Im going to start this back up because because the arguments are quite entertaining. I used to have one of those generators that ran on the outside of a bicycle tire. I was forever blowing out light bulbs so I did some experiments. Im surprised this forum does not have a separate forum for generators. First of all they are quite simple. They rotate a magnet in a coil of wire. I used a 10 amp meter hooked directly to the terminals of the generator and spun the wheel. At 10 mph it would produce 6 volts AC and 1 amp. At 40 mph it would produce 24 AC volts and 4 amps short circuited. Combining this information with the very non linear behavior of tungsten filaments I finely understood why bulbs were burning out. The resistance of the wire in the generator was less than .01 ohm so ohms law clearly did not work here. How is a generator not a pure current source as well as a pure voltage source?
Not sure I understand what your argument is… 
So you drove 64 km/h on a bicycle, while looking at the DMM on your handlebars?
Just kidding, but now you know why the incandescent => led replacements for bicycle lights are more expensive than those for regular flashlights.
The meter is probably measuring the RMS voltage and current, so the peak levels are higher and may exceed the rating of the bulb you are using.
The generator output voltage is proportional to the rotational speed (w) of the magnet. The faster a magnet passes a coil, the higher the induced voltage.
The coil has a DC resistance R; It also has an AC resistance known as Reactance due to the inductance (L) of the generator coil which also increases as the speed goes up. The Impedance (Z) is the sum of these two,
Z = R + iwL, expressed mathematical using Real and Complex components.
If you use the impedance in the calculations with the RMS values, then Ohms law will still apply and not be violated, and it appears to be about 6 Ohms based upon your measurements.
The impedance is causing limiting of the current. If you desired a constant voltage or constant current output, then you could add a rectifier bridge to convert it to DC output and some regulator circuitry to make nearly any voltage/current that you desire.
a generator doesn;t have to be either thing, it will be nonlinear at best
output depends on design, speed, load and other stuff
a light bulb is also not a pure v=ir resistor either
so it is kind of a crap shoot to tie any bulb to a naked generator output
I find it easier to swap or recharge my 18650 cells and I have still have light when I stop 
yes the led and the 1650 really revolutionized bike lights for me
i would never have been a generator guy
in olden days, i had nicads and ‘halogen’ (incandescent)
dim
short life
yellow colored light
heavy
twice what a cheap led light costs with 40 times the lumen-hours
terrible!!!
There are endless arguments on electronics forums about current versus voltage sources. This forum does not seem to want to do this but is only concerned with practical things.That is fine. A superconducting motor has a torque proportional to its impedance so its start up torque should be infinite. The moon is falling toward the earth. A bicycle wheel stands on its spokes it does not hang from them. When you slice through a cone why do you get an ellipse and not an egg shape? When my cat first encountered a mirror I told him some things are just perplexing and to not let it bother him. I turned the bicycle upside down and spun the wheel. I did try a bridge rectifier fed into 4 nicads the output voltage did not go above 6 volts with a 24 volt AC input. A nicad can take around 12 volts each before it shorts. Thats all it does it does not burn down your house like some other batteries I could mention. But since the generator and the headlight used the frame as a wire it was not a practical solution.
So do the wheel spokes have a slight motion up and down in their mounting hole as the bicycle is ridden and the wheel rotates from loaded to unloaded position?
i once worked on a telescope experiment with a superconducting motor of sorts that consisted of a quartz sphere coated in a thin layer of Niobium. The sphere was illuminated with a UV lamp to put an electric charge on the metal surface, and then it was spun up in a vacuum and held in place with no bearings using an electrostatic suspension system. All of this was done inside a liquid Helium dewar at around 4 degrees Kelvin in order to keep the Niobium in a s.c. state (~9 Kelvin, the highest of any element). Changes in motor torques were sensed and measured using a SQUID, superconducting quantum interference device, to measure changes in magnetic flux from the current loop formed by the rotating electric charge.
i’ve read a few patent applications for superconducting motors but couldn’t follow the theory of operation—it was like science fiction to me. 
It can be thought of as either (Thevenin or Norton).
Hotwire bulbs are nonlinear, and it’s just easier to think of a dynamo as a “wattage source” instead. That’s why if, say, the rear bulb pops, the front bulb pops very shortly thereafter, too.
A standard bicycle dynamo is usually StVZO approved which means using standard resistor measurement they provide 3W at 6V when running at 15 km/h, i.e. 6V 0.5A. [ Extracting more power is possible using non-resistive loads ]
Already in the 1980s many non-budget headlamps for bicycles (aftermarket and those mounted on non-budget bikes) in NL were equipped with zener diodes (2 connected in reverse direction in series, placed in parallel with the headlamp bulb). I think they were normally 6.2V zener diodes (but perhaps they were just under 6V), so with voltage drop across the other means a cutoff at ca. 7V.
In such cases where zener diodes are present you could still may have needed to replace them sooner than expected: You quickly approach the 1000 hours that you might expect from a bulb, but actually expecting 1000 hours of life out of them is optimistic because of the vibrations in use. Also even peaks of 7V means a life time decrease compared to running on 6V. I don’t think I got 1000 hours out of them but it’s a long time since I used these.
