Spectral Response of White Cree LEDs?

[justinmreina]

Hey all,
For my photodetection class I am going to analyze the spectral response of white LEDs. The main point of this project is to characterize the optical transmission through the spectrometer (which sucks !) and fully design/characterize my detection circuit.

However, being that I just can't get over Cree's LEDs and that I may just have a MCE, a XRE and a XPG lying around, I am going to compose the spectral response of these devices.

I will be measuring the transient response at 75 wavelengths across the visible spectrum. Initially I will be doing 700mA on the XRE for 30 seconds from cold @2 KS/s. Don't worry, my computer is doing it, not me! If I don't see anything, I will switch over to watching the first 100us by using an oscope (again, comp control dont worry!). At the end, I will compose these into a series of spectral responses over time, overlayed on the same graph.

So here's my big question:
Do you guys have any other tests you want me to run? I will have a calibrated photodiode measurement circuit and a great spectrometer, so if you have any thoughts let me know!

 

[blasterman]

Be curious how your graphs jive with Crees....

Also be interesting to see how Crees in the same lot and specified bin type actually vary (A gripe I have with Cree as of late) but won't tell me something I already don't know.

 

[monkeyboy]

Check out the LED museum. Craig plots the spectral response curve for most of his reviews but mainly older luxeon LED's on that website now. Just so you know what to expect.

Tint shift occurs at different drive levels of the LED. It would be interesting to see how this appears on the spectral plot. Of course it would be impossible to keep the die temperature constant in this case.

 

[justinmreina]

Ah ha! I'm going to interpret your response and clarify what I'm doing also:

- You specify steady-state spectral profiles (over wavelength) and their change with drive current. This would be an excellent test, and easy to automate. I think this is a must for me.

- I am interested in the response over time. my use right now for the lights are as a high-speed camera flash; this is the original basis for my test.
​

Sweet! I will definetly include this as a test for my system.

Thanks to everyone and keep 'em coming,
Justin Reina

*That LEDmuseum has a lot of info, wow.

 

[bshanahan14rulz]

if you have LEDs to "burn," perhaps a color analysis at very high currents like, in a flash application where you can pulse higher current through the LED as long as it's for a very short amount of time. Be interesting to see if the phosphor can react fast enough to colorshift or if it will still be the regular white.