At the time of writing, the changeover to digital television in the UK is well under way and is due to be completed in the summer of 2011, and the changeover in Australia that started in 2001 should be complete by 2013. The change was completed in the USA in 2009.

The methods of coding the picture information for digital transmission are very different from those used in the older (analog) system, but many of the basic principles of television are much the same. In this chapter we will concentrate first on the older analog system because these receivers can have a long life and may still be in use (using a digital set-top box converter) over the lifetime of this book, and look later at the differences brought about by digital television.

Oddly enough, television has had a rather longer history than radio, and if you want to know more of its origins there was a book called Birth of the Box, now out of print, which dealt with the fascinating development of this subject. Almost as soon as the electric telegraph (invented in 1837) allowed messages in Morse code (invented in 1838) to be sent along wires, inventors tried to send picture signals, and this activity was spurred on by the invention of the telephone. Sound transmission along wires is simple, because it requires only that the sound waves be converted to electrical waves of the same frequency. A picture, however, is not in a usable wave format, but by the middle of the nineteenth century the principle of scanning a picture had been established.

Suppose, for example, that you drew a picture on a grid pattern such as in Figure 8.1. If each square on the grid can be either black or white, you could communicate this picture by voice signals to anyone with an identical grid by numbering each square and saying which squares were white and which were black. It may sound elementary, but this is the whole basis of both television and fax machines (which are almost identical in principle to the first television transmitters and receivers). This grid could be described as a 12 × 12 picture, but for television we need a much larger number of squares, typically about 500 × 300. Currently, computers use 640 × 480 as a minimum, with 800 × 600, 1024 × 768, or higher, more often used. Analog television more often specifies only the vertical resolution in terms of the number of lines (currently 625 in the UK and Europe, 525 in the USA), though not all of the lines carry picture information. This makes it difficult to compare the performance of analog television receivers with computer monitors. HD television displays can use 1024 × 600 for 16:9 pictures (see later).

Early television systems, patented around 1870 by Nipkow and Rosing in St Petersburg, and subsequently constructed by John Logie Baird in the late 1920s and early 1930s, used mechanical scanning (with rotating discs or mirrors) and were confined initially to still pictures of around 30 scanned lines, in black and white, with no shades of gray. Television as we know it had to wait until electronic components had been developed, and in particular, the receiver cathode-ray tube (CRT) and the camera tube.

By the early 1930s, however, Philo Farnsworth in San Francisco had demonstrated that electronic television was possible (at almost the same time as Baird had made a sound broadcast declaring that CRTs could never be useful for television). Also at that time, a form of radar had been used to measure the distance from Earth of the reflecting layers of charged particles in the sky (the Heaviside and Appleton layers). These two forms of technology were to play decisive roles in the development of electronics in the twentieth century.

The attraction of the CRT is that it can carry out scanning by moving an electron beam from side to side and also from top to bottom, tracing out a set of lines that is called a raster. Nothing mechanical can produce such rapid changes of movement as we can obtain by using an electron beam, so that most of the television pioneers realized that this was the only method that would be acceptable for reasonable definition, though the cinema industry of those days (which backed Baird) still believed that mechanical systems might prevail. By the 1930s, it was becoming acknowledged that any acceptable form of television receiver would certainly use a CRT, and the main research effort then went into devising a form of CRT that could be used for a television camera. The principles of such a tube were: Without going into details, then, a camera CRT must allow the image projected on to the face of the tube to be converted into electron charges, and these charges are scanned by an electron beam from which it must produce a signal current or voltage that can be amplified so as to represent the brightness of each part of the scanned picture.

We can look now at a block diagram of the studio and transmission side of black-and-white television in the days when CRTs ruled. We will look later at the additions that have to be made for color.

Television transmission at its simplest starts with a television camera, which is fed with synchronizing pulses from a master generator. Light from the scene that is to be transmitted is focused through a lens on to the face of the camera tube, and this light image is converted into an electron charge image inside the front section of the tube. The scanning electron beam discharges this, and the discharge current is the video signal. This is the starting point for the block diagram of Figure 8.6. The video signal, which is measured in microvolts rather than in millivolts, has to be amplified, and the synchronizing pulses are added. The generator that supplies the synchronizing pulses also supplies the scans to the camera tube. The complete or composite video signal that contains the synchronizing pulses is taken to the modulator where it is amplitude-modulated on to a high-frequency carrier wave.

