Lighting-up from 3.9W to 5.4W using a common 6V-3W dynohub

[ConRAD]

Based on some tests performed on a common 6V-3W dynohub (Shimano DH-3D32) span at an equivalent of 20 km/h on a 28” wheel and connected to a variable resistive load, I plotted Power vs Load Resistance getting a maximum of power at about 30 Ohm load.
At this point I tried several filament lamp configurations aimed to get closer as much as possible to the above indicated optimal resistance.
In the WIRING DIAGRAM below all labelled powers and relevant resistances are the nominal ones whereas in the next table all reported numbers are actually measured values.
CONCLUSION: using a 3-ways single pole selector (S1) it seems to be possible to have the following lighting modes:
S-0 - OFF
S-1 - Common Head and Rear Light combination mode (normal mode)
S-2 - Reinforced double front light configuration mode (flood mode)
REMARK: as said above all tests have been performed with very common tungsten filament lights, it’d be interesting to repeat the same test either with alogen filament lamps and led lights.


edit above : changed the color of the lamps

[rjb]

The pilom web site has some more graphs and info about using multiple leds.


http://pilom.com/BicycleElectronics/DynamoCircuits.htm

[Brucey]

nice... (except for a minor point; the back light goes out when the dual headlights are used.... )

The main reason most front lights are the way they are is because of the German dynamo regulations; there is no real reason why the generators that meet that requirement cannot produce more power under altered circumstances and indeed most of them will do that.

BTW I have considered a circuit whereby at low speeds a single headlight is powered, but at a certain speed the system automatically switches over to a dual main beam setup. In theory it could be done quite neatly and efficiently using a few mosfets.

cheers

[ConRAD]

... the pilom web site has some more graphs and info about using multiple leds ...

I strongly suspect that those very generous power numbers, though measured with instruments, ARE NOT [W] BUT [VA], in other words they do include the reactive component of the power going to the capacitor[s].
Conversely, in my table above since I've been using filaments lamps only the whole power is to be regarded as ACTIVE POWER

[Greystoke]

I've done this on both my touring bikes with normal bottle dynamos.
low speed one 3w halogen lamp, higher speed two 3w halogen lamps in series. I worked out the capacitor required to phase shift the power so the second lamp comes on at lower speed.