I am exploring building a detector that would tell me the color temperature of flourescent lighting (in other words it should be able to tell the difference between a light that is 6500K or 10000K). Any ideas on where to start?

 

Typically, for detecting certain wavelengths, photodiodes sensitive to the region of interest are coupled with sharp narrow optical bandpass filters. Then, the photodiodes are put through transimpedance amplifiers, which you can tune in order to equalize responses of other diodes with a known source.

Steve

 

Does anyone know of any IC's that do all of this?

 

I think TI or Burr-brown used to make a photodiode and transimpedance amplifier in a single package. What you're asking for isn't exactly common-place enough to have an ASIC, so it looks like you're going to have to design it.

Steve

 

TAOS make light-to-voltage converters in TO-92 packages. Mouser has them in their catalog. Their voltage out is proportional to the illumination, with response from far UV into IR.

You will need good filters for the color temp determination, but the relative outputs of at least two of the devices should give dependable results. The level of some red frequency to another blue frequency might give you the color temp. You might need a calibration source.

 

I didn't see that company before, they make some nice stuff
I am still pretty confused about colour temperature, I have been reading into it.. I don't really understand what 10000K means for example, or what wavelength(s) of light corresponds to that..

Steve

 

I stumbled on this site:

http://www.techmind.org/colour/coltemp.html

Color temperature describes the spectrum of radiation from a black body at that degree Kelvin. The site has some nice graphics. It is not a single wavelength, but more like a continuum with different intensities at the different wavelengths. John