White Led efficiency. Got an Idea.

[JonSidneyB]

I understand that generally while led's are not as efficient as many other colors. Not sure why this is true but I am sure that there is a good reason. I also understand that while led's are often not any more efficient than an incandecent. I thought I remembered that if the three primary colors of light are mixed, that a white light results. To confirm my memory, I did a search and my memory was confirmed. http://home.att.net/~B-P.TRUSCIO/COLOR.htm
Now I would think that if a light using three or multiples of three leds in the primary colors, a brighter light would result than using three white led's. So why is this not being done? I would love to see a comparison of a light using three white led's vs a light that is using one each of the three primary colors. Can someone tell me if my thinking on this is flawed let me know. Maybe this can be the road to a more efficient light.

 

[Quickbeam]

I may be talking out of my butt here, but if I remember right, I saw in either a post or a webpage that there were a number of factors necessary to get the right combination of colored LEDs to make white light. The voltage requirements of the 3 LEDs (red, green, and blue) would all need to be different in order to put out the proper ammount of light needed for your eyes to percieve the light as white. The red needed to be very bright, the green dim and the blue a medium brightness (if I recall correctly). This was due to the efficiencies of the different colored LEDs and the eye's "perception of" and "sensitivity to" the different colors. Of course, every LED is a little different, meaning that each light would have to be custom color matched in order to make the light "white". Also, the voltage requirements for bright, efficient red LEDs are different from the green and blue. You would need to create a custom driving circuit or strange combinations of resistors in order to get it right - and it would be different for each light made - a real pain in the neck....

 

[lightuser]

This idea could be utilized if not for its output power at least for reducing the blue tint from the white LED arrays. Anyone know what could be done to implement that, especially using a production light like the Trek 1900? I'm sure you get into amperage and beam shape problems when you open that can of worms. The "perfect caving light" (linked to one of the posts) is an attempt, but I don't recall any more info than that

 

[Steelwolf]

I remember seeing in one of the links, a circuit design which allowed room to put in a couple of amber LEDs. This offset the blue tint of the white LED and gave the entire light a slightly more incandescent feel to it.

But I believe the power provided to the amber LEDs were separate from that provided to the white LED array.

 

[X-CalBR8]

How about this for an idea? You could figure a rough estimate of how many of each color LED to add to 3 arrays, have 1 power supply with 3 different connections running to each of the 3 arrays. Of course the 3 arrays would have to be all intermixed to get the right effect. Now, put a variable resistor on each of the 3 arrays and you can dial the brightness that you want from each array. Of course there is one potential drawback to this. The LEDs might have different efficiencies and put out different amounts of light when they are cool as opposed to when they are hot so such a flashlight could possibly require constant adjustment, or just wait till it warms up a minute and then adjust it. The big plus to such a flashlight is like JonSidneyB said, that it would be very efficient and throw out much more light compared to a white LED flashlight.

So far as figuring out the percentages of each color of LED that needs to be added goes, somewhere I've got the exact percentages of how sensitive the human eye is to each of the primary colors, red, green, and blue. If I can find it, I will post the info later. From memory though, the eye is quite sensitive to green, blue is kinda in the middle and the eye is very insensitive to red. So from what I remember, I would start with the following numbers as a rough estimate if someone actually wanted to try this out. For every 1 Green LED I would add 2 Blue and 4 Reds. You would also have to take a lot of other factors into account such as beam angle and how efficient each of the 3 colors of LED are at different temperatures. Seems to me that this would make for an interesting and fun project to play around with though.