Frequency Modulation and FSK

In 1933, Edwin Howard Armstrong demonstrated a revolutionary concept that would forever change radio communication: Frequency Modulation (FM). At a time when Amplitude Modulation (AM) dominated the airwaves, Armstrong's innovation offered something unprecedented—crystal-clear audio with remarkable resistance to static and interference. When he first demonstrated FM broadcasting from the Empire State Building, listeners were astonished by the absence of the crackling and hissing that plagued AM transmissions.

Today, Frequency Modulation principles extend far beyond FM radio. From high-fidelity broadcasting to satellite communications, from radar systems to medical telemetry, FM and its digital counterpart FSK (Frequency Shift Keying) form the backbone of countless critical systems. Understanding FM isn't just about grasping a modulation technique—it's about comprehending a fundamental approach to representing information that trades complexity for reliability.

Before diving into frequency modulation specifically, we must establish a clear understanding of modulation itself—the process that makes long-distance communication possible.

Why Do We Need Modulation?

Consider human speech, which produces sound waves in the frequency range of approximately 300 Hz to 3,400 Hz (the telephony band). These frequencies are far too low to propagate effectively as electromagnetic waves through the atmosphere. To transmit voice or data over any practical distance, we need to "carry" this information on a higher-frequency wave—a carrier wave—that can propagate efficiently.

Modulation is the process of varying one or more properties of a high-frequency carrier signal in accordance with a lower-frequency message signal (also called the baseband signal or modulating signal).

The Three Fundamental Properties of a Sine Wave

A pure carrier signal is typically a sinusoidal wave, mathematically expressed as:

s(t) = A·cos(2πf_c·t + φ)

Where:

• A = Amplitude (peak value of the wave)
• f_c = Carrier frequency (cycles per second)
• φ = Phase (angular offset from a reference)
• t = Time

Each of these three properties—amplitude, frequency, and phase—can be varied according to the message signal, giving rise to three fundamental modulation families:

1. Amplitude Modulation (AM): The carrier's amplitude varies with the message
2. Frequency Modulation (FM): The carrier's frequency varies with the message
3. Phase Modulation (PM): The carrier's phase varies with the message

FM and PM are closely related—collectively known as angle modulation—because both involve changing the instantaneous angle of the carrier wave. This relationship becomes clearer when we examine the mathematics.

Frequency Modulation (FM) is a modulation technique where the instantaneous frequency of the carrier signal varies in proportion to the instantaneous amplitude of the message signal, while the carrier's amplitude remains constant.

This definition contains several crucial concepts that deserve careful examination:

Instantaneous Frequency

Unlike a pure sine wave that maintains a constant frequency, an FM signal has a frequency that changes moment by moment. At any given instant, we can define an "instantaneous frequency"—the frequency the signal would have if it continued unchanged from that moment. As the message signal rises and falls, the instantaneous frequency of the FM signal rises and falls correspondingly.

Constant Amplitude

In ideal FM, the envelope (amplitude) of the carrier remains perfectly constant, regardless of the message content. This is fundamentally different from AM, where the carrier's amplitude varies with the message. This constant amplitude is the primary reason for FM's excellent noise immunity—more on this later.

Proportional Variation

The relationship between the message signal and frequency deviation is linear. If the message amplitude doubles, the frequency deviation doubles. This linearity ensures faithful reproduction of the original message at the receiver.

Visualizing FM: The Waveform Perspective

Imagine a simple sinusoidal message signal (like a pure tone). When applied to an FM modulator:

• When the message signal is at its positive peak, the FM carrier reaches its maximum frequency (f_c + Δf)
• When the message signal passes through zero, the FM carrier returns to its center frequency (f_c)
• When the message signal is at its negative peak, the FM carrier reaches its minimum frequency (f_c - Δf)

The quantity Δf (delta-f) is called the frequency deviation—the maximum departure of the instantaneous frequency from the carrier frequency. This is a crucial parameter in FM system design.

