Electronic tech: Dynamo headlamp circuit

James writes:

On 03/01/14 12:28, Joe Riel wrote:

I've got an operational model that simulates a hub dynamo. It's written
in Modelica and simulated with MapleSim (a product I help develop). I'm
happy with the basic model---it uses a magnet coupled to an
electromagnetic converter via permeances that vary with with the
rotation of a shaft. Electrical, mechanical, and magnetic quantities
can all be probed. The biggest unknown was the magnetic coupling---how
much the permeance from magnet to claw pole varied with rotation. I
used a simple estimate and am surprised that the results are completely
reasonable, though undoubtedly not terribly accurate. The design
permits simulating the varying inductance, depending on how one models
P(phi) (permeance vs rotation angle).


That's nice. What will you use it for?

Mostly it was a learning exercise, though I may use it to
"optimize" a light controller.

--
Joe Riel

Joe Riel wrote:
James writes:

On 03/01/14 12:28, Joe Riel wrote:

I've got an operational model that simulates a hub dynamo. It's written
in Modelica and simulated with MapleSim (a product I help develop). I'm
happy with the basic model---it uses a magnet coupled to an
electromagnetic converter via permeances that vary with with the
rotation of a shaft. Electrical, mechanical, and magnetic quantities
can all be probed. The biggest unknown was the magnetic coupling---how
much the permeance from magnet to claw pole varied with rotation. I
used a simple estimate and am surprised that the results are completely
reasonable, though undoubtedly not terribly accurate. The design
permits simulating the varying inductance, depending on how one models
P(phi) (permeance vs rotation angle).


That's nice. What will you use it for?

Mostly it was a learning exercise, though I may use it to
"optimize" a light controller.


Interesting. I was considering doing the same thing, using the measured
frequency from the hub to select the PWM ratio of a buck/boost converter,
but never got past playing with it in my head. I think synchronous
rectification might be a must in this application.