Modern analog television systems modulate the signal so that peak white is represented by the minimum amplitude of carrier and the synchronizing pulses by the maximum amplitude. In the original UK television system the signals were modulated the other way round, with peak white represented by maximum carrier amplitude, but this method allowed interference, particularly from car-ignition systems, to show as white spots on the screen. When the system changed to 625 lines in 1968, maximum carrier amplitude was used for the sync pulse peaks in line with the methods used in other countries. Using this system, interference causes black spots that were much less noticeable, and the compulsory use of suppressor resistors in car-ignition systems has almost eliminated the main source of television interference.

The sound signal has been omitted from this to avoid complicating the diagram. The sound is frequency-modulated on to a separate carrier at a frequency higher than that of the vision carrier, and is transmitted from the same antenna.

An analog television receiver for monochrome (Figure 8.13) has to deal with several signal-separating actions. The signal from the antenna is processed in a straightforward superhet type of receiver, and the methods that are used differ only because of the higher frequencies and the larger bandwidth. The vision and sound carriers are received together, and so the intermediate-frequency (IF) stages need to have a bandwidth that is typically about 6 MHz, enough to take the wideband video IF and the sound IF together. The real differences start at the demodulation block. A typical IF range is 35–39.5 MHz (for amplitude 6 dB down from maximum).

Figure 8.14 illustrates the IF response, which must be broad enough to include both sound and vision carriers and sidebands, yet provide enough filtering to exclude signals from the adjacent frequencies. These adjacent frequencies arise from the mixing of the superhet oscillator frequency with the signals from other transmitters.

The demodulator is an amplitude demodulator, and at this stage the composite video signal with its sync pulses can be recovered. The effect of this stage on the IF for the sound signal is to act as a mixer, and since the sound IF and the vision IF are 6 MHz apart, another output of the demodulator is a frequency-modulated 6 MHz signal (the intercarrier signal) which carries the sound. This is separated by a filter, further amplified and (FM) demodulated to provide the sound output.

The video signal is amplified, and the synchronizing pulses are separated from it by a selective amplifier. These pulses are processed in a stage (the sync separator) which separates the line and the field pulses by using both differentiating and integrating circuits (Figure 8.15). The differentiated line pulses provide sharp spikes that are ideal for synchronizing the line oscillator, and the field pulses build up in the integrator to provide a field synchronizing pulse. The effect of the differentiator on the field pulses is ignored by the receiver because the beam is shut off during this time, and the effect of the line pulses on the integrator is negligible because these pulses are too small and too far apart to build into a field pulse.

The synchronized line and field timebase oscillators drive the output stages that deflect the electron beam, using magnetic deflection (passing scan currents through coils). The fast flyback of the line scan current causes a high voltage across the transformer that is used to couple the line scan output to the deflection coils, and this voltage is used to generate the extra-high-tension (EHT) supply of around 14–24 kV that is needed to accelerate the beam.

Meantime, the video signal has been amplified further and is taken to one terminal (cathode) of the CRT, with another terminal (grid) connected to a steady voltage to control brightness. Another set of circuits, the power supply unit (PSU), uses the alternating current (AC) mains supply to generate the steady voltages that will be used by the receiver circuits.

Though Sam Goldwyn is reputed to have said that he would believe in color television when he saw it in black and white, the idea of color television is not new, and methods for transmitting in color have been around as for as long as we knew that black-and-white (monochrome) television was possible. Color television, like color printing and color photography (both demonstrated in the 1880s), relies on the fact that any color seen in nature can be obtained by mixing three primary colors. For light, these colors are red, green, and blue, and the primary colors used by painters are the paint colors (red, yellow, and blue) that absorb these light colors. To obtain color television, then, you must display together three pictures, one consisting of red light, one of blue light, and one of green light. This implies that the color television camera must generate three separate signals from the red, blue, and green colors in an image.

All of the early color television systems worked on what is called a sequential system, meaning that the colors were neither transmitted nor seen at the same time, relying on the optical effect called persistence of vision. This means that the eye cannot follow rapid changes, so that showing the red components of a picture followed rapidly by the blue and the green will appear to the eye as a complete color picture rather than flashing red, blue, and green. We rely on the same effect to fool the eye into believing that we are watching a moving picture rather than a set of still images.

