US 2373906 A
Description (OCR text may contain errors)
Ap 17, 1945. H. MOURADIAN CABLE 'SYSTEM FOR HIGH FREQUENCYTRANSMISSIONS Filed Feb. 25, 1943 ll Sheet s-Sheet 1 FIG. 3B
IN V EN TOR.
April 19-45- E H. MOURADI AN 2,373,906
' CABLE SYSTEM FOR HIGH FREQUENCY TRAN'SMISSIONS Filed Feb. 25, 1943 11 Sheets-Shed 2 "TO TELEPHONE TELEPHONE LINE 0R LINE OR STATION I STATION LINE c T0 T0 TELEPHONE TELEPHONE LINE LINEOR RERSTATION E'IF d -P -p- INVENTOR.
H. MOURADIAN April 17, 1945.
CABLE SYSTEM FOR HIGH FREQUENCY TRANSMISSIONS Filed Febfgs, 1943 ll Sheets-Sheet 3 FIGB IINVENTOI.
April 17, 1945.
. H. MOURADIAN CABLE SYSTEM FOR HIGH FREQUENCY TRANSMISSIONS Filed Feb. 25, 1 94s 11 Sheets-Sheet 4.
' LINE IN VEN TOR.
April 1945- H. MOURADl A N 2,373,906
CABLE SYSTEM FOR HIGH FREQUENCY TRANSMISSIONS adieu n IN V EN TOR.
April '17, 1945. H. MO U RADIAN I 7 CABLE SYSTEM FOR HIGH FREQUENCY TRANSMISSIONS Filed Feb. 25, 1943 1 Sheets-Sheet 6 i (a) (b) (c) (d) p (a) (b) (c) M)! INVENTOR. I
P" 1945- H. MOURADIAN I 2,373,906
CABLE SYSTEM FOR HIGHFREQUENGY TRANSMIESIONS Filed Feb. 25, 1943 INVENTOR CABLE SYSTEM FOR HIGH FREQUENSY TRANSMISSIONS Filed Feb. 25, 1943 ll Sheets-Sheet 8 IN V EN TOR.
April 17, 1945. H. MOURADIAN 2,373,906
CABLE SYSTEM FOR HIGH FREQUENCY TRANSMISSIONS Filed-Feb. 25, 1943 ll Sheets-Sheet 9 April 1 7, 1945.
Filed Feb. 25, 1943 H. MOURADIAN. CABLE SYSTEM FOR HIGH FREQUENCY TRANSMISSIONS 11 Sheets-Sheet 10 LINE LINE
I \I v l I FUNDAMENTAL CIRCUIT LINE LINE
IN VEN TOR.
p 7, 1945. I Hmoummm 1 2,373,906
QABLE SYSTEM FOR HIG H-FREQUENCY TRANSMIS SIONS Filed Feb. 25, 1943 1n Sheets-Sheet 11 LINE r 3 AF F LINE fi! n 7 INVENTOIIQ.
Patented Apr. 17, 1945 CABLE SYSTEM FOR HIGH FREQUENCY TRANSMSSION Hughes Mouradian, EhiladelphimBa. Application February 25,1943, Serial No. 477,159
This invention relates to a method of and describes the means reqlured for eliminating inductive interference between adjacent circuits, particularly those in close vicinity to each other, such as those in cable.
In the early stages of the development of the telephone art, line circuits consisted, as well known, or single wires grounded at both ends. With the use of line circuits of this type, noise and over-hearing were unbearable. As stated by General J. J. Carty-A wire was strung from Boston to Lawrence, about 26 miles, on a tele- Anybody listening on that telephone wire could hear all the telegraph messages. It was, as though, we had a drum corps and each drummer began drumming a separate tune. The first important forward step consisted in the invention and use of the metallic circuit.
Alexander Graham Bell was awarded, on July 29, 1881, U. S. A. Patent 244,426 describing the basic principles of the metallic circuit. It was pointed out by the inventor, in a deposition taken before the Commissioner of Patents that the improvements he had devised were:
1. To place the directand return wires close together, and to place around them a common cover or envelope of such thickness that the telephone wires should always be so very much nearer to oneanother than they were to any other adjoining Wires.
2. To twist the direct and return wires around one another, so that they should be absolutely equidistantfrom the disturbing wires.
The two-wire metallic circuit of the type just described has'been the basic backbone of all telephone circuits. Additional improvements have been introduced by the use of different lengths of twist for adjacent pairs, and by the use of special manufacturing measures and installation procedures which have had for their objective the still further reduction of both the residual mutual capacity unbalances and inductive reactive unbalances between adjacent pairs.
The next step in the art, consisting in the in= troduction of the phantom circuit was a step in reverse in so far as mutual unbalances were concerned. The problems of the telephone engineer in overcoming cross-talk, babble, noise; etc. were greatly increased by the. introduction of the phantom circuit. However, the economical advantages of being able to obtain, by phantoming, an additional circuit at no additional cost for line conductors, more than compensated the disadvantages ofthe increased difliculties involved, in the way of overcoming mutual interference be"- tween telephone circuits. It is important to call attention here to the fact that the phantom ciralleviated by the use of cult uses 4 wires, just as one of the several systems disclosed in the present specifications, but
while the particular arrangement known as the phantom circuit or "quadded cable arrangement resulted in an increase of mutual interference conditions as above stated, the particular form of -wire arrangement hereinafter described has exactly the opposite property of decreasing the mutual interference.
