The detector provides
both low distortion and very high dynamic range. It can cleanly demodulate
AM signals, even with positive peak modulation in excess of 200 percent.
From Frank, WA1GFZ, "The whole point of this circuit is to compensate
for the diode knee and voltage drop across the junction. The 1N34, like
other germanium diodes has high leakage. This will reduce the dynamic range
of the circuit. The best possible configuration is to get a matched set
of diodes. HP used to sell matched sets of four, but I'm sure those days
are long gone. The next best thing would be to get a lot of them and characterize
them and select the closest three. The output of the I.F. strip on many
radios is high enough to use silicon diodes. These are easier to come by.
I have been using this setup in my home brew RX since '85 with a matched
set of HP hot carrier diodes." The circuit was adapted from a circuit
published in RF Design, in 1985.
For IFs below 500 kHz or so, use a good FET
op-amp like a TLO72 for IC 1. For higher IFs, select an op-amp
with a gain-bandwidth product well over 1 MHz for best performance. Vcc
is 9 volts. As noted above, hot carrier diodes can be used for D1,
D2 and D3. Also, Shottky diodes like 1N5711s will
also work well. Even 1N914s work well.
Adjustment: Tune in some atmospheric noise
with the AGC on the RX off. The adjust the pot (5k) for maximum noise.
You will see the noise dip up and down and get distorted as the pot is
moved past optimum.
In light of the distortion
created by "standard" diode detectors in most receivers, you
should try this circuit soon.