Fallbrook Technologies' CVT

Matt O'Toole said:

I came across this tonight, a new CVT for bicycles, and/or whatever

else:

Quoted message said:

http://www.fallbrooktechnologies.com/default.asp

Perhaps not so strangely, I don't see any efficiency figures

anywhere.

Quoted message said:

Matt O.

Hmmm... if I were still living in San Diego, I'd check them out. I
wonder if it's related to the Nupace hub: http://www.nupace.com/

I think we should take up a collection and send one of each to Chalo
and let him do durability testing on them.

Jeff

Matt O'Toole said:

I came across this tonight, a new CVT for bicycles, and/or whatever
else:

http://www.fallbrooktechnologies.com/default.asp

Quoted message said:

Perhaps not so strangely, I don't see any efficiency figures anywhere.

You'll also notice that this is a smooth steel-on-steel friction drive
that requires a huge preload to not slip under even moderate input
torque. That preload will also take its toll on efficiency. This is
not a new mechanism. It is one of those commonly used in machine tools
where the V-belt Variomatic is not desired.

[email hidden]

In article <[email hidden]>,

JeffWills said:

Hmmm... if I were still living in San Diego, I'd check them out. I
wonder if it's related to the Nupace hub: http://www.nupace.com/

They're related only in that both have fundamental follies.
Fallbrook is a friction drive and is stuck between losses and
slippage.

The Nupace is a ratchet-type "CVT". Carl Fogel dug up the
patent filing last year:

http://tinyurl.com/67o38

This is pretty much exactly something I'd sketched up 10
years before the patent, after seeing pictures of Bridgestone's
so-called CVT hub around '86. It expands on the range of ratios
availble from Bridgestone's design by essentially ganging several of
them in series, but otherwise inherits all the same shortcomings.

The input/output transfer is not smooth, though this isn't
necessarily a problem for pedal drive (comparable to having a somewhat
lumpy chainring shape).
Also, as with all ratchet-type mechanisms, it does not
actually provide a continuum of ratios. Each pawl of the drive must
slip past an integer number of teeth per rotation, and the discrete
steps in number of teeth passed means discrete steps in overall gear
ratio. I don't know how many teeth the ring gear uses, but if we
gratuitously call it 80 teeth, for the 1:1.25 range of each stage,
there's actually only 20 distinct gear ratios.
Worse, the eccentric cam design doesn't isolate cam positioning
from drive torque.

-Luns