Slightly old news as this was announced at the Eurobike show earlier this Autumn, but it's fascinating nonetheless. There is a good animation on the site that explains the technology in more detail:

 

Fallbrook Technologies

I just watched the video from the site, and I dont get it. why would changing the angle of a stationary ball alter the gear ratio?

 

Are the balls spinning, if so then they must be under tremendous pressure in their 'fluid' when stretched to their extremes.

 

I think I need to re-learn mechanics theory.

 

John

Cool! So these things are going to appear in Ellsworth bikes first, and will subsequently feature in Currie electric hybrids as well as Batavus bikes. I look forward to some reviews.

 

Q

It's a pity that they weigh as much as the average hub-motor, at the moment (over 9lbs). So, if you have one as well as a hub motor...... Also, there's no 'direct-drive'.

 

Miles

I've always wanted a CVT for hub motors, it's the obvious solution but difficult to put into practice. This one has been around for quite a while so has clearly been difficult to get good enough. Weight is a bit of a problem as is space, but the worst thing is likely to be the fluid friction and compression back pressure.

 

It will work, but may well give up in efficiency loss everything it gains in it's gear variability. It will certainly lose quite a lot of the motor's power.

 

Note to John

 

Yes, the balls are spinning and though it's not very clear in the video, the balls are contacting a wrap around part of the input and output discs. Therefore the tilt simultaneously presents a smaller diameter part of the rotating ball to one side at the same time as a larger one to the other. This is the equivalent of two gear wheels of course, a small and a large one meshing. It's a two way system of course, able to gear up and down by equal amounts.

Yes, they're very coy about efficiency figures....If they could get more than 85% that would be quite an achievement. Hence my comment about direct-drive.

 

I think this route is a more achievable solution for hub-motors:

Change-over Gear

 

Or, tapped windings; parallel-series, dual voltage switching, etc. (though they are not equivalent).

 

Miles

Edited December 1, 200619 yr by Miles

Agreed Miles, changover gears of any kind would be a better solution and far more efficient. The Heinzmann hub with it's single cog drive to the orbital gear ring is easily adapted in fact. By having a side plate with a housing bulging out ahead of the fork, there'd be room to have two different size cogs transferring the drive to the orbit, with a splined dog slider operating between them. That could be electrically operated automatically by the road speed/load situation.

 

I'm not so keen on the efficiency compromises inherent in many of the electrical solutions, though they are neat and simple. High consumption is usually the cost, though they can be very effective in power terms.

I'm not so keen on the efficiency compromises inherent in many of the electrical solutions, though they are neat and simple. High consumption is usually the cost, though they can be very effective in power terms.

 

Sure, but given that there aren't any mechanical solutions on the horizon (external motors aside), do the electrical ones have an advantage over the present situation - having to choose between a motor for speed or a motor for hills?

Agreed, often better to suffer a temporary high drain on the battery than put up with the present inadequacies of hub drives, particularly when there are only occasional hills to deal with.

 

For people like me in very hilly areas, where I spend my life either going up or down, high drain solutions would really knock the range though, so often wouldn't be practical. The Torq already only gives me 15 miles range or less, some other motors only 10 miles, so there's not much latitude for loss if a bike is to remain serious transport.

 

P.S. Li-ion batteries would often need attention though. The torq's Li-ions already suffer serious chemical exhaustion when used for sustained hill climbing, so would have to be much larger to cope with short term higher drains.

Edited December 1, 200619 yr by flecc

Flecc

 

Im interested, if pushed, and you were in the situation where you could only have 1 bike, that you had to use to cummute say 10 miles each way during rush-hour on weekdays, from your location, which bike or kit would you choose of the ones that are available to buy new today, and why?

 

John

Of the present bikes in my hilly area John, none.

 

Commuting nearly 5000 miles a year by bike in heavy traffic is a serious undertaking. Only the illegal ones would cope a bit with the hills, though not very well even then, and they bring range problems.

