Automatic Gain Control

Radio signals vary enormously in strength. A station 5 km away on 40 meters might arrive at your antenna at −43 dBm — about 5 millivolts — while a DX station 10,000 km away on the same band might deliver only −113 dBm, less than two microvolts. That is a difference of 70 dB — a voltage ratio of more than 3,000 to one. On a busy band, the ratio between the weakest signal you want to hear and the strongest one can easily reach 90 dB or more. Without a system to handle this automatically, your receiver would either blast your ears on the local station or be completely deaf to the DX contact. Automatic gain control — AGC — solves this problem by continuously measuring signal level and adjusting IF amplifier gain in real time, keeping the audio output roughly constant regardless of how strong or weak the incoming signal is.

Why AGC Is Needed

To appreciate why AGC matters, consider what a receiver without it would do. If you set the IF gain high enough to hear a weak DX station arriving at −113 dBm, that same gain applied to a local station at −43 dBm produces a signal 70 dB stronger at the audio output — roughly 3,000 times higher voltage. The audio amplifier saturates, the speaker clips, and the signal becomes an unintelligible roar. If you then turn the gain down to cope with the local station, the DX contact disappears into the noise. In practice, you would be constantly turning the RF gain knob up and down as you tuned the band — completely impractical.

The table below shows typical received signal levels for a 100 W SSB station across a range of distances and conditions. Notice how the span from a loud local contact to a weak DX contact is around 90 dB — a voltage ratio of approximately 30,000 to one:

A well-designed AGC system bridges this entire range automatically, maintaining consistent audio volume regardless of where on that 90 dB span the current signal sits. The operator does not touch any gain control — the receiver manages it continuously and invisibly. This is not merely convenient; for rapid band conditions, multi-path fading, and the natural variation as stations come and go on a frequency, it is essential.

There is a second, subtler benefit. Operators call in sequence on a pile-up, one after another at different distances. Each one has a different signal strength. With AGC, the operator hears each callsign at roughly the same volume, making it easier to copy the information. Without AGC, the operator would have to mentally compensate for vast differences in audio level between callers.

The AGC Feedback Loop

AGC is a classic negative feedback control system, identical in principle to a thermostat controlling room temperature. A thermostat measures actual temperature, compares it to the setpoint, and adjusts heating or cooling until the measured temperature matches the target. AGC measures actual IF signal level, compares it to an internal reference, and adjusts IF gain until the output level stays where it should be.

The loop operates in three continuous steps:

1. Detect: An AGC detector — typically a precision rectifier followed by an RC filter — taps the IF amplifier chain and produces a DC voltage proportional to the IF signal level. This DC voltage is the feedback signal that represents how strong the received signal currently is.
2. Compare: The detected DC voltage is compared to an internal reference level. The stronger the signal, the higher the AGC control voltage rises above the reference. Below the reference (weak or no signal), the AGC control voltage sits at a low level and the receiver runs at maximum gain.
3. Correct: The AGC control voltage is fed to the gain-controlled stages of the receiver — most commonly the IF amplifiers, and on strong signals also the RF preamplifier. This control voltage reduces the gain of those stages in proportion to how far the signal level exceeds the reference, keeping the output level constant.

Because the loop acts continuously, the receiver tracks fading signals automatically. When a signal fades by 10 dB due to ionospheric variation, the AGC voltage falls slightly, gain increases by 10 dB, and the audio output barely changes. When propagation suddenly improves and the signal jumps 10 dB, the AGC voltage rises, gain reduces, and again the audio stays nearly constant. The operator hears smooth, consistent audio while the AGC is doing the work invisibly in the background.

The gain-controlled stages use variable-gain amplifiers whose gain is set by a control voltage. Common hardware implementations include:

• PIN diode attenuator in the RF path ahead of the first mixer — varying the diode bias current adjusts its RF resistance, controlling how much signal reaches the mixer
• Dual-gate MOSFET IF amplifier where gate 2 is the gain control port — applying a control voltage to gate 2 adjusts the transconductance and hence the gain without disturbing the signal on gate 1
• Gilbert cell variable-gain amplifier — an integrated circuit structure where a tail current controls the differential pair gain, widely used in receiver ICs
• Voltage-controlled attenuator using varactor diodes or a digitally controlled step attenuator

The AGC Control Chain

The block diagram below shows where each component of the AGC system sits within the superheterodyne receiver. Understanding each block helps you diagnose AGC problems and understand why different design choices affect performance.

The AGC feedback loop. The AGC detector taps the IF signal after the selectivity filter, producing a DC control voltage that reduces IF — and optionally RF — gain as signal strength increases. The S-meter is driven from this same control voltage.

View Larger

The AGC Tap Point

The AGC detector is tapped from the IF amplifier chain, and the location of that tap matters significantly. If it is placed before the selectivity filter, a strong adjacent-channel signal could trigger the AGC and reduce the gain on the desired weak signal — making the receiver apparently less sensitive just because a strong station nearby is on the band. The correct tap point is after the main selectivity filter so the AGC only responds to signals that have passed through the filter passband. Adjacent-channel signals, having been attenuated by the filter, contribute very little to the AGC control voltage and do not desensitize the receiver to the desired signal.

