High CRI Rebel Array

[spencer]

I was thinking about making a high CRI flashlight. I know that you can get the SSC P4 high CRI but they have been around for awhile and I think making my own would be better. I have decided to go with Rebel LED's because their colour LEDs seem to fill the spectral gaps better.

Seen below is a quote from LED Museum's Spectrographic analysis thread.

Nu-Flare Rebel 90 6-Watt 210 Lumen Flashlight

Here is the full post.

It shows a Rebel LED. I believe it is a cool white LED but may be neutral. A blue rebel has a nominal wavelength of 470nm which would fill that first gap nicely. A red rebel has a nominal wavelength of 627nm and fills up the red.

Now the tricky part. What ratios of LEDs do I need? Right now I'm thinking:
2 cool white
1 blue
2 red (maybe 1 red 1 red-orange/amber)

Would that give me significantly better CRI and what would the CCT be approximately? I'm thinking of putting 10 or so in a maglite. However the ratios work will be how many I will use.

 

[jtr1962]

Although you'll need to experiment to get things perfect, as a rough guide I'd say two red LEDs and one cyan LED per ten or so whites. Cyan is generally more efficient than red, so one per ten is about all you need to fill in that valley around 500 nm. The reds will take care of the red deficiency (get 640 nm reds, even better 690-700 nm if you can find it, not 620 nm red-orange). You should have the cyan and reds on independent channels so you can adjust their drive currents until the overall effect is white. The reds and cyans combined will tend to warm up the light a bit but not much. If you start off with 6500K whites, you might end up with 5000K - 5500K.