OpAmp speed for a low-frequency transimpedance circuit

[Bigman]

Hi,

I am confused and I am sure I am getting something wrong. I have a very basic transimpdance circuit as shown in the picture. The input signal comes from a photodiode and has a max. frequency of 10kHz with an amplitude of approx. 0.1uA.
==> Based on those numbers, I thought there is no need to take a high-speed OpAmp. Everything with a GBP around 200kHz and a SR around 0,005 V/uS should be okay.

When I started to do the calculation for the compensation capacity I came across the formular: C_f=SQRT(C_I/(2*pi*R_f*f_c)) <-- True, as long as C_f >> C_I
f_c ist the unity gain cross-over frequency of the OpAmp (I assume, instead I can also take for this the GBP value).
==> Well ,,, the larger f_c, the smaller C_f. Since I want to have C_f as small as possible,  I should go with an OpAmp as fast as possible.

To go now with an OpAmp as fast as possible feels somehow wrong and I am wondering: if I take a very fast OpAmp to get the advantage of a low C_f, what would be the disadvantages?

[tggzzz]

Put the actual numbers in, for each of the  components including the capacitance of the photo diode.

BTW, you may or may not wish to reverse bias the photodiode.

Make sure the opamp's supply voltages are consistent with the circuit's operation.

[Bigman]

Put the actual numbers in, for each of the  components including the capacitance of the photo diode.

BTW, you may or may not wish to reverse bias the photodiode.

Make sure the opamp's supply voltages are consistent with the circuit's operation.

If I put in the numbers, for sure there will be no reason to go with the slower OpAmp.
I would take a reverse bias voltage into consideration, to reduce the photodiode's capacity, which could be necessary to keep the bandwith of the circuit. I agree, this will also help to keep C_f smaller ... which also is reflected by the formular shows.

.... hmmm ... I am really grateful for everybody who takes efforts to give me a hint ... but what is actually your message  . I mean, how does it help me, to decide if a faster OpAmp is/isn't the better choice?

[Marco]

Well ,,, the larger f_c, the smaller C_f. Since I want to have C_f as small as possible

Why do you wish to keep Cf small?

PS. I understand it limits the bandwidth, which might be relevant with high Rf ... but it's not like you even hint that that's the case. High Rf is similar to high gain in a normal amplifier setup by the way.

[Bigman]

Well ,,, the larger f_c, the smaller C_f. Since I want to have C_f as small as possible

Why do you wish to keep Cf small?

PS. I understand it limits the bandwidth, which might be relevant with high Rf ... but it's not like you even hint that that's the case. High Rf is similar to high gain in a normal amplifier setup by the way.

Mainly, I was only wondering "why should I take a slower OpAmp, if a faster OpAmp is even better ... but is it really better, or do I have missed something substanially, such as 'if possible, the slowest OpAmp should be chosen, in order to avoid ...'"
A motivation to keep C_f low, comes also from the noise: I think, the larger C_f, the earlier the noise-amplification will rise (<--hope I am not mistaken in that point). At the end, my (theoretical) idea is: by fine-tuning R_f, C_f, etc. I put the starting point of the noise-amplification at the frequency of my highest input signal (=10kHz). And to get rid of the noise above those 10kHz, I would put a filter at the output, which "kills" everything above 10kHz.

[Zero999]

Why not build a prototype?

What are the power supply requirements? You could probably do this with a single power supply.

A cheap and cheerful jelly bean op-amp would probably do.