Electronics Guide

Fundamentals of Analog Modulation

Modulation varies a carrier signal's characteristics in accordance with an information-bearing signal. The carrier is normally a high-frequency sinusoid that propagates efficiently through the chosen medium, while the modulating signal—audio, video, or sensor data—carries the information.

The reasons for modulating at all are practical:

• Frequency translation: Baseband signals move up to frequencies where antennas of reasonable size radiate efficiently. An efficient antenna is a significant fraction of a wavelength, so a 3 kHz voice signal would demand an antenna tens of kilometers long if it were radiated directly.
• Channel sharing: Assigning each user a different carrier frequency lets many signals occupy one medium simultaneously, the basis of frequency-division multiplexing and of all spectrum allocation.
• Noise and interference management: The choice of scheme sets how the system responds to additive noise. Frequency and phase modulation can exchange bandwidth for signal-to-noise ratio in a way that amplitude modulation cannot.
• Hardware practicality: Filters, amplifiers, and antennas are far easier to build with useful selectivity and gain at radio frequencies than at baseband.

The general form of an unmodulated carrier is:

c(t) = A_c cos(2πf_ct + φ)

where A_c is the amplitude, f_c the carrier frequency, and φ the phase. Amplitude modulation varies A_c, frequency modulation varies f_c, and phase modulation varies φ. Because frequency is the time derivative of phase, the last two are members of a single family known collectively as angle modulation.

Amplitude Modulation (AM)

Amplitude modulation varies the carrier amplitude in proportion to the instantaneous value of the modulating signal. AM was the first modulation technique widely deployed in radio broadcasting and remains in service for medium-wave and shortwave broadcasting, aviation voice communication, and citizens band radio.

Standard AM (Double-Sideband Full-Carrier)

In standard AM, the transmitted signal is:

s(t) = A_c[1 + m(t)] cos(2πf_ct)

where m(t) is the normalized modulating signal and the bracketed term is the envelope. Multiplying a carrier by a message of highest frequency f_m produces sum and difference components, so the occupied bandwidth is 2f_m—twice the baseband bandwidth—split into an upper and a lower sideband that mirror each other about the carrier.

The modulation index, or modulation depth, quantifies the degree of modulation and can be read directly from an envelope display on an oscilloscope:

μ = (A_max − A_min) / (A_max + A_min)

The index must not exceed 100 percent (μ ≤ 1). Beyond that point the envelope attempts to go negative, the detector clips it, and the resulting distortion generates wideband "splatter" that interferes with adjacent channels.

Standard AM pays a steep price in efficiency. For sinusoidal modulation the total power is P_t = P_c(1 + μ²/2), so even at full 100 percent modulation the carrier—which conveys no information—consumes two-thirds of the transmitted power, and each sideband receives only one-sixth. Typical program material has a much lower average modulation index, pushing the carrier share higher still. What that expense buys is a receiver so simple it needs no local oscillator: a diode envelope detector recovers the message directly.

Double-Sideband Suppressed-Carrier (DSB-SC)

DSB-SC removes the carrier and transmits only the sidebands:

s(t) = A_cm(t) cos(2πf_ct)

All transmitted power now carries information. The cost is at the receiver: the envelope of a DSB-SC signal no longer follows the message, because the waveform reverses phase every time m(t) crosses zero. Recovery therefore requires coherent (synchronous) detection with a local carrier matched in both frequency and phase. A phase error θ scales the recovered output by cos θ, so a 90-degree error nulls the signal entirely. DSB-SC is used where a pilot tone or a carrier-recovery loop is available, and it is the modulation on the 38 kHz difference-signal subcarrier of FM stereo.

Single-Sideband (SSB)

Single-sideband modulation transmits one sideband only, upper (USB) or lower (LSB), discarding both the carrier and the redundant second sideband. The advantages are substantial:

• Half the occupied bandwidth of AM or DSB-SC—about 2.4 to 3 kHz for communications-quality speech
• All transmitter power concentrated in the information-bearing sideband
• Less susceptibility to selective fading, which distorts AM badly when the carrier fades relative to the sidebands
• Better spectrum utilization in crowded high-frequency allocations

Compared at equal average power with a fully modulated AM signal, an SSB transmitter puts roughly six times as much power into the surviving sideband—about 8 dB—and its halved noise bandwidth is worth roughly 3 dB more, which is the origin of the familiar claim that SSB is worth several signal-strength units over AM on a high-frequency circuit.