The dificulties, above referred to, brought about by the introudction of the phantom circuit were capacity unbalance testing methods during the installation of the phantom, quadded or duplex type cables. With the advent, however, of the use of higher frequencies in the art, particularly with the introduction of carrier methods for cable transmission (range of 12,000 to 140,000 cycles per second), it became necessary to give up entirely the use of phantom facilities, and to separate incoming and outgoing messages by providing two separate cables, one for each direction of transmission, in order to reduce mutual interference to tolerable limits. In addition to the steps just indicated, extremely intricate special balancing methods are at present required, each carrier circuit being balanced against every other carrier circuit, to reduce mutual interference to a point satisfactory for commercial use. A very clear and adequate exposition of the dimculties involved and the methods used at present in overcoming these difficulties will be found in the following symposium of papers, all of which appeared in 1938 A. I. E. E.
3. Crosstalk and Noise Features of Cable Carrier Telephone Systems, by M. 'A. Weaver, R. S. Tucker, and P. S. Darnell (pages 250-261) The last paper deals in some detail with the difllculties-of cross-talk and babble with the present day type K cable carrier system and is therefore of direct interest in conjunction with the consideration of the invention disclosed in the present specification.
In the field of radio applications also, where the frequencies used are expressed in megacycles, the
\ interference. between the wiring or input and output terminals of'ampliflers and modulator has been so great as to require radical modifications in the. design of these amplifying or modulating tubes in order to avoid singing" or howling, by
keeping the input and output circuits as far apart as possible, actually at opposite poles.
I have discovered thatthe difliculties experienced with the present day two-wire and four-wire metallic circuits can be overcome, if use is made of multi-conductor systems of wires for each two-way signalling circuit, said wires being arranged on the periphery of a cylindrical core, with positive and negative polarities succeedin each other around said periphery in sequence, thereby producing a partly neutralized electromagnetic and electrostatic field in space, provided further corresponding elements in adjacent similar systems of wires are set at predetermined angles with reference to the line connecting together the center of said first system of wires to the center of said similar systems as fully described in these specifications. I'find that the minimum number of wires required to produce the partly neutralized field is four (4). The advantages of the new system, first disclosed in these specifications, increase rapidly with the number of wires provided for each two way circuit.
To produce a neutralized field, in the specific ways described fully hereunder in the present specifications, the various wires of the same twoway system have to occupy fixed space relationships to each other, in contradistinction, to cite an example, of the presentday "quadded" cable system, wherein the wires .of the quad continuously change their respective positions with reference to each other. This structural characteristic is well defined in the original U. S. A. Patent 863,969 for acable of the quadded type, the first claim of which reads:
A niethod of making telephone and telegraph cable cores, which consists in twisting together 7 pairs of individual conductors with a constant length of lay from end to end ofeach twisted pair, and again twisting the twin conductors in pairs but with a different length or lay.
A second main objective of the present invention is to secure the improved type of transmission circuit for high frequency use in closely packed wiring, as a superimposed carrier circuit Fig. 1 indicates two present day two-wiremetallic circuitsin close vicinity to each other; Fig. 2
indicates one form of the proposed system for high frequency conductors using just four wires for a two-way ,circuit; Fig. 3 indicates a second form f the proposed system of high frequency conductors using six (6) wires for a single twoas a general conception for a high frequency conductor, in which a hollow core conductor is divided into an even number of segments, by means of intervening insulation. A single two-way circuit is constituted by bridging together all nonadjacent metallic segments into two conductors; Fig. 4 indicates the method of using-a proposed system with four constituent wires in a two-way circuit arrangement; Fig. 5 indicates the,method of using the proposed system with four'constituent wires in aone-way circuit arrangement,- a second set of four constituent wires being used in the opposite direction the combination representing the equivalent of thepresent day socalled four wire system, as modified for high frequency use; Fig. 6 indicates the proposed system as applied to repeater circuits of the two-way type; Fig. 7 indicates two four-wire units, designated as A and 13, with unit A making an angle of c degrees and unit B an angle of 0 degrees with the line 0102 representing the line of centers of units A and B. Only one of the four wires in each unit is shown, i. e. wires I and 5 respectively, and the definition of the angles (p and 0 is related to these two wires as shown; Fig. 8 represents the distortion function Fup as a function of the angle (p, as fully explained in the specification; Fig. 9 shows a quartet of the fourwire units A, B,;C, and D with centers 01, O2,- O3, and 04, respectively with the centers C1, C5, CB and C13 of wires I, 5, 9 and I3, defining the angles a, 1,0 and F. This drawing serves as a basis for the analytical determination of the angles which result in zero coupling between the quartet of units A, B, C and D; Fig. 10 shows the designations used in the specifications for an as-, sembly of adjacent four-wire units; Fig. 11 indicates an arrangement whereby two two-way cir-\ cuits can be derived from a set of four-wires, one two-way circuit terminated at terminals CC representing a high grade, high frequency carrier circuit, and a second two-way circuit superimposed upon the first terminated at terminals DD with no mutual reactions between these two circuits, each using the same four wires; Fig. 12 indicates an arrangement 'whereby three two-way circuits can be derived mm a single set of 4- wires. consisting of one fundamental carrier circuit actually using all four wires terminated at terminals CC and, in addition, two two-way two-wire circuits, terminated at terminalsDl) and E135 respectively; Fig. 13 indicates a set of transpositions to be installed during the manufacturing of the cable between successive equal lengths of cable to eliminate two-wire unbalances in a quartet of four-4-wire units; Fig. 14 indicates an alternative arrangement to that shown on Fig. 13, Fig. 15 indicates the'additiona1 pattern of transpositions required to extend the freedom from mutual interference on the twowire combinations to a group of 16 4-wire units; Fig. 16 indicates the additional pattern required to further extend the freedom from interference to agroupof 64 4-wire units; Fig. 1'? indicates a system of transpositions to obtain freedom from interference between a second system of 4-wir'e way circuit, inalternative positive and negative sequencesfthe seventh and central wire being a filler only; Fig. BAindicates a third form of the proposed system of high frequency conductors using eight (8) wires for a single two-way circuit circuits depicted as terminating at terminals DD on Fig. 11 of the drawings, such circuits forming part of a group of four adjacent 4-wire units, such second system being actually additional to the fundamental. 4-wire circuit terminated at in alternative positive and negative sequences.
