60 



THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1957 



CD 2 

 Q 



- I 



Z 



< 



O 



5-^ 



_l 



LU 



cr-2 



20 



40 



60 80 100 120 140 



LINE FREQUENCY IN KILOCYCLES PER SECOND 



160 



Fig. 12 — Gain versus line frequency — Xo. 1 cable. 



Fig. 13 gives the corresponding group-to-group frequency response of the 

 deep-sea hnk. These frequency characteristics will vary a little from time 

 to time, depending partly on the change since the last adjustment of the 

 temperature equalizer. 



Note that while the slope in gross cable loss between 20 and 164 kc 

 is about 2,100 db, the net loss of the equalized high frequency line over 

 this band varies only about 1 db. 



Net Loss versus Time 



The net loss of the undersea sj^stem (cable plus repeaters) has de- 

 creased slowly since the system was laid. The decrease is approximately 

 cable shape, proportional to the square root of frequency. In a year it 

 has amounted to about 5 db at the top frequency. In the early months 

 the change was almost directly proportional to time, but later the rate 

 slowed, as shown in Fig. 14. 



The changes shown in the figure are due partly to change in cable 

 temperature and partly to slow aging of the cable. Detailed studies of 

 repeater crystal frequenc,y changes have resulted in only an approximate 

 separation into temperature effects and aging. Howe^'er, it seems reason- 

 able to assume that at the end of a one-year cycle there would be little 



CD 

 Q 



< 



> 



5 



a. 



60 80 100 120 140 



LINE FREQUENCY IN KILOCYCLES PER SECOND 



160 



Fig. 13 — Group to group frequencj' response — Xo. 1 cable. 



