, crappy, heavy CVT for a bicycle will be installed on a Wal-mart bike made of toothpaste tubing, and marketed all to hell. Thousands will buy it, and consequently give up on cycling forever.
baconbrain, Jun 11 2006

       Hmm. The concept is not new, and I've acknowledged that, but in this application and with a simplified, purely mechanical design, I think it's sufficiently different to warrant a posting.   

       I'll see if I can describe it a bit better, as I don't think it need be especially complicated or heavy at all.   

       I linked to the the howstuffworks page on CVTs, but this animated gif [link 4] may be a more straightforward example of the variator mechanism.   

       Imagine the front pair of cones connected to the chainwheel, and the rear one connected to the rear hub. The lateral movement of the cones (which forces the change in gear ratios) is produced by a linkage quite like this one [link 5]. As you pedal and the rear wheel increases in speed, the pair of spring-tensioned weights are pulled outward by centrifugal force, and the collar seen sliding along the shaft in the animation [link 5] is connected to one of the cones on each side of the variator, while the other remains fixed, as in the animation [link 4].   

       Hence, the faster you bike, the higher the transmission gears itself, increasing the effort required to pedal and allowing you to go faster.   

       The torque requirements are not especially high, and there is no shock-loading on the parts, and no need for a really positive drive, so I don't see why the components couldn't be made cheaply from injection-moulded plastic and contained in a plastic housing to keep dirt out of the drive mechanism.   

       Make sense now?
Turbofrog, Jun 12 2006

       I think a mechanism that keeps constant torque on the front crank makes more sense than something speed controlled. Climbing a hill at a certain speed should allow for different gearing than flat ground at the same speed.
NoOneYouKnow, Jun 12 2006

       //Make sense now?// Turbofrog, thanks for the explanation. It's fairly well written, but not convincing as far as applying a CVT to a bike. I feel I speak from experience.   

       I've ridden many a mile on multi-speed bikes, taught classes in bicycling, and worked on way too many cheap bikes. (My brother builds recumbents, even (very good ones).)   

       As far as transmissions go, I probably hold the world's record for miles ridden behind a Comet brand CVT on a front-wheel-drive motorcycle. I installed the bugger, tested it, watched it as I rode, maintained it and checked it. And, at the end, preferred our other, non-centrifugal, transmission.   

       The only thing I really liked about the CVT was that when I backed off the throttle it wouldn't release. The engine would backfire, sending jets of flame out the exhaust stacks, which were also right in front of me. If the CVT on a bicycle didn't release, I'd probably get off the bike and kick it to bits.   

       Bike racers want control over their transmissions. As NoOneYouKnow points out, a racer will want to be able to honk up a hill at low RPM and high torque, or jump from a pack. Non-racers won't want all that, true, but they won't want to maintain a CVT.   

       Bicycle design is probably even more highly-researched than is automobile design. Bikes range from cheap Wal-Mart flashy shit, to sturdy black Chinese bikes, to lean racing machines, to Human Powered Vehicle garage-built craziness. I'd say the chances of thinking of something new are very slim. If you want to suggest something new, I say you need to show a lot of your method of implementation. Saying that we could, or should, use something is simply wishing that we could.   

       I point out that CVTs are most widely used on snowmobiles. They are not used on motorcycles, and are only just being implemented on automobiles, but not as centrifugal cone V-belt units. Belt drives are not used on bicycles--the one example linked to is for a specialized bike that uses toothed belt, not a V-belt.   

       It isn't that no one has ever thought of putting a belt on a bike. It's that there are a lot of problems to be solved first. Halfbaked solutions to those problems would be appropriate here; custard-filled cones, ferromagnetic belts, stress detectors in the handlebars, or helium-filled weights. Or, seriously, address the fact that most CVTs vary above and below a 1-to-1 direct ratio, while bicycle gear trains vary around a 1-to-3 step-up ratio.   

