US1720616A

US1720616A - Balancing telephone cables

Info

Publication number: US1720616A
Application number: US303273A
Authority: US
Inventors: Werren Herbert Trowbridge
Original assignee: Individual
Current assignee: Individual
Priority date: 1927-06-16
Filing date: 1928-08-31
Publication date: 1929-07-09
Anticipated expiration: 1946-07-09
Also published as: GB302373A

Description

July 9, 1929. H. T. WERREN BALANCING TELEPHONE CABLES Filed Aug. 31, 1928 i i i I i Patented July 9, 1929.

UNITED STATES PATENT OFFICE.

BALANCING TELEPHONE CABLES.

Application filed August 31, 1928, Serial No. 303,273, and in-Great Britain June 16, 1927.

This invention relates to 'balancin telephone cables the object being to produce a telephone line with balanced sections, more particularly balanced loading sections, of cable in an improved manner.

In order to balance telephone cables it has been proposed to divide a length of cable into two parts of approximately equal length, to make a crossing of the wires of each pair, and

to re-join them pair to pair respectively; in this way the unbalance of each pair in one of the said parts is substantially compensated by the inverse unbalance in the corresponding pair in the other of the said parts. Furthermore, it has been proposed to reduce the disturbances between groups of adjacent conductors by first determining the diflerences in capacities of the conductors of one group in relation to those of another group and by them adjusting the capacities of the groups, for example by making suitable crossings and joining the conductors of one group to those of another group, so that the capacities tend to balance. But in the known methods of halancing telephone cables, the balancing of a given loading section consisting of a number of jointed cable lengths of a multiple twin cable is accomplished at joints between adj acent cable lengths, by crossing the conductors of quads selectedsolely on account of the magnitude of their capacity unbalances.

In the manufacture of lengths of cable each com arising a plurality of quads, the different qua s are provided withmeans by which they may be identified; for example, difl'erent quads may be differently coloured, or may carry numerals or other symbols at their ends by which they can be identified. It is obviously convenient, if,when successive cable lengths are joined together, a 'redvquad (for example) 40 in one length is connected to a red quad in the neighbouring length, a green quad to a green quad, and so on. Similarly, where the quads bear numerals such as I, II, and so on, it is convenient for quad I in the first length to be connected to quad I in the second length, for quad II in the first length to be connected to quad II in the second length, and so on. Where the quads are connected in this way, so that to each quad are joined quads of the same 5 colour, or quads bearing the same symbols, the colour scheme, or the identity of the quads, is said to be maintained througln out the cable. Where, however, the crossings of the conductorsof quads are selected solely 5 on account of the magnitudes of the capacity unbalances, thecolour scheme or the identity of the quads is not maintained throughout the cable.

Furthermore, the method of cable balancing referred to above has additional disadvantages. For example, the 2-pair quads of a cable in which four-wire working is desired must be divided into at least three groups for balancing, this reduces the number 65 of quads in each balancing group and also makes it necessary to, design in advance the complete transmission system comprisin the four-wire and two-Wire circuits require In the methods used heretofore, in which 7 the cross-talk between quads is reduced by mixing its ultimate value is always indeterminate, and depends upon the number of quads in the balancing group. In the case of large cables containing large balancing 7 groups those values may be within permissible limits, but in the case of small cables or cables containing small balancing groups there are not suiiicient possibilities of eliminating the unbalanccs and therefore the crosstalk may exceed permissible limits. Moreover a cable comprising only a small number of quads cannot be used for four-wire working. Another disadvantage of this known system of balancing is that as the resulting run of any individual quad throughout the finished cable section is not predetermined, difiiculties ma arise in subsequent repair work. The object of this invention is to overcome these disadvantages, by so arranging a cable section that quads in adjacent cable lengths are joined together in a predetermined manner (for example, to maintain the identity of the quads or to maintain the colour scheme throughout the section) quads from different cable-lengths which are ultimately to be joined together to give a predetermined run will, for brevity, be referred to then made from these unbalances of the crossings between the conductors of each 2-pair quad at each of the successive joints, similar quads being joined together throughout the section, which are required to produce the minimum residual unbalance, all crossings being made within quads, whereupon the cable lengths are jointed with the appropriate crossings to form a section which is substantially a balanced cable section. The individual quads are joined in a predetermined or der; for example, the quads are joined to maintain their colour identity throughout the cable section. v 7 As another example, the predetermined order might be such that, considering any pair of quads, in two consecutive cable lengths which are joined together, the quads adjacent to one of the said pair of quads in the first of the said consecutive cable lengths, are connected to the quads adjacent to the other of the said pair of quads in the second of the said consecutive cable lengths. In cables in which the quads are connected in this manner, the same relative spacing of the quads may be said to be preserved throughout the cable section, since,-considring any selected pair of circuits in the cable, the mutual distances between the conductors formingthe said pair of circuits will be the same in every cable length in the section. Quads spaced in this way will also hereinafter be referred to as being uniformly spaced throughout the cable. In the form of cable known hitherto, however, a pair of circuits, which were neighhours in one cable length, might well be widely spaced in a subsequent cable length, so that the mutual distances between the conductors forming the pair would vary greatly.

