DIY "Spectrometer"?

Hey everyone.

Had an idea after I saw this....

http://www.mouser.com/catalog/specsheets/TCS3404 TCS3414-A.pdf

Basically, it's a digital color sensor... but what if we programmed it so it took the data that it got (in forms of RGB) and then used it with something similar to the wavelength -> hex/RGB calculator tool, but backwards?

It would give us a decent approximation of the wavelength, no? Of course, it wouldn't isolate any wavelengths in the event that you shone more than two lasers into it (or just used a magenta-BR/red combo), but I still think it would be a neat idea.

Wolfman29 said:

Basically, it's a digital color sensor...

Click to expand...

Kind of like... I dunno... a camera? except with 99.9% fewer pixels :undecided:

My method is cheaper and simpler. Have you tried this webapp by rhd?

Well, I would hope that the photodiodes on this product would be more effective at determining exactly what the correct blend of colors is.

I have yet to try your method (don't have a grating -.-), but I think that having a box which you shine a laser into and have it output your wavelength would be kind of neat.

Though probably more expensive, you can find the spectral content of any light source with a michelson interferometer.

Or take a diffraction grating and a linear CCD sensor and you have a full spectrometer. I think the HP/Agilent multiwavelength meters use some form om a michelson interferometer.
The point was making a simple device, the the mentioned sensor can be helpfull. ILX lightwave makes power meters that also determine the wavelength by using two photodiodes. One of them has a glass filter that is monotonically decreasing in transmission with an increasing (or decreasing) wavelength so that the ratio of the two photodiodes corresponds to the wavelength. Very simple, but acurate enough
The sensor mentioned here essentially uses 3 photodiodes, but it's spectral response it flat at some wavelengths, this means it can't acurately determine the wavelength there.