On Wed, 7 May 2014, gareth wrote:

I feel that I understand something when I can explain it to
a rank amateur such that they will also understand it, but
one thing I have yet to understand is the purpose of the quenching
action in superregeneration, apart from thequenching itself.

How does the quenching of an oscillating stage yield a detector
that works for AM, FM and / or SSB, when the quenching frequency
is supersonic, and neither at the RF nor the AF frequencies?

Howard Armstrong came up with the regenerative receiver about 1914,
positive feedback making a stage get a lot more gain, and of course if you
weren't careful, it would kick into oscillation, which also showed that
tubes could be used to generate radio frequencies.

But the point where things were "just right" was critical, and was
affected by the antenna swaying in the wind, or the operator's hand coming
close to the front panel. A result was the receiver often went into
oscillation when not desired, a loud squeal in the ear.

Then Armstrong came up with the superheterodyne receiver in 1918, which
has nothing to do with it, but I like the sequence.

He had patent problems, Lee de Forest challenging his patent on
regeneration, so in 1922 before the case came to trial, he did some new
experimenting with the regen, and came across a phenomena he'd noticed
earlier but had never pursued. That was what became superregeneration.

The superregen is just a regen with quenching added. Obviously Armstrong
got the early results because the same stage could be the regen and the
do the quenching, but the quenching could be a separate stage. Indeed,
the superregen has become a kind of black box, people repeating what they
heard, explaining how the same stage can also quench, rather than explain
how the process works. SHow a separate quenching oscillator, and you see
that it is just a regen that's being "modulated". That explains why the
superregen always has such a wide bandwidth, modulate anything with a
square wave and you get lots of sidebands.

No, I still don't get what happens, but by adding the quenching, the regen
became more stable. It was never able to go into oscillation, and for a
lot of purposes that was good. They helped homestead the higher
frequencies, they weren't perfect but they were better than trying to use
a regen up there. And for a lot of consumer equipment (garage door
openers, 27MHz license free walkie talkies), they were simple and cheap
receivers that didn't need adjusting.

SUperregens are primarily for AM. FM comes from slope detection (and
since the bandwidth of a superregen is basically way too wide, they are
really only good for FM reception if the signal is wideband, like FM
broadcast. Otherwise, the narrow deviation provides too puny an audio
signal. They will not work on SSB, since they never go into oscillation,
needed to provide the beat signal to convert the SSB signal down to audio.

That said, since a superregen is just a regen with quenching, you can with
the right adjustment have a regen go into superregeneration. It's not
common, but I know one ARRL VHF receiver did that, allowing for all modes
(though in that case, they had the receiver at 14MHz and a converter ahead
of it).

Charles Kitchin about 20 years ago went back and looked at early work on
the superregen, and wrote about it for Communication Quarterly. And he
discovered that if the quench waveform and depth was adjusted, one could
have better control over the bandwidth of the receiver. I know when I
actually saw a superregen schematic with a separate quench oscillator (it
was around that time, but I hadn't heard of what Kitchin was doing), it
made sense that the wide bandwidth came from "modulating" the regen
receiver. But most of the books by that time had only the vaguest of
information.

Over the years, there were endless articles about getting rid of the wide
bandwidth, but it seems like everyone was looking in the wrong place.
High-Q circuits for 2meter receivers, apparently that helped, but still
not real narrow bandwidth. There was some article in Ham Radio in the
sixties about adding a diode to kill the hangover, that seemed to help,
but still just an incremental change. But in one of Kitchin's articles,
he claims to be able to receive 2m FM (ie narrow deviation) with the
superregen described.

Michael