Showing papers in "Bell System Technical Journal in 1924"

Certain factors affecting telegraph speed

Abstract: This paper considers two fundamental factors entering into the maximum speed of transmission of intelligence by telegraph. These factors are signal shaping and choice of codes. The first is concerned with the best wave shape to be impressed on the transmitting medium so as to permit of greater speed without undue interference either in the circuit under consideration or in those adjacent, while the latter deals with the choice of codes which will permit of transmitting a maximum amount of intelligence with a given number of signal elements. It is shown that the wave shape depends somewhat on the type of circuit over which intelligence is to be transmitted and that for most cases the optimum wave is neither rectangular nor a half cycle sine wave as is frequently used but a wave of special form produced by sending a simple rectangular wave through a suitable network. The impedances usually associated with telegraph circuits are such as to produce a fair degree of signal shaping when a rectangular voltage wave is impressed. Consideration of the choice of codes show that while it is desirable to use those involving more than two current values, there are limitations which prevent a large number of current values being used. A table of comparisons shows the relative speed efficiencies of various codes proposed. It is shown that no advantages result from the use of a sine wave for telegraph transmission as proposed by Squier and others2 and that their arguments are based on erroneous assumptions.

A reactance theorem

Abstract: The theorem gives the most general form of the driving-point impedance of any network composed of a finite number of self-inductances, mutual inductances, and capacities. This impedance is a pure reactance with a number of resonant and anti-resonant frequencies which alternate with each other. Any such impedance may be physically realized (provided resistances can be made negligibly small) by a network consisting of a number of simple resonant circuits (inductance and capacity in series) in parallel or a number of simple anti-resonant circuits (inductance and capacity in parallel) in series. Formulas are given for the design of such networks. The variation of the reactance with frequency for several simple circuits is shown by curves. The proof of the theorem is based upon the solution of the analogous dynamical problem of the small oscillations of a system about a position of equilibrium with no frictional forces acting.

Some applications of statistical methods to the analysis of physical and engineering data

Abstract: Whenever we measure any physical quantity we customarily obtain as many different values as there are observations. From a consideration of these measurements we must determine the most probable value; we must find out how much an observation may be expected to vary from this most probable value; and we must learn as much as possible of the reasons why it varies in the particular way that it does. In other words, the real value of physical measurements lies in the fact that from them it is possible to determine something of the nature of the results to be expected if the series of observations is repeated. The best use can be made of the data if we can find from them the most probable frequency or occurrence of any observed magnitude of the physical quantity or, in other words, the most probable law of distribution. It is customary practice in connection with physical and engineering measurements to assume that the arithmetic mean of the observations is the most probable value and that the frequency of occurrence of deviations from this mean is in accord with the Gaussian or normal law of error which lies at the foundation of the theory of errors. In most of those cases where the observed distributions of deviations have been compared with the theoretical ones based on the assumption of this law, it has been found highly improbable that the groups of observations could have arisen from systems of causes consistent with the normal law. Furthermore, even upon an a priori basis the normal law is a very limited case of a more generalized one. Therefore, in order to find the probability of the occurrence of a deviation of a given magnitude, it is necessary in most instances to find the theoretical distribution which is more probable than that given by the normal law. The present paper deals with the application of elementary statistical methods for finding this best frequency distribution of the deviations. In other words, the present paper points out some of the limitations of the theory of errors, based upon the normal law, in the analysis of physical and engineering data; it suggests methods for overcoming these difficulties by basing the analysis upon a more generalized law of error; it reviews the methods for finding the best theoretical distribution and closes with a discussion of the magnitude of the advantages to be gained by either the physicist or the engineer from an application of the methods reviewed herein.

