Construction of Lines for Electric Circuits

THE CONSTRUCTION OF LINES FOR ELECTRIC

CIRCUITS.

BY THOMAS D. LOCKWOOD.

A series of articles written at the present time upon electrical line construction which did not refer to aerial cables in something more than a cursory manner, would be obviously incomplete, since such cables now constitute an increasingly important element in systems of electrical communication.

The idea of bunching a number of insulated telegraph wires together, so as to economize space and facilitate construction, and of suspending them in the air, is by no means new. As early as 1795, a chronic inventor of that period, Don Francisco Salva, of Barcelona, read before the Academy of Sciences in that city a memoir describing a system of electric telegraphy, in which he refers to multiple cables as follows : "It appears, however, little short of impossible to erect and maintain so many wires; for even with the loftiest and most inaccessible supports, boys would manage to injure them; but as it is not necessary to keep them very far apart, they can be rolled together in one strong cable, and placed at a great height. In the first trials made with a cable of this kind, I covered each wire with paper, coated with pitch, or some other idioelectric substance, then tying them together, I bound the whole with more paper, which effectually prevented any lateral escape of the electricity."

It does not appear that this plan went into use at that time, or in fact for 65 years thereafter, when aerial cables were introduced by Wheatstone, and erected in London to accommodate the numerous lines working his magneto dial instruments. Wheatstone subsequently patented his arrangement of cables and his method of suspending them. (See Br. Patent, Oct. 10, 1860, No. 2462.)

Until the advent of the telephone, and the almost infinite increase of electrical wires in our cities cousequeut thereon, overhead cables were not much known in this country, and were only used in sporadic cases, the only instances of their employment known to the writer being the kerite cables of the Gold and Stock Telegraph Co., Now York, extending about a third of a mile between the operating room, 195 Broadway, and the inspectors' room, 61 Broadway; and the cables of the Law Telegraph Co., also in New York, in the district immediately surrounding their central station on Fulton street. At present their use is universal. Aerial cables have become indispensable to the telephone business because they bring a good many lines into a small space; they are light in proportion to the useful effect; and, by their use, crosses are in a great measure avoided. When cables of suitable character and construction are employed, and properly suspended, their use is to be commended. Copper has so far been universally used as a material for the conductor, and will probably continue to be so employed. The number of conductors per cable has been greatly varied, as few as five conductors having been used, while as many as a hundred are now frequently bunched in a single cable. Several sizes of conducting wire have also been made use of, Nos. 20, 22, 20, 18 and 16 being the most popular. In telephone work the length of the cables is a serious consideration, the majority of those in use being but a few hundred feet long, and arranged to extend through the densely wired district of the central station only — although occasionally we find them over a mile in length. We know from sad experience that the several lines interfere with one another when they are run in parallelism for any distance, and that the currents traversing one line set up induced currents in adjacent lines; so that words transmitted over one line are heard in receivers connected with others. This often happens when the different lines are not cabled, but the effect is intensified in cables because the different wires are necessarily so near together. To obviate this annoyance special arrangements have been adopted which we shall refer to hereafter.

Insulation also demands our attention. It is of the highest importance that we secure a material which, while forming an effectual bar to any conductive transfer of electricity between the different conductors of the cable, will have a low inductive capacity. Of course each conductor must be insulated throughout its entire length. Cables must be also mechanically protected from the effects of the atmosphere and the elements. Let us now consider the several essential features enumerated, and, inasmuch as it is almost exclusively in telephonic line construction that cables are used, it will be understood that the following ideas refer more particularly to that branch of electrical communication.