Two classical generation methods dominate. The filter method produces DSB-SC in a balanced modulator and then removes the unwanted sideband with a sharp crystal or mechanical filter, which is practical because the two sidebands are separated by twice the lowest audio frequency present. The phasing (Hartley) method feeds two balanced modulators with carrier and audio signals in quadrature and sums their outputs so the unwanted sideband cancels, avoiding the expensive filter at the cost of demanding wideband 90-degree audio phase shifters. Modern transceivers usually implement the equivalent operation digitally with a Hilbert transform and a quadrature upconverter. By long-standing amateur convention, LSB is used below 10 MHz and USB above, and USB is standard for maritime and aeronautical high-frequency service.

Because SSB has no constant carrier, its output is rated in peak envelope power rather than average power, and it demands genuinely linear amplification throughout the transmitter chain.

Vestigial-Sideband (VSB)

VSB is the compromise between DSB and SSB: it transmits one full sideband plus a vestige of the other, shaped so that the receiver's own filter restores flat overall response. This approach provides:

• Faithful reproduction of low-frequency and DC-like content, which pure SSB handles poorly and which video signals require
• Bandwidth efficiency approaching SSB
• Relaxed transmitter filter transition-band requirements compared with SSB
• A transmitted carrier available for simple envelope detection or for automatic frequency control

Analog television was the great VSB application. In the North American NTSC system, a 6 MHz channel carried video baseband out to 4.2 MHz with the visual carrier placed 1.25 MHz above the lower channel edge, leaving a vestige of the lower sideband roughly 0.75 MHz wide; the 7 and 8 MHz channels used with PAL and SECAM followed the same principle with wider video bandwidths. The technique survived the digital transition in a different form: the ATSC terrestrial television standard uses 8-VSB, a suppressed-carrier vestigial-sideband scheme carrying eight-level digital symbols with a small pilot for carrier acquisition.

Frequency Modulation (FM)

Frequency modulation varies the carrier frequency in proportion to the modulating signal's amplitude while holding amplitude constant. That constant envelope is the source of FM's chief advantage: a limiter ahead of the detector can strip away amplitude disturbances—ignition noise, lightning crashes, fading—without touching the information.

FM Fundamentals

The instantaneous frequency of an FM signal is:

f_i(t) = f_c + k_fm(t)

where k_f is the frequency sensitivity in hertz per volt and m(t) is the modulating signal.

Two parameters characterize the result:

• Frequency deviation (Δf): The maximum departure from the carrier frequency, equal to k_f times the peak modulating amplitude. Deviation depends on the amplitude of the modulating signal, not its frequency.
• Modulation index (β): The ratio of deviation to modulating frequency, β = Δf / f_m. For complex program material the corresponding figure is the deviation ratio, peak deviation divided by the highest baseband frequency.

Spectrum and Bandwidth

Unlike AM, FM is a nonlinear process: its spectrum is not a simple frequency-shifted copy of the baseband. For single-tone modulation the sideband amplitudes are Bessel functions of the first kind, J_n(β), producing pairs of sidebands at f_c ± n·f_m for every integer n. The series is infinite in theory, though the terms fall away rapidly once n exceeds β.

Carson's rule gives the practical bandwidth, the span containing roughly 98 percent of the power:

B_FM ≈ 2(Δf + f_m) = 2f_m(β + 1)

The Bessel behavior also has a useful laboratory consequence. J₀(β) passes through zero at β ≈ 2.405, so the carrier component vanishes entirely at that index. Feeding a known audio tone and raising its level until the carrier nulls on a spectrum analyzer calibrates deviation precisely, a standard method for verifying a deviation meter.

Narrowband FM (NBFM)

When the modulation index is well below unity—commonly taken as β ≲ 0.3—only the first-order sidebands are significant, and the spectrum closely resembles that of AM. Occupied bandwidth is then approximately 2f_m, and the noise advantage over AM largely disappears; what remains is the immunity to amplitude disturbances and the capture effect.

NBFM is the workhorse of channelized voice communication, where spectrum conservation matters more than fidelity: land mobile and public safety radio, marine VHF, and amateur VHF and UHF repeaters. Typical deviations are ±5 kHz on 25 kHz channels and ±2.5 kHz on the 12.5 kHz channels mandated by narrowbanding programs in many jurisdictions.

Wideband FM (WBFM)

Wideband FM uses large modulation indices, generating many significant sideband pairs and exchanging bandwidth for noise performance. Broadcast FM is the canonical example: a peak deviation of ±75 kHz with a 15 kHz audio bandwidth gives a deviation ratio of 5, and Carson's rule predicts about 180 kHz of occupied bandwidth, which sits comfortably inside the 200 kHz channel spacing used in the Americas. The broadcast band spans 88 to 108 MHz across most of the world, with Japan using 76 to 95 MHz and a legacy OIRT band at 65.8 to 74 MHz, once standard across eastern Europe, now almost entirely retired.