The void shown in the center of Fig. 3A would be occupied by filler wires devoted to low freqnencyvsystems. Fig.3B indicates afourth form, I
. carrier frequency,
talk by 21*db. at 14,000 and 30 db. at 60,000 cycles.
as'zacce upon the fundamental wire circuit, with unbalances developed which are all of the same size and sign, as shown in detail on Table XIX of the specification. Figure 19 indicates the method required to derive a balanced 8-wire two-way circuit using either the arrangement of Fig. 3A or Fig. 3B of the drawings, said two-way circuit being terminated at AA, with wires I to 8 inclusive terminated in the particular manner as shown. Fig. 19 also shows how it is possible to derive a second auxiliary or additional 8-wire circuit terminated at BB and two auxiliary 2-wire circuits terminated at DD and FF. Fig. 20 shows an alternative arrangement to that shown on Fig. 19 of the drawings, with an auxiliary e-wire circuit terminated at BB replacing the auxiliary 8-wire circuit of Fig. 19 of the drawings.
The present invention has for its main purpose the eliminating of or the r ducing to a negligible minimum the following effects:
1. Crosstalk between circuits in the same cable. 2. Babble between circuits in the same cable. 3. Noise from outside sources of interference with communications including, among others, the following: (a) Noise from heavy static on open wire taps to cable circuits.
(b) Noise from telephone and telegraph repeater oifices.
Quantitative measurements have been carried out and reported by Weaver, Tucker and Darnell (A. I. E. E. Transactions, 1938, pages 250-259) on all these effects. The near-end cross-talk is so large at carrier frequencies that it has been found necessary to route outgoing and incoming messages in two different cables as already referred to. The carrier transmission arrangement used in present practice is illustrated in Fig. 3, page 240 of the paper by Chestnut, ilgenfritz and Kenner (A. I. E. E. Transactions 1938). "it may be noted that the outgoing message is delivered to the line a level of +9 db. The received message, on a separate cable pair in a separate cable, is received from the line at a level varying between -37 db. and 59 db., depending upon the for a repeater spacing of 17 This arrangement was made practicable through the use of the special methods described in detail by WeavenTucker and Darnell in their paper already referred to. As these special methods, some quite intricate in character, have been quite clearly and adequately described by the above authors, they will not be outlined here. It may be sufllcient here to remark that, as a result of the special methods used, it has been possible to reduce far-end crosstalk as follows:
Summary-Grosstalk reduction; far end crosstalk 14,000 60.000 cycles cycles Db. Db. l Crosstalk before poling or balancing 78 --66 2 Crosstalkafter pollng or balancing 88 -83. 3 Reductionobtained; 17. 5 a
For a repeater spacing of 17 miles, the nearendcrosstalk is greater than the far-end cross- As the reductions cycles per second obtainable by the methods heretofore used in the art with non-loaded pairs could not go beyond the limits indicated in the above summary, it was tion shows how the art of communication by means of paper insulated cables, finds itself handicapped and limited when carrier transmission is made use of to increase the number of channels of speech over a given number of conductors. For a more thorough and complete understanding of these 'difiiculties, reference may be made to the several original A. I. E. E. papers cited hereinabove in these specifications and to U. S. A. 2,336,627, granted to me for a System for overcoming thermal agitation noise and inductive disturbance noises in carrier systems.
The present invention has, for one of its main purposes, the reduction of both near-end and far-end crosstalk to limits below those found possible to realize in the art when use is made of No. 19 B. 8; S. gauge cable conductors for transmission. It is nowise restricted, however, as will be amply clear by an understanding of the general principles developed herein'below in these specifications, to the use of No. 19 B. & S. gauge conductors.
I have indicated in a. companion application, which recently matured into U. S. A. 2,336,627, granted to me, one arrangement whereby greater reductions in crosstalk, babble, etc., can be obtained than shown as practicable to secure at the present time in the summary- Crosstallr reduction, reproduced hereinabove. rangement use is made of a circuit layout (see! Fig. i of U. S. A. 2,336,627) in which equal amounts of crosstalk, both far-end and nearend, on two physical pairs, are made to cancel each other. The fundamental conception first disclosed in U. S. A. 2,336,627 can be carried out in at least two different ways, both using present day type cable facilities:
I. Use, for the two physical pairs required, the two side circuits of a phantom or duplex type cable conductors. The additional requirement is that instead of providing twists of different lengths for the two side circuits, as at present provided in the art, it is necessary on the contrary to provide the two side 'circuits with twists exactly of the same length. in this manner, it is possible to obtain relative exposures for each of the two side circuits which are equal to each other with respect to all other conductors in the same cable, thus satisfying a fundamental requirement of the invention.
