All righty now; I have read all 6 pages of the conversation here, and I have a working theory that I'll get to in a second, but first let's put that photo of the internals (from back on
page 4 of this thread) up here again for quick reference:
A couple of points to clarify first:
- It does not matter which direction the generator rotates, because it generates alternating current either way, not direct current. Thus if you want to test it for output using a meter, you need to select the AC voltage setting, not the DC. The generator is rated for 6 volts and 3 watts (0.5 amps) output, so a meter set on an AC Voltage range of 0-10 volts or thereabouts is all you need. You don't even need to spin the generator constantly in either direction; simply wiggling its drive wheel rapidly by hand is often sufficient to show a meter response that tells you whether it's working.
- That shaped copper wire contact that's sort of hanging in space in the photo above is indeed the ground contact for the bulb/lens assembly, serving both bulbs. When the lens assembly is closed, it will make contact with the side flange of the center bulb socket. It actually acts as the ground path for both bulbs, because both sockets are pressed into the same metal reflector, so they can share the same ground connection. The ground path for the upper (Low-beam) bulb passes through the reflector to the outer collar of the center bulb, and then directly into the copper ground contact and ultimately back to the generator via the frame of the bike.
Okay now, the switch on top is simply a beam selector, where center position will feed power to the center (High-beam) bulb, and turning it to either side will go to the upper (Low-beam) bulb instead. The generator does not have enough power to light both headlight bulbs simultaneously, so the switch will select one or the other, but not both.
Conversely, the switch
has to allow either one or the other headlight bulb to light when in operation, because otherwise the generator output will overload and blow out the taillight bulb, which has a much lower rating (0.6 watts/0.1 amps, versus the headlight bulbs' rating of 2.4 watts/0.4 amps each). You can ride with a working headlight and blown taillight, but a blown headlight will cause the taillight to fail shortly thereafter.
So anyway, from what I see in the photos, I think there is a short circuit (to ground) at the power terminal, where the outside power lead attaches to the terminal on the lower left side of the headlight. Someone has replaced the original terminal screw with a much longer Phillips-head machine screw, seen in this outside view:
It appears that the screw is off-center in the hole, so its insulating collar or washers are not preventing contact with the shell of the headlight. It's supposed to "float" in the center of the hole, without making direct contact with the metal edges. It might also be that the power lead's crimp-on eyelet (with a red collar) is making contact with the outside surface of the headlight.
With the power lead disconnected from the headlight
and with the front lens assembly open (as in the photos), there should be zero continuity between that long threaded machine screw and the shell of the headlight. The only continuity from that screw should be via the little white wire inside, leading up to where it is soldered to the center contact of the selector switch on top. From there, it should be traceable to the center bulb contact if the selector switch is in the middle, or to the upper bulb contact if the selector switch is turned to either side.
I suggest loosening that Phillips-head screw that's acting as a power terminal connection and adjusting its position so that it is properly centered within that mounting hole. Tighten it up and then test to ensure that there is no continuity from the outside threads directly to the metal shell of the headlight. Everything else in those photos looks pretty clean and unmolested, so see if adjusting the power terminal installation to be properly centered will help, as well as verifying that the crimp-on terminal of the power lead is not accidentally touching the shell either.
Given that the taillight bulb tests good but the headlight does not, I'm thinking that the taillight has survived being the only working bulb in the circuit up to now because a short to ground at the headlight connection is allowing all the current to go through there instead.
