DESIGN OF SYSTEM — NORTH ATLANTIC LINK 45 



Signal-to-noise changes from undersea temperature variations were 

 minimized by providing adjustable temperature equalizers at both the 

 transmitting and the receiving terminals, and devising a suitable method 

 of choosing when and how to adjust them. 



Partial compensation for laying effect was carried out at the cable 

 factory by slightly lengthening the individual repeater sections. In the 

 1955 cable, where the factory compensation was based on early data, 

 the increased loss compensated for about two-fifths of the laying effect 

 at the top frequency. In the 1956 cable, which had the benefit of the 

 1955 transatlantic experience, the compensation was increased to nearly 

 twice this amount. Since the loss of the added cable is approximately 

 proportional to the square root of frequency and the laying effect is 

 approximately directl}" proportional to frequencj^ the proportion of 

 laying effect compensated varied with frequency, and a residual remained 

 which had a loss deficiency rising sharply in the upper part of the trans- 

 mitted band. 



The remainder of the laying effect, which was not compensated for in 

 the factory, as well as other initial variations, were largely compensated 

 by measures taken during cable laying at intervals of 150 to 200 miles. 

 The whole length of the cable was divided into eleven "ocean blocks", 

 each either four or five repeater sections long. At each junction between 

 successive ocean blocks, means were provided to compensate approxi- 

 matel}^ for the excess gain or loss which had accumulated up to that 

 point. 



These means were twofold. The first means was adjustment of the 

 length of the repeater section containing the junction. For this purpose, 

 the beginning of each block, except the first, was manufactured with a 

 small excess length of cable. This was to be cut to the desired length as 

 determined by shipboard transmission measurements. 



The second means was a set of undersea equalizers. These equalizers 

 were fixed series networks, encased in housings similar to the repeater 

 housings but shorter. 



Before the nature of the laying effect was known, it had been planned 

 to have equalizers with perhaps several shapes of loss versus frequency; 

 but to combat laying effect, nearly all were finally made with a loss curve 

 sloping sharply upward at the higher frequencies, as shown in Fig. 3.* 

 Because the equalizer components had to be manufactured many months 

 in advance and then sealed into the housings, last minute designs of 

 undersea equalizers were not practical. The proven-integrity principle 

 prevented use of adjustable units. Because the equalizers were series- 



* This characteristic was based on a statistical analj-sis of the data on some 

 similar cable laid for the Air Force project. 