A typical early method was the frame sequential system. Each picture frame was transmitted three times, using a different color filter for each of a set of three views so that though what was transmitted was black and white, each of a set of three frames was different because it had been shot through a color filter, one for each primary color. At the receiver, a large wheel was spinning between the viewer and the television screen, and this wheel carried a set of color filters. The synchronization was arranged so that the red filter would be over the CRT at the time when the frame containing the red image was being transmitted, so that this filter action put the color into the transmitted monochrome picture. The main snag with this system is that the frames must be transmitted at a higher rate to avoid flickering, and there are also problems with compatibility and with the synchronization of the wheel (and its size). A more realistic option was a line-sequential system, with each line being shown three times, once in red, once in green, and once in blue.

Several early television systems were devised to show still color pictures, but the first commercially transmitted color television signals were put out in 1948 by CBS in New York, using a combination of electronic and mechanical methods. The system was not successful, and a commission on color television decided that no scheme could be licensed unless it was compatible. In other words, anyone with a monochrome receiver had to be able to see an acceptable picture in black and white when watching a color broadcast, and anyone with a color receiver had to be able to see an acceptable black-and-white picture when such a picture was being transmitted (to allow for the use of the huge stock of black-and-white films that television studios had bought). This ruled out the use of frame-sequential systems, and almost every other consideration ruled out the use of mechanical systems allied to the CRT.

Radio Corporation of America (RCA) had been working on simultaneous systems throughout the 1940s, and their demonstrations in 1952 convinced the commission, the National Television Standards Committee (NTSC), that the RCA system was suitable. All other systems have taken this system as a basis, but differences in details mean that NTSC television pictures are not compatible with the PAL color system used in Germany, the UK, and other parts of western Europe, or the SECAM system used in France, former French colonies, and eastern Europe. This is also why you cannot exchange video cassettes or DVDs with friends overseas, and the hope that digital systems might be compatible has been dashed because the method of coding sound used in the USA is not the same as that used in Europe, though the vision signals are compatible. Even DVDs have to be separated into PAL, NTSC, and SECAM groups.

Let’s start with the portions that are common to all analog systems. To start with, the image is split into red, green, and blue signal components, using prisms, so that three separate camera tubes or semiconductor CCD detectors can each produce a signal for one primary color. This is the starting point for the block diagram of Figure 8.16. The signals are mixed to form three outputs. One is a normal monochrome signal, called the Y signal. This is a signal that any monochrome receiver can use and it must be the main video signal so as to ensure compatibility. The other two outputs of this mixer are called color-difference signals, obtained by using different mixtures of the red, green, and blue signals. These color-difference signals are designed to make the best use of the transmission system and are of considerably lower amplitude than the monochrome signal, because in any television signal, the monochrome portion always carries much more information. Color is less important, and all the fine details of a picture can be in monochrome only because the eye is less sensitive to color in small areas. These color-difference signals are lettered U and V.

The color-difference signals are transmitted using a sub-carrier, a method that we have looked at already in connection with stereo radio. This time, the use of the sub-carrier is much more difficult because it has to carry more than one signal. To ensure compatibility, the sub-carrier frequency must be within the normal bandwidth of a monochrome signal, so that you would expect it to cause a pattern on the screen. As it happens, by choosing the sub-carrier frequency carefully, and keeping its amplitude low, it is possible to make the pattern almost invisible. The other problem is that this sub-carrier has to be modulated with two signals, and this is done by modulating one color-difference signal on to the sub-carrier directly (using amplitude modulation) and then using a phase-shifted version of the sub-carrier, with phase shifted by 90°, and modulating the other color-difference signal on to this phase-shifted sub-carrier.

If you remember that a 90° phase shift means that one wave is at zero when the other is at a maximum, you will appreciate that adding these two modulated sub-carriers together does not cause them to interfere with each other. The doubly modulated sub-carrier can now be added to the monochrome signals (remember that the sub-carrier has a considerably lower amplitude) and the synchronizing signals are added. This forms the color composite video waveform which can be modulated on to the main carrier. What this amounts to is that the amplitude of the modulated signal represents the saturation of color and the phase of the modulated signal represents the hue.

Synchronizing is not just a matter of the line and field pulses this time. A color receiver has to be able to generate locally a copy of the sub-carrier in the correct phase so that it can demodulate the color signals. To ensure this, ten cycles of the sub-carrier are transmitted along with each line-synchronizing pulse, using a time when the normal monochrome signal has a gap between the line-synchronizing pulse and the start of the line signal (Figure 8.17). This color sync interval is called the back porch of the synchronizing pulse. The illustration also shows how some colored bars appear on a signal, as viewed by an oscilloscope (see Chapter 17). The sub-carrier in these signals is represented by shading. If the sub-carrier were not present, these bars would appear in shades of gray.