       Saying that you want to put a centrifugal cone, V-belt CVT on a bicycle is a wish, not a solution. I'm going to keep saying that you haven't addressed even half of the issues involved. I'd be eager to try it out if you build it. But I'm saying "if", not "when". And thinking it will never be popular.
baconbrain, Jun 12 2006

       Automotive CVT's use 'pusher' belt systems. In order to get enough grip, there needs to be a lot of friction. I doubt that plastic would be a good choice there due to it's low surface toughness and low friction.   

       CVT's are a bit better than automatics due to hydraulic losses and gear optimization, but they're a toss-up with manual transmissions.   

       The inertia of your weight-lever system would be a detriment towards increasing speed.
RayfordSteele, Jun 14 2006

       There was a bike transmission design that came out quite a few years ago, before the Internet. The "Yankee Bicycle" featured something called a "rim band brake", as well as an automatic transmission. I found a link partially describing this bike, of which one which was found in the trash, so you can be sure they were actually produced (search for the word "yankee" at the linked page).   

       Anyway, at the rear wheel of this bike was a single sprocket, and a lever with a spring and a chain-tensioning sprocket.   

       The automatic transmission stuff was entirely part of the front sprocket. This mechanism was featured in an old article in Popular Science, and at the time I thought it was really cool, and wanted one, and so studied it pretty thoroughly, and can, I think, describe it reasonably accurately here.   

       Part of the front sprocket assembly consisted of a circular plate with six radial slots (straight from center toward rim), and an adjacent circular plate with six sprialling slots. I think there may actually have been two of the radial-slots plates, with the sprial-slots plate sandwiched between them, but it could just as easily work the other way, a radial-slots plate sandwiched between two spiral-slots plates. In this description I will assume the first sandwich.   

       When oriented properly, you will see six smallish holes through the three plates, where the spiral slots cross the radial slots. Rotating the spiral-slots plate, while holding the radial-slots plates stationary, causes the holes to move from the center toward the rim of the circlar plates (and vice-versa if the spiral-slots plate is rotated the other way). The overall assembly keeps the two radial-slots plates fixed relative to each other, and a strong torsion spring is attached to the spiral-slots plate. The "default" position for the six holes is near the rim of the circle.   

       Through the six holes are placed six small axles, and on those axles are six small sprockets. If you wanted to draw a picture, just do a simple hexagon with a modest circle at each corner, to represent these sprockets. At least two of these sprockets (at opposite corners of the hexagon) must be ratcheted, so they can only rotate one way. The bicycle drive chain goes around the outer parts of some of the six sprockets, as well as the rear-wheel sprocket and chain-tensioner, of course. Note that the "size" of the "front sprocket" of this bicycle is the result of the chain going around the outside of several of those six small sprockets.   

       Are you ready to roll? When you apply force to the pedals, you will be applying twisting/torsional force to the radial-slots plate. The six axles and sprockets are carried along, and the chain starts to transmit power to the rear wheel (because of the ratchets). If there is little resistance, the rear wheel will start to rotate, and the bike will move.   

       If there is resistance, however, the sprial-slots plate will start to twist, tightening up the torsion spring. This twisting motion will cause the six spocket axles, thanks to the radial-slots plates, to move toward the center of the circular plates, toward the center of the pedal axle. This effectively shrinks the overall diameter of the front sprocket! And slack in the chain, as that happens, is taken up by the rear tensioner.   

