This is a normal single ended 4:1 multiplexer feeding four detection
caps as was done in the NC2030. It does have a 3 db pad between the
receiver front end filter and the detector to allow for a flatter
wideband response. The detector caps do all the integrating. I have
read about folks trying to do the integrating around an op-amp, but I
have not tried that yet.
The 3 db pad reduces the receiver sensitivity to -128 dbm. The
receiver could be another 3 db more sensitive yet with more op-amp
gain, but the gain is high enough as it is. The noise floor in my
Presonus Firebox rises by only 2 db when the receiver is turned on,
which means that the receiver noise floor is slightly less than the
noise floor of the Firebox (equal would give a 3 db rise), so MDS of
the net system could be 3 to 4 db better if the receiver noise floor
were to be raised perhaps 6 db above the sound card noise floor.
I would like to try it out the active integrator for myself and see
what it looks like. It is certainly an interesting alternative
implementation, but it is not clear to me that it provides benefit. I
might model it in Spice to get a rough feeling for it, but I have
found that for this type of detection circuit, the simulation results
do not match well with real life. That is a shame, because it is much
faster to work with a simulation that with real hardware.
Bottom line, the FF SDR has normal Tayloe Detector/QSD quadrature
detector followed by a very low noise audio pre-amplifier.
- Dan, N7VE