Considering first the conductor, and bearing in mind that we have already stated the usual material, and the usual sizes employed, we may reasonably inquire whether any size is better than another, and whether experience has demonstrated any number of conductors in a single cable to be a better working number than any other. In the first place, it does not appear that any special note need be taken as to the size and number of conductors in cables of less than 500 feet or thereabouts in length. The only evils to be looked after in such limited stretches are those of leakage and induction. These we will look into presently. For cables of considerable length — certainly for any which are over 1,000 feet in length — the size of the conductor enters materially into the state of the case. We are fully aware that many aerial cables have been suspended with conductors of Nos. 26 and 22 size B. W. G. The fact need not prevent us from giving it as our decided opinion that such sizes are much too small to do good work. The element of conductivity is much more nearly related to successful telephony than many people suppose, as is proved by the ease with which telephonic conversation was carried on over the heavy copper-plated wire of the Postal Telegraph Co. between New York and Cleveland and Chicago, and more recently on the copper line of the American Bell Telephone Co. between Boston and New York. At one time there was a great tendency among telephonic constructors to build lines of very light steel and iron wires, of all sizes, from No. 14 down to No. 20, the latter size being much used in some of the country exchange districts, and having the comparatively enormous resistance of about 287 ohms per mile. In addition to the extreme fragility of such construction, and its delicate and temporary character, it was found that although the cry was frequently raised "that resistance did not trouble the telephone," resistance did actually become quite a cogent factor as soon as extra territorial lines began to multiply, and subscribers to one exchange begun to converse with subscribers to others. Consequently the use of any iron or steel wire of a gauge smaller than No. 14 has been discouraged, and even that is in our opinion too small. But why are these considerations introduced here? Is not this an article on cables? We shall see the relevancy of the foregoing remarks when we find that the approximate resistance of a No. 26 wire of copper (even when pure, which it never is) is 150 ohms per mile ; that the resistance of No. 24 averages 99 ohms per mile ; that of No. 22 about 64 ohms, and that of No. 20 about 38 ohms, the smallest of these resistances thus being considerably greater than that of No. 12 galvanized iron wire. If now it be wrong to use an iron wire of high mileage resistance, it seems but logical to assume that to use a copper wire of equal resistance per mile is also and equally wrong, and that therefore any of the above sizes are too small to be employed, except as temporary expedients and for very short cables. The experience of the writer leads him to believe that a No. 18 B. W. G. copper wire is the smallest that should be generally used as a cable conductor. Moreover it should be remembered that as the conductivity increases it is much easier to maintain good insulation, and that therefore the insulation resistance of a No. 18 wire (the wire itself averaging about 23 ohms per mile) may fall to a much lower point without reducing the practical efficiency of the line, than could possibly be the case with a smaller conductor. It will no doubt be objected by some that it is injudicious to increase the size of telephone wires, because by so doing a greater surface is exposed to inductive interference ; induction, all things being equal, being proportional to the superficial extent of the conductive service exposed to the disturbing wire or conductor. To this, it may be replied, that the advantage of high conductivity gained by increasing the size of the wire preponderates so greatly over the disadvantage of the slightly increased inductive surface that the latter need scarcely be regarded. Let us suppose a case: We have a copper wire, say No. 31 gauge, the diameter of which is .010 of an inch, or 10 mils, the resistance per mile being about 546 ohms, and its weight per mile 1.597 lbs. The circumference of any circle being to the diameter approximately as 22 is to 7, the circumference of our supposed wire proves to be a little over .0314 of an inch, and it thus has a surface of about 1,989 square inches, or nearly 14 square feet per mile. The conductivity, or conversely the resistance, depends on the sectional area, or what is the equivalent of the area, the weight. Squaring the diameter, we find the sectional area to be approximately 100 mils. Let us now take another wire having a diameter of exactly double that of the first (we find by Table IV, Prescott's Electricity, that No. 25, B. W. G., answers our purpose). Now the resistance of No. 25 copper wire is given as 136.4 ohms per mile, and it has a diameter of 20 mils ; it consequently has a circumference of .0628 of an inch, and a mileage surface of 3,979 square inches, avoiding fractions, or 27.56 square feet. The sectional area we find, however, by the same table to be 400 circular mils, and the weight in pounds per mile is 6.3888. Thus we plainly see that while the surface has only doubled, the sectional area and the conductivity have quadrupled, showing most conclusively that it is advantageous to use reasonably large conductors, and that the gain in conductivity is by no means counterbalanced by an equal loss, arising from retardation due to electrostatic capacity, or by any inductive influence depending upon the extent of the surface of the conductor. Now, as to the number of conducting wires which may be comprised in a single cable. As has been said, 100 wires have frequently been cabled together. This would seem to be too large a number if the best results are desired. In the first place, with a smaller number, there is less temptation to use a wire which is too small; second, the thickness of insulating covering may be greater without unduly increasing the size of the entire cable, which is an advantage, because the thicker the insulation the lower its specific resistance need be, and the less the induction, whether static or dynamic. Moreover, when there is an extremely large number of wires all in one cable, the conditions are very favorable to retardation, because each wire is surrounded by a great many others, all of which are connected to earth at both ends, which tends to give a high electrostatic capacity to each wire. This is of course lessened if we employ a smaller number of wires, as the mass of the outside metal coating of each wire is thus diminished, and the resistance of said mass is increased.

These considerations, then, point to the conclusion that there is another limit (in addition to the size of the cable) to the number of wires which should be included in a single cable of a quarter of a mile or upwards in length. The writer's opinion is that 50 should be a maximum.

(To be continued.)

1. Vid Fahie's History of Electric Telegraphy to the year 1837; p. 104. London and New York, 1884.