That is to say any individual quad in a cable length is connected with the similar quad in the most appropriate cable length (not necessarily adjacent) by a suitable arrangement of the crossings of the wires at the joints between the intermediate cable lengths. The method is preferably applied without the use of auxiliary balancing condensers, but if desired these may be used in conjunction with the method described.

In practice the method of balancing according to the invention may be carried into effect in the following manner. The capacity unbalances of the 2-pair quads both within the quads and between adjacent quads are determined in each of the lengths of cable of which for example there may be eight which are ultimately to constitute the cable section. These values for any given quad (for example quad N o. 1), are then scheduled in the order of the magnitude for the most important circuit with the corresponding numbers of the cable lengths, eight in all. For example it may happen that the capacity unbalances, for example, between a phantom and a side circuit, in decreasing order of magnitude show that the lengths of the quad to be connected to obtain a progressive reduction of the capacity unbalance along the lengths should be con nected in the

order

1, 5, 7, 4, 2 and so on instead of 1, 2, 3, 4, 5 and so on. But in practice the cable lengths will be laid in the

order

1, 2, 3, 4 and so on, so that the joints of any quad must actually bemade between

lengths

1 and 2, 2 and 3, and so on. In order to accomplish this and eliminate the capacity unbalances it is therefore necessary to find the crossings required. Accordingly from the schedule (for example for quad No. 1) of capacity-imbalances in numerical order of magnitude is compiled a second schedule with the numbers of the lengths in irregular sequence and the conventional references A B C D identifying the individual wires of a quad are inserted for each item and are arranged to show the crossings required between the quad lengths numbered and arranged in irregular sequence together with the calculated residual unbalances as successive lengths are joined together.

When this schedule is thus completed it is then possible to determine by inspection the actual crossings required in practice between the conductors of this quad at each of the joints between consecutive lengths, that is to say in regular sequence) to elimmate the capacity unbalances wit-bin thisquad.

So far only the capacity unbalances within a quad have been considered. These capacity unbalances in the case of a 2-pair quad arranged with two two-wire side circuits and a phantom circuit, as is known, usually comprise five quantities conventionally designated in the art as follows :Side 1 to side 2, phantom to side 1, phantom to side 2, side 1 to earth, and side 2 to earth. But there are also capacity unbalances between laterally adjacent quads which for accurate balancing of a cable section must be considered. For example the capacity unbalances between two such adjacent 2-pair quads (quad No. 1 and quad No. 2) as is known, usually comprise nine quantities conventionally designated in the art as follows Phantom 1 to phantom 2, phantom 1 to pair 3, phantom 1 to pair 4, phantom 2 to pair 1, phantom 2 to pair 2, pair ,1 to pair 3, pair 1 to pair 4, pair 2 to pair 3, pair 2 to pair 4. Now in accordance with the invention after the values of the capacity unbalances within quads, for example to take two quads (quad No. 1 and quad No. 2) have been set out in the form of schedules, and the 4 required crossings for one quad (quad No. l) have been determined, a second ,schedule is made of the capacity unbalances between the two quads (quad No. 1 and quad No. 2) and of the capacity unbalances within the second quad. From this schedule a final determination of the crossings at the successive joints in the second quad (quad No. 2) is made to compensate for both the capacity unbalances within that quad and also the capacity unbalances between the two adjacent quads.

