Torque1st
Enlightened
Some reading I found that may be of interest to some:
http://www.nocturnolightscapes.com/viewarticle-en_diploma.htm
http://www.nocturnolightscapes.com/viewarticle-en_diploma.htm
White LED Characteristics
- Thread starter Torque1st
- Start date
Help Support Candle Power Flashlight Forum
Kier
Newly Enlightened
Thank you for that. It very nicely explains the characteristics we've come to expect from our yellow phosphor + blue led = white leds that are in our lights. It also helped me understand the low colour rendering ability of the white leds.
The testing that the author performs on angle of incidence and efficacy of the phosphor layer is interesting if you could build a reflector to mix them better. I wouldn't think it'd be worth the effort
The testing that the author performs on angle of incidence and efficacy of the phosphor layer is interesting if you could build a reflector to mix them better. I wouldn't think it'd be worth the effort
2xTrinity
Flashlight Enthusiast
I have noticed this myself, especially with the Cree LEDs -- the "warmer" light is emitted to the sides, as the blue light coming from the die passes through a greater amount of phosphor materials that the light that exits head-on.
This produces the effect of reflector-based lights having a "warmer" hotspot, with cooler spill, which IMO is the best way to do it, as the majority of the blue receptors in the eye are in the peripheral vision, with the central vision almost entirely composed of red/green receptors. Of course, for area illumination (not a flashlight) that creates problems, however, area illumination is still better off with typical fluorescent lighting, IMO.
Interestingly, they mentioned near-UV based LEDs with separate RGB phosphors. This setup makes very little sense IMO, one of the advantages of using a blue + yellow, or blue + separate red/green phosphors is that there is NO UV content, which would be important for something like museum lighting etc.
I personally have found that the best color rendering from LEDs so far has been achieved by finding LEDs with a "greenish" cast to them (stronger phosphor emissions, weaker blue spike), then adding separate red emitters to the mix. Even this does not solve the problem of phosphor variations from one part to another, which are apparently very difficult to control.
It seems like some sort of RGB emitter with three separate dice will eventually be the best way to solve all the problems mentioned so far -- namely color rendering, and control of tint. I believe Cree submitted a patent application a while back for a RGB emitter with the separate dice stacked one on top of the other, as opposed to side-by-side. An RGB chip that didn't have color mixing problem would really be great, and woudl allow a lot more freedom both in flashlights, and in general use.
This produces the effect of reflector-based lights having a "warmer" hotspot, with cooler spill, which IMO is the best way to do it, as the majority of the blue receptors in the eye are in the peripheral vision, with the central vision almost entirely composed of red/green receptors. Of course, for area illumination (not a flashlight) that creates problems, however, area illumination is still better off with typical fluorescent lighting, IMO.
Interestingly, they mentioned near-UV based LEDs with separate RGB phosphors. This setup makes very little sense IMO, one of the advantages of using a blue + yellow, or blue + separate red/green phosphors is that there is NO UV content, which would be important for something like museum lighting etc.
I personally have found that the best color rendering from LEDs so far has been achieved by finding LEDs with a "greenish" cast to them (stronger phosphor emissions, weaker blue spike), then adding separate red emitters to the mix. Even this does not solve the problem of phosphor variations from one part to another, which are apparently very difficult to control.
It seems like some sort of RGB emitter with three separate dice will eventually be the best way to solve all the problems mentioned so far -- namely color rendering, and control of tint. I believe Cree submitted a patent application a while back for a RGB emitter with the separate dice stacked one on top of the other, as opposed to side-by-side. An RGB chip that didn't have color mixing problem would really be great, and woudl allow a lot more freedom both in flashlights, and in general use.
Torque1st
Enlightened
I am glad you guys found the information interesting. I had always wondered why there were blue rings at the outer edge of some of my lights. Color rendition and efficiency will continue to improve.
Actually and RGB based light source will have no better color rendering than a Blue+Yellow phosphor. In fact, it will likely be worse as it is missing so much spectral content with only three narrow lines. You can add in one or two other colors of LED and get much better color rendering. Yes when you look at the LED it may look perfectly white, but when you look at the reflection off other items, it is not so good.
There are high CRI leds such as the LuxeonI warm white which has 90CRI, but the efficiency is not so good because of the phosphors used.
The new neutral white LEDS are a more pleasing light I am finding. Does anyone else have any experience with this?
Using near UV leds is for the advantage of having near perfectly consistent color as all the light output that is visible will be from the phosphor. There is no ratio of blue to yellow to worry about (which varies with the blue wavelength).
Semiman
There are high CRI leds such as the LuxeonI warm white which has 90CRI, but the efficiency is not so good because of the phosphors used.
The new neutral white LEDS are a more pleasing light I am finding. Does anyone else have any experience with this?
Using near UV leds is for the advantage of having near perfectly consistent color as all the light output that is visible will be from the phosphor. There is no ratio of blue to yellow to worry about (which varies with the blue wavelength).
Semiman
