In article
,
wrote:

On Jan 13, 1:03 pm, Ryan Cousineau wrote:
In article
,

wrote:
On Jan 13, 3:16 am, " wrote:
http://www.google.com/patents?id=Pj5HAAAAEBAJ

What happens here?

As drawn, nothing happens. There's no mechanism to compensate for the
varying chain length, and the chains are either going to bind or
skip. Later proponents of elliptical chain rings have realized that;
a) The sprockets need not also be elliptical, b) There does not need
to be two chains, and c) There needs to be chain slack and some kind
of spring loaded tensioner.

Interestingly enough, c) is not true!

http://sheldonbrown.org/gunnar/index.htm

I'd have to fuss to figure out if the chain length is precisely constant
for any position of the rings (and if Biopace vs. true ellipses
matters), but this is an effective demonstration that the chain length
change is negligible.

--
Ryan Cousineau /
"My scenarios may give the impression I could be an excellent crook.
Not true - I am a talented lawyer." - Sandy in rec.bicycles.racing

It's hard to tell without any dimensions, but those chainrings looked
much more eccentric than Biopace. There's no way with a gearing other
than 1:1 for the chain length to stay constant for two elliptical
rings. I believe that the 2:1 arrangement can be lined up to avoid
the extremes though.

I'm referring not to the design in the patent (which is clearly insane)
but to Sheldon's bike as a demonstration that you don't need a spring
loaded tensioner (or, I believe, notable amounts of chain slack) to run
a Biopace ring and a round cog.

--
Ryan Cousineau http://www.wiredcola.com/
"My scenarios may give the impression I could be an excellent crook.
Not true - I am a talented lawyer." - Sandy in rec.bicycles.racing