Amplitude modulation (AM) is a technique where the strength (amplitude) of a high-frequency carrier wave is varied in direct proportion to the instantaneous voltage of a lower-frequency information signal. When you apply AM to a circuit, you are fundamentally shifting a low-frequency baseband signal (like audio or sensor data) up to a high-frequency passband so it can be radiated by an antenna or transmitted through a specific bandpass filter, which requires multiplying two signals together rather than just adding them.

The Math on the Bench: A Worked Numeric Example

To understand AM, you have to look at the power distribution. A common mistake homebrew builders make is assuming a 100W transmitter outputs 100W total when modulated. Let us run the numbers for a standard bench setup.

Assume we have an RF oscillator generating a 1.000 MHz (1000 kHz) carrier with an unmodulated carrier power ($P_c$) of 100 Watts. We inject a 5 kHz sine wave audio tone into the modulator with a modulation index ($m$) of 0.8 (80% modulation depth).

  • Upper Sideband (USB) Frequency: $f_c + f_m = 1000 + 5 = 1005$ kHz
  • Lower Sideband (LSB) Frequency: $f_c - f_m = 1000 - 5 = 995$ kHz
  • Total Bandwidth: $2 imes f_m = 10$ kHz

The total transmitted power ($P_t$) is calculated using the formula $P_t = P_c imes (1 + rac{m^2}{2})$. Plugging in our numbers: $P_t = 100 imes (1 + rac{0.64}{2}) = 100 imes 1.32 = 132 Watts$. The carrier still accounts for 100W, but each sideband now carries 16W ($P_c imes rac{m^2}{4}$). This means your final RF amplifier and power supply must be sized to handle the 132W average, and significantly higher peak voltages, without saturating.

What Amplitude Modulation AM Changes in a Real Circuit

In a purely linear circuit, if you wire an audio amplifier and an RF oscillator in parallel, you just get a composite waveform—the two frequencies simply add together. This does not create modulation; it creates a messy mix that a bandpass filter will just strip apart.

The Non-Linear Requirement: To actually achieve amplitude modulation, the circuit must force the two signals through a non-linear component. This is typically a diode or a bipolar junction transistor (BJT) biased near cutoff or saturation. The non-linear transfer curve mathematically multiplies the carrier and the audio signal, generating the sum and difference frequencies (the sidebands) required for AM.

Following the non-linear mixing stage, the circuit requires a tuned LC tank (an inductor and capacitor in parallel). This tank circuit acts as a bandpass filter, shorting out the unwanted harmonic mixing products and leaving only the carrier and the first-order sidebands to be fed to the antenna or the next amplifier stage.

Where You Meet This in Practice

While commercial music broadcasting has largely moved to FM or digital, AM remains critical in several high-reliability and industrial applications:

  1. Aviation VHF Communications (118.000 - 136.975 MHz): Air traffic control uses AM, not FM. FM suffers from the 'capture effect,' where a stronger signal completely drowns out a weaker one on the same frequency. In AM, if two pilots transmit simultaneously, the controller hears a heterodyne squeal and can tell that two stations are stepping on each other, which is a vital safety feature in crowded airspace.
  2. Industrial Displacement Sensors: Linear Variable Differential Transformers (LVDTs) and eddy current probes use AM. A high-frequency AC excitation signal (often 1 kHz to 10 kHz) is fed into a coil. As a metallic target moves, it changes the magnetic coupling, which physically modulates the amplitude of the carrier. The receiver circuit then demodulates this AM signal to determine exact physical displacement down to the micron.
  3. Amateur Radio (SSB): Single Sideband (SSB) voice transmission, heavily used by ham operators on HF bands, is derived directly from AM. The transmitter generates a full AM signal, then uses sharp crystal or mechanical filters to strip out the carrier and one of the sidebands, saving 75% of the transmitted power and halving the bandwidth.

Bench Scenario: Overmodulation and the Burned Final Amplifier

Theory is clean, but the workbench is unforgiving. Here is a real-world scenario demonstrating what happens when AM math is ignored during hardware setup.

The Setup: I was building a 10-meter (28 MHz) low-power AM transmitter for a ham radio project. The final RF amplifier stage used a classic 2SC1969 RF power transistor, biased for a 10W carrier output into a 50-ohm dummy load. The audio modulator was a simple op-amp circuit driving the base of the transistor.

The Numbers: The 2SC1969 has a maximum collector-emitter voltage ($V_{CEO}$) rating of 35V, and the circuit was running on a 13.8V DC supply. At 100% modulation ($m=1.0$), the Peak Envelope Power (PEP) is four times the carrier power ($10W imes 4 = 40W$ PEP), and the peak RF voltage swing at the collector reaches roughly twice the DC supply voltage (approx 27.6V peak), safely under the 35V limit.

The Outcome: During testing, I cranked the audio gain potentiometer too high to get a 'louder' signal, pushing the modulation index to $m=1.3$ (130% overmodulation). The transmitter sounded distorted on the receiver, and a spectrum analyzer showed massive 'splatter' (adjacent channel interference).

What Went Wrong: The audio peaks drove the base of the 2SC1969 far too hard. During the negative peaks of the audio envelope, the modulator tried to drive the RF amplitude below zero. Because a transistor cannot output negative RF voltage, it slammed hard into cutoff. This abrupt clipping caused massive harmonic generation and violent voltage spikes due to parasitic inductance in the collector lead. The collector voltage swung past the 35V $V_{CEO}$ breakdown limit, causing avalanche breakdown. The silicon junction punched through, shorting the collector to the emitter, instantly turning the transistor into a dead short that blew the 5A inline fuse and destroyed the part.

Common Confusions: AM vs. FM and Power Metrics

When discussing amplitude modulation AM, two specific confusions trip up hobbyists and junior technicians:

  • AM vs. FM (Frequency Modulation): People often assume FM is just 'better AM.' In reality, they manipulate entirely different properties of the wave. AM varies the envelope height while keeping the zero-crossings (frequency) perfectly spaced. FM varies the spacing of the zero-crossings while keeping the envelope height perfectly flat. FM requires limiter circuits in the receiver to strip off amplitude noise, whereas AM receivers use simple envelope detectors (a diode and a capacitor).
  • Carrier Power vs. PEP: A transmitter rated for '100W AM' means 100W of unmodulated carrier power. As shown in our math example, 100% modulation requires the final amplifier to handle 400W of Peak Envelope Power (PEP). Sizing a heatsink or power supply based only on the 100W carrier figure will result in thermal runaway or voltage sag the moment you key the microphone.

Frequently Asked Questions

Why does AM sound noisier than FM?

Most environmental and electrical noise (lightning, motors, switching power supplies) manifests as sudden spikes in voltage amplitude. Because an AM receiver's detector is specifically designed to track changes in amplitude, it faithfully reproduces this noise as loud static. FM receivers track frequency changes and use internal limiters to clip off amplitude spikes, rendering them largely immune to this type of impulse noise.

Is AM still used in modern digital systems?

Yes, but in a more complex form called Quadrature Amplitude Modulation (QAM). Instead of just modulating the amplitude of a single carrier, QAM modulates the amplitude of two carriers that are 90 degrees out of phase with each other (the I and Q signals). This allows digital systems like Wi-Fi, cable modems, and 5G to pack multiple bits of data into a single symbol, vastly increasing data throughput while relying on the same foundational amplitude-varying physics.