Energy Efficiency

[ltiu]

I'm just a lay person here, I do not have a PhD in electrical engineering.

Did a little research:

http://en.wikipedia.org/wiki/Luminous_efficacy#Examples_2

Seems like there is plenty of room to grow (or glow) since most LED are using only about 10% of the amount of energy being put in and turning this into visible light.

Interesting back of the envelope calculation, from this website:

http://www.allaboutbatteries.com/Energy-tables.html

The AA battery has approx. [FONT=Verdana, Arial, Helvetica, sans-serif]2.60 watt/hours of energy each.

The [/FONT]Fenix L2D CE has 2 AA.

It supposedly will produces 135 lumens max for 2.4 hours, according to website:

http://fenix-store.com/product_info.php?products_id=195&osCsid=2692f18873bb2263f878903017e11801

So 2.6 watt/hours multiply by 2 AA divide by 2.4 hours = 2.1667 watt/hours, the amount of energy the Fenix L2D CE uses per hour.

Since it is supposed to produce 135 lumens, we divide this by 2.1667 watt/hours = 62.307 lumens/watt. Right there in the ballpark!

Max theoretical efficiency is 683 lumens/watt, according to wikipedia.

So 62.307 lumens per watt divide by 683 lumens/watt * 100 = 9.122% energy efficiency. Again, right in the ballpark, although at the high end of it as per wikipedia's claim of 3.8% to 10.2% efficiency for LED.

The other 90.87% of the energy is WASTED!

I wonder what is the most energy efficient LED flashlight out there?

 

[cave dave]

try the same calculations at the 10 lm setting and get back to me. Way more efficient.

 

[ltiu]

... 13.9% efficiency at the 9 lumen setting with a runtime of 55 hours ... more efficient but not much more.

 

[2xTrinity]

Seems like there is plenty of room to grow (or glow) since most LED are using only about 10% of the amount of energy being put in and turning this into visible light.

This isn't true. The best white turn more like 30% of the energy input into visible light. The problem is, the eye doesn't perceive all visible wavelengths the same -- If you had red, green, and blue LEDs of equal power and equal efficiency, the eye will perceive the green one as being the brightest. The lumen metric tells you how much visible light is being produced, weighed by the eye's sensitivity to different wavelengths.

Max theoretical efficiency is 683 lumens/watt, according to wikipedia.

This only applies to light at 555nm, a monochromatic yellow-green light where the eye has its peak sensitivity. Citing efficiency numbers the way that they do is the wikipedia is very deceptive. In the case of a UV lamp for example max theoretical is 0 lumens per watt, that means it is definitely inefficient as a light source, even though it may be very energy efficient in terms of converting electricity to UV. Max theoretical efficiency for decent-quality white light is more like 300 lumens/watt, not 683.

 

[ltiu]

If this is true what you say 300 lumens/watt. Then we are getting close to the theoretical limit:

http://www.theinquirer.net/default.aspx?article=29472

The holy grail is to turn every single watt into light!

If only my car can do that, I mean turn every single energy in a gallon of gas into forward motion

 

[nerdgineer]

I've always tried to keep track of the energy efficiency of flashlights, usually based on comparing FLRs numbers and eyeballing the area under the FLR runtime curves.

Among production lights, Fenix appears to provide the most light per battery for each type of light they make. Some others come close with some types of batteries, but Fenix seems best overall.

Sadly, FLR is closed now so it's become harder to compare the new crop of DX/Kai Cree/Seoul lights with the wealth of historical data at FLR...:sigh:

 

[SilverFox]

Hello Ltiu,

When running your figures, keep in mind that the 2.6 watt hours is based on a discharge rate of 25 mA. At higher discharge rates, that figure is greatly reduced.

For better accuracy, you should determine the current draw of the light, then determine the watt hours available from the battery at that current draw.

Tom

 

[ltiu]

I did say "back of the envelope calculation" didn't I?

But thanks for the suggestions. I know the calculation was fat fingered all the way but the numbers are good approxx.

 

[OhMyGosh]

The figure I have seen around is that LED's are expected to peak out around 150 lm/watt over the next 3-5 years.