Generating FM: Direct and Indirect Methods

Direct FM varies the frequency of an oscillator itself, classically by placing a varactor diode across the tank circuit of a VCO. Deviation is easy to obtain, but the free-running oscillator drifts, so the design must lock the average frequency to a crystal reference with a phase-locked loop whose bandwidth is set below the lowest modulating frequency.

Indirect FM, the method Edwin Armstrong developed, starts from a crystal oscillator and applies a small phase modulation to it, integrating the audio beforehand so that phase modulation becomes frequency modulation. The achievable deviation at the modulator is tiny, so a chain of frequency multipliers scales both the carrier and the deviation up to the required values. Frequency stability is excellent because the source is a crystal. Contemporary transmitters more often synthesize the modulated signal digitally and present it to a quadrature modulator, which sidesteps the trade entirely.

Threshold and Capture Effects

FM's noise advantage is conditional. Below a carrier-to-noise ratio of roughly 10 to 13 dB at the demodulator, noise impulses begin to dominate the instantaneous phase and the output degrades abruptly rather than gracefully—the threshold or "click noise" effect. Above threshold, output signal-to-noise ratio improves with the square of the modulation index. Extended-threshold demodulators such as PLL and feedback detectors buy a few decibels of margin, a technique long exploited in satellite links.

The capture effect is the second distinctive FM behavior: when two signals share a frequency, the limiter and discriminator suppress the weaker one almost entirely once it falls a small margin below the stronger. The capture ratio, typically 1 to 3 dB in a good receiver, specifies how small that margin can be. The result is clean reception in the presence of co-channel interference, in contrast to the heterodyne whistles and mixed audio an AM receiver would produce.

Phase Modulation (PM)

Phase modulation varies the carrier's phase angle in proportion to the modulating signal:

φ(t) = k_pm(t)

Since instantaneous frequency is the derivative of instantaneous phase, PM and FM are two views of the same mechanism. Applying m(t) to a phase modulator after integrating it produces FM; applying the derivative of m(t) to a frequency modulator produces PM. The practical distinction lies in the spectrum: for PM the deviation is proportional to the amplitude of the modulating signal alone, whereas for FM the modulation index falls with increasing modulating frequency. PM therefore spreads high-frequency baseband content more widely than FM does, which is one way to view pre-emphasis—an FM transmitter with pre-emphasis behaves like a phase modulator over the emphasized part of the band.

Analog PM is rarely deployed on its own for voice, but the modulator is everywhere. Phase-shift keying, quadrature amplitude modulation, and every constellation-based digital scheme are discrete-valued phase modulation, and PM is used for telemetry and deep-space links where phase-coherent operation and a residual carrier for tracking are both required.

Quadrature Amplitude Modulation (QAM)

QAM modulates amplitude and phase together by summing two DSB-SC signals whose carriers are 90 degrees apart:

s(t) = I(t) cos(2πf_ct) − Q(t) sin(2πf_ct)

The in-phase (I) and quadrature (Q) components remain separable at a coherent receiver because the two carriers are orthogonal, so two independent baseband signals occupy the bandwidth a single DSB-SC signal would need. The penalty is a strict demand on carrier phase accuracy: any phase error in the recovered carrier leaks one channel into the other as crosstalk.

Analog QAM had one enormous deployment. Composite color television encoded the two chrominance difference signals as I and Q on a suppressed color subcarrier—3.579545 MHz in NTSC, 4.43361875 MHz in PAL—with a short burst of reference subcarrier transmitted during each horizontal blanking interval so the receiver could rebuild the carrier in phase. PAL added a line-by-line reversal of one component that averaged out phase errors, correcting the hue shifts that plagued NTSC. Motorola's C-QUAM AM stereo system used a related idea, and the quadrature modulator is now the standard front end for essentially all digital radio.

Noise Performance Compared

The schemes can be ranked quantitatively. For single-tone modulation, a fixed received power, and additive white Gaussian noise, textbook analysis gives a figure of merit—the output signal-to-noise ratio divided by the ratio a baseband channel of the same power and message bandwidth would deliver:

• Full-carrier AM with envelope detection: 1/3 at 100 percent modulation, and worse at lower depths, because most of the power sits in the carrier
• DSB-SC and SSB with coherent detection: 1, meaning neither gains nor loses relative to baseband transmission
• FM above threshold: 3β²/2, so performance rises with the square of the modulation index

At the deviation ratio of 5 used in FM broadcasting, the FM figure of merit is 37.5 against 1/3 for fully modulated AM—a factor of about 112, or roughly 20 dB. That margin, together with pre-emphasis and the capture effect, is why broadcast FM sounds quiet where AM on the same path sounds noisy. It is bought with six times the occupied bandwidth of an AM signal carrying the same 15 kHz baseband, and it is forfeited entirely below threshold—the essential trade of all wideband angle modulation.