WVith quads jointed in this manner each section of the cable will be substantially a balanced section. The crossings for quad No. 3 are determined in a similar manner taking into consideration the capacity unbalances between quad No. 2 and quad No. 3, and the capacity unbalances Within quad N o. 3. The remaining quads are dealt with similarly.

A method according to the invention will now be described with reference to numerical examples and the accompanying diagram- 'matic drawing in which Fig. 1 is intended to illustrate a loading section of a cable made up of eight standardlengths of cableyonly two 2-pair quads being shown. connections between the first five lengths Fig. 1 shows the the usual way a diagram of the network of effective capacities by which cross-talk is produced.

The lengths of cable are indicated by

numbers

1, 2, 3, to 8. The capacity unbalances within quad N o. 1, within quad N o. 2, and between the quads are measured for each of the lengths numbered 1 to 8, and are then set out in schedules. The withiirquad unbalances for quad No. 1 are then arranged in decreasing order of magnitude (irrespective of sign) for the phantom to side circuits. It may however, be convenient to depart slightly from this arrangement if any of the unbalances are abnormally large, as described below. The within-quad unbalances for quad No. 2 are arranged in a similar manner. The between-quad unbalances for the two quads are left in the numerical order of the lengths.

These three schedules are shown below SCHEDULE A.

Quad N0. 1 within-quad unbalances-1nmF.

Length No. sl-si PH-S1 PH-s, s,-E Sg-E SCHEDULE B.

Quad N0. 2 within-quad 1mbaZancesmmF.

Length N0. SrS PH-Si PH-Sz Sr]: Sz-E SCHEDULE C.

Length PH| PHr PHi- PH:- PHz- Pn- Pn- Prz- Pri- NO. P112 Pr; Pl; Pn Plz Pra Pr-i PM PM It will be seen that the within-quad unbalanees shown in Schedules A and B have been arranged principally in the order of magnitude of the unbalances of the phantomside circuits, but that this arrangement has been slightly departed from, referred to above, in the case of Schedule B. In the case of Schedule A the magnitudes of the unbalances run as follows: 58, 25, 23, 22, 22, (all in- PIT-S 17 (PI-LS 8, 6, (both in PH-S,). In the case of Schedule B, the order is preserved for lengths Nos. 25: thus- 55, 45, 45, (all in PH-S 41- (in PH-S 35, (in PH-S but length No 5 is followed by length No 7, despite the fact that the larger phantom-side unbalance in length N0. 6 (namely, 30 nnnF.) is greater than the larger phantom-side unbalance in length No. 7 (namely. 27 nnnF.), because the difference between the two is small and the remaining imbalances in length No. 7 (namely, 17, 22, 22, 23, mmF.) are considerably greater than the remaining unbalances in length No. 6 (namely, 12, 12, 1 0, 3 mmF.) The remaining length of Schedule B will be seen to follow in the order prescribed.

Now examining first the schedule for quad No. 1 and determining the required crossings, which are indicated by rearranging the order of the conventional references A B C D, we arrive at a selection of connection for quad No. 1, as shown in the followingschedule,

which also shows'the reduced magnitude of the residual unbalances SCHEDULE D.

Quad N0. 1.

