View Single Post
Old 09-09-07, 09:08 AM
  #7  
n4zou
Scott
 
n4zou's Avatar
 
Join Date: Jun 2006
Posts: 2,393

Bikes: Too Many

Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Likes: 0
Liked 1 Time in 1 Post
Originally Posted by Fibber
It is interesting that Mag-Lite powers their "3 watt" LED with two AA batteries (3v assuming alkaline cells). I wonder what emitter they are using, and what voltage / current it is rated at for strong output and long life? At 3v and 350ma, I'd call it a 1+ watt emitter. Are you sure that is all it is drawing?
That’s what my meter is telling me but considering how hard it is to read current accurately the 350mA reading I get is most likely not correct. This reading was obtained when checking the current draw with two fresh AA batteries. I think it's very close to 500mA when using the dynamo and bridge rectifier circuit. I tried using the flashlight without the LM-317 voltage regulator and found it came close to loading down the dynamo enough to keep the voltage below 3-volts DC. This was when using just a bridge rectifier and capacitor with just the Mag-Lite flashlight. I had a Cyclocomputer hooked up so I could read the wheel speed and at 8 MPH the dynamo with bridge rectifier and capacitor could easily exceed 3 volts output to the flashlight if a were to exceed that speed. There are times I am exceeding 25 MPH so without the LM-317 voltage regulator the LED and chip in the flashlight would have burned out. With the dynamo loaded with 3-watts of filament bulbs (original configuration) the illumination and dynamo output would peak at 12 MPH. I didn't want to find out just what the peak voltage would have been with the LED flashlight and end up burning it up. That’s something else that makes me think my current measurements are off. There is a buck boost LED driver chip driving the LED in the 2 AA version of the Mag-Lite flashlight. I tried to dissemble the flashlight to find out just exactly what they're using. They did very well when assembling this flashlight to prevent anyone from doing that without destroying it in the process!
Originally Posted by Fibber
Other possibility for you would be to run two lights in series to make use of all of the 6v output.
I don’t want to do that! I want it as simple and uncluttered as possible.
Originally Posted by Fibber
By comparison, the Element "3 watt" flashlight uses three AAA batteries (4.5v assuming alkalines). The emitter is a III on the star backing. According to the Philips/Luxeon spec sheet, the emitter does best in 1000ma (1 amp) configuration with 3.75v applied. So that is technically closer to 4 watts consumed to get the rated 70-80 lux output. To drop the overvoltage one would assume that there is either a puck regulator or a resistor in there, but if there is, it is well hidden and not obvious. As an experiment, I threw in three Everyready AAA NiMh 850ma rated rechargeables, and the light output compared very favorably to the other light running the alkalines. Yet now we only have 3.6 volts applied, and presumably less available current.
I tried rechargeable batteries in my Mag-Lite as well. That’s when I figured out they had a chip in there somewhere. It ran just fine for 15 minutes and then shut down. I reloaded with alkaline batteries and it worked fine for the remaining 20 miles back home. When I got home I checked the rechargeable batteries, which showed 80% charged. The chip shut down due to the lower voltage supplied from the rechargeable batteries.

Originally Posted by Fibber
If I do decide to harvest the emitter and optics from two of these for a homebrew project, I will have to pay attention to applied vs rated voltage. Using the AC dynamo as a 'charger' (like an automotive alternator) with full bridge rectifier and maybe the LM series regulator as you showed to supplement a battery pack was what I had in mind.
I've been playing around with some rechargeable batteries as well. I found that by just placing 2 AA ni-mh rechargeable batteries between the bridge rectifier and the flashlight the batteries act as a voltage regulator and prevent the voltage to the flashlight from exceeding about 2.8 volts. Whats very interesting is the fact that it made no difference if the batteries were at full charge or completely dead. The only problem is the current produced; well over the 180mA suggested charging rate printed on the side of the batteries! I am in the process of ordering a 1.5-watt star LED from mouser.com and a few 1N5818 diodes and capacitors as well to start playing around with the circuits here.
http://www.pilom.com/BicycleElectron...moCircuits.htm
I don't think you really want to "harvest" parts from perfectly good flashlights when you can get emitters from mouser.com.
Just go to www.mouser.com and enter the mouser part number
859-lopl-e011wa that is a 1.5-watt star mounted LED for $4.96. Be sure and click page 84 for even more choices including 3 and 5 watt LED's.

Last edited by n4zou; 09-09-07 at 09:14 AM.
n4zou is offline