A fair amount of the circuitry of an analog color television receiver is identical to that of a monochrome receiver. The superhet principle is used, and the differences start to appear only following the vision detector, where the video signal consists of the monochrome signal together with the modulated sub-carrier. The U and V color-difference signals have to be recovered from this sub-carrier and combined with the monochrome signal to give three separate R, G, and B signals that can be used on the separate guns of the color tube.

A simplified outline of a receiver, showing the usual superhet stages as one block and neglecting differences between systems, is shown in Figure 8.20. The video signal has the 6 MHz sound signal filtered off, and the vision signal is amplified. At the output of this amplifier, other filters are used to separate the sub-carrier signal, which (in the UK) is at 4.43 MHz, with its sidebands, and the monochrome or luminance signal, with the sub-carrier frequencies greatly reduced by filtering, is separated off and amplified. The sub-carrier ( chrominance) signal has to be demodulated in two separate circuits, because we have to recover two separate signals, U and V, from it. This requires a circuit in which the unmodulated sub-carrier frequency, in the correct phase, is mixed with the modulated sub-carrier signal. This type of circuit is called a synchronous demodulator.

A pulse is taken from the line timebase and used to switch on gates, circuits that will pass signal only for a specified interval. One of these gates allows the color-synchronizing ‘burst’ to pass to the local sub-carrier oscillator to maintain it at the correct frequency and phase. This oscillator output is fed to a demodulator whose other input is the modulated sub-carrier, and this has the effect of recovering the V signal. The oscillator output is also phase shifted through 90° and this shifted wave is used in another demodulator to recover the U signal. The U, V, and Y (monochrome) signals are then mixed to get the R, G, and B signals that are fed to the CRT guns.

That is the simplest possible outline, and it is quite close to the original NTSC system for receivers. The PAL receiver incorporates more complications because at the camera end the V signal was inverted on each line, and the receiver needs to be able to combine the V signals from each pair of lines by storing the information of each line and combining it with the following line.

This is done using the block diagram of Figure 8.21, which shows part of the PAL receiver concerned with color decoding. One output from the crystal oscillator is passed to an inverter circuit, and this in turn is controlled by a bistable switch. The word bistable, in this context, means that the switch will flip over each time it receives a pulse input, and its pulse inputs are from the line timebase of the receiver, so that the switch is operated on each line. The action of the switch on the inverter will ensure that the oscillator signal is in phase on one line and will reverse (180° phase shift) on the next line, and so on. Also shown in this diagram is an arrangement for an identification ( ident) signal, taken from the burst signal, that makes certain that the bistable switch is itself operating in phase, and not inverting a signal that ought to be unchanged.

The sub-carrier signals are fed to a time delay. This is arranged to delay the signal by exactly the time of one line, and the conventional method originally was to use a glass block. The electrical signals were converted into ultrasonic soundwave signals, and they traveled through the glass to a pickup where they were converted back into electrical signals. The delay is the time that these ultrasound signals take to travel through the glass block, and the dimensions of the glass are adjusted so that this is exactly the time of a line. More modern receivers used electronic digital delay circuits, using the principle of the shift register (see Chapter 11).

The delayed signal (the sub-carrier for the previous line) is added to the input signal (the sub-carrier for the current line) in one circuit and subtracted in a second circuit so as to produce averaged signals for the U and V demodulators that will contain no phase errors. Note that averaging by itself would not correct phase changes; it is the combination of averaging with the phase reversal of the V signals that ensures the correction.

In addition, there are embellishments on the circuits, and one of these is the color killer. When the transmitted signals are in monochrome, any signal at the sub-carrier frequency would produce color effects at the receiver, so that these circuits must be switched off for transmissions (old films, for example) that contain no color. This is done by using the color burst signals, so that when a color burst is present, a steady voltage from a demodulator will ensure that the color circuits are biased on and working. In the absence of the burst signal, the color circuits are biased off, ensuring that any frequencies around 4.33 MHz (the color sub-carrier frequency for the UK) in the picture signal do not cause colors to appear.

We will look at the digital television systems that have completely replaced the analog system in Chapter 16.