We use the phrase radio waves to mean electromagnetic waves that are transmitted across space (not just through air), as distinct from waves conducted along wires or in tubes ( waveguides). The existence of radio waves was predicted by the mathematician Clerk Maxwell in 1864 and later discovered experimentally by Heinrich Hertz in 1887, but no use was made of the waves until early in the twentieth century when Marconi (along with Popov, Tesla, and several others) used them for communications. They have come a long way in a period of about 100 years.

We now know fairly well what happens. When we cause a voltage to exist between any two points in space, the space itself is distorted, and we refer to the effect as an electric field. A varying voltage causes a variable distortion which appears to us as magnetism; we say that a magnetic field exists, meaning that this piece of space causes magnetic effects. This is a changing magnetic field, however, and its effect is to generate another voltage some distance away, which in turn causes a magnetic field and yet another voltage field, and so on. The unique feature of the waves is that the direction of oscillation of the magnetic wave is at right angles to the direction of oscillation of the electric waves (Figure 6.1).

These alternating fields are waves of both electricity and magnetism, called electromagnetic waves. They travel away from the point where they were generated at a speed of around 300,000,000 meters per second (often written as 3 × 10 ⁸m s ⁻¹), which is also the speed that has been measured for light, a strong hint that light is itself an electromagnetic wave. No air or any other material is needed to transmit these electromagnetic waves, though they can pass through insulating materials and they can also be conducted along metals and other conducting materials and through tubes called waveguides. Electromagnetic waves travel in straight lines, they can be reflected, refracted, and focused, and they can also be polarized (see later); all effects that are familiar from our knowledge of how light behaves.

In the early days of radio, all academic experts believed that long-distance communication by electromagnetic waves was impossible because of the curvature of the Earth. They would have known better if they had followed the work of Oliver Heaviside, who worked out that the effect of the radiation from the Sun would be to split atoms of gas in the upper atmosphere (the ionosphere) and provide a conducting layer that would reflect waves. Heaviside, however, was regarded as an eccentric engineer and was scorned by academics (does that sound familiar?), and much of his work was not published until after his death. Marconi and others disregarded the academics and pressed on with radio contacts over greater and greater distances, culminating in the transatlantic transmission in 1901. Academics then had to rediscover all that Heaviside had done and pretend that they had discovered it themselves. We will come back to the effects of the ionosphere later.

All electromagnetic waves travel at the same speed in space, and the speed is almost the same in air, though slower in denser materials. In addition, these waves have a measurable wavelength, meaning the distance from the peak of one wave to the peak of the next. As for water waves and sound waves, the speed, frequency, and wavelength are related by the equations: with f meaning frequency in hertz (Hz), v speed in meters per second (m s ⁻¹), and λ wavelength in meters (m). For example, if you generate a signal at 1 MHz, then its wavelength in space is 300,000,000 divided by 1,000,000, which is 300 m. A wave at 100 MHz, such as we use for FM radio, has a wavelength of 300,000,000/100,000,000 which is 3 m. Table 6.1 shows a few examples of frequencies and wavelengths.

It is not difficult to generate and launch waves in the frequency range that we use for radio. Anything that produces an alternating voltage will generate the signals, and any piece of wire will allow them to radiate into space. We find, however, that a launching wire, which we call an antenna (aerial in the UK), works best when its length is an even fraction (such as ½ or ¼) of the wavelength of the wave, or several wavelengths long. Lengths of half a wavelength or quarter of a wavelength are particularly useful, and a very efficient system, called a dipole, has been used for many years. These tuned aerials are not so efficient (and are difficult to construct) for the very long wavelengths, and they really come into their own for wavelengths of a few meters or less. Though we still make considerable use of the very long wavelengths, most of the communications equipment (FM radio, television, mobile phones) that we use nowadays depends on using the short wavelengths. The limitations of short wavelengths for long-distance communications are overcome by using satellites, so allowing a straight-line path for signals to and from the satellite.

