552 



BELL SYSTEM TECHNICAL JOURNAL 



the other half-section so that the unbalances in the two sections tended 

 to compensate. Table 1 shows for a typical type K repeater section 



TABLE 1 



Measurements of Mutual Inductance Unbalance (G) and Capacitance 



Unbalance (C) Before and After Poling on Quads in the 



Philadelphia-Philadelphia KN Section of the 



New York-Philadelphia E Cable 



the unbalance measurements before poling and the final results after 

 the poling splice was made. These measurements were made at voice 

 frequencies, since as discussed in a previous paper ^ satisfactory re- 

 sults were obtained at these frequencies. It will be noted that poling 

 reduced markedly the unbalances in the quads in the section shown. 

 This is particularly true of the mutual inductance unbalances, indi- 

 cated in the table by G, the reduction of which is important in reducing 

 crosstalk at carrier frequencies. 



After carrying out this and the balancing operations described later 

 for the reduction of crosstalk, it was still necessary to take other steps 

 to reduce the within-quad crosstalk. The recurrence of within-quad 

 coupling between carrier systems which may be assigned to two of the 

 pairs in a 20-pair carrier complement, has been reduced by means of 

 a splicing plan so worked out, that two given carrier systems will 

 operate on the same quad as infrequently as possible. This was ac- 

 complished by splitting the quads at the ends of each carrier repeater 

 section on a planned basis. The plan is shown in Table 2, Nineteen 

 types of splices are shown. This table is used as a guide in performing 

 the splice between the balancing bays and the input sealed test ter- 

 minal. For example, in performing splice type 8, sealed test terminal 

 jacks K-1 are connected to the pair designated 8 at the balancing bay 

 cable terminal, jacks K-2 to the pair designated 16, jacks K-3 to 



