Radio Broadcast (May 1929-Apr 1930)

RADIO BROADCAST. formula db = 20 log E, and from the calibration of the vacuum-tube voltmeter, E was obtained for various points. Thus it was possible to measure just how many db one peak was above another. Curve a in Fig. 6 shows the response curve for an Airchrome loud speaker; b is the curve of an air-column loud speaker; c is that of a cone loud speaker; d is that of a dynamic loud speaker, and e is that of the same dynamic with the condenser across the loud speaker input. The curves indicate the resonant peaks caused by the component parts of the loud speakers and units. By using filters consisting of condensers and chokes, various peaks can be removed without affecting the response at any other point. The first resonant peak of fair size indicates the fundamental response of the diaphragm. Value of Curves Inasmuch as the microphone was not calibrated with the equipment used, the tests, as previously explained, have been made only on a comparative basis; that is, the performance of a loud speaker and the response in one loud speaker can be compared to another only when the tests are made in the same laboratory. However, by using a condenser microphone and a thermophone, the responsograph can be calibrated easily on an absolute basis, in which case the curves obtained in one laboratory would be comparable with those obtained in another one. Unless a pure sine wave is impressed at all times, the shape of the curves for any loud speaker would not be the same from time to time and the tests would mean nothing. All of these tests were made using a sinusoidal input and do not indicate the performance of the loud speaker when music or speech is impressed. This accounts for the fact that, although a loud speaker very often gives a good response throughout the entire range of frequency, it does not perform well on mixed frequencies. This is due to the fact that the lower frequencies carry most of the speech energy while the articulation and intelligibility depend upon the higher frequencies. In studying the behavior of loud speakers it was observed that changing the shape of the diaphram or its mass affected its response. It was also observed that when a loud speaker had a condenser across its input the high frequencies were cut off. This gave the low frequency or bass notes a predominance and speech then became difficult to understand because the articulation decreased when the high notes were cut off. Using the responsograph, it was also possible to study the effect of baffling loud speakers, the response varying with the size of the baffle. It was also possible by means of the responsograph to determine the cabinet resonance of any compartment in which the loud speaker was placed. Another interesting experiment consisted of testing the efficiency of FREQUENCY IN CYCLES Fig. 6 — Sample response curves made with five different types of loud speakers. Note: These curves are not intended to indicate the level at which the speakers are operating but only the relative response at various frequencies. 1 11 J — "~! 1 — II — 5 — lh^ % \ 7H C "1 rM l f 4|l|ljl|lf Fig. 3 — Schematic diagram of the a.f. amplifier used in making loud speaker response measurements. fal r^-tlL Variable Condei Amplifier Micro -Am meter Audio Frequency Oscillator Speaker Vaccum' Tube Voltmeter Fig. 5 — View of a responsograph of the type used for making comparative loud speaker measurements. Fig. 4 — Schematic diagram showing the set-up of apparatus when making loud speaker comparisons. various band-pass filters as well as determining how sharp the cut-off's were. Another use to which the responsograph could be put is to measure the overall characteristics of a radio set and loud speaker combination. For this purpose an r.f. oscillator modulated by means of a beat-frequency oscillator would be used. With this apparatus it would be possible to observe just how cutting off sidebands affects the a.f. system. It would also provide a means of perfecting the radio set so that there is nearly constant frequency response throughout the entire range. The author wishes to acknowledge the use of apparatus and the assistance which has been rendered by Professor Andres and his staff of engineers at the Temple Laboratory. BOOK REVIEWS Telephone and Power Transmission, by R. Bradfield and W. J. John. Published by John Wiley & Sons, Inc., New York City, 234 pages. Price: $5.75. This new book will take its place in the telephone library along with other standard texts on the theory underlying the transmission of intelligence along wires. It was written by R. Bradfield, late of the engineering department of the British General Post Office, and W. J. John, lecturer in electrical engineering at East London College. This book is not like that of K. S. Johnson's Transmission Circuits, a practical book, but is one dealing with the mathematical theory of the propagation of electric waves on wires. It furnishes its own mathematics but to prevent throwing a scare into any potential reader, let us hasten to set down the aim of the authors, "to write for that vast majority of students and practical engineers who are engineers first and mathematicians only in so far as their profession demands." The authors realize that hyperbolic functions look more formidable than they really are, and that a telephone engineer has little need for the mathematical concepts on which hyperbolics are based. An engineer wants to know how to solve problems by their use — and to serve that end, examples and problems are included in the text. There are chapters on general theory of transmission, the application of theory to actual telephone transmission, measurements of line constants, transmission of power on wires, effect of transformers on voltage drop in transmission circuit, traveling waves in transmission lines, etc. More than half of the book is taken up with the transmission of power on wires, and for anyone contemplating the transmission of radio programs on power wires, this part of the book is useful. There are a number of tables in the book on such subjects as sizes and resistances of standard-gauge wires, attenuation of various types of lines, mathematical tables of hyperbolic functions, etc. It is interesting to note that the decibel (d b) or the transmission unit does not seem to be mentioned in the book and that losses and attenuation is still spoken of with regard to the mile of standard cable. — K. H. • JULY 1929 • • 153