A dipole antenna (Figure 6.2) for a wavelength that you want to use consists of two rod sections, each one-quarter of a wavelength long. If you imagine the inside portions connected to a signal, the voltage on the antenna rods will be in the pattern of a wave, and when the voltage is almost zero at the center it will be a maximum at the ends. This is the type of antenna that is very familiar to us and is used for television and FM radio reception, with the rods cut to the correct quarter-wavelength size for the nearest set of transmitters. The dipole is most efficient for the wave that is exactly of the size, but this is not too critical, particularly if the rods of the dipole are thick. The points where the ( feeder) cable is attached correspond to the points of maximum signal current so that the antenna behaves like a low resistance as far as signals of the tuned frequency are concerned.

Dipoles are also polarized. Polarization, as far as radio antennae are concerned, means the direction of the electric field of the wave, and this is often fixed by the design of the transmitting antenna as either vertical or horizontal. If the transmitter uses a vertical dipole, the signal will be vertically polarized, and the receiver antenna should also be vertically polarized. If the transmitter uses horizontal polarization, the receiver antenna should also be horizontally polarized. Different polarization directions are used to reduce interference between transmitters that use the same (or close to the same) frequency, and in some cases because one direction of polarization provides better reception conditions in difficult terrain than the other.

Early radio transmitters used electrical sparks to generate signals, which were a mixture of many frequencies, mainly in the very high-frequency (VHF)/ultrahigh-frequency (UHF) range, and at a short range these were seen to cause sparks to appear across the two sections of a receiving antenna. More sensitive ways of detecting the received signal soon followed, and a device called a coherer, consisting of metal filings in a glass tube, was used extensively, along with headphones, in the early days of radio as a way of detecting signals that were too feeble to make a spark.

For some time, large land-based transmitters used mechanical generators, alternators, working at about 20 kHz, and with huge long-wire antennae, so that their signals could span the Atlantic. These, and spark generators, were used until radio vacuum tubes were invented around 1906, though spark transmitters were still in use when the Titanic sank in 1912. An interesting point is that her distress signals were picked up in New York by a Marconi radio operator called David Sarnoff (who later founded the Radio Corporation of America when the US government threatened Marconi's with the equivalent of nationalization).

Reception of radio waves is simple enough: make a dipole or a single-wire antenna, and detect the small currents at the center terminals. Detection is a snag, though. At the start of the twentieth century, there was nothing that could amplify a signal at 20 kHz, and devices such as earphones could not respond to such higher frequencies or the much higher frequencies that were generated from sparks. The only way that the presence of the wave could be detected was by what we now call rectification, using some device that would allow the current to flow only one way and then detect the direct current (DC). The coherer with its metal filings was the first attempt to make such a device, but later it was discovered that crystals of the mineral galena (chemically, lead sulfide) could be used. The crystal was placed in a metal holder connected to earphones, and an antenna was connected to a fine wire, called a cat's whisker, jabbed against the crystal (Figure 6.3).

When a radio wave hit the antenna, the alternating current flowed only one way through the coherer or the crystal, and the headphones gave a feeble click, which sounded reasonably loud if you happened to be wearing the headphones. If the radio waves had been switched on and off again quickly, the sound in the headphones was a short double-click; if the wave was sustained for longer the gap between clicks was longer. By switching the transmitter on and off using a Morse key, a skilled listener could read the short (dot) and long (dash) intervals between clicks and write down the Morse code message that was being transmitted. This was the basis of radio as we used it right into the 1920s.

Neither coherer nor crystal was really satisfactory. The coherer had to be tapped regularly to settle the filings into place, and the cat's whisker had to be moved over the crystal surface to find a ‘sweet spot', a place where the action was most effective. This movement of the cat's whisker often had to be repeated, usually just as you particularly wanted to get the important part of a message.

Morse code served radio well and is still useful for emergency purposes because it requires the absolute minimum of equipment and it is the most efficient use of radio; even a low-power long-wave transmitter can have a very long range using Morse. It is not exactly useful for entertainment purposes, however, or for long-distance telephone calls, and though a method of carrying sound signals over radio waves had been demonstrated early on (on Christmas day in 1906), it took a long time to make this a commercial proposition (in 1920, in fact). The use of radio waves to carry the information of sound waves requires what we call modulation, and in the early days of radio this was always the type we call amplitude modulation ( AM).

