History of the Atlantic Cable & Undersea Communications
The illustration shows a 1.7" core diameter Telcothene and air-dielectric coaxial cable manufactured by Submarine Cables for the GPO and laid by Monarch (4) and Alert (3) in November 1947. The 1948 Anglo-Belgian cable was of a similar design.
“Telcothene” was Telcon’s name for their polythene dielectric material, first used by the company for cable insulation in 1938.The cable incorporated an inner conductor 0.45" in diameter, composed of copper tapes on a Telcothene cord, with another Telcothene cord 0.2" in diameter helically applied to the conductor. This cord supported a Telcothene tube of 0.4" wall thickness, thus producing a core of 1.7" diameter.The advantages of this construction lay in its low inductive capacity and attenuation, which enabled much higher carrier frequencies to be employed than with the solid insulated core, and therefore a greater number of speech channels to be carried over the one conductor. The 85 nm cable allowed simultaneous working of 84 circuits.
(Illustration and information from The Telcon Story)
This cable, first used on the 1947 Anglo-Dutch route, was a major technical advance, and a 1948 article in the Telcon house magazine gives more details of its design, construction, and laying.
THE ANGLO-DUTCH TELEPHONE CABLE
No one connected with the submarine cable manufacturing side of our Works will forget the excitement, not to say worries and anxieties which we experienced when this large Telcothene cable (known to us as "The one-point-seven") started to emerge from the covering shop. We remember also the difficulties experienced with jointing and with the application of the outer conductor tapes and armouring. This cable, ultimately laid between Aldeburgh in Suffolk and Domburg in Holland, was a triumph of design. manufacturing skill, and dogged determination to make a good job, by an excellent team at Telcon Works.
That this was recognized by the authorities concerned is clear from the honour conferred upon Jointer Lower, who received the B.E.M. (Civil Division) for his personal contribution, and also on behalf of his colleagues, without whom the job could not have been accomplished. We are all proud of the team spirit which this job engendered. and we hope to each even greater heights of co-operation in the future.
Nearly all readers of this magazine are familiar by now with the appearance of the large coaxial cable, shown “telescoped” in the photograph above, but only a few know the background of design, experimental work, and testing that went into the job under extreme pressure of speed and anxiety. Perhaps therefore a few words describing this work will he of interest, not only as applied to this particular contract, but as an example of the spade work necessary before any such task can be tackled.
In 1945 the Post Office Engineers invited us to consider the possibility of designing and making a cable which would provide not less than 85, and preferably 120 speech channels from England to Holland over a distance of 82 nautical miles. (Note: 120 speech channels, generally known as one super-group, means in common parlance 120 complete circuits. Each person speaking requires one circuit and, in consequence, a conversation requires two. Thus a super-group provides 60 telephone conversations.)
Our 1.7 in. cable is the sixth to he laid between this country and Holland, and some idea of the technical advances made by it can be realized when it is stated that its traffic carrying capacity (or its number of speech channels) totals more than that of all the other live cables put together, and this in spite of its comparatively light weight. This is 19 tons per nautical mile compared with the 23 tons per nautical mile for the No. 2 cable, which permitted only 7 simultaneous conversations.
A brief investigation of factory conditions indicated that we should be limited to a maximum diameter of 1.7 in. for the core. Anything greater than this meant alterations to machines which would have taken too long to accomplish, but with this size limitation the circuit requirements could not be met by the ordinary solid type of Telcothene insulated core. An air space near the centre conductor was known to make the project possible but was not favoured at first, because little was known of the ability of a hollow Telcothene construction to withstand pressure, and the only previous example of a fully plastic air spaced cable (gutta percha), made by us in 1898, had failed.
Theoretical studies and experimental work were promptly initiated to elucidate the following points on air spaced structures:
(I) With 1.7 in. as the maximum diameter of the core what minimum air space was necessary to provide the required speech channels? (Note: Air is the best known dielectric and, therefore, if air replaces Telcothene, the cable is better from a speech carrying standpoint.)
(2) What minimum wall thickness of Telcothene would be necessary to resist the sea pressure at North Sea depths and prevent the collapse of the tube?