The Modulation Index

One of the most important parameters in FM is the modulation index (β), defined as:

β = Δf / f_m

Where:

• Δf = Maximum frequency deviation
• f_m = Maximum frequency of the modulating signal

The modulation index determines many characteristics of the FM signal, including its bandwidth, spectral content, and signal-to-noise ratio performance. FM systems are often classified as:

• Narrowband FM (NBFM): β << 1 (typically β ≤ 0.5)
• Wideband FM (WBFM): β >> 1 (typically β ≥ 5)

A rigorous understanding of FM requires mathematical precision. Let's develop the equations that govern FM signal generation and analyze their implications.

The Carrier Signal

An unmodulated carrier is represented as:

c(t) = A_c · cos(2πf_c·t)

Where A_c is the carrier amplitude and f_c is the carrier frequency.

The Message Signal

For analysis, we often consider a single-tone (sinusoidal) modulating signal:

m(t) = A_m · cos(2πf_m·t)

Where A_m is the message amplitude and f_m is the message frequency.

The FM Signal

In FM, the instantaneous frequency f_i(t) varies with the message signal:

f_i(t) = f_c + k_f · m(t)

Where k_f is the frequency sensitivity of the modulator, measured in Hz per volt (or Hz per unit amplitude of the message signal).

Substituting the message signal:

f_i(t) = f_c + k_f · A_m · cos(2πf_m·t)

The term k_f · A_m represents the maximum frequency deviation Δf:

Δf = k_f · A_m

Deriving the Complete FM Expression

The instantaneous phase φ(t) of the FM signal is the integral of the instantaneous frequency:

φ(t) = 2π ∫ f_i(τ) dτ

Integrating:

φ(t) = 2πf_c·t + (k_f · A_m / f_m) · sin(2πf_m·t)

The FM signal is then:

s_FM(t) = A_c · cos[φ(t)]

s_FM(t) = A_c · cos[2πf_c·t + β · sin(2πf_m·t)]

Where β = Δf/f_m = k_f · A_m / f_m is the modulation index.

The Significance of This Expression

This equation reveals several profound insights:

1. The carrier amplitude A_c remains outside the cosine argument—confirming that FM maintains constant envelope.

2. The phase contains a sinusoidal term β·sin(2πf_m·t)—this time-varying phase deviation is what creates the frequency modulation.

3. The modulation index β appears as a scaling factor—it determines how much the phase deviates, which in turn determines the frequency deviation.

4. For single-tone modulation, the phase deviation equals β radians—this is the peak excursion of the instantaneous phase from the carrier phase.

Bessel Functions and FM Spectrum

The FM signal expression contains a cosine of a cosine, which cannot be simplified using elementary methods. To determine the frequency spectrum of an FM signal, we expand it using Bessel functions of the first kind:

s_FM(t) = A_c · Σ J_n(β) · cos[2π(f_c + n·f_m)·t]

Where:

• J_n(β) is the Bessel function of the first kind of order n
• The summation extends from n = -∞ to n = +∞

This expansion shows that an FM signal consists of the carrier frequency plus an infinite number of sidebands spaced at multiples of the modulating frequency. This is fundamentally different from AM, which produces only two sidebands.

Understanding FM fully requires contrasting it with AM—the modulation technique it was designed to improve upon. The differences between these two approaches illustrate fundamental tradeoffs in communication system design.

Amplitude Modulation Recap

In AM, the carrier amplitude varies with the message signal:

s_AM(t) = A_c[1 + m·cos(2πf_m·t)] · cos(2πf_c·t)

Where m is the modulation index (0 ≤ m ≤ 1 for standard AM).

Key AM characteristics:

• The carrier frequency remains constant
• The envelope follows the message signal
• Bandwidth = 2 × f_m (twice the highest message frequency)
• Carrier power is wasted (typically 2/3 of total power)

The Fundamental Tradeoff: Bandwidth vs. Noise Performance

FM's superior noise immunity comes at a cost: increased bandwidth. This is one of the fundamental tradeoffs in communication theory, and understanding it illuminates why different applications choose different modulation schemes.

