Understanding AGC in Communication Receivers

[erece]

What is the purpose of delayed automatic gain controller (AGC) in communication receiver ?

 

[skeptic2]

Delayed AGC is a bit of a misnomer. The AGC is delayed, not in time but in voltage. All it means is that the signal level must reach a specified value before an AGC voltage is developed. The purpose is to prevent the AGC voltage from reducing the receiver gain for very weak signals.

 

[Bob S]

What is the purpose of delayed automatic gain controller (AGC) in communication receiver ?

See figure 1 in this paper.

http://www.eecg.utoronto.ca/~kphang/papers/2001/martin_AGC.pdf

The low pass filter, with an integration time constant of the order of 1 Hz, controls the forward gain of the variable gain amplifier. The time constant introduces the AGC delay.

Bob S

 

[skeptic2]

Bob, I still don't agree. The only reference to delayed AGC in your reference is in the second sentence on page VI which says, "AGC systems that include a reference voltage inside the control loop are referred as delayed AGC." Unless I am missing something, I don't believe the addition of a reference voltage will cause a time delay in the AGC action. Its purpose is to disable the AGC at low signal levels.

I am not aware of any AGC systems in which a time delayed AGC is desired. True, in AM, SSB or QAM, the desired amplitude modulation needs to be filtered out of the AGC and that results in some time delay, but the delay introduced by the filtering is not desired. Time delayed AGC results in overshoot and if severe enough, oscillation.

 

[Bob S]

Bob, I still don't agree. The only reference to delayed AGC in your reference is in the second sentence on page VI which says, "AGC systems that include a reference voltage inside the control loop are referred as delayed AGC." Unless I am missing something, I don't believe the addition of a reference voltage will cause a time delay in the AGC action. Its purpose is to disable the AGC at low signal levels.

I believe the AGC is needed at all input signal levels to maintain a constant audio output signal level.

I am not aware of any AGC systems in which a time delayed AGC is desired. True, in AM, SSB or QAM, the desired amplitude modulation needs to be filtered out of the AGC and that results in some time delay, but the delay introduced by the filtering is not desired. Time delayed AGC results in overshoot and if severe enough, oscillation.

The AGC signal control voltage is supposed to be an average over a second or two. Causality requires that averaging produces a delay. Anything faster than this will distort the audio signal. This is why there is an RC low-pass filter in the feedback loop. Too much gain in the feedback loop will make it unstable.
Bob S

 

[skeptic2]

The AGC signal control voltage is supposed to be an average over a second or two. Causality requires that averaging produces a delay. Anything faster than this will distort the audio signal. This is why there is an RC low-pass filter in the feedback loop. Too much gain in the feedback loop will make it unstable.
Bob S

For amplitude modulated signals this is true and many FM receivers don't use AGC, just limiting. However I once designed a nearly instantaneous AGC to be used in a TDMA receiver to compensate for changes in signal level in successive time slots. That AGC used a sample and hold at the peak of each cycle of the signal in the IF.