Length No Cross s-s PEI-S1 PH-s, s-m e-E,

1 ABOD -5 -58 +28 +44 +31 4 ABGD +20 +25 -5 -44 -20 +21 -33 +23 +2 2 ABOD -1s +23 +3 11 -11 +3 -10 +26 -11 -3 6 DCAB +9 +8 -22 +22 -0 +12 -2 +4 9 7 DCBA- -20 6 -22 -27 +33 --s 8 -1s -22 +24 5 BADC +30 +10 +11 -21 -12 +22 +2 1 -43 +12 8 CDAB -13 +5 +8 +8 +9 +7 +7 -58 3 BACD -7 -e -2 +45 -12 Residual withinuad unbalanees. q +2 +1 +5 +8 The capacity unbalances within quad No. 2 and between quads Nos. 1 and 2 are now arranged in the order shown in Schedule B, and the appropriate crossings determined for quad N o. 2 by inspection and calculation as well known in the art so as to reduce at the same time both the within-quad unbalances for quad N o. 2, and the unbalanees between the two quads. In this way we obtain Schedule E as follows the crossings for quad No. 1 being obtained from Schedule D SCHEDULE E 70 the connections for the two quads which have been considered.

The lengths of cable may now be jointed, the necessary crossing between the conductors Within quad unbalances Between quad imbalances quad No. 2 quad 1quad 2 Length Quad Quad 1 2 PH PH PH PH PH P P P P a 1- 1" 1- r 2' 1'1 1 1'1- 1:- S S PH S1 PH S S s E2 Pllz Pr; P14 Pm PI: P1; P1; P1; PH

2 ABCD AB CD +2 -15 +55 +60 +8 -18 +12 +4 -10 8 +17 +20 -3 4 A B C D 0 DB A +21 +45 -40 +35 -20 +5 -8 -29 +2 -5 -48 -20 .+25

+23 +30 +15 +95 -12 -13 +42 -4 -39 6 +12 -23 0 +22 3 B A C D B A C D +18 +18 +45 -51 +72 +23 -4 +5 -21 8 0 +32 -49 +41 +48 +60 +44 +60 +10 +7 8 -34 -27 +4 -23 +32 -27 1 ABCD DCBA -18 +8 -41 -49 -54 -6 +10 +45 +10 -15 -29 -15 +36 +12 +23 +56 +19 -5 +6 +4 +17 +37 -24 -42 -25 -38 +68 -15 5 B ADC DC AB -36 -35 +12 +20 +98 3 +20 +7 +24 -13 8 +5 -49 +33 13 +21 +31 +15 +104 +1 +37 +44 0 -55 -33 -33 +19 +18 7 DC BA DCBA +17 -27 -22 +28 -22 +28 +36 +5 8 +14 +1 -15 +15 +13 +4 6 +0 +38 +82 +29 +73 +49 -8 -41 -32 -48 +34 +31 0 DC AB DC AB +12 -12 -30 3 -10 -2 -38 -2 -21 +17 -20 +30 0 -30 t +10 -18 -21 +35 +72 +27 +35 +47 -29 -24 -52 -18 +34 +1 8 CDAB DCAB -12 +2 +25 -30 +33 +6 -15 +9 +8 -2 +31 +2 +6 +5 I +4 -16 1 +4 +5 {+105 +33 +20 +50 21 -26 -21 -16 I +6 It will be understood that the same method 80 i is a plicable to the remaining 2-pair uads in t e cable, the capacity unbalances within each quad and between adjacent uads being measured and the results schedu ed and rearranged to give the crossing required.

Where auxiliary balancing condensers are employed, the values of these are determined in the ordinary way after the jointing of the cable lengths has been performed. Consider, for example, quad No. 2. The efiective capacities between the several wires A, B, C, D and ground, and the efiective mutual capacities between the airs of wires are represented by the con ensers a, b, c, d and w, m, 2, as shown in Figure 2. The unbalances referred to, in the schedules given above, under the headings S -S PH-S ,PHS S,E and S,E are given by the following expressions, respectively, in terms of these efi'ective capacities:

S E 0 d The residual values of these unbalances are, from Schedule E,

SE ==c- (13 105. From the last two equations it appears that if a condenser 12 of capacity 5 mmF. be connected between wire B and ground, as shown in Figure 1, and if a condenser d of capacity 105 mmF. be connected between wire D and ground, the unbalances S,E, and S E will be reduced to zero. Similarly, if a condenser 10 of capacity 6 mmF. is connected between wire A and wire C, and a condenser 2 of capacity 10 mmF. is connected between wire A and wire D, it will be seen that the unbalances given by the remainder of the expressions above will be reduced to zero. The procedure outlined does not differ in principle from that commonly used in balancing by means of condensers, and it is thought that the forgoing example will make the matter clear to those skilled in this art.