To start with, we have to change the sounds, whether speech or music, into electrical waves, using a type of transducer called a microphone. Microphones were, thanks to the use of telephones, reasonably familiar devices in 1906 (in the USA at least), so that the only obstacle to broadcasting in these early days was knowing how to make a radio wave of 100 kHz or more carry a wave, the audio signal, of a much lower frequency, about 300 Hz. The simplest answer is to make the amplitude of the radio wave rise and fall in sympathy with the lower (audio) frequency (Figure 6.4). This is what we mean by amplitude modulation: the amplitude of the carrier wave is controlled by the amplitude of the audio signal.

Amplitude modulation had been possible even in the early days of spark transmitters, but only with difficulty and using primitive equipment. The use of radio vacuum tubes, or just ‘tubes', in the USA, made efficient modulation possible, because the vacuum tube (called a valve in the UK) allows a large current to be controlled by a small voltage. In addition, another tube can be used as an oscillator, generating the radio waves and replacing the mechanical or spark type of generators. The first radio station intended for entertainment started operation in November 1920 at Pittsburg, Pennsylvania. The usual way of modulating was to control the DC supply to the output tube of the transmitter, using another tube, and applying the amplified audio signal as the input of this tube.

When a radio wave is modulated in this way, it is no longer a simple sinewave. If we could graph the waveform we would find the type of shape illustrated in Figure 6.4, but there is another effect that is shown only when we graph the signal amplitude against frequency rather than against time.

Suppose, for example, that we modulate a 100 kHz radio wave with a 10 kHz sinewave signal. Analyzing the frequencies that are present shows that three main waves now exist. One is at 100 kHz and is the carrier wave, the wave that is generated by an oscillator. There is also a wave with a frequency of 90 kHz, called the lower sideband, and one with a frequency of 110 kHz, called the upper sideband. These sidebands appear only when the carrier is being modulated, and their amplitude is always less than that of the carrier (Figure 6.5). In general, if the carrier frequency is F and the signal frequency is f, the sidebands are at F− f and F+ f, so that the total bandwidth of the transmitted wave is from F− f to F+ f, making a bandwidth of 2 f.

If we modulate the carrier with a signal that consists of a mixture of frequencies up to 10 kHz in bandwidth, the sidebands will show a pattern such as appears in Figure 6.6, showing a spread of sideband frequencies rather than just two single frequencies. In each sideband, the frequencies closest to the carrier frequencies are due to the lower frequencies of the modulating signal, and the frequencies furthest from the carrier frequency are due to the higher frequencies of the modulating signal.

All we need to carry the information is one of these sidebands, so that both the carrier frequency and the other sideband are redundant. There are ways that can be used of eliminating the carrier ( suppressed carrier transmission) and one sideband ( single-sideband transmission) from the transmitted signal, but for entertainment purposes the additional complications at the receiver make these methods undesirable, though both stereo radio and color television make some use of these methods for additional signals (using sub-carriers; see later). Digital radio is quite another matter, as discussed later.

At the receiving end, vacuum tubes were still a far-off dream for the few people in 1920, mainly amateur enthusiasts, who could get hold of crystals and headphones and rig up an antenna. Some 1000 listeners probably heard the 1920 US presidential election results in this way, and in later years this number multiplied rapidly. Even in controlled and restricted Britain listeners experimented, trying to hear the trial transmissions from the Marconi workshop in Writtle (Essex) and even from continental broadcasts when conditions and geographical position suited. In 1922, a private company, the British Broadcasting Company (BBC), was formed and it set up nine transmitters, each with a distinctive call sign of digits and letters. Table 6.2 shows these nine with opening dates and frequency allocation.

By the 1930s, radio as we know it was a reality. The BBC had been nationalized so that the government could control it and tax it, but in the USA broadcasting remained free in every sense and consequently the USA remained at the front of technical progress. Small tubes became available and were used in receivers, and by the 1930s loudspeakers were normally used in place of headphones so that the whole family could listen in. Amplitude modulation was used for transmissions with frequencies in the medium waveband of about 530 kHz to 1.3 MHz, and the receivers developed from the simple one-tube pattern to multitube types, some of the expensive models featuring an added electronic record-player to make a radiogram. By the late 1930s a few lucky people in the London area could, if they could afford it, buy a combined radio and television receiver (though television was confined to a couple of hours in the evening, and a receiver cost as much as did a small house at that time).

