Originally I thought we would put a whole bunch of LEDs, in a circular pattern, into the light-shell. But since the lens wouldn't come off, we needed to sneak the LED unit in "the back door," which limited us either to one large or two small LEDs. Olivier, who knows more about electronics than I do, informed me that it would be better to use two small LEDs, since otherwise we'd need to use a resistor and waste some current. (I suppose this has something to do with AC, but I'm not sure.) So we ordered two tiny Luxeon "Rebel" LEDs and one larger "K2," just in case the Rebels didn't work. Both are "red-orange," even though my red lens seems to do a good job making white light red.
The next step was to machine the body for the LED unit. Our idea was to machine a cylinder out of plastic to the inside diameter of the "entry point" on my taillight. Then we would drill a hole through the centre, which we would then tap to accept a hollow chainring bolt. Then we'd machine a little piece of aluminum to act as a heatsink and a mounting plate for the LEDs. I wish I had photos of this process, but I don't—anyway, it all went as planned.
Then last night we hooked everything up. Olivier cut slots in the plastic and aluminum to allow for the passage of wires, and he soldered the LEDs together. Then (after some difficulty) we epoxied the lights to the aluminum plate, and the aluminum to the plastic. Then we passed the wires through, and soldered a ground wire to the inside of the chainring bolt (it will ground itself through the frame.) The photos will no doubt explain all this much better than I could have. We the light up to my Schmidt wheel via Olivier Cyo front light, and let me tell you—it's BRIGHT. Perhaps even too birght—but we'll see!
Many thanks for M. Scholten for his excellent ideas, and his offer of doing all this work for me!
The plastic "body" of the light, with the headless triple chainring bolt in place. It attaches to a seattube braze-on thus.
The same piece from the other side. It's hollow, like the bolt. The slot is cut to allow the wires to make it through.
Olivier soldering the ground wire to the inside of the bolt. This required filing off some chrome plating...
The LED unit partially inserted into the light housing. I'm going to cut off at least half of that "tube."
7 comments:
Why did you wire the two LEDs in series?
If you wire them in parallel with reverse polarities one LED would light up each half cycle. However, without looking at the specs I'm not sure if those LEDs could handle the current from the dynohub, so maybe you'd need to do two on each side for a total of 4. That would undoubtedly be overwhelmingly bright.
Looks good though!
I think the current going to the taillight is already DC as it is connected in parallel to the headlight and it's fancy electronics? (save for the standlight function).
Adam the reason two is "better" is because each LED runs at 3V so two of them use all of the 6V from the hub. A single LED would require a resistor to drop the 6V to the to 3V required.
So did you try to fit it in?
I didn't stick it ALL the way in, but I did get it in to the plastic part -- so I think it will fit fine without any filing (though hopefully without cracking the glue!)
The above discussion is way out of my league -- you two can settle it!
The output of the Schmidt headlight is AC. The dynohub is a constant current source of 500 mA.
It's the Supernova that puts a DC output to the taillight.
Huuum... well perhaps some testing and redesigning is needed!
Looks like we'll be experimenting... wired in series on an alternating current, the LEDs would apparently ALWAYS flash! Imagine riding on my wheel through a night ride!
Post a Comment