> I don't see how it would work as a diff and trans both, based on the bicycle
> CVT.
[quoted text clipped - 7 lines]
> (as the bike obviously uses some sort of cable to the handle grip)
> Interesting though.
That's not how it works, It does not work like a clutch, it is truely
a variable-ratio system. The key is where each of the wheels contact
the ball in relation to the axis of the ball itself. At the extremes,
one contacts close to the axis of the ball, and the other contacts
near the ball's "equator".
It could be made into a combination with a differential, it would just
have to have one output shaft of the Diff go THROUGH the entire, with
the transmission revolving around the through-shaft. Mechanically
there is nothing preventing that from happening. That would end up
being an assembly of about a gazillion different parts, it would get
expensive pretty quickly.
That being said, it would be hard for this to work automatically,
there would have to be some mechanism to do the actual shifting. I
would imagine if one could reliably measure the RPMs of the motor and
have something operate a servo based on that, and have the servo
adjust the system. I do question however, how much torque this would
be able to handle before the rings start slipping on the balls, no
matter what scale it is.
Coyoteboy - 10 Apr 2007 02:59 GMT
> That being said, it would be hard for this to work automatically,
> there would have to be some mechanism to do the actual shifting. I
[quoted text clipped - 3 lines]
> be able to handle before the rings start slipping on the balls, no
> matter what scale it is.
I cant remember who or when but a CVT has been used on an F1 car in testing
(not allowed in racing, it was a proto) and it dealt with them fine. Could
well have been a different design but AFAIK all CVT designs use friction to
transmit torque?
GTD - 10 Apr 2007 20:53 GMT
> > That being said, it would be hard for this to work automatically,
> > there would have to be some mechanism to do the actual shifting. I
[quoted text clipped - 8 lines]
> well have been a different design but AFAIK all CVT designs use friction to
> transmit torque?
Maybe true, but consider the amount of money put in to one of those
cars, it's not surprising. My doubts were in this specific design, as
I know they work elsewhere. Ford currently has one in the All-Wheel-
Drive version of the 500, as does Can-Am in their DS250 (http://
www.brp.com/en-US/Products/CanAm/Showroom/DS250.htm)
The Ford version is a steel-belt design I believe, I've seen a cutaway
type of picture of it, probably in the dealer's brochure for the 500.
Wasn't enough info to see how it works though. I believe (possibly
quite wrongly) that Can-Am's design is a belt-and-pully type, much
like those on snowmachines and variable-speed milling machines / drill
presses, no new technology there.