A generalization of the Reciprocal Theorem

Abstract: THE Reciprocal Theorem, an interesting and extremely important relation of wide applicability, which was discovered by Lord Rayleigh, is stated by him in the language of electric circuit theory as follows: “Let there be two circuits of insulated wire A and B, and in their neighborhood any combination of wire circuits or solid conductors in communication with condensers. A periodic electromotive force in the circuit A will give rise to the same current in B as would be excited in A if the electromotive force operated in B.”1

“The stethophone,” An electrical stethoscope

Abstract: AUSCULTATIO is commonly practiced by means of the ordinary stethoscope, a device with which th e physician is able to study sounds produced within the heart, lungs, or other portions of the body and to determine whether such abnormal conditions exist as are evidenced by abnormal sounds. Of particular importance are the characteristics of the normal heart sounds, heart murmurs, breathing sounds and râles.1 It is well known that the intensity of certain of these sounds is not in itself of fundamental significance, that, for example, certain very faint murmurs may represent serious organic lesions; hence it is of pathological importance that these sounds be heard and understood.

Transmission characteristics of electric wave-filters

Abstract: The transmission loss characteristic of a transmitting network as a function of frequency is an index of the network's steady-state selective properties. Methods of calculation heretofore employed to determine these characteristics for composite wave-filters are long and tedious. This paper gives a method for such determinations which greatly simplifies and shortens the calculations by the introduction of a system of charts. Account is taken of the effects of both wave-filter dissipation and terminal conditions. The method is based upon formulae containing new parameters, called “image parameters,” which are the natural ones to use with composite wave-filters. A detailed illustration of the use of this chart calculation method is given and the transmission losses so obtained are found to agree, except for differences which in practice are negligible, with those obtained by long direct computation. In the Appendix are derived two sets of corresponding formulae which are applicable to a linear transducer of the most general type, namely, an active, dissymmetrical one; the one set contains image parameters and the other set recurrent parameters. An impedance relation is found to exist between the four open-circuit and short-circuit impedances of a linear transducer even in the must general case. Reduction of these formulae to the more usual case of a passive linear transducer is also made, those containing the image parameters being especially applicable to the case of composite wave-filters.

A dynamical study of the vowel sounds

Abstract: THE study of the vowel sounds presents a problem which has interested scientists and scholars in varied fields. A knowledge of their nature is of fundamental importance not only in communication engineering but also in acoustic science, phonetics and vocal music. From the earliest theories and the rough experiments of Willis (1829) and Helmholtz (1859) to the later measurements of D. C. Miller (191(5) steady progress has been made toward the accurate determination of their characteristics.

Theorems regarding the driving-point impedance of two-mesh circuits

Abstract: The necessary and sufficient conditions that a driving-point impedance be realizable by means of a two-mesh circuit consisting of resistances, capacities, and inductances are stated in terms of the four roots and four poles (including the poles at zero and infinity) of the impedance. The roots and the poles are the time coefficients for the free oscillations of the circuit with the driving branch closed and opened, respectively. For assigned values of the roots, the poles are restricted to a certain domain, which is illustrated by figures for several typical cases; the case of real poles which are not continuously transformable into complex poles is of special interest. All driving-point impedances satisfying the general conditions can be realized by any one of eleven networks, each consisting of two resistances, two capacities, and two self-inductances with mutual inductance between them; these are the only networks without superfluous elements by which the entire range of possible impedances can be realized; the three remaining networks of this type give special cases only. For each of these eleven networks, formulas are given for the calculation of the values of the elements from the assigned values of the roots and poles.

The transmission unit and telephone transmission reference systems

Abstract: Consideration is given to the method of determining and expressing the transmission efficiencies of telephone circuits and apparatus, and of the desirable qualifications for a unit in which to express these efficiencies. The “transmission unit” described in this paper has been selected as being much more suitable for this purpose under present conditions than the “mile of standard cable” which has been generally used in the past.

