Amplitude Modulation (AM) is a technique where the instantaneous amplitude (voltage or power) of a high-frequency carrier wave is varied in direct proportion to the instantaneous amplitude of a lower-frequency information signal. In a physical circuit, AM changes the outer envelope of the RF waveform while keeping the zero-crossings (the fundamental frequency) strictly locked. This process takes a baseband signal—like a 1 kHz audio tone or slow analog sensor data—that cannot efficiently radiate from a reasonably sized antenna, and impresses it onto a high-frequency RF carrier (like 1 MHz or 433 MHz) so it can be transmitted over the air.

To visualize this without getting lost in abstract math, think of a high-speed strobe light flashing at a steady 1,000 times per second (the carrier). If you put a motorized dimmer on the bulb that opens and closes the shutter once per second (the signal), the flash rate never changes, but the brightness of each flash pulses at 1 Hz. That brightness variation is the AM envelope.

The Math on the Bench: A Worked Numeric Example

Let's put this on the workbench and look at the actual voltages and frequencies generated when you mix two signals in a nonlinear multiplier or mixer IC (like the SA612 or an analog multiplier).

Carrier Signal ($V_c$): 1.000 MHz sine wave, 10V peak.
Modulating Signal ($V_m$): 1 kHz audio tone, 5V peak.

The first critical metric is the modulation index ($m$), which defines how deeply the carrier is being varied. It is simply the ratio of the modulating peak voltage to the carrier peak voltage:

m = V_m / V_c = 5V / 10V = 0.5 (or 50% modulation depth).

When you view this on a spectrum analyzer, you will not just see the 1.000 MHz carrier. The nonlinear mixing process generates sidebands at the sum and difference of the two frequencies:

  • Upper Sideband (USB): $f_c + f_m$ = 1.001 MHz
  • Lower Sideband (LSB): $f_c - f_m$ = 0.999 MHz

Here is where AM reveals its biggest inefficiency, which every RF designer must account for: power distribution. The carrier itself contains zero information; it is just a reference. The actual audio/data lives entirely in the sidebands. The power in each sideband is calculated as $P_{sb} = P_c \times (m^2 / 4)$. With $m = 0.5$, each sideband contains only 6.25% of the carrier power. The total transmitted power is $P_t = P_c(1 + m^2/2)$, meaning at 50% modulation, 88.8% of your transmitter's power is wasted just blasting the unmodulated carrier. This is why AM is rarely used for battery-powered, long-range telemetry unless simplicity of the receiver is the primary constraint.

Where You Meet This in Practice

While most people associate AM with medium-wave (MW) broadcast radio, the underlying physics of amplitude modulation shows up across modern electronics and hobbyist workbenches:

  1. Aviation Communications (118–137 MHz): Air traffic control uses AM, not FM. Why? Because FM suffers from the 'capture effect,' where a stronger signal completely drowns out a weaker one on the same frequency. In an emergency, if two pilots transmit simultaneously on an AM channel, the controller hears a heterodyne squeal and can distinguish that multiple parties are talking, rather than just hearing the louder transmitter.
  2. Analog FPV Video (5.8 GHz): First-person view drone pilots using analog video systems rely on AM for the luminance (brightness) carrier. The video signal directly modulates the amplitude of the 5.8 GHz microwave carrier, allowing for ultra-low latency compared to digital encoding.
  3. Cheap 433 MHz RF Modules: The ubiquitous blue/green RF link modules found in Arduino starter kits use ASK (Amplitude Shift Keying) or OOK (On-Off Keying). This is simply AM driven to the digital extremes: 100% modulation for a logic '1' and 0% modulation (carrier turned off) for a logic '0'.

Real-World Scenario: Overmodulating a 433 MHz Telemetry Link

Theory is clean; the bench is messy. Here is a walkthrough of a common failure mode when building custom AM telemetry systems.

