[Laser] discrete vs op amp sensitivity

[F1AVYopto at aol.com]

Hello  Art 
Very  interesting subjects. 
We are  waiting your reports. 
Be careful to  compare optical front ends with well define precisely primary  
requisites. 
If the goal  is only the sensitivity limits without bandwidth consideration 
it is probably  the PGP front end the best because: 
All the  Operational Amplifiers have a differential structure that gives 
always two FETs  at their inputs.  One of these FETs  is an extra noise source. 
To fix the  gain and the bandwidth, the classical solution is the use of a 
feed back  resistor. This resistor is a major noise source. However another 
solution is the  feed back at the cold side of the photodiode. It is for example 
the “PIF  circuit” from Kalevi Hyyppa and Klas Ericson. 
Some  experimenters work again to improve the PGP principle.  The FET choice: 
2N5457 or better BF861A.  The functional point for this FET: The higher the 
drain current the lower the  noise.  Two or four FETs in parallel  structure 
can reduce the noise again. 
The noise of  a sound car could seem a negligible factor because most of the 
time the input  signal is strongly pre amplified by low noise external devices 
before to be  sampled? 
It seems the  problem is the same if you put a very good HF preamplifier at a 
noisy receiver  input. 
The external  first stage is known to be determinant??  
73 Yves  F1AVY