"David J. Windisch" wrote in message
news:3f13ee31_1@newsfeed...
Hmmmm. Been a long time since I thought about the r-d.

IIRC #1: It (the reciprocating detector) would detect SSB and c-w, and it
was silent between SSB bursts and c-w characters.

Hmmmmmm. My RD implementation behaves like a BFO, no silence like you
recall. Heterodynes 5kHz and higher with a wide IF. Olberg's hr 74/09
article says "Through regenerative feedback Q1 and Q2 form a simple
oscillator operating at the filter centre frequency". When not locked onto a
carrier ("nonsynchronous mode") he said, "the reference level is no longer
completely amplitue controlled by the input signal". That seems to suggest
that as a locked on signal becomes weak, the amplitude of the reference
signal will diminish until lock is lost and rise when it becomes unlocked.
He went on "but it does have signal-induced phase fluctuations", whatever
that means. I do notice a tendency for the reference signal (BFO) to pull on
signal strength and also on keyed carrier or bass speech energy close to the
lock-in range. The latter can sound like a lf resonant belch. The narrower
the lock-in range, the less susceptible it would seem, but then with no
better than 50Hz tuning steps and some drift to contend with, we can't be
too narrow.


IIRC #2: It was best preceded by mode-suited, steeply-sloped filtering.

I recall speculating, back then, that its operating principle was to
generate the bfo signal by ringing a bfo tank, offset from the incoming
signal, with the bursts of incoming signal. When the tank ceased
oscillating, the r-d output was silent.

A mode-suited, steeply-sloped filter ahead of the r-d would help keep its
bfo tank from being rung by unwanted (off-frequency) signals.

It would certainly prevent them beating against the product detector's self
oscillation. The filtering of the reference signal inside the RD is supposed
to be fairly narrow but not so narrow as to make lock difficult. The
narrower the filter, purportedly the greater the immunity to impulse noise.
In my current implementation, I use a ceramic resonator paralleled by a
white top 455kHz IF L-C resonator. Olberg showed just an inductor in
parallel with a quartz crystal, the inductor apparently chosen to resonate
with the crystal's interelectrode capacitance at the crystal's frequency. My
LC ratio is much lower and is perhaps not as effective at suppressing
wideband noise but I can tune for a tighter lock range than Olberg
recommended.


The bfo offset required for these modes, and thus non-synchronousness of
the
bfo with the incoming signals, I think, might help to support, to some
degree, the speculation above, as well as to explain the use of
"reciprocating", rather than "synchronous", as part of the name.

The "reciprocating" nomenclature seems to be related to switching the path
between opposite phase legs of the product detector on the carrier phase
reversals that occur with zero crossings of the modulation envelope of a
suppressed carrier AM signal. Perhaps he borrowed it from "reciprocating
engine" metaphor. The inventor, Badessa, did not use the term in his patent.


Thinking about it now, I simply can't recall any attraction other than
novelty ;o)

It does seem to me to be an inexpensive synchronous/nonsynchronous detector
for all AM modes (and possibly NBFM) that can be easily constructed for low
frequencies, eg 455kHz. I'm sure my parts cost is only about $10 compared to
$150 for a kit that uses the Sony ICF2010 parts. I'd like to try a 2010 as
many extol the virtues of its SD and see how they compare. My main complaint
with the RD (or, at least, with my RD) is the audio null that occurs at one
point in the lock range.

73, Dave, N3HE
and 73 to you.
Tom VE3MEO