DESIGN OF SYSTEM — NORTH ATLANTIC LINK 37 



sible without adding a repair repeater. Therefore it was decided to let 

 n = 3. Since the repair repeater is a 2-tube repeater while the regular 

 repeater has 3 vacuum tubes, ?i = 3 corresponds to about 5 repair re- 

 peaters per cable. 



To determine the maximum voltage E„i , it was necessary to consider 

 blocking capacitors and earth potentials. 



Based on laboratorj^ life tests, the blocking capacitor developed has an 

 estimated minimum life of 36 years at 2,000 volts. It is estimated that 

 life varies inversely as about the fourth power of the voltage. The po- 

 tential actually appearing on these capacitors is determined by the dis- 

 tance of the repeater from shore, the power potential applied to the sys- 

 tem, and the magnitude and polarity of any earth potential. 



Earth potential records on several Western Union submarine tele- 

 graph cables were examined. These covered a continuous period from 

 1938 to 1947, including the very severe magnetic storms of April, 1938, 

 and jNIarch, 1940. It was judged rea.sonable to allow a margin of 400 

 volts (200 volts at each shore station) for magnetic storms during the 

 final years of life of the system. 



With an assumed maximum voltage of 2,300 volts on the end repeaters 

 due to cable-current supply eciuipment, and 200 volts per end as allow- 

 ance for the maximum opposing earth potential which the system would 

 be permitted to offset without automatic reduction of the cable current, 

 the voltage across the end repeater in late years of life (i.e., after line 

 current had been increased to offset aging) would normally be 2,300 and 

 would infrequently rise to 2,500. On the rare occasions where earth po- 

 tential would rise above twice 200 volts, the cable current would be some- 

 what reduced and the transmission affected to a reasonably small extent. 

 In the early years of life when the cable-current supply voltage would 

 normally be about 2,000 volts, an opposing earth potential of twice 500 

 volts could be accommodated without affecting cable current; according 

 to the telegraph cable records this would practically never occur. With 

 conditions changing in this way over the years, the life of the blocking 

 capacitors in the end repeaters was calculated to be satisfactory. 



Accordingly E^ was established as 2,300 volts. 



The system length L was estimated as about 1 ,985 nautical miles and 

 the ultimate current I was estimated as 0.250 amperes, with a correspond- 

 ing ultimate Erep of about 62.8 volts. 



Substituting in (3), 



A^ = 55 



N-n = 52 working repeaters 

 S = 37.4 nautical miles. 