II. Use, for the two physical pairs required, a so-called spiral-4 quad. The requirement that the two physical pairs should have the same length of twist is automatically satisfled by the spiral-4 construction. Adjacent spiral-4 quads would of course be provided with diiferent lengths of twist.
made exactly the same by construction as nearly as practicable. In the present application, dis- In this aT-Q closure is made for the first time in the art of an arrangement whereby instead of developing equal exposure effects of the same sign, on two physical pairs (or 4 wires) and then eliminating these effects by opposing them to each other at (a) The outgoing and return conductors, instead of being oppositely located to each other in the 4-wire system, are on the contrary adjacent to each other; On Fig. 2 of the drawings of the present specifications, current on conductor I finds its return on conductor 2 and the current on conductor 3 finds its return on conductor 4. This is indicated by the sequence of on Fig. 2 of the drawings. See
also Fig. 11 of the drawingswhere in the fundamental circuit CCf the instantaneously currents fiow through conductors l and 3, and the currents through conductors 2 and 6. Thus, the angular displacement of the conductors carrying outgoing and return conductors in the present invention is 90 and not 180 for the 4-wire arrangement.
(b) The various systems of 2n wires in individual groups in the same cable are set at "predetermined angles with reference to the lines connecting the respective centers of said 21!. wire systems. Again referring to Fig. 2 of the drawings, the above means that the angles o ando of corresponding elements i and 5 of two adjacent 4-w'lre systems with reference to the line of centers O1, 02 of said system are taken equal toeach other, the common value being determined in the specific manner first disclosed in the present specifications.
This point is made clearer by reference to Fig. of the drawings, in which A represents any 2n-wire system of the type first described in these specifications and similar balanced systems a're designated by B, C, D, etc.. stacked as shown. It will be noted that a set of arrows is'also shown on Fig. 10, all such arrows being parallel to each other.- The directions of the arrows with reference to the line joining the centers of units or tem in said row. The full meaning of this point 4 (c) The "wires, in any one 2n-wire system, are not twisted together (as in present-day practice with paper insulated cables) but, on the contrary, are maintained parallel to each other at the predetermined angles already defined in some detail hereinabove, which produce no coupling or no inductive effects" between adjacent 2n-wire systems.
Since the freedom from interference is due in part, in the case of the present invention, to the use of pre-selected fixed relative locations of any one of the 2n-wire systems with reference to all other similar systems, such freedom would be destroyed if these optimum relative locations wer changed, as they would be in case the wires of the 2n-wire systems were twisted. 2n wire spiralling may be used, but in this case, all 2n wire systems must have the same spiral length and corresponding elements be properlydisposed.
When conditions a, b and 0 have been satisfied,
no matter how short a length of the 2n-wire sys- 40 ing of an inherently balanced system.-
To summarize the above, it may be stated that the basic foundation of the invention consists, broadly, in cons'tructinga cable with a multiplicity of separate cores, each core consisting of 2n-wires, where n is an entire number greater than unity. such wires in each core being symmetrically disposed around the periphery of a perfect circle, as shown on Fig. 2, Fig. 3', Fig. 3A and Fig. 33, all 2n wires of each separate core being used cooperatively to form a, single speech systems A; B, E, -F in the horizontal direction'or A, C, I, K in the vertical direction indicate what is referred to as predetermined angles in the claims. Wires bearing the same numbering such as l, or occupying the same relative positions in each of the 2n-wire systems, are referred to as corresponding elements in the specification and in the claims. For instance, on Fig. 9 of the drawings, depicting a set of adjacent 4-wire systems, wires I, 5, 9 and 13 are, by definition, "corresponding elements. In practice, such wires will be identified by special colored paper insulation or binders, or by any other suitable identifying marks.
The predetermined angle is again, by definition, the angle between a designated element such as wire I in a 2n-wire system and the line joining the centers of a row of adjacent systems or carrier circuit, all of said wires being maintained parallel to each other at predetermined angles. thus be formed of n wires. The wires designated as are joined together to form one side of a "composite conductor, the wires marked forming the second side. The joining of said wires is best carried out at repeater points,
though nothing prevents or precludes such com- I wire system of the present invention, using wires l, 5, 3, I, 2, 8, 4, 8 would be terminated as a single circuit at AA. 1
' The testing of the wires for maintenance purposes is, however, most conveniently carried out if the points where the component wires of a 2nwire system are joined together are restricted to repeater points or to other points where specially provided terminating and testing facilities can be made available.
Each side of said single circuit would.
I have also-discovered that where 211 wires are used in accordance with the provisions of the present specifications, the electrostatic and elec tromagnetic unbalance figures decrease very rapidly as 12 increases numerically. With just two wires, I find it is actually impossible to construct certain wires in each cable reel are provided with a transposition of the 7 type as shown on Fig. 13 and Fig.
