Uhm, ok ..... my old calibrator was another of the crazy projects i built years ago, for the units i had at my old workplace, after i've seen how much money they asked for recalibrate our Coherent unit
..... so i just had this crazy idea: "what if i make a laser unit stable enough, that can be used as reference, and that permit me to self-check my instruments for different wavelenghts ?"
"Model number 1" was simply using green and red ..... then i modified it, when PHR803 sleds becomed available at decent prices, and converted it in a 4-wavelenghts unit ..... now it's IR - red - green - BR (just cause this way i can check for different absorption properties for the reading plates).
It's not a single-beam assembly, i had no reasons for mix the beams like in a projector, it simply have the beams side by side (except for red and IR, where i used the double diode from the PHR) ..... and the beams are not focused for infinite distance, they simply project a "spot" of few millimeters at a specific given distance ..... but it worked for years without problems, for me
About the photodiodes, you need to use them in the feedback branch of the driver, just for keep the output stable, so is not so important the linearity and response curve, cause is not used as measuring sensor ..... it just regulate the current, inversely proportional to the light it take (in poor words, when the light increase, it lower the current til the light go back to the preset point, and when the light decrease, it increase the current).
There are a lot of possible circuits, for use a photodiode as feedback sensor, from the simplest ones using a pair of transistors, to the more complex ones with op-amps and temperature compensation ..... i can draw some examples if needed, but you can also found them already posted here and there in internet, probably.