An electrical frequency analyzer

Abstract: An apparatus has been developed by means of which it is possible to measure and obtain a permanent record of the frequency components of an electric current wave. The device has two frequency ranges: 20 to 1250 cycles and 80 to 5000 cycles; the amount of power required does not in general exceed 500 microwatts; and the time necessary for making a record is about 5 minutes. An attachment is provided which permits of the making of simultaneous harmonic analyses of two complex waves in the same length of time. In principle, the process consists in feeding the complex wave to be analyzed into a selective network, the essential feature of which is a sharply tuned circuit whose frequency of tuning is controlled by varying the capacitance in small steps with a pneumatic apparatus similar to that in a player piano. A maximum of response of the circuit occurs at each frequency of tuning which coincides with a component of the complex wave. An automatic photographic recorder of the response to each frequency of tuning is provided by means of which the frequency and magnitude of each component of the complex wave may be obtained. For convenience of operation, an automatic control apparatus is provided, so that it is only necessary to connect the complex source or sources to be analyzed and press a starting button. The completed record of the analysis is delivered after the machine has passed through the entire range of frequencies. The application has so far been principally to problems in the communication field such as the analysis of performance and distortion at audio frequencies of vacuum tube and mechanical oscillators and amplifiers, analysis of complex telephone waves and speech sounds, and the effect on a complex wave of transmission through electrical and acoustic apparatus. In the power field many applications are obvious, such as for example, quantitative comparison as to frequency content of the voltage and current supplied to and delivered by transformers, voltage and magnetic flux studies in generators and motors, commutation, and the effect of wave-shapes in power transmission line problems and control apparatus.

Mathematics in industrial research

Abstract: THE necessity for mathematics in industry was recognized at least three centuries ago when Bacon said: “For many parts of nature can neither be invented [discovered] with sufficient subtility nor demonstrated with sufficient perspicuity nor accommodated unto use with sufficient dexterity without the aid and intervening of mathematics.” Since Bacon's time only a very small part of nature has been “accommodated unto use,” yet even this has given us such widely useful devices as the heat engine, the telegraph, the telephone, the radio, the airplane and electric power transmission. It is impossible to conceive that any of these devices could have been developed without “the aid and intervening of mathematics.” Present day industry is indeed compelled, in its persistent endeavors to meet recognized commercial needs, to make use of mathematics in all ot the three ways pointed out by Bacon. The record of industrial research abundantly continus his assertion that sufficient subtility in discovery, sufficient perspicuity in demonstration, and sufficient dexterity in use can be achieved only with the aid of mathematics.

The building-up of sinusoidal currents in long periodically loaded lines

Abstract: Important information regarding the excellence of a signal transmission system is deducibile troni i knowledge of the mode in which sinusoidal currents “build-up” in response to suddenly applied sinusoidal electromotive forces, since on the character and duration of the “building-up” process fiepend the speed and fidelity with which the circuit transmits rapid signal fluctuations.1 The object of this note is to disclose and discuss general formulas and curves which describe the building-up phenomena, as a function of the line characteristics and the frequency of the applied e.m.f., in the extremely important case of long periodically loaded lines. The formulas in question are approximate but give accurate engineering information and are applicable to all types of periodic loading under two restrictions: (1) the line must be fairly long, that is, comprise at least 100 loading sections, and (2) it must be approximately equalized, as regards absolute steady-state values of the received current, in the neighborhood of the applied frequency. Fortunately these conditions are usually satislicd in practice in those cases where the building-up phenomena are of practical engineering importance. Furthermore, the formulas to be discussed supply a means for the accurate and rapid comparison of different types of loading in correctly engineered lines.

High frequency amplifiers

Abstract: IN this paper, a simplified mathematical treatment of the theory of high frequency amplifiers is presented, and the theory is verified by experiment. This method of mathematical analysis provides a ready means of predicting the performance and action of an amplifier from a knowledge of the fundamental constants of its circuit and places the design of high frequency amplifiers on a precise and rational basis. The paper also includes a description of various methods for quantitatively determining the amount of amplification at high frequencies.

