Low noise op amp

I seems the photodiode series resistance is pretty high - I'm still researching this - so if my incident light varies between curves (2) and (3) of fig.2-2 of JK17PWGBDR's Hamamatsu document, even if my opamp does its job of forcing its negative input to 0V, the voltage at the "true" photodiode will vary between VDC and VDC', and the current through the series resistance, which I can read, will vary in the corresponding proportion, which is too low to my taste.

Jean-Pierre Coulon

ce,

Cool. Gotta love it when things work. Have fun.

With JFETs like the BF862, the IDSS isn't that well specified, but nothing bad happens if you just stick a bunch in parallel. You might not get exactly the same performance as if they were all servoed to run at exactly I_DSS, but you still win fairly big by putting 2-10 in parallel.

Cheers

Phil Hobbs

Temperature shifts are the big killer, as you point out. If PD leakage is a genuine problem (it almost never is, at least with silicon), then the OP might need to heat sink the PDs, but I've never needed to. It's thermoelectric voltages that are the problem electrically.

Optically, etalon fringes are the big killer. A 1-inch polarizing beamsplitter cube can give you 12%/kelvin temperature drifts--yes, folks, that's 120,000 ppm. Strategic misalignment is your friend here.

Cheers

Phil Hobbs

Better ? in terms of ? I notice the so called low noise OpAmps, where some sub-nV per root-hertz are specified, are having a rather high bandwidth. Also observe that these low noise specifications are tied to a rather low impedance, usually below 100 Ohms, or less. This is physics though and cannot be cheated. However what can be improved is the bandwidth, if not really required.

Rene