(3) At what rate would water penetrate the Telcothene tube? Telcothene is commonly considered to be impervious to water, and for all general purposes it is so. In common with all such materials, however, including gutta percha, Telcothene does allow very small quantities of water to pass through it. The rate of this diffusion is very slow indeed–slower, in fact, than through many other plastics, but the actual diffusion rate was not known when these studies were made.
(4) What changes in normal conductor design could be made to improve the efficiency of the cable without unduly impairing its flexibility, having in mind the need for picking up the cable and re-laying it should it he damaged at any time, or should it be decided to insert submerged repeaters at a later date?
(5) Owing to the increased difficulty of extruding a uniform hollow tube of Telcothene, should we be able to manufacture an air-spaced core sufficiently free from diametrical irregularities to prevent troubles in the sensitive and powerful amplifiers which must operate at the two ends of the cable?
It should be borne in mind that a coaxial cable telephone system must be looked upon as being in every way similar to a multi-wave wireless system, in which each conversation is allocated a wave length, coupled with a tuning system at the far end to separate the circuits. Further, each wave length must be separated from its neighbour sufficiently to prevent interference. The major difference between the two systems is merely that the electrical impulses are confined in the one case to the cable, whereas in the other they travel in free space. It will be obvious that with a system involving 80 to 100 circuits, each on a separate wave length, the cable must operate uniformly over a very wide range of frequencies. In the case under consideration a maximum frequency of 1 megacycle per second was required, which, although low for wireless operation, is very high for cable operation over such a long length. Submerged amplifiers would have been of assistance here, but were not sufficiently developed at this time.
(6) How should the joints in a hollow cable be made to ensure keeping the sea water out and maintaining the insulation, while guaranteeing to provide bulkheads in the tube? Without such bulkheads, damage to the cable (for instance by trawling operations) might allow the sea water to enter the air space and travel along the inner conductor to a great distance.
(7) With these and other points in mind, would the ultimate design provide not only an efficient telephone system for immediate operation, but also, would it enable us to satisfy the British and Dutch Postal Authorities that they could rely on trouble-free operation for many years?
The answers to these questions were very urgently required, for it was essential that the cable should be laid in 1947 to provide the circuits necessary for the expansion of international communications.
These problems were not solved in the order given. If they had been development work would have been much simpler, but the following were the ultimate findings upon which the final design was based:
A 1.7 in. cable having an air space of approximately 20 per cent. would provide the required speech channels. This resulted in the adoption of a Telcothene tube having a wall thickness of 0.4 in., which was found adequate to withstand the sea pressure at the depths involved. (Note: A maximum depth of about 35 fathoms, at which the hydrostatic pressure is about 100 lb. per square inch, was to be encountered). Experiments indicated that such a cable would be safe as regards pressure to a depth of approximately 75 fathoms, but at greater depths than this compression of the air space due to the sea water pressure would cause deterioration of operational efficiency.
The rate of penetration of water through the Telcothene was a very serious problem. First of all it was necessary to devise a means of measuring diffusion because it was so low that normal methods were ineffective. When these measurements were completed the calculations indicated that in fresh water the cable might, under the worst possible conditions of temperature and pressure, commence to fill with water after about three or four years, and in any case the humidity of the air inside the cable would reach high values at the end of only two years. Electrical measurements showed that it was necessary to deposit liquid water before changing the communication characteristics of the cable. That is to say, a merely humid atmosphere in the cable was not harmful, but obviously, as soon as liquid drops of water appeared inside the cable, its electrical characteristics would deteriorate. Much time was expended on considering ways of providing water absorbing substances in or near the air space in order to maintain the low humidity, and it is possible that in the future one or more of these methods may be adopted. In the present case, however, the comparatively shallow depths in which the cable was to be laid, and the fact that sea water is salt and not fresh, provided the necessary safeguards.
It is rather difficult to give a simple explanation of the processes involved, but briefly, fresh water from the sea passes through the Telcothene wall into the air space until the latter is almost saturated with moisture vapour. At this point, the salt in the sea begins to play its part and tends to “suck out” moisture from within the tube. While this “sucking” action is only slight, it is nevertheless sufficient to counteract that last little bit of water ingress which would result in the cable becoming waterlogged. The phenomenon which apparently causes the moisture to come out of the cable into the sea is connected with what is known as osmotic pressure.
