The servicing of high-fidelity ("hi-fi") audio equipment demands of the service technician and amateur enthusiast a more critical sense of appraisal of audio reproduction than that required in the case of ordinary sound equipment, where quantity rather than quality is the dominating factor. While the experimenter-enthusiast will be fully aware of this higher critical faculty, and will generally possess it, the man whose job it is to service domestic electronic equipment for a living, taking hi-fi equipment in his stride along with radio and television receivers, may not have such a sensitive ear. If he is not closely acquainted with the foibles of the modern hi-fi enthusiast, the service technician may well be excused his doubts and irritation on encountering the insistence of such an enthusiast that a high degree of distortion is being produced by his apparently excellent amplifier! Accustomed to a standard of reproduction based on years of servicing radio sets of considerably limited audio fidelity, the technician may feel that the enthusiast's request for service of equipment which even in its alleged faulty condition is capable of reproduction of a high order, is far from warranted. This problem of differing standards can make life exceedingly difficult for the service technician when he starts to undertake the repair of hi-fi equipment. To be really successful at the job, the technician must himself develop "hi-fi" standards of judgment. This is usually automatic, as anyone dealing with the servicing of hi-fi equipment works in close liaison with the enthusiasts who operate it.
A technician new to the field quickly becomes initiated, and quickly realizes that (for instance) where a close-tolerance 50k resistor is stipulated by the maker, replacement cannot be made satisfactorily with a 47k resistor of mediocre tolerance, as can often be done in less exacting equipment with little adverse effect.
Hi-fi equipment does not just happen. It is created in the laboratory by a large number of small points being given a great deal of attention. The net result is hi-fi. Slight disturbance to one or more of these small points, either as the result of alteration in value of a component or unskilled service, will unbalance the design and possibly cause distortion. To the uninitiated service technician the distortion may hardly exist, but to the hi-fi perfectionist a world of difference will be discernible. The technician will have to use instruments on which to base his judgment of reproduction; listening tests waste time and lead to frustration.
Essentially, there are three types of hi-fi enthusiast. First, there is the music lover who wishes to play his favorite records with the minimum of distortion. This type is less technically exacting, since a reasonable quality of reproduction is sufficient to re-create in his mind the atmosphere of the concert hall-slight distortion thus goes unnoticed. Then there is the technical perfectionist whose observations are keenly focused on the various responses of the equipment. This type may not possess a highly developed aesthetic interest in music, but he is able to judge with curious accuracy just how much harmonic distortion there is, how the equipment is handling transients, whether or not additional damping would go to improve the overall results and similar technical matters. He obtains great satisfaction from listening to sounds of large magnitude with little distortion, and when he says that distortion is present it is most desirable for the technician to agree with him-until he can prove otherwise, of course! Finally, there is the type who is a compromise between the other two he represents the large majority of hi-fi enthusiasts, who are enthusiasts because they are not only technically interested in obtaining distortion-free reproduction, but are also interested in music in itself.
It is as well for the technician new to the hi-fi world to familiarize himself with these three types of enthusiast; this is equally as important as having the technical know-how, for anyone actively engaged in the servicing of hi-fi equipment will soon become aware that he has to be something of a psychologist as well as a technician-and it is most desirable to know one's subject. This is because sound is a subjective thing, and since it is this in which we are ultimately interested, it is most important to learn a little about it and its effects before moving on to more objective technical matters.
SENSATION OF HEARING
Sound is the stimulus which when applied to the ear gives rise to the sensation of hearing. It is not wholly true to consider sound as emanating from any particular source. Sound is essentially a function of the listener's ear, nervous system and brain. There would be no sound from an explosion, for example, occurring in a place without an ear, nervous system and brain to record it, though there would be considerable air disturbance, to say the least.
