NORCAL NC2030 Manual page 111

detection is not simply a sample. I think of the detector as being an integrating detector. Each input
capacitor sees the incoming signal for only ¼ of an RF cycle. During this ¼ cycle, the voltage on the
detection capacitor is charged via the R of the system impedance as seen at the input to the multiplexing
IC. If the detector input were to be fed directly from a 50 ohm signal generator, the capacitor would be
charged by the input signal through the 50 ohm system. Mathematically, this process looks like an
integral. If you were to take the integral of a ¼ cycle of sine wave centered about its peak, you will find
that the peak detection voltage will be ~90% of the open circuit voltage of the generator output.
The integrating effect makes this detector unique in that, unlike other mixers, this detector produces a
difference only and not the typical sum and difference products a mixer normally produces. In a DC
receiver, the difference produces the desired audio signal, while the sum products would normally just
be thrown away.
A typical diode mixer, normally provides a 50 ohm termination and a 6 db signal loss. This means that
with a 1v pk-pk RF generator open circuit voltage, the 50 ohm diode mixer termination attached to the
50 ohm generator cuts the signal down to 0.5v pk-pk, while the 6 db conversion loss on top of this yields
an output of 0.25v pk-pk.
This new "Tayloe detector" on the other hand will look like an open circuit close to the detection
frequency, and will produce an output voltage of 0.9v pk-pk, a value quite close to the open circuit
generator voltage. Moving a ways away from the detection frequency, the detector starts to look like a
short circuit, thus rejecting unwanted off frequency signals.
Thus the new detector does have some conversion loss, contrary to the claim of some folks. I think their
confusion arises from looking at the voltage at the input to the detector, which is a point on the far side
of the system impedance across which the voltage drop actually occurs. The 90% detection voltage can
be observed experimentally by measuring the open circuit voltage of an RF generator, then connecting
the generator to the detector input and measuring the detector capacitor audio outputs. An alternative
analytical approach is to model the detector in Spice and looking at the voltage at various points in the
circuit.
The audio preamplifiers use 100 ohm input resistors. These resistors are not used as detector impedance
terminations, but rather to isolate the detector capacitors from the op-amp "virtual input" + and –
terminals. The effect of the 100 ohm resistors is to allow better detector bandpass characteristics, thus
helping the off-frequency reduction of large signals. The use of the 100 ohm resistors into the preamp
op-amps are a tradeoff between sensitivity and large signal performance since the introduction of any
resistors in the low level signal path produces noise which reduces sensitivity. If large signal
performance were not a consideration, these resistors could be eliminated and sensitivity would
subsequently be somewhat improved.
The gain of the audio preamplifier is 67 or 36.5 db. The gain of the op-amp is the feed back resistance
(10K) divided by the input resistance (100 ohm resistor + 200 ohm detector output). This gives a
voltage gain of 33.3. However, the differential input of the detected signal gives another 2x gain to
bring the total signal gain to 66.6.
Norcal NC2030 v5 12-18-05
Page 111 of 128