A severe problem with the use of AM and the medium waveband was interference between stations. Medium-wave broadcasts can be picked up over quite long distances, up to 1000 miles (over 600 km), and where two stations used frequencies that were close, they set up an interference that could be heard as a whistle at a receiver tuned to either station. The whistle, in fact, is at the frequency that is the difference between the two transmitted frequencies. For example, if one station transmits at 600 kHz and another at 605 kHz, the whistle is heard at a frequency of 5 kHz. As new radio transmitters were introduced all over the world, it became quite impossible to share out frequencies on the medium waveband, and interference became the common experience for all listeners, particularly after dark when the reflecting layers in the ionosphere moved higher and reflected waves that originated from greater distances.

A way of getting around these problems had been devised by an inventor who, of all that ill-treated fraternity, was the least fortunate. Edwin Armstrong developed frequency modulation (FM) to eliminate the effects of interference, both artificial and natural, and built a radio station (now a working museum) in the Catskill mountains just outside New York to prove the point. As the name suggests, the low-frequency audio signal is carried by altering the frequency of the radio signal (Figure 6.7), with the amplitude kept constant. Since interference of all types only alters the amplitude of radio waves, FM can be virtually free from interference unless it is so severe as to reduce the radio wave amplitude to almost zero. FM is used universally today for high-quality broadcasts and for short-range local radio because if two FM transmitters are using almost the same frequency, a receiver will lock on to the stronger transmission, ignoring the other one and with no whistle effects. There is strong resistance to the proposal to turn off FM broadcasts all over Europe because of the poor coverage of digital radio (digital audio broadcasting or DAB) in so many parts of the UK. FM broadcasting in the UK did not start until 1955, though there were some commercial FM transmitters in the USA by 1941.

Stereo sound from discs and from tape was well established by the 1950s, but stereo sound broadcasting took a little longer. Stereo sound at its simplest means that two microphones are used for a recording, each picking up a slightly different signal. When these two signals are used to power separate loudspeakers, your ear detects a subtle difference in the sound; the effect on the ears is like that of 3D cinema on the eyes. Stereo recording on tape uses separate tracks for the signals; and on disc uses two tracks on one groove. We will look at these systems in Chapter 7.

Stereo broadcasting could be achieved by using two separate transmitters, one for each channel, but this would be very wasteful and the stereo effect can be disturbed if the two frequencies are subject to interference at different times or to different extents. Though the first experiments in stereo broadcasting in the UK (in the 1950s) used separate frequencies, it was obvious that stereo broadcasts would not succeed unless two conditions were met. One was that a single frequency must be used and the other was that users of mono receivers should still be able to pick up a mono signal when they tuned to a stereo broadcast. As it happened, these were very similar to the conditions that had been imposed on color television systems in the USA (see Chapter 8) in 1952, and similar methods could be adopted.

The system that evolved in the USA and which was adopted with only a few modifications in the UK was one that used a sub-carrier. As the name suggests, this is a carrier, which can be modulated with audio signals, and which is then itself used to modulate the main carrier. The frequency of the sub-carrier must be somewhere between the frequency of the main carrier and the highest frequency of the signals. The other important step was how to use the left (L) and right (R) channel information. The solution was to modulate the main carrier with the sum of the channel signals, L + R, which allowed any receiver tuned to the correct frequency to pick up this signal, which is a mono signal.

The stereo information was then contained in another signal which was the difference between the channels, L − R. This difference signal has a much smaller amplitude than the L + R signal and is needed only by stereo receivers, making it suitable for modulating on to a sub-carrier.

Figure 6.12 shows the frequency bandwidths of a stereo transmission before modulation on to the main carrier. The bandwidth that will be needed by the mono signal (L + R) is 15 kHz, and whatever other frequencies we use must not overlap this set. A single low-amplitude 19 kHz signal is also present; this is the pilot tone and it is used at the receiver (see later). The L − R signals are amplitude modulated on to a 38 kHz sub-carrier, and this sub-carrier is then removed (making it a suppressed carrier), leaving only the sidebands of the modulated signal along with the pilot tone. These sidebands also need a bandwidth of about 15 kHz on each side of the 38 kHz mark. This set of different signals in different frequency ranges is used to frequency-modulate the main carrier of around 100 MHz.