Carrier and Sideband Suppression

Suppressed-carrier systems begin with a modulator that produces no carrier output, then remove the unwanted sideband if single-sideband operation is required. The building blocks are:

• Balanced modulators: Diode ring mixers, dual-gate FET stages, or Gilbert-cell multipliers form the product of carrier and message. Because the output is proportional to the product, the carrier appears only through component mismatch; practical circuits achieve 40 to 60 dB of carrier suppression, often with a trim adjustment to null residual feedthrough.
• Sideband filters: Crystal, mechanical, or ceramic filters with steep skirts remove the unwanted sideband from the balanced modulator's DSB-SC output, typically at a fixed intermediate frequency that is then mixed to the operating channel.
• Phasing networks: Two balanced modulators driven in quadrature, their outputs summed, cancel the unwanted sideband without a sharp filter. Suppression is limited by amplitude and phase matching—a 1-degree phase error or 1 percent amplitude error holds unwanted-sideband rejection to roughly 40 dB.

At the receiver, the missing carrier must be regenerated. Costas loops and squaring loops derive it from the sidebands themselves; pilot-tone systems transmit a small reduced carrier for the receiver's phase-locked loop to track, the approach used for FM stereo, color television burst, and many satellite links.

Demodulation Circuits

AM Demodulation

Standard AM uses envelope detection: a diode rectifies the intermediate-frequency signal and an RC network follows the envelope. The time constant must be long enough to smooth the carrier but short enough to track the fastest audio component—too long and the detector produces diagonal clipping on high-frequency, deeply modulated peaks. Adding a small forward bias or using a synchronous detector avoids the diode's threshold and improves performance on weak signals.

DSB-SC and SSB require a product detector fed with a locally generated carrier. In an SSB receiver this is the beat frequency oscillator, and its accuracy determines intelligibility. Because the error shifts every audio component by the same number of hertz rather than by the same ratio, harmonic relationships are destroyed: tuning high makes speech sound thin and quacking, tuning low makes it sound growling and bass-heavy, and errors beyond roughly 50 Hz are clearly audible.

FM Demodulation

Every FM detector converts frequency variation into amplitude variation, and all of them are preceded by a limiter that removes amplitude noise first:

• Slope detection: Tuning a resonant circuit slightly off frequency so its skirt converts deviation into amplitude change. Simple, but linear over only a narrow range.
• Foster-Seeley discriminator: A center-tapped transformer and balanced diodes produce an output proportional to the phase difference between primary and secondary. Highly linear, but it responds to amplitude changes and therefore demands a good limiter ahead of it.
• Ratio detector: A rearrangement of the Foster-Seeley circuit with a large stabilizing capacitor that provides inherent amplitude limiting, at the cost of some linearity and output level.
• Quadrature detector: A phase-shift network and a phase detector, easily integrated on a chip. This is the standard implementation in consumer receiver ICs.
• Phase-locked loop: A loop that tracks the incoming frequency, its error voltage becoming the demodulated output. Linearity is set by the VCO rather than by a tuned circuit, no alignment is needed, and threshold extension of a few decibels is available.
• Digital and software-defined detectors: After quadrature downconversion and sampling, the demodulated output is the time derivative of arctan(Q/I), computed exactly and without drift. This is now the usual approach in new equipment.

Automatic Gain Control (AGC)

AGC holds the receiver output roughly constant while the input varies over many decades. A detector senses the signal level, usually at the intermediate-frequency stage or the demodulator output, and feeds back a control voltage that adjusts the gain of the RF and IF amplifiers.

The defining parameters are:

• AGC range: The span of input levels over which output stays within a stated tolerance, typically 60 to 100 dB, obtained by distributing control across several stages.
• Attack time: The response to a sudden increase in signal, on the order of 1 to 10 ms in a communications receiver. Too slow and a strong signal blasts through before the loop reacts; too fast and the loop pumps on speech peaks and static crashes.
• Decay time: The response to a decrease, deliberately much slower—commonly a few hundred milliseconds to several seconds—so that the gain does not rise between syllables and amplify background noise. Many SSB receivers offer selectable fast, medium, and slow settings, sometimes with a "hang" interval that holds gain constant for a fixed period before releasing.
• AGC threshold: The input level at which the loop begins to act; below it the receiver runs at full gain to preserve sensitivity.

