I've been riding with my Sanyo dynamo hubs for a while now, and as far as I can tell, I don't notice the drag at all. I notice the vibration, but that's different. Nominal output is 3W of electrical power, so let's say it's 50% efficient. So it's drawing 6 watts of my kinetic power. While I tool around gently on my mostly flat streets, how much does that add up to? Is there a way to put this in perspective? I can lift the bike up and spin the wheel, and I notice it stops before a regular wheel would, but it's not hard to spin the wheel. This leads me to believe that the cost of running these is vanishingly close to zero.

Thoughts?

I was curious too, so I plugged a couple of numbers into this:
An interactive, model-based calculator of cycling power vs. speed

With the default parameters for rider weight, frontal area, etc., the difference between 150W and 144W is 18.53 mph - 18.24 mph = 0.29 mph. Not surprising you can't tell the difference. At 100W vs. 94W, it's 0.40 mph. Over a 25 mile ride, that amounts to about 2.5 minutes. That would matter in a race, but is seems pretty negligible under most other circumstances.

Your numbers help, especially 0.29 mph. I weigh about 170 lbs. I'm not terribly strong or terribly weak. I don't often ride at 18 mph, though. I think my typical cruising speed is about 14 mph.

I put my numbers in. I said my bike weighs 30 pounds, since it often does, or more. I don't know anything about my frontal area and air drag, so I left the default values in. So I need about 100 watts to keep that speed up. So hmm, still, the dynamo, if it truly sucks 6 watts, is slowing me down by 1/4 mph. That is truly imperceptible.

If I take a century ride with a bunch of friends, how much more tired will I be than they are at the end of the day? I'd imagine lots of other factors will bear on that more than the drag of the dyno hub.

Any thoughts about the accuracy of my 50% efficiency guess?

I don't even know if the electrical output is actually 3 watts. What kind of rating is that? Average? Based on some assumed speed?

6 watts is .008 HP so not much. The 3 watt rating should be max. If you are using a 1 watt LED and your 50 % guess is correct then it would take 2 watts to turn it. There will also be a slight drag even if the light is not on. It's not enough to worry about unless you are racing.

6 Watts is not noticeable. I ride with a power meter, basically always and 100 W is not even a recovery effort.

The info that stuck in my mind for calculating how much energy is lost, was that a good dyno hub was like adding 6 feet of ascent per mile. Although post 20mph the game changes for the worse with most of the dynohubs on the market.

I'm using Sanyo hubs, which have more drag than the others. The difference is even greater without an electrical load. I use this as an excuse to keep my lights on in the daytime. The light might make me safer, and the kinetic cost is negligible. I don't exceed 20 mph often, and when I do, I don't sustain it. I suppose on the very rare occasion I join a bunch of A riders and cruise at 20, I can switch off the light or bring a different bike.

This discussion reinforces my decision to use dynamos. I'm pleased beyond expectation with them.

Does anyone have an estimate of how much old sidewall or bottom bracket dynamos consume? More, of course, but even when I use those, they don't wear my body out. I wouldn't run one in the day without absolutely needing the light. I'd say the noise is more of annoyance than the drag.

6W is the amount of energy needed to raise 1g of water a bit over 1 degree Kelvin in one second.


So 6W is not much of anything, as already said.

I think @mibike has it right with his estimate, since I've seen some charts showing the typical drag of 2-3 watts (hub dynamos). To be honest, to me it does seem like a lot to give up when you're looking at .25 mph difference. It takes some work to raise your threshold power enough to increase your speed that much. In theory, thinking about it. But then I get some perspective when I realize that I haven't aired my tires up in a week, or two weeks currently, and that costs me quite a bit more than that and I don't seem to care in practice. Or buying a faster tire will gain the 6 watts, and that's never happened either. I could probably find a couple of watts in my drive train. Even though I think I care, in practice it's obvious I don't. If you're at all like that then I think it looms larger in our minds than it does in actual practice.

[MENTION=227213]wphamilton[/MENTION], I like the way you put that. In theory, we could say it's significant, but in practice it isn't. And as von Neumann said, in theory, there's no difference between theory and practice, but in practice there is.

I like both of these comments.

Chris Juden, technical editor of the CTC, ran drag tests on bottle, BB and hub dynamos.

"All of the generators were easier to turn than riding up a 1 in 300 slope. Another way of putting that is a rise of 18 feet per mile; and there's quite a cluster of sidewall and hub-driven models around the 1/500 line, or 10 ft per mile. If that's a hill I'm a Dutchman!" - CJ

I wouldn't use a bottle generator for P-B-P, but for commuting IMO they're a competent bit of kit.

One in 300 is tiny. Now I'm surprised I can feel it. I'm pretty sure I'm not imagining it. The noise might cause a placebo effect.

The small difference is largely due to how power needed to overcome air resistance is proportional to speed cubed, and at speed most power goes into overcoming air resistance. Because of that, doubling your power only increases your speed to the cube root of 2 or by about 26% on level ground, and halving your power only slows you by about 20% there.

It's a bit different going up a steep hill, however, as there most energy goes into overcoming gravity rather than air resistance. There, if you lose six watts and drop from 100W to 94W, the difference will be approximately linear, so you'll be about 6% slower -- so it'll be more significant there, but still relatively small.

The 50% efficiency guess is probably fairly close to accurate.

Every now and then I'll realize I have the light on in daylight, and not the full light but the daytime running lights. So I'll turn it off. Does it make a difference? Nah, I can't tell. But one day I was in Prospect Park at night and since it is a closed road without any obstacles I turned off the fully illuminated light, and yes I could tell the difference. Would it make a difference in my riding, probably not, but I could tell.

I'm using a Shimano hub and Cyo premium light, no taillight.

And in a plug for dynamo lights in general, and my Cyo in particular, people have been commenting on how much light the thing throws off onto the road, even the guys with the light cannons.

[MENTION=13059]zacster[/MENTION], you're running a 20" front wheel, right? Might that make a difference? Is the hub designed for small wheels? If not, you're spinning it fast.

I just put a B&M headlight on one of my dynamo-hub-equipped bikes. Man, that thing looks bright, compared to the halogen bulb it replaces.

No, not a 20", a 26". The difference between 26, 27 and 700c is pretty negligible in terms of output.

Agreed.

My view is a little more pessimistic. Say you average 90 watts output in a moderate pace ride. 6 watts extra is an increase in effort of about 7% to maintain the same speed. To a small motor like a human, 6 watts lost is not insignificant. If you have to pedal 7% harder than a friend you are riding with you likely will tire out faster if you are of similar ability. Also if it is equivalent to 10 feet of climb per mile (I think 10 is more realistic than 6 for most dynos) then on a century ride you climb an extra 1000 foot hill which is something in my book.

Bravo.

OK, so maybe I shouldn't use a dynamo hub on a century ride if I'm not feeling confident about my ability compared with my riding companions. But for commuting and errands, it's nothing to worry about.

100% of my riding these days is commutting and I leave my lights on low during the day, LightOn! brand. When I was skinnier and riding faster I'd turn the dyno lights off going up hill relying on helmet lights then switch the dyno on on flats. I can sense the difference but it's not a significant one . If I was doing fast group riding I wouldn't use one but for solo riding and giving cars one less excuse to turn into me from the opposite lane I'm all for it.

That sounds about right to me. Only worry about it when that 5 - 7% might matter like when you plan to be riding near max speed for extended periods or on a distance ride that will be near your limits. Otherwise think of it as training resistance to build your endurance faster!

Some of the better dynos I think might even be in the 60% range for mechanical efficiency.