14 of the drawings. All other unbalances are of negligible order as shown hereinbelow. Where 8 wiresare used, in the particular manner fully describedhereinbelow, and as illustrated on Fig. 19 and Fig. 20 of the drawings, all unbalances are eliminated when separate cores,
each consisting of 8 wires as per Fig. 3A are assembled together in, accordance with Fig. 10 of the drawings. Thus, without making 11:01, which would automatically result in a zero electromagnetic and electrostatic field throughout space (but which would be quite-impossible to realize in practice) it is possible to secure an inherently balanced cable system with an '8-wire layout as per Fig. 3A of the drawings, and an almost but not quite perfectly balanced system with a 4- wire layout as per Fig. 2 of the drawings.
Where the inductive couplings can be annulled as shown 'hereinbelow, bothfar-end and nearend crosstalk can be eliminated. Babble, which is .the resultant crosstalk from a multiplicity of inducing sources in the same cable, will also cancel out. As to the other remaining aspects of the primary objective of the present invention, i. e., the elimination of noise from outside sources of interference with communications, these will evidently also cancel out if all circuits in the same cable make use of transmission system first fication.
In view of the outstanding fact that the building of an inherently balanced cable communication circuit requires a multiplicity of wires, which disclosed in this speci- .45 the balanced type forming the second main objective of the invention, is covered in considerable detail in so far as 4-wire balanced systems are concerned, in Tables XI to XIX, inclusive.
In the case of 4-wire systems, the auxiliary circuits are circuits DD, shown on Fig. 11, or circuits DD and EE' shown in Fig. 12. It will benoted that the arrangement shown in Fig. 11 of the drawings for DD is a 4-wire arrangement of the series type. It is clearly evident that the secondaries of windings l, 3 and 2, 4 could be bridged in parallel instead of in series, and equivalent results obtained thereby. This alternative arrangement has not been illustrated in the drawings. In the arrangement shown in Fig. 12 of the drawings, actually two (2) 2-wire circuits have been derived which do not interfere in any way with the fundamental circuit CC of the d-wire type first disclosed in these specifications, nor do they interfere with each other.
In the case of 8-wire systems, as per Fig. 3A of the drawings, while six (6) additional circuits appear possible, actually the number of mutually non-interfering circuitsds more strictly limited. The arrangement of Fig. 19 of ,the drawings shows, in addition to the fundamental 8-wire circuit AA', using all 8 wires ii to 23, inclusive, the
following mutually non-interfering superimposed or auxiliary circuits". 7 (a) A second 8-wire circuit terminated at BB.
(b) A 2-wire circuit terminated at DD. (0) A 2-wlre circuit terminated at FF.
Alternative arrangementsto circuits DD and FF are the circuits EE' and GG'. From thestandpoint of best overall practical performance, it may be preferable in some cases to further restrict the above arrangement to the combination-of circuits BB and DD or BB and FF as an alternative, thus obtaining 3 circuits as a total,
AA with the 8-wires, l to 8 inelusive.
The arrangement of Fig. 20 of the drawings is alternative to that illustrated by Fig. 19 of the drawings for an 8-wire arrangement. It includes,
in addition to the fundamental circuit AA, the
following mutually non-interfering'superimposed or auxiliary circuits:
is a decided practical disadvantage, I have car- 1 ried out investigations to determine whether the wires used for a balanced circuit could be utilized to secure additional or auxiliary communication circuits, thereby decreasing the primary costs of furnishing service by meansof cable facilities, using high frequency carrier systems. The term superimposed circuit is frequently used in the art instead of auxiliary or additional circuit, as above stated. I have discovered that lt'is' definitely' practicable to secure such superimposed circuits, though these, I find, will be of poorer quality from the standpoint of crosstalk and babble' than the fundamental inherently balanced circuits hereinbefore described. However, if these auxiliary circuits'are restricted for use tothe shorter haul traffic between points along the route'of the cable,- they would still meet therequirements of present-day telephone service.
Under these conditions, all such superimposed circuits would be obtained at no additional costfor line conductors and thereby would greatly de- 4 crease the-overall costs of. providing speech facilities along the route of the cable. This point,
, standpoint of practical principles first disclosed in these specifications. 1
Alternative arrangements to circuits DD and FF are the circuits EE" and GG'. Again, it may be preferable in some cases to restrict the 2-wir'e auxiliary circuits to either DD or FF. Attention is called to the need for providing the transposition T between BB and CC on Fig. 19 of the drawings, as otherwise the crosstalk between adjacent auxiliary 8-wire circuits would be intolerable instead of zero. This is due to the fact that the couplings produced onfBB' and C0 by immediately adjacent similar circuits, are of equal and opposite sign and cancel each other if connected as shown.