Variants serve particular needs. Delayed AGC leaves the RF stage at full gain until the signal is well above the noise floor, preserving noise figure. Keyed or gated AGC samples the level only during a known interval—the sync tip in a television receiver, the interval between pulses in a radar—so that modulation content does not corrupt the measurement. Placing AGC ahead of a sharp IF filter risks having a strong adjacent-channel signal desensitize the receiver, so modern designs derive the control voltage after the filter wherever possible.

Automatic Frequency Control (AFC)

AFC compensates for frequency drift caused by component aging, temperature change, or oscillator instability. The circuit measures the frequency error between the received signal and the receiver's expected intermediate frequency, then generates a correction voltage that pulls the local oscillator back into alignment.

In an FM receiver the discriminator already provides the error signal: its DC output is zero only when the signal sits at the center of the passband, so filtering that output and applying it to a varactor in the local oscillator closes the loop. AM and television receivers use a separate discriminator or phase detector on the carrier. The loop must be slow enough to ignore the modulation and fast enough to follow drift, and it needs a defined capture range, since an AFC loop can lock onto a strong adjacent signal and hold the receiver there.

Frequency synthesis from a crystal or temperature-compensated reference has reduced but not eliminated the need for AFC. It remains important wherever the transmitter itself may be off frequency—low-cost portable transmitters, Doppler-shifted signals from satellites and aircraft, and the frequency-offset conditions common in cable and satellite receivers.

Squelch Systems

Squelch circuits mute the receiver output in the absence of a wanted signal, sparing the operator a constant hiss. Several mechanisms are in common use.

Carrier-Operated Squelch

The simplest form opens the audio path when received signal strength exceeds a threshold, often reusing the AGC voltage. It is easy to build but cannot distinguish a wanted transmission from interference or a nearby carrier, and its threshold must be reset as conditions change.

Noise Squelch

The standard method in FM receivers exploits a property of the FM detector: quieting. Noise energy just above the voice passband, typically sampled in a band around 6 to 10 kHz, is large when no carrier is present and collapses when a signal captures the receiver. Rectifying that band and comparing it with a threshold gives a decision that tracks signal quality rather than raw strength, so the squelch opens for a usable signal and stays closed for a strong but noisy one.

Continuous Tone-Coded Squelch System (CTCSS)

CTCSS adds a low-level subaudible tone, drawn from a standardized set spanning roughly 67 to 254 Hz, beneath the voice. The receiver decodes the tone and opens only for the matching value. Because the tone falls below the 300 Hz corner of the audio passband, it is inaudible in normal use, and a high-pass filter keeps it out of the loudspeaker. CTCSS lets several user groups share a channel without hearing one another and reduces nuisance openings from distant co-channel stations. It provides no security of any kind—it is an addressing scheme, not encryption, and every user still shares the channel.

Digital-Coded Squelch (DCS)

DCS replaces the analog tone with a continuously repeating low-rate digital word transmitted in the same subaudible region. It offers many more distinct codes than CTCSS and decodes faster, and it is widely supported alongside CTCSS in land mobile and amateur equipment.

Pre-emphasis and De-emphasis

Noise at the output of an FM discriminator is not flat: its power spectral density rises with the square of the baseband frequency, so the top of the audio band suffers far more than the bottom. Program material, by contrast, usually has less energy at high frequencies. Pre-emphasis and de-emphasis exploit that mismatch.

Pre-emphasis

The transmitter boosts high-frequency baseband content by a first-order network rising at 6 dB per octave above a corner frequency. Standard time constants are 75 μs in North America and South Korea and 50 μs in Europe and most of the rest of the world, corresponding to corners near 2.1 kHz and 3.2 kHz respectively. The boost cannot be unlimited, since deviation is capped by regulation, so broadcast processing must limit the treated signal to keep peak deviation within its licensed value.

De-emphasis

The receiver applies the exactly complementary roll-off, restoring flat frequency response while attenuating the noise that the discriminator produced in the same region. The net improvement in weighted signal-to-noise ratio is commonly quoted at roughly 12 to 13 dB for the 75 μs characteristic. A receiver using the wrong time constant is audibly wrong: 50 μs de-emphasis on a 75 μs transmission sounds bright and hissy, and the reverse sounds dull.