The source of any stimulus producing the sensation of hearing is always in some state of vibration. This can be demonstrated by the piano string, the tuning-fork or, to keep in line with our present theme, the cone or diaphragm of a loudspeaker. The vibration may be so slight and so rapid that it is not visible, or it may equally be so large and relatively slow as to be easily observed, as in the case of a loud mains hum affecting the cone of a loudspeaker. It is of little purpose in trying to alleviate the latter condition by securing the speech coil of the loudspeaker cone to the magnet pole piece with good-quality glue-a condition which was once observed by the author when investigating for lack of signals! (However, when questioned, the owner was true to principle in remarking, "but I got rid of the terrible hum which was caused by this cone thing vibrating." A true story!) In the case of an organ pipe and other wind instruments, the source of the stimulus is a column of air. This can be realized from the considerable agitation of fine dry sand on a piece of paper when brought over the mouth of the pipe. The same effect can be observed by placing the sand-laden paper over the vent of a vented loudspeaker enclosure when the system is fed with low-frequency signals to which the vent is tuned, or resonated. In fact, it represents a good method of discovering the vent resonant frequency assuming that an audio generator is at hand to feed a variable audio signal to the loudspeaker--and the free resonance of the loudspeaker cone. In the latter case, of course, the sand-covered paper is held over the loudspeaker cone. The reason for the agitation is that the air is moving in and out of the pipe or vent rapidly, and so sets the paper vibrating.
In many cases the vibration can be felt by placing a finger on the string or loudspeaker cone. It is surprising how sensitive the finger can be in this respect; some engineers check for mains hum by lightly placing the finger on the loudspeaker cone. Air vibration can also be felt. Standing in front of a large loudspeaker fully loaded to, say, 10 watts of low-frequency signal readily illustrates this fact.
Any stimulus of sound (in future we shall refer to it as sound in terms of both cause and effect) may vary in three ways, that is, infrequency, loudness and quality or timbre. The number of complete vibrations made by a sound producing device in one second is called the frequency and determines the pitch of the resulting note. As an example, the string corresponding to bottom A in the piano vibrates at 27 5 hz.
The loudness of a note or sound is governed by the amplitude of the vibration which, of course, determines the energy applied to the ear. The quality or timbre, which distinguishes between notes of the same pitch sounded by different instruments, results from the presence of harmonics in the make-up of the sound. For the present, these can be considered as subsidiary vibrations whose frequencies are exact multiples of the fundamental vibration.
AUDIBLE FREQUENCY RANGE
As the frequency is reduced, the resulting note eventually becomes resolved into the separate impulses of which it is composed. As the frequency is increased, however, the note becomes very shrill, and at about 15,000 hz it is little more than a hiss. The high-frequency limit of audibility varies widely with different individuals. The limit is usually higher with young people, often extending to the region of 20,000 hz, while with increase in age the limit may fall to some 9,000 to 10,000 hz. Some people are highly conscious of the 10,000 hz note produced by television receivers, while others, usually older people, are not at all disturbed. The high-pitched squeak of a mouse is often inaudible to people in their fifties, but often very disconcerting to young people.
At this point it should be made clear that a person who is virtually deaf above, say, 7,000 hz is still able fully to appreciate music containing harmonic components extending well above this figure. It is still necessary for hi-fi equipment employed by such a person to be capable of reproducing all frequencies to the limit of the audio spectrum (the frequency range of good quality equipment usually extends well beyond the accepted audio range, for technical reasons which will be explained later). Tests have revealed that distortion-free reproduction of music containing harmonic components up to some 18,000 hz gives the sensation of considerable mutilation, when passed by way of a filter which chops off all frequencies above 7,000 hz not only to a person whose hearing is unimpaired up to 18,000 hz, but also to one who is essentially deaf at 7,000 hz.
The reason for this, as we shall appreciate better later on, is that a large part of music is composed of steep, rapidly occurring wavefronts (transients), produced by harmonic components of the fundamental frequencies of the various instruments. Cutting the higher frequency components has the effect of spoiling the desirable steepness of the wavefronts as well as reducing the overall amplitude of the sound. Since transients are responsible for the "attack" attributable to music, destroying these in a way that impairs the corresponding accelerations of the wavefronts is obvious equally to persons with and without extended frequency range.
There is another important characteristic of the human ear which gives the impression of dissimilar volume to sounds of different pitch. The sensitivity of the ear rises to a maximum in the region of 3,000 hz, and falls off at frequencies above and below this range.
THE DECIBEL AND THE PHON
While the ear is considerably sensitive to small changes in pitch of a sound it is much less sensitive to changes in amplitude (volume). Instead of following a linear law, the sensitivity of the ear to changes in volume is logarithmic. This simply means that the impression a listener receives when a sound of certain volume is suddenly increased is proportional to the logarithm of the ratio of the energy or power of the two sound levels. The common logarithm of the ratio of two powers gives their relationship in be/s. In mathematical form
Nb= log10 (P2/P1).
This holds for a decrease in power as well as for an increase in power, so that when P2 is less than Pl the value of Nb becomes negative.