The possibility of utilizing auxiliary circuits in addition to the fundamental circuit using 21ewires, is an important development from the application of the basic It should be noted that these auxiliary circuits, in order to meet the present-dayrequirements of a maximum of 1000 units of crosstalk between cir- 2n-wires is actually a balanced 4-wire the angle of rotation is ,lengths of cable. If, for instance,
cults in the same cabledevoted to toll service, require the use of a pattern of transpositions as illustrated by the diagrams of Fig. 13. to Fig. 18, inclusive. Th e transpositions will be, in practice, provided at the splicing of the wires between successive cable reels. The working out and the application (if such patterns of transposition, unless restricted to a few combinationsmay become unmanageable, as well known to those method consists in rotating all wires in each sep-- arate core consisting of 2n-wires. by an angle equal to Y around the geometrical center of each separate 2 Thus, where-the fundamental circuit of circuit,
'% =90 degrees angle of rotation is 45,-etc.- 'Such an arrangement can be best carried .at during the manufacturing of the cable, since no complicated pat- '3 mental circuit is a'Zn-wire system. I have designited by the symbol 2" on the diagrams in Fig. 14, Fig. 17 and Fig. 18, the rotational arrangement or the transposition of the wires. Thus, wires 1, 2, 3 and 4 of the -wire inherently balanced system are transposed with wires 2, 3, 4
In the case of the 8-wire balanced system, the
4, I. If reference is made to Fig. 2 of the drawings, which shows the location of the wires in a single one of the separate cores in the cable, it
will be seen that the change of positions-of wires conceivably be called a barrel type transposition. Thus wire I is connected to wire 2, etc.,
and such .transposition" is equivalent to a rotation of90 around the center 01 of the sepabalanced auxiliary circuits by ineans of the'Z type transpositions applied on all wires as above described, and such use is supplemented by the,
application of the X-type transposition system in conjunction with the comparatively simple patterns indicated on Fig. 13 toFig. 18, inclusive, it becomes feasible to develop auxiliary circuits which will be sufiiciently well balanced to meet the requirements of present-day telephone service, provided further the length of such circuits is restricted in accordance with the indications givenhereunder.
In the case of the i-wire inherently balanced circuit of Fig. 2 of the drawings, where two auxiliary two-wire circuits are derived as per Fig.
12 of the drawings and the transposition patterns of Fig. 14, Fig. 15 andFig. 16 are used, a balanced cable section will include 64 successively spliced reels. If, as is-customary, the length of each reel is about 116 of a mile, a balanced cable section will have a length of 6.4 miles with the transpositions of the X" type installed between I reels at the cable splices; The arrangement'just described will leave a maximum unbalance be-- tween the auxiliary 2-wire circuits of 10' micromicrofarads per mile. It will be noted, by refer-.
ence to Tables XVI andXVII, that when use is made of the Z type transpositions as iirstdis-v closed in these specifications, the unbalance be-" tween 2-wire circuit, l, 3 df one unit, such as unit A, andthe unbalance of a second 2-wire circuit, such as circuit 5, 3 of anadjacent unit is zefo. Therefore, ifwe furtherrestrict ourselves in the utilization of two-wire possibilities and arrange if these 'Y type transpositions are changed into I, 2, 3, I to2, 3,4, I is equivalent to what might. the "XY type, also as described hereunder in these specifications, thena cable reel can be manrate core consisting of wires I, 2, 3, 4, since the angle subtended by wires 1 and 2, i. e., the angle O1, 2 is 90. 1
The "2 type barrel transposition must be applied, during the manufacturing process, to equal Z type transposition on all wires for the inherently balanced system as per Fig. 2 of the drawings. plings is of the order of 10 db. This reduction applies only to the auxiliary circuits-such as circuit DD of Fig. 11 of the drawings or, to the two 2-wire circuits DD and EE' of Fig. 12 of the drawings. This will be clear by reference to Tab1es XVI and XVII in comparing the unbalance. figures under the third column marked Section a with those under the last column.
When use is made of the general method of the cable reel The reduction obtained in inductive cou-,
unbalances will have been removed as between units of any one sub-group such as A, B, C, D-
orcE, F, G, H as per Fig. 10 of the drawings. Under these conditions, either a single two-wire auxiliary circuit can be obtained with only 1 or '2 micro-microfarads unbalance per mile and therefore of better grade than the two two-wire circuits of Fig. 12 of the drawings, using the same number of cable reels in sequence to secure a balanced cablesection as per'p'atterns of Fig, 14, Fig. 15 and Fig. 16, or the length of the balanced section can be reduced by omitting the transposition pattern of Fig. 16 of the drawings.
The observations Just made as regards the amtiliary two-wire circuits which can be obtained with a cable core asper Fig. 2 of the drawings apply with equal force'to the auxiliary two-wire wire unbalances of the auxiliary 4-wire circuit terminated at DD can be reducedare shown in Table XVIII and illustrated in Fig. 1'7- of the 4 l drawings. In the case of the 4-wire auxiliary. cir reducing. crosstalk between the inherently un- 7 cuit of Fig. 20 01 the drawings terminated a t BB'.
' I have .found that an auxiliary circuit of this type In order to help in an understanding of the manner in which the difiiculties due to capacity and mutual reactance unbalances are removed or reduced to negligible amounts with the l-wire two-way system first disclosed in this specification, the following analysis for a two-wire circuit is presented as a basis of comparison. Fig. 1 of the drawings shows two adjacent No. 19 B. 8; S. gauge pairs. It is assumed that wires 6 are subject to interference from wire 2. There will be two different efiects, as well known, when an 1. An induced current effect upon wires 3, l due to the reactance unbalance between the system of wires 2 and 3, t.
2. A direct potential effect, due to capacityv unbalance between the system of wires 2 and The capacity unbalance effect is computed using the Fender-Osborne formula (ElectricaLWorld, vol. 56, p. 667, also Electricity and Magnetism by Pender, page-40, 1919 edition), giving the mutual capacity between two wlres- (1)' C= 2m microfarads per mile cosh=1 where =spe cific dielectric capacity D=distance between centers of wires considered. r=radius of wires.