Companding Techniques

Companding—compressing before transmission and expanding after reception—improves the apparent dynamic range of a noisy channel. A 2:1 compressor followed by a matching 1:2 expander restores the original dynamics while pushing quiet passages well above the channel noise, which the expander then attenuates along with the program. Applications include:

• Broadcast audio processing: Compression and limiting raise average modulation and therefore loudness and coverage, while peak limiting keeps deviation legal.
• Two-way radio: Voice-grade compandors improve intelligibility on marginal links, and analog cordless and mobile telephony used them widely before digital systems arrived.
• Analog noise reduction: The Dolby and dbx systems applied companding to magnetic tape and to broadcast, with dbx using broadband companding and the Dolby systems splitting the audio into bands and acting mainly on quiet, high-frequency content.

The engineering difficulty is tracking. The expander must mirror the compressor closely in gain law, attack, and release, or the result is audible pumping and breathing; level errors anywhere in the path between the two translate directly into gain errors. Analog companding largely disappeared from new consumer designs once digital coding made wide dynamic range available without it, but the compressor and limiter remain fixtures of every broadcast air chain.

Stereo and Multichannel Encoding

FM Stereo Broadcasting

The FCC approved the combined General Electric and Zenith pilot-tone system in April 1961, and it remains the world standard. Its design goal was backward compatibility: a monophonic receiver must recover a correct mono signal while ignoring the stereo information entirely. The composite baseband that modulates the FM carrier contains:

• L+R signal (0 to 15 kHz): The sum of the channels, which any mono receiver demodulates directly
• 19 kHz pilot tone: A phase-locked reference transmitted at roughly 8 to 10 percent of total modulation
• L−R signal (23 to 53 kHz): The difference signal on a DSB-SC subcarrier at 38 kHz, exactly twice the pilot frequency and phase-locked to it
• Subsidiary Communications Authorization (SCA) subcarriers: Optional narrowband FM subcarriers, usually centered at 67 kHz and 92 kHz, carrying reading services, background music, or data; the complete multiplex baseband extends to about 99 kHz
• Radio Data System (RDS/RBDS): A low-rate data channel on a 57 kHz subcarrier, the third harmonic of the pilot, supplying station name, program type, and traffic information

The decoder recovers the pilot with a narrow filter or a phase-locked loop, doubles it to 38 kHz, synchronously demodulates the L−R subcarrier, and matrixes the results: L = (L+R) + (L−R) and R = (L+R) − (L−R). The cost of stereo is noise. Because the difference signal occupies the high end of the baseband where discriminator noise is greatest, and shares the available deviation with the sum signal and the pilot, stereo reception is on the order of 20 dB noisier than mono for the same received signal. Receivers therefore blend toward mono as signal strength falls, and stereo separation—typically 30 dB or better on a strong local signal—degrades gracefully rather than becoming unusable.

Analog Television Sound

Analog television carried its audio on a separate FM aural carrier offset from the visual carrier by 4.5 MHz in the North American system. The BTSC multiplex added stereo and a second audio program using a subcarrier structure locked to the horizontal line rate, the same compatibility strategy as FM stereo transplanted into a television channel.

AM Stereo Systems

Five incompatible AM stereo systems—from Motorola, Harris, Magnavox, Belar, and Kahn—were placed before the United States Federal Communications Commission around 1980. In 1982 the Commission declined to pick one, leaving the choice to the marketplace. Motorola's C-QUAM (Compatible Quadrature Amplitude Modulation), which phase-modulates the carrier with the difference signal while amplitude-modulating it with the sum signal so that an ordinary envelope detector still recovers correct mono, won that contest decisively. After Congress directed the FCC to settle on a single standard, the Commission designated C-QUAM in 1993. By then most broadcasters and receiver manufacturers had moved on, and AM stereo never approached the penetration of FM stereo.

Subcarrier Systems

A subcarrier is a carrier modulated onto another carrier, letting one transmission carry several independent services. Applications include:

• FM broadcasting: SCA channels for reading services for the blind, background music, foreign-language programming, and utility telemetry, plus the RDS data channel
• Television: The color subcarrier, the aural carrier, and the second audio program channel, all sharing one 6 to 8 MHz channel
• Telemetry: IRIG-standard FM/FM telemetry, in which many low-bandwidth sensor channels each modulate a separate subcarrier oscillator and the summed baseband frequency-modulates the RF carrier—long the standard for flight test and range instrumentation
• Program distribution: Analog audio subcarriers on satellite television transponders, once the standard way to distribute radio network programming, now displaced by digital methods; modern satellite radio services are fully digital

Subcarrier planning is an exercise in budgeting. Every subcarrier consumes part of the main carrier's deviation, so adding one reduces the modulation available to the primary program and degrades its signal-to-noise ratio. Spacing must leave guard bands wide enough that intermodulation products between subcarriers, and their beats with the pilot and program content, do not fall back into an active channel.

