294 BELL SYSTEM TECHNICAL JOURNAL 



effects involved are cumulative with distance and the accuracy involved 

 in the determination of the various characteristics is therefore set by the 

 maximum distance over which the system is designed to operate. For a 

 distance of 4000 miles the total loss at the top frequency of 60 kc. will be 

 approximately 16,000 db, the attenuation difference between the top fre- 

 quency of 60 kc. and the bottom frequency of 12 kc. nearly 6000 db if the 

 cable is at about the average temperature, 55°F. The range of variation 

 in loss with temperature, assuming aerial cable over the whole distance, 

 will be about ±8 per cent of the total at 60 kc. It is desired to correct 

 these frequency differences in loss and variations with temperature so 

 accurately that individual channels will be constant to within ±2 db. 



Prior to the beginning of experimentation with cable carrier systems 

 limited use had been made, in connection with carrier systems operated over 

 open-wire lines at frequencies up to 30 kc, of conductors in relatively short 

 entrance and intermediate cables. The available data, however, were 

 quite inadequate for the cable carrier problem. Accordingly, an extensive 

 series of tests was undertaken. Reels of standard toll cable were placed 

 in temperature controlled rooms where the extreme temperature variations 

 of the mid-west could be substantially duplicated (the actual laboratory 

 temperatures ranged from just below 0° F to 120° F) and measurements 

 were made to determine the changes in the parameters of the cable ac- 

 companying these wide temperature variations at frequencies from 1 kc. 

 to 100 kc. and higher in some cases. Certain of the tests even studied the 

 effect of varying the humidity content of the cables. Further measure- 

 ments were then made on suitable lengths of pairs in actual commercial 

 cables in which carrier systems were to be installed. These results cor- 

 roborated and extended the data from the laboratory measurements; the 

 subsequent operation of equalizers, regulators, etc., based upon these data, 

 showed no essential discrepancies. 



The present paper, after referring to the types of toll cables employed for 

 the new carrier systems, outlines the methods employed in determining their 

 characteristics both in the laboratory and in the field, summarizes these 

 characteristics for t^'pical 19-gauge cable at frequencies up to 100 kc. and 

 finally extends them to frequencies as high as 700 kc. for 16 and 19-gauge 

 cables. 



Types of Carrier Toll Cables 



The type K carrier system has been designed so that it may be applied 

 to existing cables, thus in many cases avoiding the installation of expensive 

 new cables. Most of the standard toll cable in the Bell System contains 

 chiefly 19-gauge paper insulated conductors in "multiple twin quads," i.e., 

 two conductors are twisted together to form a pair and two pairs twisted 



