- #1

3dB

- 3

- 0

I have a receiver chain consisting of the following (where all conversion factors are assumed to be scalar and constant under the conditions in which I use them):

1) An APD (avalanche photodiode): Input: Optical power P1in [W]; Output: Current I1out [A]; Conversion factor: C1 [A/W].

2) A transimpedance amplifier: Input: Current I2in = I1out [A]; Output: Voltage U2out [V]; Conversion factor: C2 [V/A = ohm].

3) A monolithic high-frequency amplifier, 50 ohm differential input. (~50 ohm resistor connected across the amplifier inputs, negative input is AC-coupled to GND, positive input is AC-coupled to the transimpedance amplifier output.) Use Z0 = 50 ohm. Input: P3in [dBm]; Output: P3out [dBm]. Gain factor is C3 [dB].

My question is: How should I convert the input optical power to an electrical power, at the input of component 3) above? (Or more specifically, how do I convert U2out to P3in?)

I want to use power expressed as dBm when making the necessary calculations, since I will treat the receiver as a RF system. For example, the amplifier gain is expressed in dB.

Assume I have the input optical power and all conversion factors, no point in confusing the argument with a bunch of numbers.

**************************

My thoughts on the matter are as follows:

I am confident that the output from the transimpedance amplifier is U2out = P1in * C1 * C2, and the units seem to be correct.

But the only way I can think of to convert U2out to P3in is to take P3in = (U2out)^2 / Z0.

I have a gut feeling that this is incorrect, since (look at the formulas) this means that the original input power is squared. In my case this is a problem since the required dynamic range of the system, when calculated at the input (the APD), is _half_ of the dynamic range when calculated at the 3:rd component.

Since I do not care about the DC components of U2out, the Z0 resistor seems to be where the voltage drop occurs (but I suspect I should treat the differential input of the 3:rd component as a parallel connection of the external matching resistor and the input impedance, but that would only introduce a factor of 2 in the calculations).

I can't find the error. I suspect the root cause is in the C1 conversion factor, [A/W].

Can anyone please point me in the right direction? Or if I am already pointed in the right direction, tell me why? ;-)