[Elecraft] AGC Independent S-Meter?

[Ron D'Eau Claire]

Dr Megacycle wrote: 
Well, I have late 1950s Hammarlund HQ170 whose S-meter works just  
fine with the AGC/AVC turned on or off. As far as I have been able to  
ascertain, the S-meter functions identically with or without the AGC/ 
AVC turned on. It was designed to do so. 

---------------------------------

The HQ170 S-meter rectifies the AVC signal separately in one diode section
of a 6BV8 (that's a tube, folks! <G>), but it will not provide the same
readings with the AVC turned off! In fact, in the HQ170 manual says, "...the
'S' meter circuit is connected to the separate AVC diode section of V8
(6BV8) and gives an indication of all types of signals in all positions of
the AVC, however the 'S' meter calibration is valid ONLY WITH AVC POSITIONS
SLOW, MEDIUM OR FAST and not in OFF position, although it will indicate and
may be usable in MANUAL position."  (Page 9 of the HQ-170 Communications
Receiver Instructions and  Service Information manual. The emphasis is in
the manual text.) 

Let's take a look at how an S-meter works, starting with some background. 

It all started long ago with the superheterodyne invented by Col. Armstrong.
That allowed a lot of amplification in a receiver that was easy to tune and
without the howls and squeals that, even today, are used to signify one is
tuning in an "old time" receiver! That basic superhetrodyne format is still
the standard today in almost all receivers, including those in the Elecraft
rigs. 

With the amplification the superhetrodyne provided came a nuisance: when
listening to a weak station, if a strong station came on frequency it'd
blast your ears! 

For CW signals the solution was easy. One added a "hard limiter" to the
audio channel. If a really strong signal was encountered without warning, it
was clipped of at some preset maximum volume (below the threshold of pain,
hopefully). Most operators used the limiter at all times in case a huge
signal came on frequency. 

The problem with a hard limiter is that is clips off the signal - turns a
sine wave into an almost-square wave. The fact that the end result sounded
like a buzz-saw was of no consequence. The current interest in soft, pure,
sine wave tones when listening to CW is a recent pursuit that has many old
timers chuckling. Everyone listened to CW signals that sounded a lot harsher
than the sidetone from the stock K2 and it was (and still is) "music to
their ears". 

But phone transmissions were a whole different story. Distorting voice (or
music) with a hard limiter was a serious problem. A means to prevent
blasting listeners out of their chairs without distorting the audio was
needed. Automatic Gain Control (AGC) a.k.a Automatic Volume Control (AVC)
was the answer. AVC (or AGC) automatically controls the gain of some of the
stages of RF amplification in the receiver to turn down the signal and avoid
overload and "blasting". It works like this. Phone signals were Amplitude
Modulated (AM). That is, they consist of a steady carrier with sidebands
containing the audio modulation. Two rectifiers are used in the receiver.
One recovers the audio from the sidebands and the other, the AGC detector,
produces a d-c voltage proportional to the strength of the carrier. The d-c
voltage it produces is used to control the amplification of the stages in
the receiver ahead of the detector. The stronger the signal, the more
voltage produced, the more voltage produced, the more the amplification of
the stages ahead of the detector are turned down. 

That produced a much-reduced change in loudness in the speaker or phones
when tuning from a weak to a loud signal. The AGC circuit quickly became
standard in virtually every superhet used to receive AM phone signals, from
the console radio in the living room of the 1920's and 30's to the car radio
of the 1940's to today's radios. A huge range of signals could be received
with only nominal changes in volume - something easily controlled with the
audio gain control. 

At some point some smart guy (or gal) realized that the AVC voltage changes
in proportion to the strength of the incoming signal. That's the whole idea
of the AVC! The stronger the signal, the greater the voltage! If we measure
that voltage, we can show on a meter the relative strengths of various
signals. The "S-Meter" was born!

But AGC or AVC was only useful for AM phone reception. For CW reception we
needed a beat-frequency-oscillator (BFO). The BFO is almost on the same
frequency as the signal at the detector in order to produce the audio beat
frequency we hear. The BFO is a huge, locally-generated signal, compared to
the CW signal. It was impossible to keep the BFO out of the AVC detector.
The relatively huge BFO signal made the AVC system react as if it was tuned
into a very strong signal at all times, and so the AVC turned the receiver
gain to minimum and kept it there. So, for decades, superhetrodyne
communications receivers had a switch to turn the AVC off for CW reception
and we continued to use the manual RF gain control and a hard limiter to
protect our ears. CW operators never looked at an "S-meter". 

But some tinkerers wanted to have AVC for CW too. That interest grew as AM
was replaced by SSB. SSB, like CW, requires a strong local BFO signal, so
even though it was "phone" the AVC in the receivers couldn't be used. The
trick was to rectify a sample of the signal to see how strong it was without
letting the BFO get into the AVC. Two ways were developed. 

One was to rectify a sample of the audio signal *after* the detector. That
produced a d-c AVC voltage proportional to the signal strength. It was
called, for obvious reasons, "Audio AVC" (or AGC). It did well for SSB but
had a bothersome drawback for CW. Remember, the d-c AVC voltage is produced
by simply rectifying a sample of the signal. Rectifying an audio tone of,
say, 600 Hz, it takes much longer for the AVC voltage to develop than when
rectifying, say, and I.F. of 4 MHz. That caused a slight delay in the
"attack" or time to turn down the volume when a strong signal appeared on
frequency, resulting in an annoying "pop" in the speaker or phones. One
approach was to let the AVC voltage return to high gain only slowly after
once detecting a strong signal, so that if it was a CW signal or an SSB
signal with a pause in the speech, the gain would not return to full volume
before the next code element or SSB word came through. That helped, but it
meant the receiver was "deaf" to weak signals for a while after the strong
signal was silent. Still, audio AVC is simple and effective and often used
today. The Elecraft KX1, for example, uses audio AVC.  

The other approach was to sample the signal well before the detector and
BFO, where the signal could be isolated and avoid the BFO sneaking in. One
common way to do that was to have two mixers feeding two, separate I.F.
amplifiers operating on two different frequencies: one was for the signal
and the other for the AVC. That way, the AVC detector was tuned to a
frequency far removed from the BFO so the BFO won't interfere with it. That
allows the desired fast attack time since the AVC voltage is produced by
rectifying a signal at radio frequencies instead of audio frequencies. The
Elecraft K2 uses this approach. 

No matter how AVC voltage is developed, it's the AVC (or AGC) voltage that
drives the S-meter.

It's possible to use a receiver to measure signal strength that doesn't have
AVC. We might rectify the audio output and look at it on a meter to see
changes in the signal strength, or we might sample the I.F. and rectify it.
But the usefulness of such readings without the extended dynamic range
provided by an active AVC is very limited for on-air communications
purposes, unless one operates the RF gain control manually. If we change the
RF gain control, we lose all sense of calibration of the S-meter. That's why
you don't see S-meters offered in communications receivers that continue to
work when the AVC is turned off. 

Ron AC7AC