[geeks] Electric bike?

[Francois Dion]

On 7/18/06, Sheldon T. Hall <shel at tandem.artell.net> wrote:
> Spake Francois Dion ...
> >
> > Anybody built their own electric bike or has one?
> >
> [snip]
> >
> > I picked up a Kollmorgen flat servomotor which would fit perfectly in
> > the "V" above the bottom bracket but I haven't found a good way for
> > the transmission.
>
> Can you do it with a single-speed set-up?  I know SFA about electric bikes,
> but try me on regular-ole road bikes up to the mid eighties ...
>
> If a single speed setup would work, how about a centrifugal clutch on the
> motor driving a belt on the _left_ side of the bike?  Don't modern hubs have
> some sort of mounting for disk brakes on the left side of the rear hub?
> Even if not, it would be easy enough to use an old double-sided track hub to
> fit the "other" end of the drive belt.  Or even have a machinist make you a
> hub, using an old Phil or Hi-E unit for a pattern.
>
> Alternatively, just hire a local bike racer and have him push you.

Hehehe... I had a good laugh there! I used to ride with some friends
in college and when one would get tired, we'd push the person. It
worked.

Actually, I'm not doing this for speed. I can do > 30mph on my road
bike (80s Peugeot with huret-mavic-stronglight setup) and about 25-26
mph on flat with the mountain bike with city slickers (not sustained
and definitely not going uphill). Commercial electric bikes such as
the Giant Lafree and Wavecrest Tidalforce are limited to the ebike
speed limit (25mph in most state). Above that they have to be
registered as mopeds (up to 35mph i believe). They are also >60lbs and
cost > $2000.

For optimal range, my mountain bike with friction drive (think
Velosolex) is setup for roughly 20 mph. I've set it up to do
regenerative braking too (hence, it doesn't go much faster than that
even with a good assist). I made a variation to the circuit, for peak
output and went on a track and pulled the bike up to 36.5 mph. I had
done a few fake "runs" on my mag trainer first, to test different
types of roller. The ceramic like stone roller shrapnelled at that
speed... So I'm sticking with knurled steel, altough an aluminium
roller, knurled similarly would be even better.

Now back to your question: does a single speed setup work?

Yes. In fact practically every commercial unit is setup single speed.
Some hub motors are 2 speed with internal gears. Then you have the
Cyclone, Optibike and Giant Lafree that uses the chain and benefit
from multi speed. Great to go up hills obviously.

The hub mounted pulley on the left side is what a "Currie" drive is
typically. There are many variations of that, but again single speed.

http://www.peltzer.net/ebike/
http://www.americanchaindrive.com/

More efficient than friction drive, but they suck going uphill unless
you have a really big motor and large heavy lead acid (sealed SLA AGM
type) batteries (NiMh are not good at outputting large current). I
have a Currie electrodrive motor (a rebranded Kollmorgen) with
integral clutch (really cheap from that auction site) and have played
with this a bit.

Based on my different tests, I've come to be really interested in the
power thru the chain type of drive.

One guy has an interesting solution with an intermediary shaft, but it
requires some pretty intrusive modifications :)

http://www.jstraubel.com/ebike/ebike.htm

I've just been trying to figure a mechanical way to drive through the
regular chain. Hmmm... maybe an electrical interface would work:
replace the bottom bracket by a generator and power the motor that is
driving the chain :)

I'm actually half serious, this could work...

Due to the complexity of thru the chain and chain drive in general,
the kit industry is pushing hub motors like crazy, but that is just
not the most optimal way.

BTW, What kind of road bike you got?

Francois