Noise Behavior in AM:

• Noise affects the amplitude of the received signal
• Since information is carried in amplitude, noise directly corrupts the message
• The signal-to-noise ratio (SNR) at the output roughly equals the SNR at the input

Noise Behavior in FM:

• Noise primarily affects amplitude (noise has random phase and amplitude)
• Since FM information is in frequency, amplitude noise has minimal effect
• A limiter circuit can clip the amplitude, removing most noise
• The output SNR can exceed the input SNR—especially for high β systems

The FM Improvement Factor

For wideband FM, the output SNR improvement over baseband is:

SNR_out / SNR_in ≈ 3β²(β + 1)

This shows that increasing the modulation index (and hence bandwidth) provides better noise performance—the essence of the FM advantage.

For example, standard FM broadcasting uses:

• Δf = 75 kHz (maximum frequency deviation)
• f_m = 15 kHz (maximum audio frequency)
• β = 75/15 = 5

The improvement factor is approximately 3 × 25 × 6 ≈ 450 or about 26.5 dB better SNR compared to AM with the same carrier power!

Generating an FM signal requires hardware that can vary the output frequency in response to an input voltage (the message signal). Several methods have been developed, each with distinct advantages and applications.

Direct FM Generation

In direct FM, the modulating signal directly controls the frequency of an oscillator. The most common approaches include:

1. Voltage-Controlled Oscillator (VCO)

A VCO is an oscillator whose frequency is proportional to an input voltage. When the message signal is applied to the VCO's control input:

• Positive message amplitude → frequency increases
• Negative message amplitude → frequency decreases
• Zero message → center frequency (f_c)

VCOs are simple and can achieve large frequency deviations, but they suffer from frequency drift and may not maintain precise center frequency without additional stabilization.

2. Reactance Modulator

A reactance modulator uses a transistor or vacuum tube circuit that appears as a variable reactance (capacitance or inductance) to the tank circuit of an oscillator. The message signal controls this apparent reactance, thereby controlling the oscillator frequency.

3. Varactor Diode Modulator

A varactor (variable capacitance) diode changes capacitance with applied voltage. By including a varactor in the frequency-determining circuit of an oscillator, the message signal (applied as reverse bias) controls the oscillator frequency. This method is common in modern solid-state FM transmitters.

Indirect FM Generation (Armstrong Method)

Edwin Armstrong developed an ingenious indirect method that overcomes the frequency stability problems of direct FM. The key insight is the relationship between frequency and phase:

Frequency is the rate of change of phase

Mathematically: f(t) = (1/2π) × dφ/dt

This means:

• If we integrate the message signal and then phase-modulate the carrier, we get FM
• The crystal oscillator provides excellent frequency stability
• The narrowband FM from the phase modulator is then frequency-multiplied to wideband FM

Armstrong Transmitter Architecture:

1. Crystal Oscillator: Generates a stable carrier at a sub-harmonic of the final frequency
2. Integrator: Converts m(t) to ∫m(t)dt
3. Balanced Modulator: Creates phase modulation
4. Combining Network: Produces narrowband FM
5. Frequency Multiplier Chain: Multiplies both carrier frequency and frequency deviation
6. Mixer: Translates to final carrier frequency
7. Power Amplifier: Provides transmission power

Why Frequency Multiplication?

The phase modulator can only achieve small phase deviations (narrow β). Frequency multipliers (×2, ×3, etc.) multiply the frequency deviation proportionally:

• If Δf = 100 Hz after the phase modulator
• After ×8 multiplication: Δf = 800 Hz
• After another ×12 multiplication: Δf = 9,600 Hz

This allows achieving the large frequency deviations needed for wideband FM while maintaining crystal-controlled frequency accuracy.

FM detection (demodulation) is the process of recovering the original message signal from an FM wave. The demodulator must convert frequency variations back into amplitude variations that represent the original message. This is more challenging than AM detection, requiring circuitry that responds to frequency rather than amplitude.