The whole range of hearing corresponds to a change comprising 13 bels, that is, starting at a power or intensity near the threshold of hearing to a point where the intensity begins to be painful. As 13 bels is too rough a scale for ordinary use, each bel is divided into 10 decibels (db). Thus, the difference in level between two powers (P1 and P2) in decibels is given by N db = 10 log10 (P2/P1). This expression holds for any change of power, electrical as well as acoustical. Clearly, the ultimate effect of any change of electrical power in a hi-fi amplifier, for instance, is to produce a change of acoustical power from the loudspeaker. It is as well to become familiar with the decibel, as it crops up frequently in audio work.
As an example, suppose an amplifier delivering l watt into a loudspeaker is adjusted to promote an increase of I watt. The output is now 2 watts.
Although the effect can be realized from the statement that "the power has doubled", there is little point in saying that "the power has increased by I watt" unless, of course, it is first clearly indicated that the original power was l watt. It is much better to say that "the power has increased by 3 db". Thus, doubling the power is equal to a 3 db increase (3ยท01 db, to be precise), and halving the power is equal to a 3 db decrease. In the latter case it is usually said that a -3 db power change has occurred.
A change of 2 db, equal to a power of 3 watts being increased to 4.75 watts or decreased to 1.9 watts, for example, is just about discernible by the average person, while a change in level of 1 db is hardly perceptible to the ear.
The decibel is also extensively adopted to compare two currents or voltages. When used in this way it must be ascertained that the resistances (R) in which the currents (I) and voltages(E) operate are the same. When this is the case: N db= 10 log10 (I22 //1 2 ) or 10 log10 (E22/ E12), these being equal to 20 log10 (I2//1) and 20 log10 (E2/E1). When the resistances are not equal due allowance has to be made:
N db= 20 log10 (I^2//1) + 10 log10 (R2/R1) and N db= 20 log10 (E2/E1) + 10 log10 (R2/R1).
TABLE 1.1 CONVERSION OF DECIBELS TO POWER AND VOLTAGE/CURRENT RATIOS
The decibel, as we have already seen, is essentially a unit for measuring relative powers, so when it is employed to express current and voltage gains and losses, allowance has to be made for the fact that power varies by the square of the change of current or voltage. For example, an increase in current or voltage by a factor of two results in the power being increased by a factor of four.
When N db is known, the power, current and voltage ratio can be found as follows: P2/P1 = antilog N db/10, /2//1 = antilog N db/20 and
E2/E1 = antilog N db/20.
Decibel tables save the toil of making complex calculations, samples being given in Table 1.1 and Table 1.2. Table 1.1 gives conversion of decibels to power and voltage/current ratios. Figures not given in the table may easily be calculated. For example, if two db figures are added, their corresponding power or voltage/current ratios must be multiplied. Table 1.2 gives conversion of power ratios to decibels.
The apparent loudness of any tone is related to its pitch or frequency as well as to its amplitude or intensity. The phon is the unit of loudness level actually appreciated by the ear, and represents about the limit of difference in loudness of which the ear is sensible. At a frequency of 1,000 hz, the loudness level of a pure tone in phons is equal to the number of decibels above the reference power, though this does not hold with any other frequency. It is this apparent non-linear loudness level over the audio spectrum which has recently encouraged the use of "loudness" controls on hi-fi amplifiers. As we shall see later, they function essentially to increase the bass response as the volume is reduced.
TABLE 1.2 CONVERSION OF POWER RATIOS TO DECIBELS
FIG. 1.1. Simple harmonic motion, such as that produced by a tuning
fork, can be represented by a sine wave.
FIG. 1.2. The two sine waves (a) are related in that one has twice the
frequency of the other. In (b) the waves are compounded to give a composite
wave. Here the sound is no longer pure, but has a high second-harmonic
content.
HARMONICS
Most vibrating bodies execute a simple harmonic motion, giving a pure tone, or the vibration is composed of a combination of simple harmonic motions, giving rise to overtones, which are usually related in frequency to the fundamental. The sine wave (Fig. 1.1) is representative of simple harmonic motion, such as that produced by the vibration of a tuning-fork.
It is the presence of overtones or harmonics which is responsible for the difference in the quality between the sounds produced by the various instruments of an orchestra. The human voice is also rich in harmonics, and since the harmonic content differs between individuals, it is often a simple matter to pick out a certain person by his voice. This is not always the case when contact is by way of the telephone, since this instrument is not wholly responsive to high-order harmonics, its high audio-frequency range being considerably limited, and causing a change in the quality of a voice. This effect is aggravated by speaking through a cloth held in front of the micro phone mouthpiece.