Moreover, it will be understood that the method accordin to the invention may be applied by first balancing parts of a cable .section, for example by groups of 4 lengths,

connecting the lengths in each group with appropriate crossings and then balancing the groups and connecting them with appropriate crossings to obtain a substantially balanced sect in whilst maintaining the identity of the quads throughout the cable section. The capacity unbalances for the groups may I be determined if desired in the field.

One advantage of the method according to the invention is that it enables a telephone system to be designed for long distance workmg where four-wire repeater workin is essential, with a relatively small num r of circuits, as quads suitably placed apart throughout the length of the line can be readily chosen. Another a'dvanta e is that cross-talk between quads can be re uced considerably below the values which are normally obtained by mixin the quads as in the prlor art on account of t e fact that similar. uads are adjacent through the whole cable, t us enabling the cross-talk between these quads to be reduced by measurement of the unbalance and suitable crossings of the wires or pairs.

It is to be understood that the method according to the invention may be applied in a similar manner to a non-phantomed cable. The method according to the invention can be applied in conjunction with the present known methods of balancing, for example in a cable having a centre, first and second layers, the centre and first layers may be balanced by mixing and the circuits used for two wire working, whilst the second layer may be balanced in accordance with the invention and the circuits used forfour wire working.

I claim:

l. A method of balancing a section of an electric signalling cable comprising a plurality of 2-pair quads divided into lengths, according to which the capacity unbalance within and between quads are first measured in each of the lengths of cable which are ultimately joined to constitute a cable section, a determination is then made from these unbalances of the crossings in or between pairs of conductors at each successive joint, conductors being joined together throughout the section which are required to produce the minimum residual unbalances, whereupon the cable lengths are jointed with the appropriate crossings to form a section which is substantially a balanced cable section.

2. A method of balancing a section of an electric signalling cable comprising a plurality of 2-pair quads divided into lengths, according to which the capacity unbalances both within the quads and between adjacent quads are first measured for each of the lengths of cable which are ultimately joined to constitute a cable section, and a determination is then made from these measurements of the crossings between the conductors of each 2-pair uad at each successive joint, similar quads eing jointed together throughout the section, which are re uired to produce the minimum residual un alances, where- 'upon the cable lengths are jointed with the appropriate crossings to form a section which is substantially a balanced cable section.

3. A method of balancing a section of an electric signalling cable, comprising a plurality of 2-pair quads divided into lengths, accordin to which the capacity unbalances both wit in the quads and between adjacent quads are first measured for each of the lengths of cable which are ultimately joined to constitute a cable section, the appropriate crossings for individually balancing the quads and the appropriate crossings between the said groups of conductors are determined from said measured unbalances and the'joints made, the same relative spacing of the quads being maintained throughout the whole cable section.

4. A method of balancing a sect-ion of an electrical signalling cable, comprising a pluralit of 2-pair quads divided into lengths which are afterward 'oined, according to which the capacity un alances both within the quads and between adjacent quads are first measured for each of the lengths of cable which are ultimately joined to constitute a cable section and, as determined from said measured unbalances, a selection of crossings is made at some of the joints for the unbalances between quads, and a selection of crossings is made at the remaining joints for the unbalances within quads substantially as described, the crossings being made within quads.

5. A method of balancing a section of an electric signalling cable comprising a plurality of 2-pair quads divided into lengths which are afterward 'oined, according to which the capacity un alances both within the quads and between adjacent quads are first measured for each of the lengths of cable which are ultimately joined to constitute a cable section and, as determined from 'said measured unbalances a selection of crossings is made at some of the joints for the unbalances within quads, and a selection of crossin s is made at the remainin joints for the unbalances both within and between quads, substantially as described, the crossings being made within quads.

6. A method of balancing'a section of an electric signalling cable according to claim 2, wherein the capacity unbalances between the circuits are further reduced by the inser tion of auxiliary balancing condensers.

HERBERT TROWBRIDGE WERREN.