It was also ascertained that the tendency of the air space to absorb water increases with the pressure of the water outside, that is to say, with the depth at which the cable is laid. Calculations demonstrated that the cable would tend to become waterlogged only at depths exceeding 140 fathoms. Below that depth the tendency of the water to pass into the Telcothene tube is more powerful than the "sucking" action of the salt water, and eventually the cable would be rendered useless.
The conductor design was thought to be comparatively simple as it was ascertained by calculation and experiment that copper tape applied lengthwise (in rather the same way as a cigarette paper) provided the necessary conductivity. This, however, meant that the conductor was inflexible (like a cigarette) and necessitated ensuring that the core was never bent to a diameter of less than 4 ft. 6 in. which was, however, considered a reasonable minimum for such a heavy type of cable. It is interesting to note here that several obstacles arose while making the trial lengths of this cable, and whilst conductor design was originally thought to be a simple point, it ultimately turned out to be the most difficult to overcome, and was the cause of much delay and experiment during the running of the contract.
Experimental runs indicated that uniformity, whilst not so good as for a solid core, was sufficient to meet the requirements of the amplifiers at the ends of the cable, and rigid specifications to this end were laid down by the Post Office Authorities.
After hundreds of trials and much experimental work, both in the Laboratory and on the part of the jointers in the factory, a satisfactory method of making joints, which complied with the requirements demanded in (6) above, was evolved. The process was extremely lengthy and required carefully controlled conditions. It can best be described by reference to the photograph below, from which it will be seen that the inner conductors were connected together by means of a sort of cage, the Telcothene being injected right from top to bottom of the joint in one operation.
These joints which, as stated, also acted as bulkheads, were found by test to resist hydrostatic pressures of 1,000 lb. per square inch when tested internally from one side of the bulkhead to the other.
With these various answers to hand it was possible to predict with reasonable certainty that the cable should be expected to operate unimpaired for a very long time, but in view of the extremely revolutionary ideas involved, a conservative estimate for the life of the cable was placed at ten years. It is known, however, that should the number of circuits fall off after this period, the cable will still continue to function but with reduced efficiency. Time, of course, will tell, but it is confidently expected that the cable will still be operating efficiently at the end of 30 to 40 years. Long before that time it is proposed that the number of circuits in the cable shall he increased by the use of submerged amplifiers or repeaters, and it is probable that a description of these repeaters, which arc now being developed, will be given in a later issue of this magazine.
In order to complete this picture, a brief word is necessary about laying the cable. Shore end sections were laid out well in advance of the main operation as it was impossible for the large new cable ship “Monarch” (9,000 tons) to lay these in the shallow water involved. The whole length of the main cable, weighing 1,400 tons, was placed in two of the “Monarch’s” tanks and the ship steamed from Greenwich on November 24. She buoyed the whole route on the next day in quite unpleasant weather and a word of praise should be given to the officers and crew on the excellent navigation and technical work performed during the completion of this job. The ship returned to Aldeburgh in a full gale and anchored for the night, which, for some of the landsmen on board was a distinctly uncomfortable one. On the 26th the gale had subsided somewhat and the first splice was made to the English shore end. This took many hours and the ship commenced to pay out cable early on the morning of the 27th. The final splice was begun and completed late on the 28th. During laying tests were made continuously from the Aldeburgh repeater station and the cable’s circuits were used for telephone communication during the operation.
Telcon and Post Office Engineers have once more made history and have been pioneers in the opening up of new fields in submarine telephone communication. Such cables will be of inestimable value in future in maintaining a 24-hour communication service in secrecy and under all conditions of weather. When the development of undersea repeaters has been completed the number of circuits as well as the length of span from land to land can be increased within limits to satisfy the requirements of the communications authorities. It is, therefore, not unreasonable to state that all ideas of the limitations of submarine telephone communication between the British Isles and the rest of the world have been altered. Belgium, Denmark, Holland, Norway and Sweden can now all be linked to these islands by multi-channel telephone cables, and who shall say when the much-talked of cable between Europe and the U.S.A. will be completed?
Note: Anyone desirous of studying this subject in greater technical detail should refer to the articles in Electrical Review, December 19, 1947, and in British Plastics, January, 1948.
Last revised: 8 May, 2016