Peak Detection vs. Average Detection

The AGC detector can be designed to respond to either the peak level or the average level of the IF signal. Peak-detecting AGC responds to the instantaneous peak envelope and is fast to react; it is commonly used in CW receivers where each character pulse represents a brief burst of energy. Average-detecting AGC responds to the mean amplitude over a time window; it is better suited to AM and SSB where the instantaneous envelope varies widely with the audio modulation. Using a peak detector with a slow AM signal can cause the AGC to ride up on loud audio peaks and then drag down during quiet passages, creating audible distortion. Most modern receivers use a combination — peak detection for fast attack, average detection for steady-state level control.

Where the AGC Controls Gain

In a well-designed receiver, the AGC applies gain reduction to the IF stages first, and only extends back to the RF preamplifier for the strongest signals. The reason is noise figure: the RF preamplifier is the first stage and has the greatest influence on the receiver's overall noise figure. If the AGC reduces RF preamplifier gain early, it worsens the noise figure and reduces sensitivity for weak signals that may coexist with a strong signal on an adjacent channel. By keeping the RF gain at maximum for as long as possible and managing dynamic range with IF stage gain reduction, the receiver maintains good sensitivity while still protecting downstream stages from overload.

Attack Time and Decay Time

Two time constants govern how quickly the AGC feedback loop responds to changes in signal level. These time constants are set by the RC filter in the AGC detector circuit — a larger capacitor produces a longer time constant (slower response), and a smaller capacitor produces a shorter one (faster response). The two constants are called attack time and decay time (also called release time).

The attack time must be long enough that normal audio amplitude variations within a speech syllable do not trigger gain reduction, but short enough that a genuinely strong signal is controlled before the audio amplifier overloads. A typical prescription for SSB and CW is an attack time in the range 10–30 ms. This is fast enough to prevent an audio blast when a new station appears, but not so fast that the AGC chases individual audio peaks.

The decay time has a more complex effect. If it is too short, the AGC releases gain quickly during the natural pauses between words in an SSB transmission. During those pauses, the IF signal drops to noise level, the AGC releases gain to maximum, and the background noise rushes up loudly — a very noticeable and fatiguing artifact called noise rushing or AGC breathing. If the decay time is too long, a very strong station that disappears from the frequency leaves the receiver at greatly reduced gain for several seconds, during which any weaker signal that appears is temporarily inaudible. A decay time of 500 ms to 1 second is a common compromise for SSB voice reception.

AGC timing diagram. When a strong signal appears at t = 0.5 s, the AGC control voltage rises quickly (fast attack, ~20 ms), reducing gain and holding audio level constant. When the signal disappears at t = 2.0 s, the control voltage decays slowly (slow decay, ~800 ms), allowing gain to recover gradually. A brief audio dip is visible as gain recovers.

View Larger

AGC Threshold and Delayed AGC

The AGC threshold is the minimum signal level at which the AGC begins to reduce gain. Below the threshold, the receiver operates at maximum gain — the AGC is inactive, and the full sensitivity of the receiver is available for weak signals. Above the threshold, the AGC activates and reduces gain proportionally to keep the output level constant.

Setting the threshold correctly is important. If the threshold is too low — say, at the noise floor level — the AGC activates on noise itself, continuously reducing gain and degrading weak-signal sensitivity. On a quiet band with a low noise floor, this means the receiver is actually fighting itself, applying gain reduction when no useful signal is present. If the threshold is too high, moderately strong signals get through without gain control, causing distortion before the AGC engages. A good threshold setting is just above the noise floor of the band being used, allowing full sensitivity for weak signals while controlling gain for medium and strong ones.

Delayed AGC is a refinement of this idea. Instead of applying a smoothly varying control voltage as soon as the signal exceeds the threshold, the delayed AGC applies no gain reduction at all below the threshold, and then applies a stepped or gradual control above it. The "delay" refers to the fact that the gain control action is delayed until the signal is strong enough to warrant it — it does not mean a time delay. Below the threshold, the receiver runs at maximum gain exactly as if the AGC were switched off. Above the threshold, gain reduction is applied proportionally. This maximizes sensitivity for weak signals while still providing excellent control for strong ones. Delayed AGC is the standard design in quality HF receivers.

Hang AGC for SSB and CW

Standard slow-decay AGC does a reasonable job on SSB, but there is a fundamental tension: the decay time must be long enough to avoid noise rushing between words, but not so long that the receiver stays deaf after a strong station disappears. Hang AGC resolves this tension with a different approach. Instead of decaying continuously at a fixed rate after the signal drops, the hang AGC holds (hangs) the control voltage at its peak value for a fixed hold time — typically 0.5 to 1.5 seconds — and then releases gain quickly rather than slowly.

