LED taillight on an AC bike

[JimF]

Last year I found a 6-volt LED taillight bulb that is capable of working in either a postive ground vehicle or a negative ground vehicle.

An AC bike is simply a negative ground bike half the time and positive ground the other half of the time. All this happens at high speed as the AC alternator's sinusoidal wave swings the voltage polarity with engine RPMs.

The LED worked, it wasn't that bright but it did work. Better than a light bulb too because I wasn't blowing up incandescent bulbs from vibration or over-voltage. Over voltage can occur because the voltage is unregulated out of the alternator.

I did have to re-wire the brake light so the ignition coil would not be grounded through the brake light filament (standard Ducati wiring on a AC bike) because I no longer had a filament for the ignition's ground current to travel through when I switched to LEDs. This required changing the type of brake switch, from SPST (single pole single throw) normally closed, to the type of switch put on DC bikes which is SPST normally open.

Toward the end of the year on a fall ride somebody told me I had no taillight. The next time we all stopped I took my bike's tailpiece off and the bulb was gone. The BA15 bulb base was still in the socket, but the guts of that LED bulb somehow disintegrated and fell out. I think it was held together with epoxy, and perhaps temperature, time and vibration conspired to break the epoxy bonds.

I have since made my own LED taillight bulb out of three architectural grade high power Luxeon Rebel LEDs that are current driven.

IMG_0814.JPG

My LED bulb is ground specific, it has to have a negative ground. Fortunately my taillpiece is in fiberglass so I put a forward biased diode in the ground wire coming off the dual-filament bulb socket and going to the motorcycle frame which transforms both the running light and the brake light AC voltage into pulsating DC voltage, 50% duty cycle.


This is my LED bulb on the test bench, all three of its emitters lit and 100% DC voltage from a power supply, that is to say no half duty cycle as will be the case when installed on the AC bike.

BA15 on the test bench.jpg

Below is a photo of the bulb with only the running light on, that is to say only one of its three emitters is lit, and that emitter is being powered only half the time. It's daytime, but near dusk. Still plenty of light in the sky, you can see the streetlamp is not yet on.

250_LED.jpg

Finally I have a taillight bulb that works even in the daytime.

brake_on.jpg

Brake light on.

[Jordan]

Nice work, Jim.

I once put a single white LED behind an amber lens. When lit up, it looked - white!
Weird. I put an orange coloured LED in to restore normality.

Jordan

[JimF]

Hi Jordan,

Here is my non-physicist understanding.... and I could be wrong.

Incandescent light bulbs look white because their light covers the entire additive light spectrum; so when all those colors are added together it appears to be white light.

Placing a red translucent lens between your eyes and the incandescent bulb allows only light in the red color spectrum to pass through the lens so your eyes are only seeing red wavelengths and thus red light. Same for orange, yellow, green lenses etc.

LEDs only emit light at one color wavelength. So with LEDs it is best to match the LED color to the color of the lens it will go behind. This combination will allow the LED's light to pass through the lens. There is some wiggle room in all this. A red translucent lens will allow other wavelengths to pass, but they are inhibited. And a LED radiates a range of light to either side of its intended wavelength, it just drops off exponentially.

White LEDs are different than white incandescent bulbs. While the light appears white to our eyes, it does not contain the full color spectrum. What you saw with the white LED behind the amber lens was some subset of the white LEDs wavelengths with some of the non-amber wavelengths filtered out.

On a similar note, with the LED traffic signals that are appearing near me I sometimes don't see the green lights very well because I have a translucent amber sun shield on the top third of my helmet visor. When I am looking through the amber sun visor it filters out a lot of the green colored wavelengths.

Jim

[JimF]

http://www.motoscrubs.com/LED_taillight ... rcycle.htm


I put together this page for reference use by anyone wanting to put a LED taillight on their AC electrical system Ducati single.


If anyone uses this let me know if there are any errors, omissions, confusing text, etc. and I will try to improve the page.


Jim

[Jordan]

Hi Jim,
I don't like batteries on bikes much, especially if they don't have a starter motor. It's good to see solutions for good batteryless operation.
Is it necessary to use a separate diode, when the LED is a diode? It would conduct half the time.
Some bikes with AC systems use a clipper diode for regulation - actually 2 zener diodes in series and connected "back to back".
As LEDs are current sensitive, and voltage is proportional to current, some regulation should be a good thing.