       The smaller-diameter sprocket assembly means that your "gear ratio" has changed, you have "increased mechanical advantage" in applying turning force to the rear wheel. And when the resistance diminishes, that caused the front sprockets to move closer together, the torsion spring smoothly (as in "continuously variable:) restores them to the rim of the assembly, once again maximizing the overall gear ratio.
Vernon, Aug 21 2007

       Vernon, it has probably been said before, but it'd be really helpful if you could illustrate your posts/idea.
xaviergisz, Aug 21 2007

       I don't know if this has been said before. Belt drives have much lower efficiencies than chain drives. This is less of an issue for a car with an engine (although the Volvo and Daf cars that had them returned very poor mileages per gallon), but it's a big issue for a bike. Even worse than putting fat tyres on a bike that spends its whole life on decent roads.
Cosh i Pi, Aug 21 2007

       Vernon, how did the 'Yankee' get around the problem when changing gear: the chain will become tight when the circumference is increased. If the outer planetary sprockets are directional, then I can see how it would account for changing down to a smaller diameter, but not the other way. Interesting, none-the-less.
Ling, Aug 21 2007

       [Ling], if the chain tensioner at the rear wheel takes up slack when the front sprocket assembly shinks, it can certainly give back that slack when the assembly expands again. But if you are talking about the shrinking/expanding distance between the front sprockets, I'm sure the answer is in the non-ratcheting of 4 of them. Also remember that the pedals WILL rotate some even if the back wheel doesn't move, allowing the slack here go "get out" of the front assembly, so the rear tensioner can get it. When expanding, it probably doesn't happen too fast for the needed slack to be acquire-able.
Vernon, Aug 21 2007

       though I'm a big fan of CVT i think this would add "drivetrain" weight. this added weight would by far neglect any benefit that would come out of the use of a CVT on a bike.   

       but if you were to have it on your bike (maybe for novelty purposes) i would think it would be much easier to manually control the CVT instead of having it controlled automatically.
F_R_O_G, Aug 21 2007

       Bicycle weight is not critical except for professional race bikes. If you think about it, the weight of the bike is next to nothing anyway compaired to your weight. Lets say it costs even as much as 10 extra pounds for a CVT, that is a 5% increase in total weight for a 180lb man and a 20lb bike.
James Newton, Aug 27 2007

       if you have a 20lb bike than you have a performance bike that you would have to put down some money for. unless your 6 years old....   

       at least on a car any added drivetrain weight is like adding 10 times that amount in car weight. this should translate over. so a CVT adds 5lbs to the drivetrain and 5lbs to the frame of the bike. so we add it up. 5lbs x 10 = 50lbs + 5lbs = 55lbs. so now take 55lbs and put it on your bike and see what happens after 10 miles.
F_R_O_G, Aug 28 2007

       The Yankee bicycle does not have an actual auto transmission. It has 9 positions (diameters) into which the front sprocket expands and contracts. There is a lever by the hand grip when depressed will up-shift or down-shift the gear.   

       Like pie wedges, the spocket is composed of 6 parts that move in and out. Only two of the wedges have teeth. Opposite of each other, they each have only 7 teeth. They engage the chain during the left and right powerstrokes, respectively.   

       To up-shift push the lever and forward pedal to the desired gear, then release. To downshift push the lever and back pedal. One nice thing about the design is that if you come to a stop while in a higher gear, you can down shift to 1st gear without moving an inch.   

       The rim band brake is unique. The rear wheel rim has a channel that goes around its perimeter. Riding loosely in the channel is a coated cable, (I'm not sure of the material). When the hand brake is applied the cable puts resistance on about 65% of the wheel perimeter. There is no front brake.   

       I'm suprised that no other bike makers have picked up that brake system. It offers superior braking ability, and actually grabs better in wet conditions.   

       Popular Science gave Yankee a Product of the Year nod back in the early 90's. Shortly afterward Nordic Track bought it and sold it under their name changing the color from teal to dark blue. It may have been ahead of its time . That bike should have sold alot better.   

       I bought my Yankee in early 1992 after reading about it in PS, and still take it for a ride once in a while.
Bill J, Sep 13 2007

       [Bill J], thanks for the info. Perhaps what I described can therefore actually qualify as an automatic and continuously variable bike transmission.   

Vernon, Sep 14 2007


back: main index

business  computer  culture  fashion  food  halfbakery  home  other  product  public  science  sport  vehicle