Why remove the sub-carrier and supply a 19 kHz sinewave signal? The sub-carrier contributes nothing to the signal; it carries no information. It does, however, require transmitter power, and removing the sub-carrier makes an appreciable saving, allowing the transmitter to carry more of the useful sideband signals. The L − R signals cannot easily be demodulated, however, without supplying a 38 kHz signal in the correct phase, and this can be supplied locally at the receiver providing there is a small synchronizing signal available. This is the purpose of the 19 kHz pilot tone, whose frequency can be doubled in the receiver to 38 kHz and then used to correct the phase and frequency of the local oscillator in the receiver. The transmitter power used for the 19 kHz tone can be small, because the amplitude of this signal is small. As a bonus, the 19 kHz wave can be used at the receiver to turn on an indicator that shows that a stereo signal is being received.

Figure 6.13 shows the block diagram for transmission. The separate L and R channel signals are added and subtracted in a circuit called a matrix to give the L + R and L − R signals, and the L + R signal is used to frequency-modulate the main carrier. A master 38 kHz oscillator is used to provide the carrier for the L − R signals, and the 38 kHz signal is also used to provide a 19 kHz signal for the pilot tone, which is also frequency-modulated on to the main carrier. The modulated L − R signal is passed through a filter which removes the 38 kHz sub-carrier, so that the sidebands can also be frequency-modulated, along with the pilot tone, on to the main carrier. The modulated main carrier is then amplified and used to supply the transmitting antenna.

Figure 6.14 is the block diagram of the stages following the demodulator for a stereo FM receiver. The early stages of the receiver are exactly the same as they would be for a mono FM receiver, right up to the point where the FM signal is demodulated, providing three sets of signals. A low-pass filter separates off the L + R main signal and a resonant circuit separates out the 19 kHz pilot tone, and this is used to control the phase and frequency of a 38 kHz oscillator. Finally, a high-pass filter separates out the sidebands of the L − R signals.

A circuit called a phase-sensitive demodulator has inputs of the 38 kHz carrier frequency (obtained by doubling the 19 kHz pilot tone or by using it to synchronize an oscillator) and the L − R sidebands, and its output is the L − R signal itself. The L + R and L − R signals are fed into another matrix circuit, producing the L and R signals at the outputs. If you wonder how this is done, think of what happens when you add the inputs: and also subtract them: so providing the separate L and R signals.

Snags? The stereo signal is more easily upset by interference, particularly car ignition and other spark interference. A mono signal can use a low-pass filter to remove much of the effects, but this would remove the L − R signals from a stereo transmission. In addition, the noise level of a stereo transmission is always higher because the bandwidth is greater, and noise depends on bandwidth, the only noise-free signals are those with zero bandwidth, and signals with no bandwidth carry no information. Many FM receivers will automatically switch to mono reception when the noise level is high.

In fact, because the transmitters for digital radio in the UK use the old ITV transmitters (in the days before UHF television) the coverage is poor. This was not such a problem for television because houses had antennae on the roofs. Digital radio receivers that use a simple telescopic whip antenna receive a weak signal, and this can cause a most objectionable sound like boiling mud. In addition, digital radio transmitters often compress the sound signal (see later) so much that the quality suffers compared to a good FM receiver. Some misleading advertising suggests that digital provides a better quality, but this has certainly not been borne out by experience.

Digital radio was supposed to be at its best in cars, where its use eliminates the need to keep retuning to a local FM station as the car leaves one transmitter region and approaches another. The main advantage of digital radio is that a large number of digital radio transmitters can be used in a small amount of bandwidth without interference problems. This is more of an advantage for the broadcasters than for the listeners. Car manufacturers have been very reluctant to make the change, and have had to be bullied into it. The only real use for a digital car radio is that you can cruise around to find where the signal can be picked up, and car manufacturers will probably play safe and install radios that can be switched to FM when the digital signal is unobtainable or of low quality.

Another possibility for digital radio broadcasting is the use of broadband Internet. This obviously requires a computer, and the sound from computer loudspeakers is fairly poor unless you have invested in high-quality units. With WiFi (Internet signals distributed from a small wireless transmitter in your house) the signals can be relayed to audio equipment, providing a signal quality that can be superior to the DAB signal. If you want to check out what can be achieved, look at the website http://www.shoutcast.com/, which is just one of the many sites offering radio by Internet. At the time of writing Internet radios (not part of a computer, but needing a WiFi signal) were just being advertised in the UK.