Hi-fi amplifiers must be capable of responding fully to all high-order harmonics, and themselves must not be responsible for the introduction of harmonics which are not present in the original sound.
Harmonics consist of notes having 2, 3, 4, etc., times that of the fundamental. The violin, for example, is rich in harmonics at twice and five times the fundamental note to which the string is tuned. The amplitude of the harmonic is also important, and is relatively large in the case of a violin.
In Fig. 1.2 (a) two sine waves representative of simple harmonic motion, one of which has twice the frequency of the other, are given individually, the higher-frequency one being the second harmonic of the lower-frequency fundamental. In Fig. 1.2 (b) the sum of the two waveforms is given graphically, it being obtained by adding the ordinates of the fundamental and second-harmonic waves. A waveform such as shown in Fig. 1.2 (b), being obtained at the output of an amplifier as the result of a pure sine wave input (Fig. 1.1), would indicate most forcibly that the amplifier itself is producing a very large degree of second-harmonic distortion. Apart from being revealed on the screen of an oscilloscope, the distortion would be readily detected, since the ear is capable of recognizing the two sounds, even when they are compounded to form the wave of Fig. 1.2 (b).
TRANSMISSION OF SOUND
Any sounding body causes the surrounding air to be alternately com pressed and rarefied in sympathy with the vibrations. As long as the vibrations occur, a wave of high pressure is followed by a wave of low pressure and again by a wave of high pressure, and so on. Compression and rare faction waves are thus radiated in all directions from the sounding body at 1,088 feet per second, and an eardrum within range will be caused to vibrate in exact sympathy.
Air is the chief medium for the transmission of sound waves, as is clearly revealed by the classic experiment of extracting the air from a bell-jar in which is placed a sounding electric bell. As the amount of air in the jar becomes smaller, the sound of the bell gets weaker. To a lesser degree, all material substances can transmit sound waves. A wood rod, for example, is sometimes used to detect mechanical noises in a car engine. One end of the rod is held in contact with the ear while the other end is held in close contact with the region of the engine being checked for noise. The wood rod serves to transmit the sound waves in this case.
Sound waves in air are known as longitudinal waves. This term simply indicates that the particles of the wave-carrying medium travel backwards and forwards in a path whose direction is the same as that in which the wave is travelling. Electromagnetic waves, on the other hand, are known as trans verse waves, indicating that the particles of the medium travel in paths at right-angles to the path of the wave as, for instance, the waves upon the surface of water.
Sound waves cannot directly be represented by a sine curve, since the particles of the wave-carrying medium remain in a straight line, being com pressed and rarefied as we have seen. Nevertheless, it is possible to represent diagrammatically, to scale, longitudinal waves by means of a sine curve.
The result is similar to the sine wave in Fig. 1.1. Such a wave possesses four distinct characteristics, which are (1) amplitude, (2) frequency, being the number of complete cycles per second emanating from the sounding body, (3) the velocity at which the wave travels from its source, and (4) wavelength, being the distance between each consecutive peak. The wave will also be endowed with the shape created by harmonics of the fundamental frequency (Fig. 1.2 b). It is important to remember the relation between wavelength, frequency and velocity which, irrespective of the form of the wave, is expressed as the velocity being equal to the product of the frequency and wavelength, or velocity (V) ~ frequency (f) times the wavelength()..). The wavelength can be found by dividing the velocity by the frequency, i.e., ), feet = 1,088// This expression can be useful when investigating for standing waves in the listening room, as well as for other purposes.
Fig. 1.3. Showing at (a) how sound waves are inclined downwards when
the wind is in the direction of the sound, and at (b) how the waves are
given an opposite tilt when the wind is against the sound. Similar refraction
occurs due to temperature variation of the atmosphere.
It sometimes happens that the service technician, hi-fi enthusiast or sound engineer is called upon to supply sound reinforcement in the open air -- at a fete or garden party, for example--when the question may arise of the effect of wind upon sound waves from the loudspeakers. When the wind is fairly strong it is desirable to place the loudspeakers (with due consideration to the other factors involved) in relation to the listeners so that the sound is travelling with the wind. This is not because the wind affects the intensity of sound, though the velocity would be changed.
The reason that the sound is more clearly heard when it is travelling with the wind than if there were no wind, and vice versa, is that the sound waves are tilted as the result of increasing wind velocity with increasing altitude.