The Setup: You are building an analog telemetry link for a remote weather station using a raw 433 MHz SAW oscillator transmitter module and a simple super-regenerative receiver with a diode envelope detector. You want to send a 0–3V analog wind speed voltage directly via AM, bypassing the need for an ADC and microcontroller on the TX side.

The Numbers: The TX module's RF output stage expects a 1V peak carrier drive. You AC-couple your 0–3V wind sensor signal through a capacitor to the modulation pin, but you forget to establish a proper DC bias point. Your modulating voltage ends up swinging symmetrically from -1.5V to +1.5V.

The Outcome: Your modulation index hits $m = 1.5$ (150% modulation). This is known as overmodulation.

What went wrong?
When the modulating signal exceeds the carrier amplitude ($m > 1$), the RF envelope is driven to zero and the carrier phase abruptly inverts by 180 degrees. The receiver's simple diode envelope detector doesn't understand phase; it just rectifies the absolute voltage. The phase inversion creates a sharp 'V' shape (a cusp) at the zero-crossing instead of a smooth curve. This sharp edge injects massive high-frequency harmonic distortion into your recovered wind speed voltage. Furthermore, the sharp time-domain edges cause spectral splatter, blasting RF noise across adjacent 433 MHz ISM channels and potentially violating FCC/CE emission masks.

The Fix: You must clamp the modulating signal. Add an op-amp limiter circuit (using diodes in the feedback loop) to ensure $V_m$ never exceeds $0.8 \times V_c$. Keeping $m \le 0.8$ (80% modulation) guarantees the envelope never hits zero, preserving the smooth curve for the receiver's RC low-pass filter to track accurately. For a deeper dive into receiver envelope tracking, the All About Circuits RF textbook provides excellent schematics for diode detector time-constant tuning.

What People Commonly Confuse With AM

When troubleshooting RF circuits or studying for a ham radio license, a few concepts frequently get tangled up with the core AM modulation definition.

  • AM vs. FM (Noise Immunity): People assume AM is 'worse' than FM universally. AM is highly susceptible to amplitude noise (lightning strikes, brushed motors, switching power supplies) because the noise directly alters the envelope. FM rejects this amplitude noise via limiter circuits in the receiver. However, AM requires vastly simpler receiver hardware and narrower bandwidth for the same baseband audio quality.
  • The Carrier vs. The Sidebands: Beginners often tune a narrow bandpass filter directly over the carrier frequency to 'clean up' an AM signal. If you filter out the sidebands, you filter out the information. You will recover a perfectly clean, completely silent, unmodulated carrier. The sidebands are the signal.
  • AM vs. SSB (Single Sideband): As noted in the power math above, AM wastes massive energy on the carrier and transmits the audio twice (USB and LSB). SSB is a derivative of AM where a balanced modulator cancels the carrier, and a crystal filter removes one sideband. It is technically AM, but stripped of its redundant components for maximum power efficiency. The Electronics Tutorials communication section offers a great visual breakdown of the DSB-SC (Double Sideband Suppressed Carrier) spectrum compared to standard AM.

Frequently Asked Questions

Can I demodulate AM without a dedicated RF IC?
Yes. The simplest AM demodulator is a single germanium or Schottky diode (like a 1N60 or BAT85) followed by a parallel resistor and capacitor. The diode chops off the negative half of the RF cycles, and the RC filter smooths the remaining positive peaks, tracing the audio envelope. This is the exact circuit used in crystal radios.

Why does my AM signal sound distorted when the volume is too high?
This is the audio equivalent of the 433 MHz overmodulation scenario described above. If the audio amplifier drives the modulator past 100% ($m > 1$), the envelope hits zero and 'flat-tops' or phase-reverses, creating harsh harmonic distortion that sounds like crackling or fuzz on the receiving end.

Is PWM (Pulse Width Modulation) a type of AM? No. PWM varies the duty cycle (time-domain width) of a square wave pulse, keeping the amplitude constant. While both are modulation techniques used to encode information, PWM is fundamentally a time-based digital encoding scheme, whereas AM is a continuous analog variation of voltage amplitude in the frequency domain.