Powering LED bike headlamp

[**DONOTDELETE**]

I have been fooling around with this idea for awhile. I used to use halogen battery lights but got tired of the batteies going dead when I needed them most. Generators have power as long as you are moving, but traditional 2.4W lights are too dim. What about powering an LED array with a generator? Could you get more light that way? Or make a commuter road light with a long battery life? The problem is adapting the nominal 6VAC/500mA output to the 3.6VDC/380mA power needs of my 19 LED array. So far I have tried running the rectified, filtered gen output through a LM317 but the regulator eats so much power that the gen can't keep up. A transformer would step down the voltage and step up the current, but I can't find an appropriate size. Are there any other options?

 

[MrAl]

Hi there,

You can simply try the series resistor
solution. Start with a 10 ohm 1 watt
resistor in series with a 1 amp Schottky
rectifier in series with the LED's.,
assuming they are all wired in parallel.
If thats not bright enough, write back here
with the results of what bightness you
seem to get, or try going down to 5 ohms
with the Schottky diode connected also.
If you cant get the Schottky easily,
try a standard 1N4004 rectifier diode first
and see how it goes.

--Al

 

[Harrkev]

The LM317 acts as a "smart" resitor. It simply adjusts its resistance to maintain a constant current. As such, it is rather inefficient if Vin is much bigger than Vout. It's advantage is that it is cheap and easy.

If you don't mind spending a little money, the EternaLight Versalux module should work at 6V DC. It is expensive, but powering it is a no-brainer, as it already has a regulator installed, as well as a brightness knob.

If you want to use simple LED's, and obtain the maximum efficiency, then you need to create a step-down regulator. There are several companies that produce these chips (includine Linear Technologies and Maxim), but you will have to know what you are doing (or be very patient and willing to experiment).

I hope this helps. If you give us your background and abilities, we, as a forum, can taylor a solution to fit

 

[PeLu]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Mcd:
What about powering an LED array with a generator? Could you get more light that way?
<HR></BLOCKQUOTE>

If you are happy with your LED's light distribution (most bikers want to have a more focused light) you could do that. Try to go to more LEDs, so that you get 500mA all together (including your rear light, new rear lights may use less than the usual 50-100 mA). You will come to some 24 LEDs. Use half of them for each half cycle, using only one additional diode per string. You may use a Schottky diode. If you are a daredevil, you may leave out the rectifying dioedes at all. All the white Nichia LEDs should accept at least 5V of reverse voltage. And at 5V forward voltage, your LEDs will be damaged anyway .-)

Don't put any resistors in series, they are of no use and will not reduce current.


<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>A transformer would step down the voltage and step up the current, but I can't find an appropriate size.<HR></BLOCKQUOTE>

Forget about the transformer, it will not work as you want it.

Which generator are you using?

 

[MrAl]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by PeLu:

Don't put any resistors in series, they are of no use and will not reduce current.
<HR></BLOCKQUOTE>

What makes you say this PeLu?
He said he had a 6 volt ac output right?

--Al

 

[The_LED_Museum]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by PeLu:
If you are a daredevil, you may leave out the rectifying dioedes at all. All the white Nichia LEDs should accept at least 5V of reverse voltage. <HR></BLOCKQUOTE>

I'm not sure I'd want to leave out that diode. The last bike generator I fooled with could pump out at least 15-20V open circuit just by hand-spinning the little wheel thing on the end, while it was rated for 6V with the light bulb load attached to it.

 

[PeLu]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Stingmon:
I'm not sure I'd want to leave out that diode. The last bike generator I fooled with could pump out at least 15-20V open circuit just by hand-spinning the little wheel <HR></BLOCKQUOTE>

You could easily get much higher voltages open circuit. But we don't have open circuit voltage at any regular situation.
If you look at one half cycle you have hafl of the LEDs in forward and half of them backwards. The forward LEDs will eat up the current. And at the 0.5A over the forward LEDs (what's the forward voltage for this 60mA maximum per LED?) there will never be a voltage which the backward LEDs could not withstand.
Actually I never sacrified a white LED to look about it's actual maximum backward voltage. Would be something for a LED museum guy..... .-)

 

[PeLu]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by MrAl:
He said he had a 6 volt ac output right?
<HR></BLOCKQUOTE>

He said he is using a bicycle generator.