If the main bulb dies then the zener diodes may or may not be able to dissipate enough power to keep the taillamp alive. The zener diodes used weren’t designed to dissipate 2.4W but due to factors such as airflow they may have stayed cool enough in such cases.
Possible problems: If there was a bad contact of the front light then the taillamp would quickly burn out. Burning out of the taillamp was not an issue in headlamps with zener diodes in any other situation than that and if the zener diodes burned out after the headlamp burned out.
That sounds like a really cool experiment. Where does the telescope come in? I don’t see being able to drive anything with a floating sphere. And no wonder telescopes are so expensive. Did you ever hook in a flux capacitor?
Err, is this true?
Doing a thought experiment, I mentally removed the bottom spokes from a wheel and applied weight.
The top spokes, under tension, supported my imaginary bike.
Removing the top spokes and supporting the wheel with the bottom spokes, and applying weight.
The thin spokes, under compression, bent and wheel hub collapsed.
On my imaginary bike.
Is a coat hanger a single spoke? If so it can easily support a 20lb weight in tension (a top spoke).
Yet can’t come close to supporting 20lbs in compression even if perfectly balanced.
Am I missing something? I am an old fart who perhaps is thinking too much…
All the Best,
Jeff
Naw, the bike “hangs” on the top spokes.
Reduce it to its basics, one spoke on top, one spoke on the bottom.
That’s why suspension bridges hang from the cables, vs being pushed up from underneath.
Now, things like wagon-wheels have a contribution from both, being rigid.
Not convinced?
Tweels.
IMO it doesn’t really matter if one person argues that the wheel hangs from the top and another argues that it stands with support from below. It won’t work without both, plus a few more.
jeff, the thinking is that they are all under tension
” tension in the spokes on top of the wheel doesn’t increase when the wheel is under load, (it would if the wheel hung from these spokes) but the tension in in the bottom spokes does go down. Have a friend sit on a bike and measure the tension yourself. …And read Jobst’s book.”
Back up the spoke thinking a little…notice that in nearly all wheels the spokes are laced tangentially from hub to rim. Radial spoking - straight line (ish) from center to circumference is much less common and always much less durable, apples to apples with the same components. Don’t discount the rim design, either…matters a lot and it all works together…plus the skill of the builder (or machine) often makes or breaks it (literally, and sooner or later).
Jobst’s treatise is excellent, a bit outdated but for the most part foundationally just as relevant as it was in the first edition. Bicycle wheels are pretty amazing really. Cheap, heavy, overbuilt wheels can conceal a host of shortcomings in materials or build quality, but as people have pushed the edge toward lighter and lighter wheels, the art has really progressed (and I’m not even considering carbon fiber, etc).
Even a strong, stiff wheel has give to it and is never really round. As it rolls, it’s flexing out of round with every chord of every revolution, and thus the tension of spokes, rim, rim joint is always flexing. I don’t know where the applied force ends up, but it’s not radial (or not directly so, not all of it). This does depend a bit on the diameter (spoke length) as well as the lacing pattern…tangential lacing calls for longer spokes of course, and the more tangential you go, the longer they get, changing lots of things in the physics of things. Then you can try to account for straight gauge wire spokes vs. double- or triple-butted spokes, flat bladed spokes…plastic or composite spokes…even the nipples and their seats in the rim.
Then in a wheel that is offset to account for the width of gears or disc rotor mounts, everything changes some more. Mount a tire and fill it with high pressure air, more changes. 
Yeah, spoke-crossing is to help keep the rim from twisting around the hub.
Ever see a dragster take off, with the big fat wheels actually wrinkling around the sidewalls? The spoke layout is supposed to try to minimise that.
“Angular rigidity”, I guess you can call it.
F’n timestamps on every f’n image, it seems. Just goggle “dragster sidewalls wrinkle”.
Agree, but a bike wheel using spokes of string would still support weight from the hub on a stationary wheel. Cutting the top strings would let it collapse.
A bike wheel is a marvel. As are wire wheels on some sports cars of yesteryear.
Now a wagon wheel is the exact opposite. The spokes are ridged. And are in constant state of compression when the wheel is in proper shape.
The spokes are a slip fit into the hub and felloes. Then the whole thing is held in compression by the steel tire.
On a loose wagon wheel you can see the top spokes shifting out of position as the weight shifts from lower spokes as the wheel turns.
Had an uncle in back in the 60s(?) who had an Jag with wire wheels. He would con me into helping him polish them in the summer.
He worked for the phone Co. They were putting in the system for the new CIA shop in Va. Also the Airport at the end of the road.
The long road was not officially open then. He would let the Cat loose on the way to work and back. First time I ever went over 100 in a car. I was sworn to silence from telling my folks. Never did tell them.
BTY this Aunt and Uncle knew little boys. For Christmas they would give an Erector set of some other sort of toy that required interaction and imagination. None of this “Here’s a nice sweater and socks for little Jeffrey” KaKa…
All the Best,
Jeff