One of the first things we learn upon entering the world of electronics is how to design an op-amp circuit with a specified gain. It’s not particularly difficult and, even after we become familiar with all the nuances and imperfections associated with amplifier circuits, we can still confidently design systems that require an output signal that is equal to the input signal multiplied by a fixed gain.
But what happens when this entire paradigm falls apart? What can we do when the fixed parameter is not the gain of the amplifier but the magnitude of the output? A fixed gain can produce a constant output amplitude when the input amplitude is known and unchanging, but this is not always the case and, furthermore, sometimes the input amplitude is highly variable.
Closing the Loop
The solution here is something called automatic gain control, abbreviated AGC. We can intuitively conclude that there really is no way to achieve this in an open-loop system—the amplifier circuitry must have knowledge of the output amplitude in order to properly adjust the gain. It follows, then, that AGC requires feedback. It also (unsurprisingly) requires a variable-gain amplifier (VGA).
The following is a (very) basic architecture for an AGC system:
The output of the VGA is fed not only to the next device in the signal chain but also to measurement circuitry that determines the amplitude of the output and adjusts the gain accordingly. The amplitude measurement is performed by the detector block, and different types of detectors are used—the four standard detector types are envelope (or rectifier), square-law, true-RMS, and logarithmic.
Adapting to Change
Like other closed-loop feedback systems, an AGC can “lock onto” the input signal such that gradual changes in input amplitude will have minimal effect on the output. However, an AGC cannot instantaneously adapt to rapid changes; actually, extremely fast response time is not desirable because this would make the AGC circuit overly sensitive to noise or to intentional variations in the amplitude of the input signal (i.e., amplitude modulation).
The term “attack time” refers to an AGC circuit’s response to increases in input amplitude, and “decay time” refers to its response to decreases in input amplitude. The following plot from Analog Devices compares the attack and decay behavior for the four standard detector types (for some reason, “LINBNV” is the abbreviation for an envelope detector).
Image courtesy of Analog Devices.
As you can see, the system’s response requirements should be taken into consideration when choosing the detector type.
AGC for RF Rx
AGC is a critical aspect of RF receiver design. The energy density of electromagnetic radiation decreases with the square of distance. Thus, the RF signal strength at the receiver varies drastically depending on how close the receiver is to the transmitter. AGC ensures that the received signal is consistently amplified to a level that allows for efficient processing by the demodulation circuitry.
In this age of highly integrated, expertly designed, widely available analog and mixed-signal ICs, it’s not likely that you will ever need (or want) to design your own AGC system (which is by no means a simple process). However, it’s good to be familiar with the basic techniques and concepts. If you’re interested, an abundance of additional information is available in a design tutorial from Analog Devices.