"Pete KE9OA" ) writes:
I did do a version of that quasi-sync detector, using a Philips SA637
digital FM receiver chip. I took one of the limiter outputs and fed it back
into the LO input of the on-board mixer. The 455kHz I.F. signal was applied
to the RF input of the on-board mixer, and the audio was taken from the I.F.
output of this mixer. Not bad, although the demodulated audio level was
fairly low (about 50mV p-p). I was actually considering using that detector,
until I discovered the appnote for the Analog Devices AD607.

Pete

Of course, what we've tended to see is an all or nothing thing when it
comes to AM (with carrier) detection. A simple diode detector, or
a PLL based synchronous detector (with or without phasing for sideband
selection). In some ways, it's been driven by market not by curiosity.
In the early seventies, those Signetic 56X series of analog PLLs came along,
and simple synchronous detectors (or outright receivers) were all over
the place. Most of the people applying them in hobby circles hadn't
woke up one day and said "we need a better AM detector", it was because
all of a sudden one could get the function on a single IC. So there was
a novelty, if nothing else. Various communication type ICs came along
at that point, and people were interested in "all mode" detectors from
them. Since a double balanced mixer was common to most of such ICs
(the MC1496, various FM detectors, the LM373, various Plessey ICs
and let's not forget Ralph Burhan making a Loran C receiver out of
an IC made for simple AM/FM broadcast reception), it certainly caused
an interest in simple quasi-synchronous or outright synchronous
detectors. But again, it often seemed to be "how can we get the most
out of this IC" rather than "let's try to improve the lowly AM
detector".

So there have been very little instances of biased diode detectors
(I think one of the Drake receivers used one, and Rohde showed a
few in his Ham Radio articles a quarter century ago, and I've seen some
in Wireless World in years gone by). Whether or not that is a useful
path, there never was the level of interest in trying such things that
there was for synchronous detectors, even before the latter became
easy with ICs.

And while at one point the limiter feeding the product detector
was described as a "synchronous detector" and worthy of experimentation,
as fancier detectors became more common, they are dismissed as merely
quasi-synchronous and maybe not even worth the effort. Yet, they
are a middle ground, maybe offering some increase in performance
without the extra circuitry of a PLL based detector. I can point
to a 1955 (I think) article in QST where someone built a fancy
receiver, with two parallel IF strips. One was AM bandwidth,
the other CW bandwidth. But, there was switching at the outputs
of the IF strips, so the CW strip could feed the product detector
at the output of the voice IF, and get some level of increased performance.

Let's not forget (to the original poster) that the whole point
of going beyond a diode detector is to improve operation with weak
carriers compared to the sidebands, or for that matter with adjacent
channel interference. The synchronous comes about because one needs
the reinserted carrier right where the incoming carrier is, but
except for that need, it's all about mixing the incoming signal down
to audio, just as in sideband reception. With a strong signal, the diode
mixer works fine for that, mixing the sidebands with the carrier. But as
the carrier weakens, or an adjacent signal becomes stronger to "capture"
the detector, the simple diode detector suffers.

Long before before CQ ran the article about the synchronous detector
in 1958 or so, most or all hams who tried for better AM reception tried
other things. Like boosting the carrier of an incoming signal in
reference to the sidebands, "exalted carrier". The common, or maybe
only means, of doing this was by using a Q-Multiplier, which could
provide a narrow peak for the carrier, but a relatively sloppy
skit so it didn't attenuate the sidebands too much. Nobody talks
about this anymore, even though we've seen in this thread the comment
"simple and good" don't apply to synchronous detectors. But how
much improvement is needed, versus how much circuitry one is willing
to accept?

Michael VE2BVW