Modulation Analyzers and Measurement

Modulation analyzers characterize transmitters and diagnose faults. A modern instrument is usually a vector signal analyzer that digitizes the RF, downconverts it to I and Q, and computes every parameter in software—so a single capture yields amplitude, frequency, and phase views of the same event.

AM Measurements

• Positive and negative modulation depth, measured separately, since asymmetry reveals distortion
• Carrier frequency and stability
• Sideband symmetry, which exposes incidental phase modulation
• Total harmonic distortion of the demodulated audio
• Carrier and unwanted-sideband suppression for DSB-SC and SSB
• Spurious and out-of-band emissions against the applicable emission mask

FM Measurements

• Peak and RMS frequency deviation, and modulation index
• Residual FM, hum, and noise with modulation removed
• Baseband frequency response and verification of the pre-emphasis curve
• Harmonic and intermodulation distortion
• Stereo separation, crosstalk, and pilot frequency and injection level
• Occupied bandwidth measured against the emission mask

The Bessel-null method described earlier provides an absolute deviation reference that depends only on the accuracy of the audio frequency, which makes it the standard way to verify a deviation meter's own calibration. Broadcast installations supplement bench instruments with continuous modulation monitors that alarm on overmodulation, pilot loss, or carrier failure.

Practical Applications

Analog modulation continues to serve across broadcasting, safety-of-life communication, and instrumentation.

Broadcasting

• AM medium-wave broadcasting: 526.5 to 1606.5 kHz with 9 kHz channel spacing in ITU Regions 1 and 3, and 525 to 1705 kHz with 10 kHz spacing in ITU Region 2, the Americas
• Shortwave broadcasting: AM in allocated segments between roughly 2.3 and 26.1 MHz, still used for long-distance and international service
• FM broadcasting: 88 to 108 MHz in most of the world, 76 to 95 MHz in Japan, with 200 kHz channel spacing in the Americas and 100 kHz commonly used in Europe
• Analog television, where it remains in service, using VSB video with an FM aural carrier

Communications

• Aeronautical voice communication: AM from 118 to 137 MHz, with 25 kHz channels and 8.33 kHz channels in much of Europe. AM is retained deliberately, because an overlapping transmission produces an audible heterodyne rather than being silently captured, so controllers can detect a blocked call.
• Marine VHF: FM on channels defined by ITU Radio Regulations Appendix 18 between 156 and 162 MHz, including channel 16 for distress and calling
• Land mobile and public safety radio: narrowband FM across VHF and UHF allocations, increasingly alongside digital systems on the same channels
• Amateur radio: SSB for long-distance high-frequency work, FM for VHF and UHF repeaters, AM by preference among vintage-equipment operators
• Citizens band: AM and SSB on 40 channels from 26.965 to 27.405 MHz in the United States
• High-frequency maritime and aeronautical long-distance service: upper-sideband SSB voice

Specialized Applications

• Aerospace and range telemetry using FM subcarrier oscillators for analog sensor channels
• Instrumentation links using voltage-to-frequency conversion, where frequency modulation provides galvanic isolation and immunity to level errors along the path
• Analog magnetic recording, in which luminance in videotape formats and wideband data in instrumentation recorders were stored as FM so that tape amplitude variations did not corrupt the signal
• Navigation aids such as VOR, which encodes bearing in the phase relationship between a 30 Hz reference conveyed on a subcarrier and a 30 Hz variable component

Design Considerations and Best Practices

Selecting and implementing an analog modulation scheme means balancing several competing requirements.

Spectral Efficiency

SSB and VSB use the least bandwidth for a given message, and wideband FM the most. The available channel plan usually decides the question before any other consideration does: a 12.5 kHz land-mobile channel simply cannot hold a broadcast-quality FM signal.

Power Efficiency and Amplifier Class

Constant-envelope FM permits saturated Class C or Class E amplification at 70 percent efficiency or better, which is why handheld FM radios achieve useful battery life. AM and SSB carry information in the envelope and therefore require linear Class A, AB, or B stages with far lower efficiency and strict intermodulation limits. High-power AM broadcast transmitters recover much of that loss with digital or pulse-step modulation schemes that switch the final stage while synthesizing the envelope.

Noise Performance

Above threshold, FM output signal-to-noise ratio improves with the square of the modulation index, so raising the deviation ratio is a direct exchange of bandwidth for quieting. The link budget must keep the demodulator above threshold under worst-case fading, because performance below it collapses rather than degrades.