Limiter: The First Stage

Before frequency detection, FM receivers typically include a limiter—a circuit that clips the FM signal to a constant amplitude. This:

• Removes any amplitude variations caused by noise or interference
• Maintains the frequency information, which is unaffected by clipping
• Provides the "capture effect" where the stronger of two signals dominates

Common FM Detection Methods

1. Slope Detection (Simple but Imperfect)

The simplest detector uses a tuned circuit slightly off-center from the carrier. The frequency response slope converts frequency variations to amplitude variations. An envelope detector then recovers the message.

Advantages: Simple construction Disadvantages: Nonlinear, limited dynamic range, susceptible to amplitude variations

2. Balanced Slope Detector (Foster-Seeley Discriminator)

Uses two slope detectors on opposite sides of the carrier frequency. The difference between their outputs is:

• Positive when frequency is above f_c
• Negative when frequency is below f_c
• Zero at exactly f_c

This provides better linearity and amplitude rejection than a single slope detector.

3. Ratio Detector

A modification of the Foster-Seeley discriminator that includes an additional capacitor for amplitude limiting. The name comes from the fact that it responds to the ratio of voltages across two halves of the circuit.

Advantages: Built-in amplitude limiting, good for consumer FM receivers Disadvantages: Lower output, limited capture ratio

4. Phase-Locked Loop (PLL) Detector

A PLL detector uses feedback to track the incoming FM signal:

1. A VCO generates a signal at the estimated frequency
2. A phase detector compares VCO output to the received FM signal
3. The phase error controls the VCO to minimize the difference
4. The VCO control voltage is the demodulated message signal

Advantages: Excellent linearity, inherent noise immunity, easily integrated Disadvantages: Requires careful design for acquisition and tracking

5. Quadrature Detector

The quadrature detector multiplies the FM signal by a 90°-shifted (quadrature) version of itself. The product contains a DC component proportional to the frequency deviation.

Advantages: Good linearity, simple IC implementation Disadvantages: Requires precise 90° phase shift network

One of FM's most remarkable characteristics is the capture effect—a phenomenon where an FM receiver completely suppresses a weaker signal in the presence of a stronger one on the same frequency. This behavior has profound implications for both the advantages and limitations of FM systems.

Understanding Capture

When two signals of different strengths are present on the same FM channel:

• AM receivers: Both signals combine, producing audible interference between the two
• FM receivers: The stronger signal "captures" the receiver, and the weaker signal is almost completely suppressed

For a typical FM receiver with a capture ratio of 1-2 dB, a signal just 1-2 dB stronger than an interferer will completely dominate, with the interferer contributing negligible output.

Why Does Capture Occur?

The capture effect arises from the limiter stage in FM receivers:

1. Two FM signals of different amplitudes sum together
2. The resultant signal has an envelope that varies as the signals move in and out of phase
3. The limiter clips this envelope to a constant amplitude
4. The limiting action suppresses the weaker signal's contribution to the instantaneous frequency
5. After limiting, the stronger signal's frequency dominates

Capture Ratio Definition

The capture ratio is the minimum difference in signal strength (in dB) required for the stronger signal to completely suppress the weaker one. Lower capture ratios indicate better receiver performance:

• Excellent FM receiver: 1.0 - 1.5 dB capture ratio
• Good FM receiver: 1.5 - 2.0 dB capture ratio
• Average FM receiver: 2.0 - 3.0 dB capture ratio

Practical Implications

The capture effect explains why:

• FM provides clear reception even with moderate interference
• FM stations can reuse frequencies geographically
• FM radio has largely replaced AM for high-fidelity broadcasting
• Digital modulation schemes (which are variants of FM/FSK) dominate modern communications

We have established a comprehensive understanding of Frequency Modulation—the theoretical foundation upon which all subsequent topics in this module will build. Let's consolidate the key concepts:

What's Next: Frequency Shift Keying (FSK)

With this solid foundation in analog FM, we're ready to explore its digital counterpart: Frequency Shift Keying (FSK). FSK applies FM principles to digital data transmission, using discrete frequency deviations to represent binary data. You'll see how FM's noise immunity makes FSK an excellent choice for reliable digital communication—from modems to wireless sensor networks to spacecraft telemetry.