Limitations of AM
AM plainly works — it reached most of the planet and it is still on the air. So why did engineers spend two decades looking for something else? Four complaints. Three are fixable. One is not.
A 50 kW Electricity Bill
At m = 1 a 50 kW medium-wave station radiates 33.3 kW of carrier that conveys nothing. The sidebands get 16.7 kW — and since they are mirror images, only 8.3 kW of that is non-redundant. One sixth of the bill is doing the job.
Efficiency Peaks at One Third
Divide the sideband power by the total and this is what you get. It rises with m and tops out at exactly 1/3 when m = 1. There is no modulation depth at which standard AM is efficient.
Where the Power Goes
Real Programme Material Is Quieter
Speech and music spend most of their time well below their peaks. At an average depth of m = 0.3 the efficiency is 0.09/2.09 = 4.3% — a handicap of −13.7 dB before the signal has travelled one metre.
Noise Is the Signal
Thermal noise, lightning, a switching supply, a brush motor — every disturbance adds to the wave, and adding changes amplitude. AM put the message in the amplitude. So the detector has no way, even in principle, to tell an envelope change the transmitter made from one a lightning strike made.
There Is No Limiter
An FM receiver clips the incoming wave to a constant height and throws every amplitude variation away, noise included, because none of its message was in the amplitude. Put that circuit in front of an envelope detector and it erases the message and keeps the noise.
Fading That Is Selective
Plain fading only makes a station quiet, and the AGC handles it. But an HF delay spread of about 1 ms puts the channel’s coherence bandwidth near 160 Hz — and a carrier and its sidebands are kilohertz apart. They fade independently.
The Carrier Fades, m Doubles
Let the carrier fade 6 dB while the sidebands do not. Its amplitude halves, so the modulation index the receiver sees becomes 2m. A well-behaved m = 0.6 arrives as 1.2 — over-modulation and envelope reversal, with nothing changed at the transmitter.
Every Impulse Is a Click
Commutator motors, dimmers, switching supplies, power lines, cheap LED drivers. All broadband, all impulsive, all strongest at exactly the low frequencies AM broadcasting lives on. An impulse is a large brief amplitude event — the worst possible shape of interference for a detector that reads amplitude.
No Capture Effect
Two AM signals on one frequency simply add, because that is what amplitudes do. You hear both stations plus a whistle at the difference between their carriers, and nothing in the receiver can pick one. FM’s limiter suppresses the weaker almost completely — so AM’s answer has to be regulatory: licensed spacing, protected contours, power cuts at sunset.
Redundancy, Not Width
AM is narrow — 2fm beats everything in Module 4. The waste is that the two sidebands are mirror images, so DSB spends 2fm to deliver fm. And the bill is paid in fidelity: a 10 kHz channel halved leaves about 5 kHz of audio, which is why AM sounds dull.
What a Fix Would Need
- Put the information where amplitude noise cannot reach — frequency or phase
- Keep the envelope constant, so a limiter can flatten it safely
- Spend all the power on the message — no idle carrier term
- Offer a trade: a knob that buys noise immunity with bandwidth
- Degrade to one usable station rather than two ruined ones
Carson argued in 1922 that frequency modulation only widened the band. He was right for narrowband — and Armstrong showed in 1933 that going deliberately wide changes the answer.
When the Defect Is a Feature
Aviation voice is AM partly because there is no capture effect. If two aircraft transmit at once, both are heard — garbled, whistling, obviously overlapping. Under FM the weaker aircraft would simply vanish, with no indication it had ever spoken.
Key Takeaways
- Efficiency is m²/(2+m²) — peak 1/3, and only 4.3% at a real m = 0.3
- That carrier buys the diode receiver: it is a trade, not pure waste
- The irreparable defect: noise is amplitude, and so is the message
- Selective fading is the same defect — a faded carrier doubles the effective m
- No capture effect, so co-channel signals just add
- The bandwidth complaint is redundancy, not width — SSB fixes it, coherently
Module 3 Complete!
You have built amplitude modulation from the idea to the equation, taken it apart with a diode, and totalled up its bill. Everything on the fix-it list points the same way: keep the amplitude constant and put the message in the frequency.