This matches how SSB speech actually works. A station transmitting on SSB produces signal during each syllable and gaps of near-silence between words and between phrases. The hang time is set to be longer than a typical inter-word pause (around 0.3–0.5 seconds for normal speech) but shorter than a typical transmission pause between sentences. This means:

• During the transmission, the AGC holds its control voltage steady through all the natural pauses between words — no noise rushing, no breathing
• The audio level stays consistent throughout the entire transmission
• When the station finishes transmitting and a longer pause occurs, the hold time expires and gain returns quickly to full — ready for the next station to respond

Hang AGC vs standard decay AGC receiving an SSB transmission with natural speech pauses. Standard decay AGC allows gain to partially recover during inter-word pauses, causing background noise to rush up (left). Hang AGC holds the control voltage steady throughout the transmission, then releases quickly when the station finishes transmitting (right).

View Larger

Hang AGC is particularly valuable for CW reception. CW characters are short bursts separated by gaps. With standard slow-decay AGC and fast CW, the AGC might partially release between every character, causing the noise floor to fluctuate in rhythm with the keying. With hang AGC set to hold for the length of a character space or longer, the AGC locks onto the signal level and holds it steady for the entire over, releasing only during the longer gaps between exchanges.

Many modern transceivers offer selectable AGC modes: FAST (short decay of 100–200 ms, useful for rapid CW or AM monitoring), SLOW (long decay of 1–2 s, good for SSB in stable propagation), and AUTO or HANG (hold-and-release, optimal for SSB and CW). Some high-end transceivers allow the operator to independently adjust the attack time, hold time, and decay time through menus, giving complete control over AGC behavior for different operating conditions.

The S-Meter

The S-meter (signal strength meter) that appears on every receiver — as a physical needle meter on older equipment or a bar graph display on modern transceivers — is driven directly from the AGC control voltage. As the received signal increases and the AGC control voltage rises to reduce gain, that same rising voltage deflects the S-meter upward. The S-meter therefore reads the strength of the signal that is being received and controlled by the AGC, displayed on an agreed scale of S-units.

The IARU (International Amateur Radio Union) standard defines S9 as a received signal level of −93 dBm into a 50-ohm antenna input on HF, corresponding to approximately 50 microvolts. Each S-unit represents 6 dB, which is a factor of 2 in voltage. Above S9, signal strength is reported in dB over S9 — S9+10 dB, S9+20 dB, S9+40 dB, and so on:

A practical point worth understanding: S9 to S9+10 dB is a 10 dB change — that is a 10-fold increase in power, which is a substantial difference. Moving from S9 to S9+40 dB is a 40 dB change — a 10,000-fold increase in power. When you hear operators say “you are 20 over 9” they mean the received signal is 20 dB above the S9 reference, or roughly 100 times the power of an S9 signal. These are enormous differences, which is why AGC must handle such a wide range.

The S-meter only gives a meaningful reading when the AGC is active and controlling gain. If the AGC is switched off or the signal is below the AGC threshold (full gain, no control action), the S-meter may not deflect normally. Also note that the S-meter accuracy depends on the AGC detector linearity and the quality of the receiver's calibration — consumer-grade receivers often show S9 several dB higher or lower than the IARU standard. Use S-meter readings as a relative guide, not an absolute measurement. For precision signal level measurements, use an instrument such as a calibrated signal generator and measure receiver sensitivity separately.

AGC in Digital and SDR Receivers

Software-defined radios and modern digital receivers implement AGC quite differently from analog superheterodyne receivers, but the same fundamental principle applies: measure signal level, compare to reference, adjust gain. The hardware and software details differ significantly.

The most critical point in SDR design is the ADC overload problem. An analog-to-digital converter has a fixed full-scale input voltage. If the RF or IF signal exceeds the ADC's full-scale range, the converter clips — the top of the waveform is sheared off, creating severe harmonic distortion and intermodulation products that corrupt not just the strong signal but all weaker signals in the passband simultaneously. Unlike an analog IF stage that clips gradually and somewhat gracefully, an ADC clips abruptly and catastrophically. A hardware RF attenuator or automatic gain stage must therefore prevent the ADC from ever overloading, even before the digital AGC has time to respond.

Many SDR receivers use a two-stage AGC approach that manages this correctly:

1. Hardware stage — a step attenuator, switchable RF preamplifier, or variable-gain LNA controls the analog signal level into the ADC. This stage responds quickly to large sudden signals and protects the ADC. It is controlled either by a hardware circuit or by software commands to the RF hardware.
2. Digital stage — DSP scales the digital samples after the ADC to the target amplitude within the available numerical dynamic range. This stage provides fine adjustment, handles slow fading, and implements the precise attack and decay time constants that give good audio quality.

Because SDR AGC is entirely in software, it can be far more sophisticated than analog AGC. An SDR can apply different AGC settings to different decoded signals simultaneously, can use predictive algorithms that anticipate signal level changes based on recent history, and can implement frequency-selective AGC that prevents a strong CW signal from reducing gain across an entire SSB passband. These capabilities are simply impractical in analog hardware but trivial in software.

AGC Parameter Summary