Radio Broadcast (Nov 1926-Apr 1927)

A Fundamental Analysis of Loud A Radio Club of Is Required to from a Loud Paper Discussing the Signals a Loud Speal ce — Factors Determining the Quality of Output in JOHN F. NIELS Engineering Dept., F. A. D. Andrea, Inc. CENTURIES before we dreamed of modern loud speaking equipment, the natives of Africa had crude systems of communication in the form of cocoanut shells connected by a taut string, which acted as a medium for transmitting sound vibrations. Outgrowths of these crude systems of communication are our present telephone network and radio systems. The primary function of any communication system is to transmit intelligence and entertainment. In considering the operation of any such system, the reproduced sounds may be referred FIG. I to as having two properties, i.e., intelligibility and naturalness. In radio broadcasting, the communication system is supplemented by entertainment, and the property of naturalness, therefore, increases in importance in the reproduced speech. Moreover, the transmission of music imposes even more severe requirements upon a communication system because of the wide range of frequencies and intensities required for proper appreciation. It is the purpose of this paper to present in a popular fashion a few of the fundamentals of operation of one link of such a system, namely, the acoustic reproducer. The fact that there is ample opportunity for distortion in those parts of a radio system preceding the loud speaker is apparent from a consideration of the change the speech wave must undergo between studio and loud speaker. Although there is equal opportunity for distortion in the transmitter and receiver, transmitter distortion is generally negligible. The fact that a broadcasting station serves a large volume of receiving sets would point to this state of affairs, since the receiver must be far cheaper than the transmitter. Quality in the receiving set itself is affected by the sharpness of tuning of the radio-frequency stages, by the time constant of the grid leak-condenser combination of the detector circuit, by the characteristic and power capacity of the audio amplifier, and finally, by the loud speaker itself. Before considering in detail the characteristics of loud speakers, let us digress for the moment and consider the nature of the signal it must reproduce. In general, the loud speaker should reproduce faithfully both speech and music, each of which presents its own peculiar problems. Speech consists in general of two fundamental types of sound, namely, continuents and stops, and their combinations1 (see bibliography on page 590). The former are those produced by a continuous flow of air, such as the letters F, S, etc., while the latter consists of those sounds produced by a sudden stoppage of air, such as letters, P, B, and M. For instance, analyzing the word "Past," we notice, in Fig. 1, that " P" appears as a transient indicated by a high broad peak; "A," a nearly continuous frequency of approximately 800 cycles varying in amplitude; "S" and "T" are of a high frequency character, of low amplitude, and continuous. It is readily seen from the nature of the word that, in order to obtain perfect reproduction, the loud speaker must reproduce frequencies of an extreme nature with proper relative amplitude and without time lag. If a loud speaker is inefficient at the upper extreme of its frequency spectrum, it is generally noticed that "S," "T," and other high-frequency combinations, are either missing entirely or are of low relative intensity. In addition, a loud speaker may have resonant peaks at certain frequencies which may so exaggerate some sounds as to completely mask others. Speech energy is distributed over a frequency band of about 50 to 10,000 cycles per second and, in general, has a maximum between 1 w J V ft l CELLO ORGAN PIPE "C" i X K \ 1000 2000 3000 4000 5000 FREQUENCY FIG. 2 A TYPICAL MODERN CONE LOUD SPEAKER TROMBONE ORGAN PIPE C" FIG. 3 150 and 250 cycles. The frequency distribution of speech energy is shown graphically in Fig. 2. Music is characterized by various harmonics which may be of larger or smaller amplitude than the fundamental. These harmonics distinguish the same note in the same octave as played on different instruments. Fig. 3 shows the difference between a cello organ pipe "C" and a trombone organ pipe "C". It is readily seen that a wide frequency spectrum must be faithfully reproduced to enable the listener to distinguish between different instruments. Musical sounds are characterized by being sustained at definite frequencies for comparatively long periods and by having the change in pitch take place in definite musical intervals called, thirds, fifths, octaves, etc. Musical notes are usually very rich in harmonics; in some instances, as in the case of the cello, the harmonics may even exceed the fundamental frequency in intensity. Musical energy is distributed over a frequency band of from about 16 cycles to something over 10,000 cyles