Impedance of loaded lines, and design of simulating and compensating networks

Abstract: A knowledge of the impedance characteristics of loaded lines is of considerable importance in telephone engineering, and particularly in the engineering of telephone repeaters. The first half of the present paper deals with the impedance of non-dissipative loaded lines as a function of the frequency and the line constants, by means of description accompanied by equations transformed to the most suitable forms and by graphs of those equations; and it outlines qualitatively the nature of the modifications produced by dissipation. The characteristics are correlated with those of the corresponding smooth line. The somewhat complicated effects produced by the presence of distributed inductance are investigated rather fully. In the absence of distributed inductance a loaded line would have only one transmitting band, extending from zero frequency to the critical frequency. Actually, however, every line — even a cable — has some distributed inductance; and the effect of distributed inductance, besides altering the nominal impedance and the critical frequency, is to introduce into the attenuating range above the critical frequency a series of relatively narrow transmitting bands — here termed the “minor transmitting bands” — spaced at relatively wide intervals. The paper is concerned primarily with the impedance in the first or major transmitting band; but it investigates the minor transmitting bands sufficiently to determine how they depend on the distributed inductance, and to derive general formulas and graphical methods for finding their locations and widths — an investigation involving rather extensive analysis. The latter half of the paper describes various networks devised for simulating and for compensating the impedance of loaded lines; it furnishes design-formulas and supplementary design-methods for all of the networks depicted; and outlines a considerable number of applications pertaining to lines and to repeaters.

Vacuum tube oscillators — A graphical method of analysis

Abstract: THE vacuum tube oscillator is fast becoming one of our most versatile circuits and the requirements which are being imposed upon it are constantly increasing in severity. In some cases it is asked to efficiently convert several kilowatts of direct current power to alternating current power. At other times, it may be called upon to deliver an alternating current having a frequency which shall remain constant within extremely narrow limits. It may be required to operate at a few cycles per second or at several million.

Electrical tests and their applications in the maintenance of telephone transmission

Abstract: THE installation and maintenance of the circuits in a telephone plant employed for the transmission of speech require the use of various testing schemes to insure a high grade of commercial service. Circuits are engineered and installed to meet the established standards of transmission in the most economical manner and this having been done the next step is to provide an adequate testing program. A number of the electrical tests required in this program include well known laboratory methods adapted so that they can be readily applied in the field, while others have been developed for particular use in telephone maintenance work.

Deviation of random samples from average conditions and significance to traffic men

Abstract: THE traffic executive deals with questions which lead him into the consideration of problems of widely differing natures At almost every turn he is confronted by the fact that his decisions and programs in relation to these different phases of the work must be based on records which are seldom continuous and in most cases are merely “samples” These sample records are assumed to measure the characteristics of the entire volume of facts or data of which they are taken to be representative In the use and analysis of these records there are a number of perplexing questions which come to his mind if he allows himself the luxury of a little theoretical speculation

A clock-controlled tuning fork as a source of constant frequency

Abstract: Note: The art of electrical communication employs such a wide variety of methods for the transmission of intelligence that it utilizes alternating currents whose frequencies cover the entire range between a few cycles per second and several million. With the increasing use of these methods, it becomes more and more imperative that determinations of the frequency of any alternating current may be made with extreme accuracy. In particular, recent developments in carrier current telephony and telegraphy over wires have placed exceedingly rigorous limits on the frequency adjustment of certain types of apparatus. It is many times necessary to hold such equipment as oscillators or filters to within 0.1 per cent. of given frequency values under commercial operating conditions. This means that calibrating devices used in the manufacture and maintenance of such circuits must be reliable to 0.01 per cent. and that the primary standard should be good to about 0.001 per cent. or one part in 100,000. The present paper discusses one of the methods recently developed in the Bell System Laboratory for obtaining a source of practically constant frequency with which other frequencies may be compared. It consists of a clock-controlled tuning fork making 50 vibrations per second and, as is shown, the maximum deviation of its frequency from the mean is less than one part in 50,000. A study has also been made of means for improving the constancy of the control clock and a new type of clock mechanism consisting of an electrically actuated pendulum, the impulse of which is controlled by a photo-electric cell, is suggested. — Editor.

Design characteristics of electromagnets for telephone relays

Abstract: Note: The electromagnets described are confined to relays, although the principles involved apply as well to selector magnets, clutch magnets and electromagnets in general. A treatment from the viewpoint of the telephone engineer is given of the important considerations which determine the design of the magnetic parts of relays and the economics of the winding dimensions. A knowledge of these factors as well as of the general considerations which are discussed is of great importance in the selection and application of relays to the telephone system. The operating and economic importance to the Bell System of the great number of relays required in the operation of the plant has been described in a previous paper.1

Some very long telephone circuits of the Bell System

Abstract: Recent papers1 have discussed at length the use of toll cables for the handling of certain long distance traffic. These cables, which are being used in areas of dense traffic, have been made possible by many developments in cable design, repeaters and loading coils. Coincident with these developments have gone others which are finding their application in the extensive establishment of improved open wire circuits for use over very long distances. The purpose of the present paper is to discuss some of the considerations involved in the overall design and maintenance of these very long open wire circuits. These circuits are often referred to as “backbone” circuits and supply a network of trunk lines for the entire Bell System. The most important of these routes are shown in Fig. 1.

