Originally Posted by
PlanoFuji
Apparently neither your nor the source you cited understand electronics nor basic physics. The dynamo (which provides ALL of the power) can only output 3W. You can increase voltage at the expense of current, or current at the expense of voltage; however, the total power remains the same (ignoring conversion losses).
You are powering your LEDS with 3W of power-period.
You are not likely (as in certainly not) getting the LUMENS you claim. Even the Supernova only claims a MAX brightness of 800 lumen. It doesn't actually provide the measurement when it is powered from a standard 3W hub...
Further, your setup exhibits all of the problems common with battery powered lights without the advantage of being able to draw more power than a dyno can provide. Specifically, your optics waste much of the light generated on the sky and other areas you don't need light. You would be far better off with even the cheapest dyno lights that at least have properly designed optics to make use of the limited power available in a dyno set-up.
Also, rather than misunderstanding physics and arguing about it, just pick up a cheap light meter and show us the measurements.
Plano, you are assuming that the rating of the dyno is a power limiting value. I think it's instead a guaranteed level of performance at a specific road speed - the statement of compliance with the specifications of the German road law (I can't give the citation). At a higher road speed or at a lower electrical load impedance it may put out more than that. We don't know it's maximum power, short-term or long term because of several other uncertainties: we don't know if the standard (i.e. LED light) load is a matched load for any given generator, we don't know if the generator physical design is thermally limited (affects long term power output), and we don't know at what combinations of load and speed the magnetics saturate (electrical limit of the generator circuit).
I'd say (and I cannot say I can do this at my house, I could have done it three jobs ago when I was still a power electronics engineer) test the generator by spinning the wheel at fixed speeds and loading it with known resistor loads. Observe the load waveform with a current probe (or a shunt) and an oscilloscope. Take data to indicate at what power level the current is becoming distorted, also what the RMS voltages and currents are. Since it's an AC machine, power factor must be considered in assessing output power. Just for a little more fun and understanding, probe the dyno case temperature with a thermocouple right after each test to observe temperature rise with increasing power level.
While we're at it we should measure the Thevenin voltage and source impedance as well as the Norton current of the dyno.
With all this we'd have a good understanding of what the dyno's capability is.
I suspect you're exactly correct, that Mr. Bubbles' neat setup is not doing what he thinks it is, because he does not understand how the load circuit affects the generator. But now we should find the truth.
EDIT:
I've read it all now, and I see that while my suggestion to characterize the dyno might have some merit, it could well be that Mr B's circuit is in fact delivering the power he thinks it is.
I don't see why the German law means you must put out 3 watts no less and no more, unless the German authorities were trying to eliminate extra-bright lights on the public roads. It that case the spec needs to ahve an upper and lower bound, and I don't recall that it does. I read it more than 10 years ago!
Today I'd suggest just buy a Luxos B or U.