This effect is illustrated in Fig. 1.3, where at (a) is shown how the waves are inclined downwards when the wind is in the direction of the sound and, at (b), the opposite tilt when the wind is against the sound. It must be remembered that the waves always travel at right-angles to their own planes and, under the influence of wind, their velocity is altered with increasing height.
The velocity of sound waves is also affected by temperature. A temperature rise promotes an increase in velocity, and the effects shown in Fig. 1.3 are often produced from this cause. During a hot summer's afternoon, for instance, sound waves may be tilted skywards as the result of the air temperature being greater at lower levels than at higher levels. The converse effect is often experienced when the lower air layers are at a lower temperature than the upper air layers. For this reason, distant sounds are often clearly heard on a cool. still evening, the effect being particularly noticeable over the surface of water.
BEATS
When one is slowly overtaking a noisy heavy goods vehicle in a car whose engine is not unduly quiet, a drumming or beating sound may develop and vary in frequency as the car engine is increased in speed in order to overtake the other vehicle as quickly as possible. When this effect is first experienced, one may incorrectly conclude that the back axle is due for renewal! The disturbance, however, is quite natural, being caused by a beat tone created as the result of sound waves from the two engines combining, the frequency of the beat being equal to the difference in frequency of the two sounds involved.
Such beats are sometimes produced in amplifiers, pick-ups and loud speakers, and may give rise to spurious tones, referred to as intermodulation distortion, which may or may not be harmoniously related to the tones from which they arise. The distortion usually gives considerable harshness to audio reproduction, as well as to the sound of car engines! The hi-fi technician will encounter many problems in which resonance plays a leading part. If an audio oscillator is connected across the terminals of a hi-fi loudspeaker system, and the oscillator is tuned fully over the audio spectrum from about 15 hz to 15 khz (15,000 hz), it will be found that at various frequencies different objects in the room will start vibrating vigorously in sympathy with the sound produced by the loudspeaker. (Let us hope that the loudspeaker enclosure is not subject to such disturbance.) When the sound has a frequency equal to the natural frequency of an object, then the object will vibrate in sympathy with this sound. This process is called resonance.
Heavy damping of the object, due to its design and firmness, will greatly reduce the intensity of the resonance. Loudspeaker enclosures are usually made so as to reduce their natural resonance to the minimum, though at the time of writing a speaker enclosure is undergoing development that is designed intentionally to resonate or flex at certain frequencies. The enclosure panels are designed to resonate at different frequencies as a means of damping the air column resonance within the enclosure, and so spread the effectiveness of the damping over a wider frequency range. It is reasoned that the more conventional method of acoustic damping wastefully converts sound energy at the resonant frequency into heat.
The usual arrangement, which is often adopted for hi-fi, is to use sand filled panels, or panels of concrete, for speaker enclosures. In this way complete rigidity is secured, and there is little fear of the enclosure walls flexing, even when the alternating sound pressure within the enclosure is at a high level.
Resonance effects are at their height in the small, popularly-priced record-players, often colloquially referred to as "pop boxes"-not a hi-fi term! Here the loudspeakers ( or loudspeaker) are contained within a portable housing along with the amplifier and record-player-often an auto-unit is employed. If an audio oscillator is connected across the speaker of one of these devices, things really start resonating within the box. After the case itself has ceased to resonate up to 200 hz, the valves in the amplifier take over, then the various metal levers of the record unit at about 2,000 hz, and so on.
When the instrument is used as intended, the box resonance enhances the bass response in a synthetic manner, and when "bop" records are played the other higher-frequency resonances undoubtedly merge with the general background effects. There are on the market, however, quite good portable record-players in which undesirable resonances have been damped as far as is possible. These instruments, of course, are more expensive than the single valve outfits which are produced essentially for the reproduction of popular music in current demand. Nevertheless, true hi-fi equipment is demanded for true fidelity reproduction, and portable equipment is then completely out of the question. Separate units are essential, and pieces of equipment which are prone to resonance should as far as possible be removed from the loud speaker system.
The power of resonance is illustrated by the traditional order "break step" which is given to a company of soldiers about to cross a bridge. If the troops' normal rhythmic step happened to coincide with the natural frequency of the bridge, vibrations of large magnitude would be promoted and there would exist a definite possibility of the bridge breaking up.
