What a radio wave can carry
A radio transmitter produces an oscillating electromagnetic wave. On its own, a steady wave carries no information, just as a steady hum carries no message. To send something, the wave must be changed in some way that a receiver can detect. The history of radio engineering is largely a history of better ways to make those changes, and each step needed a new kind of equipment.
Spark transmitters
The earliest practical wireless systems used a spark gap. A high-voltage electrical discharge across a gap sets off a burst of oscillation in a circuit, and the antenna radiates it as a short, rapidly fading wave. Heinrich Hertz used sparks in the late 1880s to show that such waves exist, and Guglielmo Marconi and others built them into communication systems in the 1890s. This field, then called wireless telegraphy, aimed to rival the electric telegraph without needing wires.
Spark signals were an on-off code. An operator keyed the spark to make dots and dashes of Morse code, and the receiver, often a coherer or later a crystal or magnetic detector, clicked or buzzed in response. The method worked for ships and coastal stations, but it had drawbacks. A spark makes a broad, messy signal that takes up a wide range of frequencies, so two stations near each other tended to interfere. It could not carry speech in any practical way.
Continuous waves
Engineers wanted transmitters that produced a clean wave at a single frequency, which would allow tuning and carry sound. Several approaches appeared in the first years of the 1900s.
- The arc transmitter, developed by Valdemar Poulsen and others, used an electric arc to sustain oscillation.
- The high-frequency alternator, with contributions from Reginald Fessenden and Ernst Alexanderson, was essentially a generator that spun fast enough to produce radio frequencies. It was used for powerful long-wave stations.
- The vacuum tube oscillator, developed in the 1910s, was smaller, cheaper and more controllable than either. It became the standard for the following decades.
The vacuum tube grew out of work by John Ambrose Fleming, who patented a diode valve in 1904, and Lee de Forest, whose three-element Audion followed in 1906 and 1907. Many people contributed to turning the three-element tube into a reliable amplifier and oscillator, and patent disputes over the tube and its circuits ran for years, so assigning the credit to one person is not straightforward.
Amplitude modulation
With a steady carrier available, sound could be added by varying the carrier’s strength in step with the sound wave, which is amplitude modulation (AM). Fessenden is often credited with demonstrating the principle in the first years of the century, although reliable voice transmission needed the later development of tubes. By the 1910s, researchers and amateur experimenters were sending voice and music, and by the 1920s AM was the foundation of broadcasting.
On the receiving side, a tuned circuit picks out the wanted station, a detector recovers the sound from changes in strength, and an amplifier makes it loud enough for a speaker. Two receiver developments are especially associated with Edwin Armstrong. One is the regenerative circuit, around 1912, which feeds some output back to the input for much greater sensitivity. The other is the superheterodyne receiver, from around 1918, which shifts every station to the same intermediate frequency so that filtering and amplification can be done well. Priority claims on these circuits were contested, and courts and historians have not always agreed.
AM had a basic weakness: noise changes signal strength, so static is heard as part of the program. Automatic gain control and better circuits helped, but they did not remove the problem.
Frequency modulation
Frequency modulation (FM) keeps the carrier strength constant and varies the frequency instead. A receiver can then ignore variations in strength, and most natural and man-made static is rejected. Some engineers, notably John R. Carson in a 1922 analysis, argued that FM offered no benefit compared with AM when used in a narrow band. Armstrong showed in the 1930s that a wide frequency swing gives a strong improvement in noise performance, at the price of using more spectrum. His patents date from 1933, and he demonstrated the system in 1935.
The idea of varying frequency was not entirely new to him, and some earlier researchers had considered related schemes, so the details of “who invented FM” depend on what is counted. Armstrong’s contribution is the practical wide-band method and its demonstration. Wide-band FM also supports high-fidelity sound and, later, stereo, which helps explain why FM became the preferred band for music.
The pattern behind the steps
Looking back, a pattern emerges. Each stage solved the biggest weakness of the previous one, and each depended on components that arrived from other fields:
- Spark telegraphy proved that waves could carry code across distance, but it was crude.
- Continuous waves allowed tuning and the addition of speech.
- AM made broadcasting possible, but noise stayed a problem.
- FM traded bandwidth for clarity.
Later work, including transistors and the integrated circuits that are now standard, changed the hardware without changing these basic ideas. Today, modulation schemes used in cellular networks and digital radio are far more elaborate, but they still alter the amplitude, frequency or phase of a carrier wave, or a combination of those. The vocabulary of the early radio engineers lives on in these systems.