From: Andrea Baldoni on Thurs, Aug 17 2006 1:34 pm


Hello.
I'm wondering about a thing...

When the AGC reduces the gain of an amplifier, composed by FETs (DG or SG)
or BJT, the SNR remains constant?

Maybe, maybe not...

Or the performance of the amplifier may degrade/upgrade?

Maybe, maybe not...

I'm researching about the matter and I just read that, in a BJT for instance,
emitter current is inversely proportional to the noise. So, if AGC reduces
the gain (so current), SNR degrade?

Not necessarily true. Noise, true natural noise, in a bipolar
junction transistor is dependent on MANY things besides the
emitter current. Foremost of all is the source impedance
of the BJT's driving source: SNR will increase on both sides
of the measured least-SNR current in some transistors. If you
look at the characteristics/datasheet information on several
hundred different transistors you will begin to get the idea
of how many factors go into noise generation within the
transistor. :-)

Most complex Gilbert-cell ICs (Motorola MC1590, MC1350 as
examples) will control gain by "stealing" current of a
differential-input, common-emitter stage. While the example
ICs can control gain over a 60 db dynamic range, the worst
noise generated inside those bipolar-transistor architecture
ICs occurs at minimum gain control input (maximum
amplification).

At higher and higher input signal levels, there would be more
AGC control action but the SNR would also be higher and higher.

The question arises from a thing I just noted with a HF receiver... disabling
AGC reduces (slightly) the noise (at least in FM reception)...

There isn't much FM on HF. What there is would be in
narrow-band Data mode signals. Some of that Data is a
combination of AM and PM similar to a wireline modem's
modulation.

Disabling the AGC would put the overall gain of the receiver
at maximum. Naturally there would APPEAR to be "more noise"
since the receiver would be picking up ALL noise available at
the antenna input.

Maybe just be a side effect, like noisy gain control signal...

That's also possible, but unlikely there in my opinion.

What is needed in an investigation of this is a reasonably-
well-calibrated signal generator with a calibrated attenuator.
On AM you could simply increase the signal generator output
until the total signal plus noise reaches a certain level at
the detector, usually 10 db above input of just noise alone
(common definition of SNR).