[lamenting on]
I had to learn a lot about electronics and electrotechnics the last decades. Most of my knowledge is somewhat outdated now (I was pretty good on vacuum tubes (Austrian invention!) but I think I've forgotten everything). But the main point is, not to assume that one thing has to behave like another for the simple reason it amkes the same task.
example: incandescent bulbs and LEDs both make light. When an incandescent bulb fails, it is an open circuit, therefore a LED has to behave the same (which is wrong).
The same now here.
This is absolutely not aimed on you, Mral (you annoy me only with your short lines .-) . This is my very own personal problem, but I'm in lamenting modd just now .-) I had to learn that we have to have fun here, now I'm having fun lamenting .-)
[/lamenting off]

bicycle generators actually are constant current sources and no voltage sources. They will put out constant 0.5A (more or less). Open circuit is for them the same as short circuit for a voltage source (e.g. a battery). For example, Stingmon doubts that the LEds could stand the reverse volatge because his generator puts out 20V and more open circuit. If I say that a 123 cell could easily put out 10A shorted, why will it not damage the lightbulb which only withstands 2A?
Of course the generator is not a perfect current source, if you change your load voltage, current will also change a little (like a batteries voltage will change under different loads).
As a battery can only deliver a limited current, the generator can only do a maximum voltage (at 0.5A). And this voltage depends on the generator's RPM. When cycling slow, the light is dim but it reaches it's nominal current at 12km/h. And stays somewhat constant for higher speeds. Maybe it will be at 0.6A at 40km/h.

OK, back to the actual problem:
Your load has to consume 0.5A. If it does not, you need a series resistor to get rid of the excess current.
On the other hand, if you want a brighter light, you have to put a second one in series and not in parallel. The best setup is that you use one light up to e.g. 15km/h and switch a second one in series (that means at lower speeds the second lamp is just shorted out). There is a switch for that available which works with the generator's frequency. The nice thing is that you need less light at lower speeds.

In our case I would use 2x24 LEDs if Mcd could afford that, shorting one pack out at lower speed.
Or, even better, using Luxeon Lumileds. For this setup you may put them opposed and in series. And at least two strings of it.

 

[ElektroLumens]

MCD,
Regarding a power source for a multiple LED bike light.

Recently, I designed and built a step up voltage regulator board, using the LT1070 IC chip by Linear Technology. The way the board is designed, it can be set up to 13.77 volts (55 volts is possible, or up to 5 amps). I plan to use 4 batteries for an input of 6 volts, and drive LED's in series of 3, which would be 12 volts, or 4 volts each. I can set the voltage up to 13.5 to overdrive them if I choose. As the batteries drain, the same voltage is maintained for continuous brightness, until the batteries are depleted.
A power regulator board like this might be something you might consider for driving LED's on your bike?

Wayne Johnson

 

[**DONOTDELETE**]

Hi, thanks for your replies.

FYI, it is a standard old-fashioned Union generator. (Presumably a familiar item to PeLu, as they are common in Germany) I have not checked the volts under load, but they are rated conservatively; the old ones had problems blowing out lights when you got going at a good clip. I also have a Taiwanese one rated at 16V; I have measured almost 30V under load, cranking the pedals with one hand.

Someone mentioned stepping up battery voltage. Yes, that is an option I've considered, but for reasons of range and environmental impact, my goal is to make the bike generator power a usably bright light. Maybe not bright enough for trail use, but good enough for dark roads at 15-20 MPH. And certainly brighter than current incandecents. Maybe LEDs are not the right technology for this. BTW, I started this project long before the LS came out, or I would used them. That is my next step. Problem is, you can't mount several in a tight array because of the heat sinks.

 

[JollyRoger]

I'm not sure if you would want to mount LS led's in a tight array for this application anyway. I mean, without the optics, the light is VERY diffuse and there is no "beam" of light to throw or direct anywwhere anyways. You really do need a very good reflector or the optics for the LS to have a good beam. And using the optics means that each LS led needs some space....so even though the optics aren't that big (imagine a space about 1 inch around each LS led in your array to leave room for the optics), you won't have a "tight" array. I imagine using two or three ls leds would light up the ground in front of you well enough on a dark road...

If all you want is to be "seen" by cars, then all you need is one LS led w/o the optics. It is so bright and diffuse...

 

[The_LED_Museum]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by PeLu:
You could easily get much higher voltages open circuit. But we don't have open circuit voltage at any regular situation.
If you look at one half cycle you have hafl of the LEDs in forward and half of them backwards. <HR></BLOCKQUOTE>

Actually, I failed to see where there were two opposing strings, each one using half the AC cycle. In that case, the Vr shouldn't get high enough to damage any one LED because the string on the other part of the AC cycle will be using that power instead of letting it go to waste.
That means the reverse biased string wouldn't see more than 4 or 4.5 volts across each LED at any given moment, assuming the strings of LEDs were correctly assembled and an equal number of LEDs were used in each of the two strings.

So you might as well leave that rectifier diode in the parts drawer.