 

The ideal one I would choose is one I already own but which is no longer available, a Giant Twist Lite, but with the SRAM 5 speed hub which I added.

 

The reasons are, it's light, can be cycled easily without power better than any other electric bike, feels like any good ordinary bike, and becomes exactly like a good one when the battery is out. It has refined controllability, legendary reliability and build quality and all I've done in nearly four years and 6000 miles is fit puncture proof tyres and greased the chain. I still use the original two NiMH batteries. There's also the effective mudguards and very strong full size carrier. It drives through the cycle transmission, so adding the SRAM improved the low gear from 43" down to 37" making it suitable for even 1 in 6 hills. At the same time it upped the top gear to enable the motor to drive to beyond 17 mph, with pedalling to 20 mph with ease on the flat and up to 27 with pedalling on slight downslopes. High speed handling not as good as some, but I did 38.4 mph downhill once without danger.

 

If I had to have only one bike now regardless of commuting or not, the new Torq and Quando would get the chop, I'd keep the four year old Twist.

 

For commuting, the inadequate dynamo lights would have to be changed for something more effective as usual with all production bikes. Giant followed me a year later by putting the SRAM P5 on a top of the range variant, but spoilt it a little by adding the weight of extras like suspension forks, hub brakes and dynamo, thicker frame etc, making it less like an ordinary bike and less easy to ride as one. At a pinch, that would be my second best.

 

If I commuted in a fairly flat area, I'd choose a Torq but modify it with my twistgrip mod which is an essential for maximum control in heavy traffic, fit an SRAM P5 gear hub, Magura hydraulic brakes and better lights. The 250% range SRAM P5 does away with all the unnecessary intermediate gears the 200% derailleur has and extends the gear range down to a better low gear. In practice the P5 operates on just the upper three gears nearly all the time, the lower two only necessary on much steeper hills, so riding with it is much more pleasant without constant unnecessary gears to go through, and it also has the maintenance free reliability of a hub gear. The Maguras would be a huge improvement for braking power and pleasant operation. It would end up an extremely expensive bike though, so only for one who could afford the chauffeur following with one's briefcase in the Rolls.

 

Probably not the answer you need John, but that's how things are at present.

Edited December 1, 200619 yr by flecc

I just checked the Ellsworth site, and they are already advertising their new ride bike (the first bike to use the NuVinci™ drive). It looks very pretty. The problem is, the basic version is $3000 dollars ($4000 for the signature series).

 

The Ride

 

That doesn't bode particularly well for the price of electrics based on this drive.

 

Q

Thanks for the link Quicken, but what a horrible looking thing, different for no other reason than being different. The exotic materials would have been better used in a more conventional design which time has proved superior to his sort of gimmickry.

 

I think much of that price is cashing in on difference, and forthcoming bikes with this drive will probably be much cheaper.

Here is a link to a review of the Nuvinci hub on a non-electric bike.

 

phil's velo blog | ATC NuVinci Review - Seamless Shifting works perfectly well!

"The efficiency of about 97% feels about accurate." Is that a typo... ???

Patting head and independant tummy circles

 

Thanks for the link Russ. I agree with fleccs coments about that other bike exploiting its differences just for marketing purposes. My hunch is that this CVT will remain a very expensive item of kit for some time to come.

 

I can understand how the dual input epicyclic CVT used on the Swizzbee works because I had a 'play' with an epicyclic geartrain training aid during a course I attended years ago (and through later theory) - hence the subject name. Even after reading this whole thread and viewing the video on the manufacturers website I dont see how this works though (You're not alone JohnInStockie). It just looks like a giant ballbearing where the cage attempts to rotate or tilt the balls to me and I cant see how the input and output shafts can differ in speed or even transfer power. I think they need to include the path of drive on their video so we can see exactly where the drive is transferred.

As the angle of the ball axles is altered, the contact points with the drive/driven disks, and their relationship to the axis of the balls, changes. It's the ratio of the radial distances of the two sets of contact points from the axis of the balls that gives the differential, I think.