The mutual reactance effect is computed using the following formula (see Bulletin of the U. S.
Bureau of Standards, vol. 8, No. l. p. 15) for the coeificient of mutual inductance between two wires, of length-(l) separated by a distance (d) M 2l(log 2 Zl+ g- The mutual inductance between wire 2 and a With such an arrangement, the
, alternating current is allowed to flow in wire 2.
result in the following capacity unbalance figures for the system of wires 5, 2, 3', i in function of the angle -19, when the angle c is taken equal to zero.
TABLE I CAPACITY UNsALANcE IN MICROMICROFARADS PER MILE Systems (1, 2)-(3, 4)
Capacity Mutual Capacity Angle 0 unbalance capacity unbalanc e (1) (1,
5, 770 17, 600 ll, 830 4, 900 -14, 060 9, 160 3, 910 -10, 060 6. 750 2, 500 7, 050 -4, 55D 0 0 0 2, 500 +7, 050 +4. 559 --3, 910 +10, 660 +6, 750 -4. 900 +14, 060 +9, 160 5, 770 +17, 600 +11, 830
The unbalance'values for relative angles be= tween the two systems of wires included in the range 180-360 are symmetrical to those shown in the above table, with reference to 0=l80. The corresponding data for inductive reactance is or similar type to that shown above for ca pacity unbalance,
The final valuesindicated in Table I, when plotted,- show a form .very similar to a cosine curve. An analysis carried out by Fouriers method showed that the following series could be used to represent the residual capacity unbalance figures between the systems of wires 9, 2 and it, 4 in function of the angle 0 defining the relative angle between these two systems.
Capacity unbalance of 5, 2-3, 4 equals (l0,695 cos 0+873 cos 30+219 cos The sign is used, since we assumed wire 2 to have a negative instantaneous voltage at the origin of time and this wire has the dominating influence over the system 3, 4 in view of its closer proximity for 0 zero. The analysis showed no sine'components and no terms in even powers of 0 in the cosine series. In terms of the fundapair of wires 3, 4 where wire 3 and wire t repre-- sent the two sides of a metallic circuit with distances di and 112 from wire 2, respectively, is given by thefollowing equation Q 1k M-2l(]og d2 l If-wires 3, t are twisted around each other,
this is equivalent to stating that both wire 3 and wire 4 are rotating around a common center 0:. See Fig. l of the drawings for a visual illus tration. Under these conditions when wires and Q interchange positions, both' the capacity unbalance and the reactance unbalance figures change in sign. It is also'evident that for each position of wires 3 and d, as defined by. the angle a 01' Fig. 1 of the drawings, there will be a dinerent capacity unbalance and a difi'erent reactance unbalance figure. The actual computa-' tions, when carried out both for wires l and 2,
mental component in cos 0, the succeeding higher order odd harmonics are quite small, Expressed in percentage of the maximum unbalance the above series can be putin the form- (3) Capacity unbalance=-ll,830 (.904cos 1 0+.0738\cos 30+.0185 cos 50+ The above series development was worked out on the basis that wires i, 2 were stationary and had no twist, while wires 8, l had a definite twist thereby bringing about a rotation of these wires with reference to wires 8, 2.' Under these cir- 'cumstances, for a variation of in the value oi 0 for'one-half of a cycle, the net capacity unbalance between the two systems of wires would be exactly equal to zero, provided further certain other physical factors could be kept absolutely constant, suchas distance between each of the wires 5, 2 taken separately and each of the wires 3, 4 also taken separately. With only paper insulation loosely wound around each wire as separators, this is not quite so easy to provide. Furthermore it is not feasible to provide complete twists or half-twists between successive splices of cable on all pairs in said cable. Since, in addition to the above points, the cable contains more than twoipairs and a limited number of twists lengths are available in cable manufacturing ma- 5 chinery, all pairs in a cable are provided with a J twist or lay, several such twist lengths being used for any cable. If, therefore, both wires I, Z and 3, 4 are provided with twists, then for any change in angle of wires 3, 4 there will be a corresponding change in the angle o between the line of centers 0102 and the line joining the centers of wires i and 2. For any change in the angle (p, a new table of unbalancevalues will hold similar to Table I and the simple conclusion that for a change of 180 in the value of 0 a perfect balance will be obtained'between the system of wires i, 2 and 3, 4 will hold no longer. This point can be understood more clearly if reference is made to the variation of the distance between any one wire of the first system of wires, such as wire i and any other wire of the second system as a double function of both 0 and p. The complete expression for the distance D between wire I and wire 3 is given by the equation:
where Do=O1O2, r1= distance between the center 01 and the center of wire I which is alsothe distance between the center 0 and the center of wire 3. The distances between the other wires of the system i, 2, 3, d can be obtained by suitable changes in the signs of the coeflicients of cos cos 0, &c. It is important to note that if only cos 0- had appeared in the expression for D then the whole problem would be simplified into that of one system of wires rotating relatively with reference to the second system of wires con sidered as stationary; This is not, however, the
case. Actually, in practice, as already pointed out, considerable difficulty is being experienced with the present day system of cable conductors with diflerent lengths of twist on adjacent pairs, particularly when used for carrier purposes. This question will not be dealt with in any further detail here except to point oiit that the latest quantitative measurements on this subject will be found in the Bell System Technical Journal, Oct.
