k so turns out plugging Ic into the shot noise formula is *not* correct...
i guess im free to use a simple op amp + BJT current source as long as it doesnt have some horrible 1/f noise down at 50 Hz.
>>2948377
>when I read up on this, the sources I found say a FET is always better than a BJT when it comes to noise.
i think that might just be true for amplifying weak signals, where input current noise * source impedance dominates. but is that still the case with zero source voltage noise and zero source impedance? youll have to forgive me, ive never used a FET in my life.
>use lock-in amplification
actually the reason for this circuit is to measure the noise of a photoconductor vs current; you have lots of options to separate signal from noise, but fewer options to separate noise from noise. zero options, in fact. so no fancy domain-shifting noise reduction techniques for me :^)
>>2948386
>anon also needs a low noise voltage source like LT3045 (positive) and LT3093 (negative) to generate the appropriate bias with as [low] noise as possible
"as low noise as possible" would be chemical batteries. for larger voltages, ive had a good amount of success with just a 431 + pass bjt + 2 enormous metal film caps.