Jordan

[DewCatTea-Bob]

[quote="Jordan
" I don't like batteries on bikes much,
It's good to see solutions for good batteryless operation. "

____ There's a bunch of various solutions for achieving such with most-all Duke-models, but because headlights have always consumed so very-much power, none of the involved battery-less methods have been what ya could actually consider as being "good", (since the lights have always been rather quite dim at the lower-RPMs).
__ But now that we have LED.type headlights, there's no-longer any substantial downside to employ systems which don't require a 'storage-battery'.



" Is it necessary to use a separate diode, when the LED is a diode? It would conduct half the time. "

____ That's a fairly logical point that I-myself once brought-up within an older related thread which Jim had already started (over a year ago). _ We then didn't ever get-around to discussing exactly why he thought the added valve-diode was needed, so I've assumed that it's intended function is to protect the LED from being exposed to possible excessive reverse-bias voltage, just in-case the LED's own rated breakdown-voltage isn't sufficiently high enough. _ So-then the extra/added-diode's 'breakdown-voltage' should be at-least 100v, (in the event of a full-throttle missed speed-shift).



" Some bikes with AC systems use a clipper diode for regulation - actually 2 zener diodes in series and connected "back to back". "

____ Yep, probably such as this unit... http://cgi.ebay.com/ws/eBayISAPI.dll?Vi ... 0920556582
And those AC.voltage-regulators could be useful in most-all battery-less systems, (or even possibly within battery-powered systems as well !).



" As LEDs are current sensitive, and voltage is proportional to current, some regulation should be a good thing. "

____ Certainly so, especially if the system-voltage climbs near/above the LED's rated 'breakdown-voltage',, however, a single Zener-diode is all that's needed for THAT possible use.



____ One note I'll add for now, is that the notion of creating a full-wave, (rather than merely-just the half-wave), rectified current-waveform just for the sake of merely the LED, should-not be a very worthy consideration since the LED requires so very-LITTLE power ! _ And that being the case, any possible low-RPM LED.flickering could just as well be reduced (if not entirely eliminated), simply with the addition of a relatively small capacitor (likely under 100uf - thumbnail-sized).
But adding such a capacitor would definitely also make the LED brighter, so then a voltage-limiter of some-sort may also be of use (to make life easier for the LED).

____ I'll read-through Jim's likely quite useful report sometime when I'm fully-awake (after a GOOD sleep-period), and then see if there's really any part of it that could use any tweaking at all. -- (I know that my-OWN writings most-often seem to benefit from double-rechecking my chosen post-wording.)


Enlightening-Cheers,
-Bob

[JimF]

" Is it necessary to use a separate diode, when the LED is a diode? It would conduct half the time. "

Very true. But notice in the course of this discussion several kinds of diodes have been mentioned; clipper diodes, Zener diodes, power diodes, light-emitting diodes... and we haven't begun to scratch the surface.


Each is designed to perform some task better and as a result is poor at some other task. To the logic of a LED being self-rectifying, we could also suggest that it could act as it's own zener diode and thus be self-regulating in addition to being self-rectifying and light emitting all at the same time. But there are reasons why when you look in your 400 watt stereo audio amplifier you don't see four light emitting diodes used to create the bridge rectifier that is going to supply the amplifier with hundreds of watts of audio power. Instead you are likely to see physically massive power diodes attached to a huge heat sink. Different diodes are engineered to do different things better.



"...some regulation should be a good thing."

If we accept the challenge to regulate the pulsating DC we've created to run a LED taillight, keep in mind that we are not back at the engine trying to turn the AC bike into a DC bike, rather we are at the taillight just trying to run a few LEDs without destroying them. So in addtion to the half-wave rectifier diode we will be forced to use the two seperate diodes to rectify both the running light and the brake light voltages (as they are independent of each other at the taillight) and then we incur two seperate regulators as well.

The simplicity of the experiment to run a LED taillight on a AC bike by using unregulated pulsating DC with one or two diodes discretely placed in the wiring loom quickly escalates into assembling a half-dozen or more components together, mechanically mounting the components to something, wiring them together and soldering them up, waterproofing them and securing the assembly where it will be hidden in use. Mutiply this by two if as we need to manage both the running light and brake light, ore else to implement a single regulator we have to do the DC conversion and regulation ahead of the brake light switch where the running light circuit and brake light circuit can feed off of a single regulated power source, in effect closer to the engine.