I realized this week that I had never plotted I-vs-V curves for phototransistors, and the data sheets don’t usually include them, so I measured some today.
I put an LTR-4206 phototransistor in series with a 1kΩ resistor (so that voltage across the resistor was current in mA), and drove the pair with a slow triangle wave (3.2Vpp with 1.6V offset) from my FG085 function generator, using a 470µF capacitor in parallel with the function generator to smooth out the steps.
I illuminated the phototransistor with a narrow-beam 591nm LED (33-2UYC/H3/S400-A6) from about 0.5″ away. The LED has a 68Ω current-limiting resistor to 3.3V, resulting in an IR voltage drop of 1.248V, for about 18.35mA.
I recorded the voltage across the 1kΩ resistor and across the resistor plus phototransistor using PteroDAQ on a TeensyLC.
I interposed a Brand 1.5ml cuvette full of water or a piece of white paper to get different light levels:
Making a colorimeter with a phototransistor to get good linearity probably requires a transimpedance amplifier, rather than a simple resistor, but it looks like a 2.5mA current range for the amplifier would be adequate. With a 3.3V supply, the VCE could be set to 0.7V and the gain to 1kΩ, with the output in the range 0.7V–3.2V.
Interposing a piece of paper in the light path is good for aligning the LED and the phototransistor, as the beam pattern becomes quite clear. One of the biggest challenges in making a homemade colorimeter is ensuring that the LED and phototransistor don’t get jostled as the cuvettes are added or removed.