Overdriving

Candle Power Forums

Help Support Candle Power:

Ralph_Hilton

Newly Enlightened
Joined
Jul 26, 2003
Messages
18
City & State/Province
Semmering, Austria
I've only recently become involved with LED flashlights and have just modded a couple - one with 4 LEDs and one with a 1W Luxeon.

I also bought a small keychain one LED light for GBP5 ($7). Out of interest I measured the current through the single LED and it was over 50 mA. 150% overdrive.

It would appear that the cost/efficiency ratio of LEDs doubles every couple of years.

LEDs are most often specified with 100,000 hours of life; about 11 years continuous use.

After 4 years, at most, the LED will be an antique. Thus, it seems to make sense to me to overdrive LEDs to a much shorter lifetime.

Does anyone have graphs made up showing overdrive vs. lifetime? Or some data to evaluate with about how much one can overdrive LEDs?

I'd be interested to know what lifetime I might expect from the 50mA driven keychain light.

Graphs would be best if anyone has produced them..

Tks.
 
I haven't seen any graphs to support this, but I assume that 100% overdrive will cause a 50-90% reduction in service life. This won't matter if you break it or lose it before it burns out. For intermittant duty, overdriving can be justified as long as the temperature doesn't get out of control. Note however that efficiency will degrade at higher currents. Also, I'd advise against overdriving a luxeon for few reasons; it costs more to replace than a 5mm unit, and isn't much bigger but produces 15-20 times the heat.

As for your keychain light, it's likely to meet a more gruesome end than old age (leaking battery, salt water damage, Rex's teeth, blender, tire, shall I go on?)
 
Tks. 90% reduction seems fine - 10,000 hours = 3 hours a day for 10 years.

There's a thread about 1 watt Luxeons going for $4 in the Elektrolumens forum - 30 lumens+ bin. At that price I think I could afford to blow out a couple on testing.

I had the idea of sinking an overdriven 1W Luxeon (or a few) with a computer CPU heatsink and fan.

Five of them bolted with thermal grease to a copper sink was my thought. 100% overdrive would be 3.5 amps!

Another thought was just to pull all the incandescents out of my apartment and replace with a few dozen Luxeons. With the electricity bill and the cost of light bulbs I'd be better off within a year.

As for Rex's teeth - he's a clever *******. 10 mins alone in the apartment and the fridge is bare - see http://www.fzint.net/rhilton/photos/Mphotos2.htm
No way I'd let him near my flashlights!
 
Nice angel you got there on the table /ubbthreads/images/graemlins/smile.gif
P.S. It would be very intersting to see a 60mW green laser pointer over there at the clouds /ubbthreads/images/graemlins/smile.gif
 
There's almost no way, no graph or chart, per se, to tell how much reduction/increase in the life of an LS you're going to get when you increase or decrease voltage/current to it. Too much variables in determining what's the condition for its runtime as well. If you consider that heat's its greatest enemy, than obviously the better dispersion you have (be it a better body or larger area exposure to open air), or the less current you drive through it, the most likely you'll be able to extend the life of an LS. But even that's inexact, as something like a small short could kill an LS before its time...
 
To me, it's worth the reduction in the life of the LED. I think they will be far eclipsed by other LEDs, or something else, by the time their useful life is over.
However, I'm not speaking from experience. I have only a few overdriven lights, none of them a year old yet.
 

Latest posts

Back
Top