1939, in a paper bearing the title, Experience in applying carrier telephone systems to toll cables,
by W. B. Bedell, G. BLRansom and W. A. Stevens.
- tions which will be described in detail, there will be no coefiicient'of coupling between two adjacent two-way circuits, assuming each such two-way circuit tov be formed of 4-wires, with polarities disposed alternately positive and negative. The
four (4) wires would be disposed at the four corner of a perfect square. The four constituent wires of the two-way circuit would be insulated separately and then combined into a symmetricall-wire unit or quad arrangement as shown on asraeoe resent, in combination, one side or a two-way circuit, as used for carrier purposes, and wires i and 4- would form the opposite side. Similarly,
wires 5 and I would formone side and wires 8' .and 8 the opposite side in the second unit. It is important to note that as we go around in the first unit starting with wire i and assuming wire I to have a positive potential at a given instant of time, thenwe have the following sequences in signs(+), for wires I, 2, 3 and 6 respectively. correspondingly, we hate a similar set of sequences for wires 8, 8, i and 8 respectively. It is necessary in order to understand the invention, to work out the capacity unbalance between the four wires of one unit i, 2, 3, l, and the four wires of an adjacent unit such as 5, 6, l, 8. Referring to Fig. 'I of the drawings, let 01 represent the center of the first unit and 02 the center of the second unit. Let 01 represent the center of wire i and C5 the center of wire 5. Let (p represent the angle made by 0102 with 0101" and 0 represent the angle made by 0102 with O2C5. Iflnd that the most convenient way of defining the capacity between a fixed location for wire i and any variable location for wire 5 in the second unit, is to express said mutual capacity as a function of the angle 0-01 shown on Fig. 7 of the drawings by the following Fourier series development.
5) C1s=Co+C1 cos (0-01) +03 cos 2(0-00 C3 C08 3(0-01) +C4 00S 4(0-81) equal negative value. This is almost self-evident and can be immediately checked by reference to the relation: g (6) 'n1s==(c1om 2'r.(c1o1 cos (0-01) +r1= where 11 represents the distance C202 0- 1 changes sign, cos 0-01 does not change in si and D15 retains its original numerical value.
Therefore, the capacity C15 which depends di- Fig. 2 of the drawings. The four wires would be maintained at fixed space relations with reference to each other by means of a wrapping of paper or a string spirally appliedor any other suitable insulating separators. Designating the constituent wires ofone such system as I, 2, 8 and 4, and those of an adjacent system as 5, i, 'I and D, then wires l and 3 in the first urilt would repthe center 02 of the second adiacent'umt of 4.
rectly upon the distance D15 also remains the same. If, for purposes of greater generality, the usual sine terms of a complete Fourier series development'had been included in the above expression for C15 then if 0-01 were changed in sign, all such sine terms would also change in sign, while all the cosine .terms would retain their original value intact. Since, as just demonstrated, the numerical value of C15 will not change, under the assumption that 0-01 has changed in sign without changing in numerical value, 'it follows that all sine terms in the Fourier series development have zero coeflicients. It may be recalled here that'the actual computed values of capacity relations between two adjacent cable pairs as depicted on Fig. 1 of' the drawings and" as expressed by "Equation 3 does I The above' not contain any sne terms either. demonstration asto the underlying reasonsfor the absence of the'sine terms in the Fourier series angle 0-01, the angle mad y 2 withithe straight line Joining thecenter C1 of wire I with we (see Fig. '1 or the drawings) The following tables, Table II to Table v, in
elusive, indicate the computed values of the actual capacities between wires 6, 2, 3 and 4 with wires 5 and 7, when wire 5 is at its shortest possible distance from wires i, 2, 3 and 6 respectively, and wire 5 is at its most distant location from the same wires. The situation just outlined holds when in Formula 5, e61=0. The computations referred to have Ibeen carried out for two adjacent i-wire units, built in accordance with Fig. 2 of the drawings, consisting of four 19 E. 31 S. gauge paper insulated'conductors, the distance between adjacent units being equal to .16 inch and the distance between any one conductor and the geometrical center of its corresponding 4-wire combination equal to .0468 inch.
TABLE II CAPACITY IN Mrcaomcnomssns PER MILE or WIRE 1 Wire 7 Difference TABLE III CAPACITY IN MICROMICROFARADS PER MILE or WIRE 3 Angle 1; Wire 5 Wire 7 Difference 4 TABLE IV CAPACITY IN Mxcnomcnorsnens PER MILE or WIRE- 2 Angle 4a Wire 5 Wire 6 Difierence TABLE V CAPACITY IN Mronomcaomaans PER 1mm: or WIRE 4 Wire 5 Wire 6 Difierence To secure a general relationfor capacity unbalance between two adjacent cable unit of the type described in these specifications as depicted on Fig. 2 of the drawings, and as used in the manner first shown on Fig. 4 of the drawings, we make use of the Fourier series'development given by Equation 5 in the manner outlined below, The mutual capacities of wire I an wires 5, '6, I and 8 can be written down as follows, for any position of wire 5 defined by the angle 6,
Fig. 5: I
C15 C1 608 (0-61) C2 COS 2(991)+ C cos 3(0-0 )+C4 (:03 409-01) C4 COS The above set of equations can be simplified into: