US1032658A - Method of and apparatus for electrical-wave transmission. - Google Patents

Description

E. CLEMENT.

METHOD OF AND APPARATUS FOR ELECTRICAL WAVE TRANSMISSION. APPLICATION FILED APR.10,1903.

1,032,658. Patented July 15,1912.

2 SHEETSSHEET l.

E. E. CLEMENT.

METHOD OF AND APPARATUS FOR ELECTRICAL WAVE TRANSMISSION; APPLICATION FILED APR.10,'1903.

2 SHEETS-SHEET 2.

1,032,658. Patented July 16, 1912.

EDWARD E. CLEMENT, OF WASHINGTON, DISTRICT OF COLUMBIA.

METHOD OF AND APPARATUS FOR ELECTRICAL-WAVE TRANSMISSION.

Specification of Letters Patent.

Patented July 16, 1912.

Application filed April 10, 1903. Serial No. 152,019.

To all whom it may concern:

Be it known that I, EDWARD E. CLEMENT, a citizen of the United States, residing at Washington, District of- Columbia, have invented new and useful Improvements in Methods of and Apparatus for Electrical- Vave Transmission, of which the following is a specification.

My invention relates to methods of and apparatus for transmission of energy by means of electric currents.

It has for its object the improvement of such methods and apparatus in the particulars hereafter pointed out.

a For a number of years investigations'have been systematically carried on, and the theory of design of electric circuits thereby gradually established, publications by Heaviside, Pupin, Stone, Mascart and Joubert, and others, rendering the subject a familiar one to electrical engineers. Line properties and the principles underlying the electro-magnetic theory being therefore well understood, it is unnecessary for me to go into an extended discussion thereof. For such a discussion reference may be had to patent to Stone, No, 578,275, and the patents to Pupin, Nos. 652,230 and 652,231, as well as tothe well known publications named above, also to the paper read by Prof. Pupin before'tho A. I. E. E., March 22, 1899, and published in Vol. XVI, of the proceedings of the institute, commencing with page 93.

It has been said, in regard to the problem of electrical wave propagation that, viewed on the physical side, the transmitting apparatus may be taken as a source of radiation, the receiving apparatus as a boundary of secondary radiation due to reflection, and

the wave on the line as an interference wave having for components the direct transmitted and the reflected waves. The power absorbed at the receiving end is due to the difference between the arriving wave energy and that which is reflected per unit of time. Apart from this, energy is absorbed all along the line, and to reduce this absorption to a minimum without increasing the cost of the line beyond prohibitory limits is the Ultima Thule of long distance electrical transmission engineering. The power absorbed in any element of the line depends upon the angle of lag between the current and thepotential gradient or electro motive intensity in that element. This angle depends upon the ratio of reaotance to resistance of tie element, and Dr. Pupin gives the rule that an eflicient transmission requires a line in which the reactance per unit length should be large in comparison to the resistance. I

Many attempts have been made to design a circuit which for currents of 'ven characteristics should approximate t e ideal distortionless circuit; and many suggestions have been made looking to the increase of inductance in the circuit without a corresponding increase in resistance or capacity. So far as I know, none of these suggestions have borne fruit, and none of the designs have come into use, save only those of Dr. Pupin, set forth in the patents above referred to, who distributes inductance sources at recurring intervals in his cir-' cuit, with the idea of producing a loaded line which shall become equivalent with suflicient approximation to a uniform telegraplr line when n is not infinitely large, (n being the number of impedance coils introduced in the circuit). When n is infinitely large the conductor becomes an ordinary line with uniformly distributed resist-- ance, capacity and self induction. In other words, it seems heretofore to have been practically assumedthat it is impossible to attain the ideal circuit, the principal efforts expended, and the most successful ones being along the line of artificial approximation only.

It is the principal object of my invention to produce uniformly distributed induc tance, without increasing the cost of the line beyond prohibitory limits.

In the patents issued to Sylvanus P. Thompson, November 17, 1891, Numbers 571,706 and 571,707, a system of recurring coils was described, and in the subsequent patents to Stone and to Pupin the same general idea is adhered to. In other words, the line is divided up .into sections, and the uniformly distributed capacity of each section is sought to be balanced by one or more coils; the result-being that the average inductance and capacity per unit length of line are balanced. The determination of the least number of coils per wave length or per mile which will produce the desired average result then is taken as marking the limit beyond which it is unnecessary and in fact useless to go. In other words, given certain results, and the least means for producing "dition, the size of the line conductors can be ,also kept down to approximately the present dimensions, so much the better.

I attain my objects by producing self-in ductance as an inherent line property, uniformly distributed, but intensified ever present conditions; That is, by my method the line furnishes its own self-inductance without extraneous artificial sources, or in one form, and to put it in a crude way, the entire line is one long coil. I have succeeded in keeping the resistance and capacity almost within present limits, and the construction and cost of the conductors remain substantially the same. The method is applicable to cables without substantial change in present designs, and it can be adjusted to meet the varying requirements of subway orsubmarine lines, as well as of aerial circuits.

Briefly, my method consists in loading the entire line uniformly. I may increase the iron in and about a single conductor, using it in such divided forms that the core losses shall be negligible; and I preferably make my conductors partly or wholly of iron; but the ultimate and most vital part of my method consists in giving the line conductors themselves a double functionl Most modern circuits, and particularly telephone circuits, are metallic, z. e. all-metal loops. It has been suggested, in connection with the other suggestion that iron be massed. near the individual conductors, that they be spaced apart. In aerial lines this is done to some extent, but in cables it is not. I twist the pairs, and make each conductor a core for the other, placing iron in or about the conductors, as stated, so that the prescribed conditions heretofore laid down shall be met, 2'. e. plenty of. iron to give lag, and well subdivided to avoid losses.

Arrangements of conductors typil'ying my invention are illustrated in the accompanying drawing, wherein the same reference letters indicate the same parts throughout and wherein Figure 1 represents a short length of a single conductor arrangedto give increased inductance over the forms ordinarily in use. Fig. 2 is a similar view of a modified arrangement of a single conductor. Fig. 3 showsa pair of conductors twisted on a subdivided iron core. Fig. 4 represents a simplified arrangement similar to that of Fig. 1. Fig 5 represents a air of conductors twisted on a sub-divided 11'011 core, and each made up like the conductor in Fig. 1. Figs. 6 and 7 show preferred arrangements, wherein each conductor of a pair serves as one side of a magnetic circuit, as well as one side of the electric circuit. Figs. 8 and 9 are diagrams of two-wire circuits connected according to my invention. Fig. 10 is a similar diagram of a three phase circuit. Fig. 11 is a diagram illustrating my method of forming symmetrical magnetic fluxes in opposite and mutually reacting members of a circuit. Fig. 12 is a cable made up of individual pairs such as shown in Fig. 7.

Referring to the drawings, in Fig. 1, a is a core of fine soft-iron wires, surrounded by a copper tube, which may also be of fine wires as shown. Paper insulation 0 is wrapped over this composite conductor, and a layer of fine soft;iron Wires d is then superposed on the paper, and a final covering of paper or other dielectric is put on, as shown at c. This construction results in some increase in capacity, but greatly increased inductance, even in the single conductor. The iron in core and envelop being finely divided permits rapid magnetic changes, and the copper interposed serves to reduce .the total resistance, carrying most of the current, in fact. This arrangement may in some cases be changed by making the core a of copper and the enveloping tube 7) of iron. In such case, however, the principal function of the outer tube of iron wires, d, disappears, as the copper is no longer included between two elements of a magnetic circuit, and the conductor may be simplified as shown in Fig.

4, by omitting the wires d entirely, thereby keeping down the capacity, also.

In Fig. 2 the outer tubular wire envelop d is replaced by a series of punched iron disks, (l'; threaded on the wire over the paper. Each of these is radially slit to prevent a closed circuit, and all being divided, magnetic changes are accelerated.

In Fig. 3 I have shown two solid wire conductors, to and w, preferably of copper, twisted around a divided iron core (Z made up of punchings as in Fig. 2. Each disk is radially slit and has a depression or teat in the center, so that as they are superimposed they all nest together and the twisting revolves them without bringing irregular cutting strains on the conductors. This form of cable is produced by assembling the disks and laying the wires in the grooves, then twisting the whole, and it is most satisfactorily made in a machine having the conductors on spools mounted to revolve with a circular table, while the disks feed down through a central aperture in the bottom of a hopper, the twisted cable beingas in Fig. 3, but instead'of the solid copper conductors, conductors containing iron'are employed, as in Figs. 1 and 2, the cable constituting a development of the conductors shown in those figures, made up in pairs. A

fnearly completemagnetic circuit continuousalong the entire length of the conductors is thus provided, and the effect inincreasing the inductance of the circuit is very marked.

'All the above forms of construction are important, but reliminary. The conductors shown in Flgs. 1, 2 and 4 im art increased and uniformly distribut inductance to the circuit in which they are incorporated, and-the factors of resistance and capacity may be made anything desired within limits. Thus the amount of copper in the conductors, blended with the iron strands, will determine the conductance, and although I have described this copper as being within-or without the'iron, it is perfectly feasible, practically, to mix the copper and iron wires together, ajbett er distribution of the current, or more uniform density, being thus attained than where the two metals are kept separate. Such forms of conductor, and particularly those in Figs. 1, 2, and 4, are useful anywhere, in either metallic or single conductor lines; but for metallic circuit, and-particularly for cable work, the-most important arrangement is shown in Figs. 6 and 7. In this arrangement, shown in Figs. 6 and 7, the two conductors A and B of metallic circuit are given a double function. As they have electric conductance and their electrostatic capacity, (other factors remaining constant), is a function of their length, so I give them magnetic conductance, and immediately the electromagnetic property of inductance becomes also a function of their length. In other words, since we are endeavoring to balance the electrical by the magnetic properties of the line, it seems logical to erfect the symmetry of the line by giving-1t electromagnetic as well as electrical possibilities. To surround the conductors with iron, or to incorporate "iron into a cable with them is good as an expedient, but after all it does not confer on the circuit itself any innate neutralizing powers. To obtain a perfect balance, as well as perfect uniformity in distribution of both inductance and capacity,

cated at C itappears that both must be attributes of the. line itself. So I make the line conductors A and B of iron wire, finely stranded and very soft, or I may make them of iron and copper strands mixed, or of iron strands in a core with an inclosing envelop of copper, or lastly, of copper in the core, and an inclosing iron envelop. In any case, I give the actual'line'conductors magnetic properties. With the two conductors thus made of iron and copper, I then twist them around, each other to form a pair. Observe that by thus twisting, each conductor forms a helix about the other, and each also forms a core for the helix of the other. In Fig. 6, to make this plaine'r, conductor B'is shown carried along straight for a space, as at T, and conductor A is wound closely upon it, while conductor A is then carried along straight fora similar space, as at T", and B is wound closely upon it. Each thus serves both as one side of the magnetic circuit, and

also as one side of the electric circuit. Ob-

serve also, in the performance of this dual function thatt-he direction of polarization, that is the magnetic flow due to any given impulse is in proper sequence or symmetricalin the two sides, just as the direction of current flow is in proper sequence and symmetrical.

In Fig. 7 I have shown the pair of conductors simply twisted in the ordinary manner, which I prefer, and consider quite sufficient for my purposes, although I do not wish to limit myself thereto. If desired still further to increase the magnetic retardation, fine viron wire may be wound over the pair of magnetic conductors at intervals along their length, as shown at C, C, a recurring short length magnetic circuitQbeing thus produced; or, if greater effects are desired the wire may be wound the whole, length of the conductors, as indi- If still further eflects' are desired, the iron'wire envelop of Fig. 1, or the iron disks of Fig. 2 may be used, slipped over the two conductors as one so as to completely surround and envelop them in a magnetic medium. f

It is to be noted here thatthere is no dividing line in my circuit Where one section begins and another leaves off. There are no separate windings. There are even no separate or individual turns, as each conductor twists about the other continuously and from one end to the other of the line the electromagnetic action is quite as uniform and as evenly distributed as the electrostatic or electrical action. There is thus no averaging of lumps of inductance. Any increase in length of the line, or any raising of transmitted frequencies, affects both factors in the equation in due proportion. Thus the question of different spacing of coils for different wave lengths can never arise. The self induction and the capacity of the line, being inherent and not artificial are adjusted once for all, and the proper adjustment of the conductance, capacity and inductance, in this manner, enables a close approach, not a rough approximation, but a real approach, to the ideal distortionless line.

In adjusting the constants of my line, the conductance will of course be determined by the amount of iron and of copper, in the core, which likewise aflects the'inductance. The initial magnetic effect may be governed also by the number of turns per unit length of cable in the twist of the conductors,

while the capacity may be kept down if the core conductor is kept down in size. If in Fig. 6 the core a be. made of copper, and the core sheath 5 be made of fine softiron wires, or if the iron and copper wires be mixed the diameter of the whole conductor need be but little more than at present, and the current density will be more uniform. Where it appears desirable the entire conductor may be made of iron, but in such case the size will of course be increased. It may be more desirable to use copper, and when the adjustment for balance on the line permits this copper to have, say one-half the cross section of the wires now usually employed, the external size of any cable need be increased less than oneit." third; that is, the present three hundred pair telephone cable should accommodate 5 two hundred pairs of the new construction,

1orthereabouts.

The individual twisted pairs A-B may be made up into cables as in present practice. Thus two of these pairs may be twisted together, two twisted pairs again twisted, and so on, or a series of left twists may be overlaid with a series of right twists, and so on as shown in Fig. 12. The complete cable may then be wrapped with paper and a lead or other outer covering applied. It should be noted that in making up phantom circuits, composed of metallic circuits with their conductors in parallel, the magnetic cores of both individuals of each pair act together as one core for the phantom pair,

and so 'on. It should also be noted that while I have described my twisted magnetic conductors as in close proximity, and in cables, etc; they may well be on poles, and

somewhat separated from each other. Thus in a'line, where the current flow is large, turns about each other of the nature of transposition of the wires, if the latter are not too widely separated, will obviously produce inductance in a degree depending on the periodicity and other factors.

I have not, of course, worked out details, such as insulators, supports, etc., for every kind of circuit with which this invention may be used, but Wherever alternating or other changing currents are to be transmitted, or Wherever electrical waves are to be propagated throu h material media, In invention is applica 1e, and I include al systems whatever within its scope. Taking the forms of conductor shown in Figs. 1 and 2, consisting in each case of a conducting core ab and a superposed insulated magnetic body as d or d extending along the conductor, two of these complete conductors may obviously be twisted toether to form a mutually reactive pair, as in Figs. 6 and 7. In such case the current would be carried entirely in each core ab, and its principal efiect would be to magnetize the magnetic body 0? or d of the twin conductor. That is, the actual electric conducting bodies need not be themselves magnetic, but each conductor (using the word in a generic sense) may be compound, composed of tWo conductors (specifically) one electric, a, and the other magnetic, d or J. Of course this is an elaboration of the form of conductor (0-4 (of copper and iron) shown more particularly in Fig. 4.

In order to make the method of connection' of my circuits perfectly clear, I give in Figs. 8, 9, and 10 diagrammatic representations of three forms of circuit with transmitting and receiving apparatus connected thereto.

Fig. 8 shows a circuit composed of two conductors A and B, such as those shown in Figs. 4, 5, 6, and 7, formed int-o a twisted pair and wlth a transmitting apparatus (shown as a generator) G, connected at one end, and a translating device R connected at the other. These two conductors, A and B, contain magnetic material, and each therefore acts upon the other to produce a changing magnetic flux therein, corresponding to the current changes in itself, these combined magnetic fluxes mutually co-acting to produce a large time constant in the circuit. j

Fig. 11 shows very clearly the nature of the effects in the circuit due to the twisting of iron conductors and a current flow thereth'rough. We have here a battery M, a circuit X, and a double helix Y of iron, con- 115 nected at yy in the circuit. A current flow in the circuit passing in each side of the helix around the other side, produces a magnetic condition therein, indicated by poles marked N and. S. This is similar to what 120 goes on in a line constructed according to my method, for each Wave propagated through the circuit. It should be observed, in this connection, that by reason of the helical twist of both line wires, about each 125 other, the direction of each, and the current flow therein is almost at right angles to the direction of the other at any given point, the angle becoming less as the twist of the wires becomes longer and approaching a right 1 0 angle as the twist is made shorter. This is obviously as it should be for the greatest advantage. r

Fig. 9 shows a circuit composed of two composite conductors A. and B, such as shown in Figs. 1 and 2. In this figure the electric conductors a, a, and the magnetic conductors 11,41, are distinctly separated, buteach. one a, is accompanied by its partner d, in the twist. Obviously, (1-, might in this case be entirely of copper, and 03 entirely of iron.

Fig. 10 shows my invention applied to a. three phase circuit. Here A, A and A are three conductors twisted or otherwise carried around each other continuously, asbefore, and having a transmitter or generator G at one end of the circuit and a translating device or receiver R at the other. As the E. M. F. waves succeed each other in the circuit Wires, producing symmetrical rising and falling current waves, the magnetic effects will respond symmetrically to produce the lag in the currents, for which they are adjusted. In'this case the action occurs between successive pairs formed as the phases succeedeach other, each pair for the time being acting like the simple pairsheretofcre described. The ideal line, for all purposes stated, to carry out mymethod, should have pure iron (or other magnetic metal) wires, of a suitable size to produce the desired conductance, twisted together to produce the necessary inductance; and the latter, by reason of the size and inherent magnetic nature of the line conductors themsclves,can be adjusted to neutralize the efiects due to their capacity.

Having thus described a method of attaining perfectly distributed inductance, I do not think it necessary to go into the details of winding and sizes of wire for all kinds of lines. Lines for power or telegraphic transmission will necessarily differ from those employed for telephone transmission, but in every case, given new method of symmetrizing the line, magnetically as well as electrically, the adjustment of capacity and inductance is simple. By following the methods herein laid down the transmission over any line may be rendered much better, and the distortion greatly reduced or entirely prevented, or the distance of transmission may be greatly increased, or both.

As showing one practical method of employing the invention for telephonic frequencies, the following instance is given of a cable which has been built and tested, the measurements being very carefully made. The general form of each pair is that of twisted iron-and-copper Wire, each conductor forming a core for the other. Each wire is composed of an iron core of 0.042

cm. diameter, made of the best soft Norway.

iron covered with copper to a finished diameter of 0.0834 cm. L This wire after being drawn is carefully and thoroughly annealed. The two wires of the pair are insulated from each other with paper strips of such anum-- her and thicknesses to bring the centers of the two wires 0.149 cm. apart after they have been twisted with 19 turns per lineal foot. Cable formed in this manner has a resistance of 151 ohms per circuit mile and a capacity of 0,077 mf. per mile .These constants are equal to those of a-cable formed in the usual manner of commercial copper wire with a diameter of 0.072 cm. and a disknownnow or not. I belleve that'I have successfully solved a very serious problem, and have paved the way for the solution of many others. While I do not claim that I am the first vto suggest massing solid iron about a conductor, I do believe I am the first to suggest or to use any true magnetic method, or to confer upon the line conductors, themselves, magnetic properties. I believe I am the first to present any method of attaining uniformly distributed balancing inductance in a circuit, and I shall claim the same herein as broadly as the limitations of language permit. I believe I am also the first to discover and apply the properties stated of a metallic two wire circuit or circuit doubled on itself, for any purpose. By this I mean particularly the formation of magnetic fluxes in symmetrical relations in the conductors, under the influence of electric currents in the same having like symmetrical relations. I shall therefore claim this also broadly.

I desire it to be understood that I am aware many variations may be made in the details of connection, and in the materials I have specifically described and illustrated. Thus the iron introduced in the conductors may be finely subdivided in the form of powder or dust, as suggested for another purpose by Heaviside, or the .IIOII or steel wires may be copper plated to lncrease thelr conductance, or the iron or steel wires may be in short lengths, to produce the same double subdivision shown in the disk form, both transverse and longitudinal, the copper as before stated carrying the current. All such changes, however, lie within the scope and purview of my invention.

' subdivided iron throng Having thus described my invention, what I claim and desire to secure by Letters Pat: ent of the United States is:

1. Acable for electric transmission constructed to reduce distortion, and comprising a twisted pair of wires of low resistance having magnetic properties, with iron wound spirally around said pair, substantially as described.

2. In a system of electrical wave transmission, a circuit comprising a pair of associated conductors arranged in a continuous and symmetrically changin figure so as to be continuously and mutua' y reactive, and a body of subdivided magnetic material closely following and associated with each conductor throughout its length, substantially as described.

3. In a system of electrical wave transmission, a circuit composed of a double or twin conductor arranged in a continuously interlocking figure and having each member associated with a separate body of finely out its length, substantially as described.

4. In a system of high frequency electrical wave transmission, a cable composed of a plurality of twisted conductors, each conduotor individually associated with a body of finely subdivided iron throughout its length, substantially as described.

5. A11 electric circuit composed of opposite conductin elements, each consisting of a mixture of nely stranded iron wire and copper wire, laid u in a twisted pair, substantially as descri ed.

6. A high inductance circuit for electrical wave transmission, composed of conductors containing strands of iron, said conductors being twisted together but electrically insulated from each other, substantially as described.

7. A high inductance circuit for electric wave transmission, consisting of conductors containing iron, twisted together, and iron masses uniformly distributed alon the twisted pair, substantially as describe 8. In a system of electrical wave transmission, the method of regulating the efiective inductance which consists in providin a compound wave conductor having each member containing magnetic material, and the other member carried thereabout in a helical direction, and increasing or decreasing the pitch of the spirals, or the angle of their tangents to a plane transverse to the axis, substantially as described.

In witness whereof I have hereunto subscribed my name this tenth day of April, A. D. 1903, in the presence of two Witnesses.

EDWARD E. CLEMENT.

Witnesses:

SPENCER B. Psnn'rrss, EDWIN S. CLARKsoN.

Images