Apart from the resonance effects of objects, air itself can be caused to vibrate at certain frequencies under controlled conditions. As an example, tuning-forks are sometimes mounted upon hollow boxes so as to increase the volume of sound. The normally feeble sound from a tuning-fork is considerably amplified because the size and shape of the box is arranged so that the air inside possesses a natural vibration period equal to that of the fork. Thus, both the vibration of the fork and the vibration of the air, at the particular tuned frequency in both cases, contribute to the total energy of sound applied to the ear. Such a box is known as a resonator.
This particular effect must not be mistaken for the increase in volume which can be obtained by holding the stem of a tuning-fork in close contact with a table-top or board. In this case, the table-top of board simply serves as a sounding board; forced vibration is produced by the fork, and as a consequence the overall vibration is communicated to a much greater quantity of air than when the fork is vibrating unaided.
A well-known resonator is that due to Helmholtz. It was developed some hundred years ago for the purpose of harmonic analysis of a note, and it is still used for this purpose. Such resonators consist (in the original) of a brass spherical shell on which is formed a taper containing a small hole for the purpose of inserting into the ear. Diametrically opposite is a larger opening for presenting to the source of sound.
The air in the resonator resonates to one particular frequency-that to which the resonator is tuned-and when a sound is applied, the resonator picks out and amplifies only that component of the sound to which it is tuned. In this way components of a complex note too feeble to be detected by the ear alone become easily audible and can be checked for relative strength.
Resonators of this kind are made in sets, the note of each being set to the required standard. The resonant or resounding frequency is governed by the volume of air and the area of the pick-up aperture. The frequency is decreased by increasing the volume of air or by decreasing the area of the aperture.
The phase inverter or reflex loudspeaker enclosure adopts the principles of the Helmholtz resonator at the low-frequency end of the audio spectrum.
STANDING WAVES
Resonances also occur in the listening room, as the hi-fi service technician will undoubtedly discover for himself during the process of investigating for poor results in a customer's home on equipment which has previously worked with excellent results in the demonstration room! Such resonances, sometimes referred to as eigen-tones, are produced by multiple sound reflection between the opposite walls, and occur at the frequency at which the distance between the opposite walls is exactly one half-wavelength. This condition gives rise to standing waves at the critical frequency, whilst also considerably accentuating the response at the resonant frequency. In effect, the room serves as a resonator, and the air resounds at the frequency to which the room happens to be tuned.
Further resonances occur as the result of the other two parallel walls and the ceiling and floor, and others governed by the dimensions of the diagonals.
The worst conditions occur when the room approximates a cube, with the speaker situated in the center of a wall. Apart from the chief low-frequency resonance or eigentone at a half-wavelength, others, though possibly less disturbing, present themselves at all harmonics of the basic frequency. Thus, with the main resonance at, say, 40 hz, created by a cube-shaped room with 14 ft. sides, additional resonances at 80, 160, 320 hz and so on will also result. Reciprocally, it follows that the reproduction will be exaggerated at frequencies for which the walls are a multiple of half a wavelength apart.
ELECTRICAL REPRESENTATION OF SOUND
In all forms of sound broadcasting, recording and reproduction a means must always be provided to convert the sound energy into electricity. Such a conversion device, capable of receiving energy in one form and passing it on in another form, is known as a transducer. The microphone comes under this classification.
FIG. 1.4. A sounding tuning fork held in front of a microphone will
give rise to a current waveform, as illustrated, of frequency coinciding
with that of the fork.
All microphones possess a thin diaphragm on which the sound pressure operates, and the resulting vibrations create currents of electricity which rise and fall in precise sympathy with the sound waves. For example, a sounding tuning-fork held in front of a microphone will give rise to a current waveform of frequency coinciding with that of the fork (Fig. 1.4). Similarly, a complex sound wave, composed of a number of tones and harmonic parts, will be electrically reproduced with equal accuracy. Within limits governed by the design and purpose of the microphone, the electrical output will depend upon the intensity of the sound applied to the instrument. Increasing sound intensity will result in increasing output, and vice versa. The electrical out put will also vary with the frequency of the applied sound, though for high quality work the microphone must respond evenly over the whole of the audio spectrum.
The output of a microphone is conveniently expressed in decibels relative to a fixed reference level. The reference level chosen is sometimes 0 db = 1 volt (open-circuit) with a sound pressure of 1 dyne per square centimeter.
Thus, a microphone with an output expressed as - 74 db below 1 volt/dyne/ cm^2 would have an open-circuit voltage of approximately 0.0002 volts r.m.s.