Edited January 7, 200719 yr by Miles

Yes, that's how it works ok, this is what I wrote on it earlier in the thread:-

 

"- - - the balls are contacting a wrap around part of the input and output discs. Therefore the tilt simultaneously presents a smaller diameter part of the rotating ball to one side at the same time as a larger one to the other. This is the equivalent of two gear wheels of course, a small and a large one meshing. It's a two way system of course, able to gear up and down by equal amounts."

 

That last sentence refers to the tilt changing the effective contact diameters in two directions according to whether tilt is to right or left.

 

That ball point-contact drive is what I don't like, since for the drive not to slip the contact pressures will have to be extremely high. That indicates the need for very rigid side members and the weight inherent in that, and even then I'd question how reliable this system will be over time.

As the angle of the ball axles is altered, the contact points with the drive/driven disks in relation to the axis of the balls changes. It's the ratio of the radial distances of the two sets of contact points from the axis of the ball that gives the differential, I think.

I understand the theory of that alright, I just cant see where/how that happens. It looks like the axles for each ball fit into slots in both the input and output disks, if this is the case then there can't be a difference in speed between the two. I watched the video a further 3 times but still cant see where the drive is transmitted I think I'll email the sites webmaster and suggest they modify that video to show the drive path through the assy.

I think the blue arrow path shows it better on this link, as the two diameter points of drive transfer can be seen better than in the video, which I also found poor.

Thanks for that link, flecc - much more useful than the video.....

Edited January 7, 200719 yr by Miles

I think the blue arrow path shows it better on this link, as the two diameter points of drive transfer can be seen better than in the video, which I also found poor.

Thanks flecc, that does make it clearer. My email to the company has been sent. Certainly it has the capacity to be very efficient due to its extremely small, inflexible contact area and consequently low friction. Miles, just as a train wheel on a rail wastes much less energy than a truck tyre on a bumpy road, thats the key to its efficiency.

 

My concern would be over its robustness given the extremely precise tolerances that would be needed in manufacturing and the high stresses it would need to be preloaded with. Just as trains rely on their weight for traction, those disks would have to be squashing the balls considerably if they're not to slip. The bit about the solidification of the fluid inside was also unclear. If its a lubricant then it will encourage the balls to slip on the disk, if its not, the rotating balls will wear through their axles/bearings. I have the general gist at least but I still prefer the idea of the Swizzbee arrangement (other than the fact it contains derailleur gears).

I agree on the pressures involved Flying Kiwi. While I also agree about the rail analogy, I'm far from convinced about the efficiency with that fluid pressure system. The fluid compression forces involved on all those contact points of the balls would add up to a considerable total and the pressures would generate heat, itself an indication of lost efficiency. On a bicycle that would be unwelcome, and on the larger vehicles they project it for, the pressures and heat generation would be very high. All in all, I don't think we'll be seeing that much of it in practical applications.

I'm far from convinced about the efficiency with that fluid pressure system. The fluid compression forces involved on all those contact points of the balls would add up to a considerable total and the pressures would generate heat, itself an indication of lost efficiency. On a bicycle that would be unwelcome, and on the larger vehicles they project it for, the pressures and heat generation would be very high. All in all, I don't think we'll be seeing that much of it in practical applications.

I just think of it operating without the fluid having any significant effect - after all what happens when there's fluid on a rail line (such as leaf juice)... It would depend on the viscosity and other properties of the fluid as to how much energy is taken up, just as its easier to use an egg beater in water than in treacle. I see an increase in efficiency compared to a standard bevel or helical gear train (particularly one with plastic or nylon gears). The contact area looks to be smaller than where gears mesh and I suspect it's more rolling contact and less sliding. It would certainly be alot more efficient than belt CVTs (which have been used in fairly recent low power cars).

Miles, just as a train wheel on a rail wastes much less energy than a truck tyre on a bumpy road, thats the key to its efficiency.

 

Maybe, but 97%.... I'm sorry, that figure doesn't seem credible, even at 1:1.

Edited January 8, 200719 yr by Miles