Cycling Gear Hubs and Their Efficiency

The number of electric bikes that I see on my cycling trips is growing steadily over the last few years so I started to read up on the used technologies and attributes of the different systems.

Using only derailleur gears on all of my bikes, I was surprised to see the variety and quality of hub gear boxes. For me this was something of the distant past linked to a specific two gear fold bike that my family owned back in the 70s. That one actually used backpedalling to switch between the two available gears. Spooky to even think about it nowadays.

Especially interesting is the CVP (continuously variable planetary) gear hub by Fallbrook Technologies Inc.

The principle underlying this gear box with an infinite number of translations is fascinating in itself - even though it was already described half than a millennium before our time by Leonardo da Vinci. The name of the product is a reminder of this fact.

The previous picture shows a nice overview of the basics and this YouTube video shows things in movement.

One of the more interesting questions however is not easy to find an answer to, namely how efficient such a system is in transmitting the forces. Or to put the question differently, how much of my sweat will be shed only to heat up the gear box rather than propelling me forwards. Although there are some comments that the efficiency is not "good enough for sportive drivers" quantitative measure are very difficult to find.

After some searching I finally found a very good page measuring the efficiencies of different hub gear box systems on a self made test bed. Although German only, the measurements are very interesting and I encourage you to read the whole article on the setup. Plunging ahead, here are the combined results for several hub gear systems.

Absolutely amazing for me is the efficiency of the Speedhub 500 system by Rohloff. I did not expect such a good result. It made me readjust my mental picture of gear hubs versus derailleur gears.

On the other hand, the efficiency of the NuVinci system substantiates the qualitative comments found previously. So even if it is a fascinating system I rather not want to use the hub on a bike that is using only my own power to drive. On an electric bike, the system may well be more attractive because of its continuous nature. In the end I think one would need to test drive such a bike to find out if it is so much more fun.