What photodetector parameters should I be focusing on, and in what priority, to maximise the performance of such an interferometer?

Question

I'm trying to select a photodetector for an interferometry experiment. The interferometer will be trying to measure subtle vibrations occurring at a relatively high frequency (high kHz to low MHz). It will be using a CW visible-wavelength laser with relatively low power (<50mW). I've read the Thorlabs photodiode tutorial, but I'm still unsure of how to select the photodetector. Specifically, what parameters should I be focusing on, and in what priority, to maximise the performance of such an interferometer?

Answer 1

I assume this is related to Do interferometers require equal-powered (or as equal as possible) beams being emitted at the target and the photodetector(s)?.

First off, 50 mW is a lot for a photodiode. You should have no problem detecting those light levels. I would recommend just getting a reasonably priced photodiode that works for your wavelength and bandwidth. You could get one of the variable gains ones as those allow you to play around with some settings. Best if you're able to easily plug your photodiode into your data acquisition device (an oscilloscope or something). The battery powered ones are annoying because they still give signals, just nonsense signals, when the battery is going low. I prefer ones that you plug in.

Regarding noise specs and performance etc. My advice would be to just get a cheap/reasonably priced version. You'll probably get some signal out of that. If you find your signal to noise is to low you can do some more learning/investigating and then buy an appropriate photodiode after the fact.

So, in short, my advice is to not spend too much time trying to get the perfect photodiode before you know all of the details about your system. Just get SOMETHING so that you can start looking at signals, then you can analyze and learn and make a more informed decision down the road. In the end this latter approach will typically be faster and more effective in experimental physics despite having more "steps".