Frequency Stability

Stability requirements scale with narrowness. Broadcast FM tolerates hundreds of hertz of error; a 2.4 kHz SSB channel demands accuracy within tens of hertz for natural-sounding speech. Synthesized local oscillators referenced to temperature-compensated or oven-controlled crystals, and GPS-disciplined references in fixed installations, are the usual answers.

Adjacent-Channel Performance

Transmitter filtering and controlled modulation levels keep emissions inside the regulatory mask; overmodulation of an AM transmitter is a common cause of failing it. On the receiving side, selectivity, blocking dynamic range, and reciprocal mixing—the noise a local oscillator's phase noise translates into the passband when a strong adjacent signal is present—determine whether a nearby transmitter is merely present or genuinely disabling.

Troubleshooting Common Issues

AM Systems

• Overmodulation: Produces distortion and splatter into adjacent channels. Check audio drive level and the limiter, and verify negative-peak depth separately from positive.
• Carrier shift: A change in average carrier level with modulation, indicating amplifier nonlinearity or inadequate power supply regulation in the modulated stage.
• Low modulation: Weak audio wastes coverage, since talk power scales with modulation depth. Verify audio chain gain and modulator sensitivity.
• Distortion: Trace to improper bias, excessive drive, or an antenna mismatch that loads the final stage nonlinearly across the modulation cycle.
• Poor carrier or sideband suppression in SSB: Retrim the balanced modulator null and check the phasing network or sideband filter alignment.

FM Systems

• Frequency drift: Check the reference oscillator, its temperature compensation, and the modulation loop's lock condition.
• Deviation errors: Measure with a deviation meter verified by a Bessel null, then adjust modulator sensitivity. Under-deviation produces low audio at the far end; over-deviation produces distortion and adjacent-channel interference.
• Poor stereo separation: Verify pilot injection level and phase and check stereo encoder alignment; separation is exquisitely sensitive to phase error between the sum and difference paths.
• Threshold degradation: Points to insufficient signal level, a degraded receiver noise figure, a misaligned front end, or feedline and antenna loss.
• Incidental amplitude modulation: Often traced to poor limiting or to an amplifier stage being driven into a nonlinear region, and it shows up as noise after the discriminator.

General Issues

• Spurious emissions: Check output filtering, shielding, and grounding, and look for synthesizer spurs and mixer products that land in band.
• Intermodulation: Indicates nonlinearity somewhere in the chain, sometimes in a corroded external joint rather than in the equipment. Check bias points and drive levels first.
• Hum and noise: Examine power supply filtering, the grounding scheme, and shielding, and look for ground loops between racks and between audio and RF sections.

The Place of Analog Modulation Today

Digital modulation wins where spectrum is scarce and the content can tolerate coding delay, and that describes most new systems. Analog modulation nevertheless persists for concrete reasons rather than inertia alone:

• Installed base: Billions of AM and FM receivers exist, and no broadcaster abandons an audience lightly. Where analog broadcasting has ended, it has ended by deliberate policy—Norway completed the shutdown of its national FM networks in 2017—rather than through technical obsolescence.
• Graceful degradation: An analog signal fades into noise and remains partly intelligible; a digital signal works until it does not. That behavior matters at the edge of coverage in safety-of-life services.
• Latency: Analog modulation adds essentially no delay, while coded digital voice adds tens to hundreds of milliseconds—a real problem for air traffic control and for tactical coordination.
• Detectable interference: The heterodyne produced when two aircraft transmit at once is a feature of AM that digital capture would remove.
• Simplicity and cost: An analog link needs no codec, no synchronization, and no error control, which still matters in low-cost and low-power designs.
• Hybrid operation: HD Radio in North America and Digital Radio Mondiale elsewhere transmit digital carriers alongside or within the analog signal, allowing a gradual transition rather than a cutover.

Conclusion

Analog modulation systems represent more than a century of communications engineering, from the first envelope-detected voice transmissions to the multiplexed stereo baseband that still occupies the FM broadcast band. Their three mechanisms—varying amplitude, frequency, or phase—remain the only three available, and every digital scheme in use today is a discrete-valued application of the same three.

The engineering value of studying them lies in the trades they make explicit. Full-carrier AM buys receiver simplicity with transmitter power. SSB buys spectrum and power efficiency with carrier-recovery complexity and frequency accuracy. Wideband FM buys quieting with bandwidth, and forfeits it entirely below threshold. Those exchanges reappear, in the same form, in every modern radio link, which is why a working command of analog modulation remains the foundation for communications design and maintenance alike.

Related Topics