From: "Tom" on Wed,May 25 2005 10:30 am

I agree with you and slowing the attack is the only way that I have
been able to approach stable operation with a fast release. But 20ms or
longer attack runs counter to what I understand to be the objective -
an attack speed of less than 13 ms and ideally about 1 ms. So, unless I
have this wrong, how do other receivers accomplish similar speeds
without self-oscillation?

The way my circuit operates (I think) is as follows (I'd be happy to
send a schematic to anyone who is interested) :
a) assume an impulse of signal of duration very much longer than the
attack time
b) the rectified signal is filtered of RF by a series-parallel R-C
attack network whose adjustable output feeds an emitter follower
c) the emitter follower pumps current as a low resistance source into
the release R-C network so the attack is not greatly slowed - its
output feeds the AGC driver amp
d) at some point, equilibrium should be reached - the current flow
through the release resistor and AGC driver base should equal the flow
though the emitter follower - but maybe the emitter follower pinches
off and that could be a cause of instability?
e) the signal drops, the attack network discharges at attack speed and
shuts off the emitter follower, so the release capacitor discharges
through its parallel R at release speed, the voltage to the AGC driver
falls so the AGC bias rises at roughly release speed to increase RF/IF
gain.

Having written that out, I have an idea or two I will try.

Having encountered a similar problem many years ago, I'll offer
this as a suggestion: Analyze the behavior of the total signal
amplification chain at LOW frequencies, not at the RF or IF
carrier. Know the control characteristics of the AGC voltage
input to the amplifier versus the total amount of gain of the
receiver chain. Approach the whole receiver AGC action as a
low-frequency servo loop (which is what the AGC actually does).
Think servo control systems theory.

Control systems theory is a rather abstract thing and there
probably will be no sudden bright light of understanding
switched on, but here's a bit of that: The AGC loop action
works by BOTH magnitude and phase at low frequencies.
"Nyquist" and "Bode" plots are helpful there, even though
both of those subjects are also rather abstract. In general,
if the AGC control action results in instability or even motor-
boating, the overall receiver gain - related to the control
voltage range - is too high. Adding a voltage divider at the
low-pass R-C filter of the AGC voltage input will demonstrate
that. Also, the low-frequency phase shifting in the AGC
voltage "decoupling" can upset the phase versus magnitude of
the control voltage. Note: Vacuum tube or FET RF/IF
controlled amplifiers probably use such R-C decoupling, working
only on AGC voltage; other amplifier types might have some
other form of R-C filtering at low frequencies. That low-
frequency magnitude AND phase relationship is important for
total loop stability.

What has to be considered in the AGC loop is the response
through all the decoupling newtorks between the ACG control
source and the controlled device(s). For a "non-linear"
loop (separate attack and decay times) that analysis will be
difficult. It is much easier to analyze with a Spice
simulation that has the capability to model a controlled-gain
amplifier. The whole loop at low frequencies can be modelled
that way. In starting that, forget the RF and IF components
and consider only the amplifications at low frequencies; the
source of the AGC control (detector output) may have to be
modelled slightly differently in that the detector is, in
effect, similar to a power supply rectifier. If that model
is tweaked to be stable with sudden transitions on its input,
then it will be stable at RF and IF.