314 THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1957 



With incorporation of these changes, it was known that the factory 

 lengths of cable between repeaters were adequate to keep misalignment 

 within an ocean block within reasonable limits. The system could then 

 be equalized between ocean blocks so that the signal level at the first 

 repeater of a new block would be approximately correct, and the total 

 system noise thus would fall within limits. 



This equalization was accomplished in two ways. Excess cable of the 

 order of | to 3 miles in length was provided at the top end* of each ocean 

 block. Based on measurements, this could be cut longer or shorter than 

 the nominal spacing of repeaters, so that the repeater gains and cable 

 losses would be matched at some frequency in the band. Residual devia- 

 tions in other parts of the band could then be mopped up if necessary 

 by inserting a simple equalizer, housed in a container similar to those 

 used for the repeaters. Ten such equalization points were provided in 

 each cable. 



In practice, sending levels were adjusted at the cable station to give 

 test tones at the grids of the output tubes of the repeaters which if the 

 system equalization were perfect, would be flat across the frequenc}^ 

 band, and at the proper level. These tones were measured on shipboard 

 at the end of the ocean block being paid out. The results were plotted 

 against mileage, with one sheet for each frequency being measured. 

 Because of the "laying effect" and of temperature and pressure changes 

 on the cable as it progressed to the bottom, these plots displayed a slope. 



The value of loss (or gain) to be ascertained for each frequency was 

 that which would exist when the entire ocean block was on the bottom. 

 To obtain this, it was necessary to extrapolate the cur^'es to the mileage 

 point representing the end of the block in question. The extrapola- 

 tion was required to avoid stopping the ship at the end of the block, and 

 so had to anticipate the time needed for turning over and cutting the 

 cable end at the proper point, and making one or two splices (depending 

 on whether or not an equalizer was inserted at the point in question). 



Having read the extrapolated values from the curves, these were com- 

 pared with objectives for that block junction, and the deviations plot- 

 ted. Transparent overlays, showing the net effect of each of several types 

 of equalizer combined with varying amounts of cable around the nom- 

 inal spacing, greatly facilitated the final decision as to cutting point and 

 equalizer choice. 



This implementation of the system undersea equalization represented 

 a very large part of the effort required of the testing crews during laying. 



Additional data gathered for fault location, aging studies and other 



* First end out of the cable tank. 



