1. At what point does the drag of the generator become noticeable? My SON28 hub has almost no drag, and I assume it is putting out 2.5W (5V/500mA). It it were putting out 5W (5V/1A), I assume that there would be double the drag and I would notice it. (I assume that if it could output 5W with not much drag, it would have been designed to do that.) How many extra watts can I generate without noticing it?
2. What are the efficiencies of the different generator designs? How much of my extra energy (to overcome the drag) actually makes it out of the generator in the form of watts? How does the design of the generator (hub vs bottle, or high quality manufacter vs low quality) affect that efficiency?
- Dave
One of the reasons that the extra effort is often quoted as "a foot of rise per thousand" or per mile, or a few feet mile, as George does above is that the effort added by the modern topclass hub dynamos falls into the noise level of ambient conditions (road surface, wind, rider variations) where even the most sensitive and experienced cyclist has little chance of detecting it. This is not surprising when you consider the numbers. If a cyclist puts out a constant 200W, the highest dynamo output you're talking about, 5W, comes to only 2.5% of the total, and the added effort (or loss in forward motion due to cutting the dynamo in) is thus a fraction of a fraction, below the human detection level. But no dynamo that you would consider is that inefficient. Though nobody is immune to the laws of physics, especially conservation of energy, a well-made hub dynamo has every appearance of giving you something valuable for such low cost that it is almost free, which probably accounts for the perennial popularity of the thing. See:
What you're really comparing the hub dynamo to isn't zero energy consumption, because the best non-dynamo hub will have at least some friction, but the marginal (i.e. additional) friction of fitting a dynamo hub. You're already committed to the dynamo hub's extra friction, if any, and weight (usually), because you must have lamps, but both are in negligible in the overall picture of propelling a hundred kilos of touring cyclist, bike and luggage, far, far, outweighted by convenience and safety considerations. So, what you must consider next is the frictional cost of cutting in the dynamo. This isn't as stupid as it sounds: there was a Shimano dynamo only about 20 years ago (and before that it was a Sanyo dynamo) that delivered more drag when switched off than when producing electricity! But the frictional cost of cutting in the dynamo turns out to be lost in the white noise too, too small to notice.
Conclusion: fitting a SON, Shimano or PV dynamo is a no-brainer.
None of this applies to the sidewall dynamo in the first post; if the generator, of whatever kind, doesn't work, it is dead weight. Really, what I say above applies only to modern hub dynamos and to low-friction, efficient sidewall generators that are so well made that they cost nearly as much, and in some cases more, than a pretty good hub dynamo.
