There are several DIY optical spectroscopes projects around, most too simple and fragile (even made of cardboard), just good enough for educational purposes. Others I can find are quite sophisticated, requiring custom optics or specialized sensors.
I want to make a spectroscope that just works, cheap and good enough to tell if a particular LED light source will be effective to grow plants underneath. Think of a reliable version of the simple cardboard one.
There are also professional tools to measure PAR energy, but they are in the several hundred of euros range and some depend on specific, known light sources.
Components
1×
Raspberry PI
any with a camera port will do
1×
Raspberry PI camera without the IR filter
The easiest way to use a known, characterized CMOS sensor that is available and cheap
1×
3D-Printed case
1×
DVD-R to cut a difraction network
It has to be a DVD-R, not a CD, not a BlueRay, not other types of DVD
The cardboard prototype helped eliminating some uncertainty.
It worked better than expected and I got some very promising spectra. I used my mobile phone as camera, and two known laser sources to trim/calibrate the resulting spectra image. I did all that using just Gimp and a spreadsheet. The sensitivity of the sensor is not known so it can't be compensated, but as a proof of concept, it worked.
In pink, a 532 nm green laser, in yellow a 650 nm red laser. The blue plot is some random white LED source.