Some contemporary advances in physics — III

Abstract: An electrical circuit having a natural oscillation-frequency any-where below 108 can be constructed by anyone with suitable condensers, inductance-coils, and a few feet of wire at his disposal. It can be set into oscillation by abruptly closing it when the condenser is charged, by coupling it to an audion, or otherwise; and the waves which it radiates while oscillating can be detected and measured, at least when the frequency exceeds 104. Thus it is possible to generate perceptible electromagnetic waves of frequencies up to 108, and hence of wavelengths down to 3 metres, by methods that may be called electrotechnical. Waves shorter than 300 cm., frequencies higher than 108 cycles, are not easily produced by any such method; for if one uses excessively small condensers and inductance-coils in the hope of forcing the circuit-frequency much past 108 1 or even omits coils and condensers altogether, it is found that the auxiliary apparatus, the audion, even the wires of the circuit themselves, possess capacities and inductances which can not be annulled and which hold the oscillation-frequency down. By devising oscillating systems which have scarcely any outward resemblance to the circuits of familiar experience (although a formal analogy can be established) Hertz and his successors generated electromagnetic waves of frequencies up to 1011 and wavelengths down to 3 mm. Beyond a certain gap there commences, near frequency 1012 and wavelength 0.3 mm., the far-flung spectrum of rays emitted by molecules and atoms. This interval is one of the two lacunae in the complete electromagnetic spectrum extending from 104 past 1020 cycles, which were mentioned in the preceding article of this series. Unlike the gap between the ultraviolet and the X-rays, it is not believed to be populated by rays resulting from important processes occurring within the atoms, nor do we know of any other peculiar type of radiation which should be sought within it; and perhaps the bridging of it, when finally and unquestionably achieved, will be held notable chiefly as a feat of experimental technique or a tour de force. On the other hand, so long as the gap remains unspanned, we can hardly dismiss the possibility that something in the order of nature may reserve one range of wavelengths for the “natural” rays resulting from atomic processes, and limit the “artificial” waves generable by electrotechnical means to a distant range which never can be extended to overlap the other.

Relays in the Bell System

Abstract: Note: Before they can converse people must either be brought together or virtually be brought into one another's presence by the telephone. Any telephone system must establish talking connections between its subscribers, and these connections must be built up, supervised and disconnected when desired. This work is accomplished by the use of relays of various kinds, and the speed and accuracy of service is largely dependent upon them. There are completed daily in the Bell System about 42,000,000 telephone calls. These involve the successful and accurate operation of over one and one-half billion contact connections daily. Many kinds of relays are employed in the Bell System, varying from the simple electromagnetic drop to the sequence switch, the thermionic vacuum tube and the panel selector. Today a circuit connection between two subscribers served by manual exchanges in a large multi-office district involves about 21 relays. When these subscribers are served by machine switching offices, the number of relays in a local connection may be as great as 146. It not infrequently happens that in setting up a toll connection more than 300 relays are employed. In the present paper the relay developments leading up to, and making possible the present communication system, are outlined with particular reference to electromagnetic relays. A few typical circuit applications are given with a discussion of the requirements imposed upon relays which influence their design. Several types of relays are illustrated and their distinctive features are described. The subjects of relay design, manufacture and maintenance and also telegraph relays will be dealt with in future papers.

Practical application of the recently adopted transmission unit

Abstract: THE purpose of this paper is to outline the practical considerations involved in the use ol the transmission unit (abbreviated TV), which was recently adopted by the Bell System to replace the mile of standard cable in transmission engineering work. A description of the TU, together with a discussion of the considerations which led to its adoption has been given by Mr. Martin in another article in this issue.