The output is sometimes expressed in terms of power for a stated sound pressure. The RMA rating is defined as the ratio in db relative to 0.001 watt dyne per square centimeter.
At this point it should be noted that a sound pressure of 0.0002 dyne per square centimeter corresponds to the limit of audibility of a 1,000 hz note. This in turn corresponds to zero phon, and to give the reader some idea of the loudness scale, a quiet room is rated at 20 - 30 phons, average conversation 60 phons, interior of a tube train with the windows open 90 phons, proximity to an airplane engine 120 phons, while 130 phons is approaching the threshold of feeling or pain.
SOUND REPRODUCTION
To be of practical use, the very small power available at the output of the microphone must be considerably amplified, and this has to be performed without alteration of either the character of the electrical waveform, due to the sound waves, or of the response over the entire audio spectrum. With regard to the latter consideration, however, poor acoustics of the room in which the microphone is used (the studio) can sometimes be countered by the use of a frequency-selective network between the microphone and amplifier input. For example, the exaggerated response at low frequencies due to a room of small dimensions is sometimes mitigated by the introduction of a filter network which attenuates the bass frequencies at the microphone, in relation to the higher frequencies, before the signal is applied to the amplifier.
This process is known as equalizing for room acoustics. Similarly, the equalizing function may take place somewhere in the amplifier chain.
Fig. 1.5. Curves showing how the bass and treble response of an amplifier
can be altered to suit the acoustics of the listening room.
Most amplifiers are composed of three distinct sections. First there is the voltage amplifier whose purpose is to step-up the small audio-frequency (a.f.) voltages occurring in the varying sound input to a workable level. This section may also contain equalizing networks of suitable form to cater for the various signals for which the voltage amplifier is going to serve. Next comes a tone-control section, in which controls are available for adjusting the degree of amplification of the treble and bass frequencies of the signal, usually relative to 1,000 hz.
The idea is illustrated in Fig. 1.5. It will be seen that the bass is continuously variable from -12 db to + 12 db at 40 hz, and that the treble is continuously variable from -15 db to + 12 db at 10 k hz. Having such a control of the response of the amplifier aids considerably in the correction of impaired room acoustics from the reproducing point of view. The presence of low-frequency resonances, for instance, can be prevented from over emphasizing the bass from the aspect of the listener by applying a suitable degree of bass cut and, possibly, treble lift. Conversely, some rooms may be acoustically "dead"; they have a tendency to absorb more of the lower and higher frequencies and thus seem to require more bass and treble than average rooms. Tone controls serve to correct such deficiencies of the listening room and maintain the faithful balance demanded by the hi-fi enthusiast.
Finally, the equalized, amplified and tone-controlled signal is passed on to the power amplifier, by way of a volume or loudness control, and is changed from voltage to power for operation of the loudspeaker. Some equipments have the power amplifier as a unit completely independent of the voltage amplifier and tone-control section, while other smaller amplifiers are complete in themselves. A block diagram of the three sections we have discussed is given in Fig. 1.6.
The loudspeaker is also a transducer, but it operates in the opposite way to that of the microphone; it receives an electrical representation of the sound which was applied originally at the microphone, and passes it on in the form of sound energy. We shall discuss both microphones and loud speakers in some detail in later sections.
We now have a complete picture of the whole chain of events, from the sound waves to the microphone, from the microphone through the amplifier to the loudspeaker, and from the loudspeaker to the ear. Let us always bear in mind that the results heard are a function of the mind of the individual, and that they are colored not only by the equipment used for the reproduction of the sound, but also by the studio and listening-room acoustics.
Although it is impossible to match the acoustics of the ordinary listening room with those of the concert hall, it is surprising what can be done synthetically by equalizers and tone controls, not to mention loudspeakers and enclosures!
Fig. 1.6. Block diagram of the three main sections of a hi-ft amplifier.
We have so far considered "live" reproduction of sound, that is direct from the microphone to the loudspeaker. Of course, sound can be "stored" and used when required. The most popular medium for storing sound in this way is the gramophone record.
Instead of actuating the cone of a loudspeaker to produce sound waves coinciding with those at the microphone, the microphone-amplifier set-up powers a recording head whose purpose is to cause lateral vibration of a sapphire or diamond cutting tool in sympathy with the sound energy applied at the microphone The recording head is tracked radially over the recording blank, while the blank is carried on a heavy turntable which is arranged to rotate at a perfectly even speed. The recording head is also pivoted in such a way that the cutting tool is pressing on the surface of the disk. A spiral groove is thus cut upon the surface of the disk, running from the circumference to near the center. The sound vibrations impart upon the groove a wavy lateral effect, which is clearly visible on any gramophone record. Sound is thus "stored". Playback is accomplished by rotating the disk at the same speed at which it was recorded, and by the use of a pick-up carrying a stylus having a hemi spherical tip, which rests in the V-section groove cut by the cutting tool.
The pick-up is mounted on a tone arm which is free to rotate about a center some distance from the center of the record, and is free to move only in an arc which is approximately radial with respect to the record.
The pick-up is thus carried across the record under the control of the groove, while at the same time the stylus is caused to vibrate laterally in sympathy with the lateral waveform imparted during the recording process.
The lateral vibration of the pick-up stylus gives rise in the pick-up to an e.m.f.(electromotive force) having the same pattern as the waveform on the record and, in some cases, in proportion to the velocity of the lateral movement of the stylus. This e.m.f. is applied to the input of an amplifier, is suitably equalized, and ends up as sound waves from the loudspeaker-ideally, as a replica of the sound waves applied to the microphone during the recording session.
STEREOPHONIC SOUND
A single-channel (often referred to as monaural) reproducer system, i.e., one microphone, one amplifier system and one loudspeaker system into which the single channel is working (more than one microphone and loud speaker may well comprise a single sound source from the monaural aspect), can never give true fidelity of reproduction. Highly satisfactory reproduction of an orchestra cannot be secured if all the sound is radiated from a hole in the loudspeaker cabinet. The use of two or three speaker systems does not help much in this respect when they are all connected to the output of a common channel.
The "range" of the orchestra can only be realized by the use of two or more completely independent channels. With a two-channel system, which is highly suitable for domestic use, there are two loudspeakers each fed from a separate microphone (or from a separate signal source) through separate amplifiers. The basic idea is to place the loudspeakers relative to one another as the microphones are placed in front of the orchestra, or as they were placed during the recording of the program.
In this way both ears of the listener are brought into operation in a selective sense. The orchestra appears to be spread in correct proportion across the room, between the loudspeakers, and the listener can pick out the individual instrumentalists as readily as if he were in the concert hall. The "muddiness" of the monaural system disappears completely, and a third dimensional sense of presence is created.
At present the normal method of stereophonic recording on disks is what is known as the "45/45 system" where the two stereo channels are carried in one groove. This system (described in more detail in Section 10) has taken the place of the system where one channel is recorded by the "hill-and-dale" process, as adopted 80 years ago by Thomas Edison in connection with the phonograph. With this, the cutting tool of the recording head was arranged to oscillate vertically in sympathy with the sound vibrations, so that the depth of the groove corresponded to the wave pattern of the sound; hence the term "hill-and-dale". The other channel was recorded in the same groove laterally, and a special recording head was used to modulate the groove both laterally and vertically in accordance with the two-channel signals applied. On playback, a pick-up functioning electrically opposite to that of the recording head gave two outputs corresponding to the two recorded channels. The signals were amplified independently, and were fed to the two loudspeakers to give the effect of stereophony. Hill-and dale/lateral and 45/45 stereo recordings differ essentially only in the way in which the signals in the two channels are phased.
MAGNETIC RECORDING
Wire and tape coated with a magnetic material are also used for recording. The wire or tape (wire is now rarely used) is drawn steadily over the pole of an electromagnet, the current in which is caused to follow the wave pattern of the sound. The wire or tape thus becomes magnetized, as each section passes over the pole piece of the electromagnet, to a degree dependent upon the electrical representation of the sound applied at the microphone, and a magnetic wave-pattern is imparted upon the medium.
On playback, the medium is again drawn at the same speed across the pole piece of an electromagnet which this time is not energized, but which has induced in it small voltages corresponding to the varying flux in the core as the result of the magnetized medium. The voltages, representing the recorded sound signal, are applied to the input of an amplifier and end up as sound from the loudspeaker. When required, the recording can be easily erased by passing the medium over a permanent magnet or an electromagnet which is energized by a pure signal having a frequency above the audio spectrum (30-50 khz). This system lends itself readily to two-channel operation, it being a simple matter to record one channel on one half of the tape and the other channel on the other half by the use of slightly displaced electromagnets for record and playback.
Sound can also be stored on film, on the principle adopted for the sound track on cine film, but a description of this method falls outside the scope of this guide.