362 BELL SYSTEM TECHNICAL JOURNAL 



cal considerations indicate that if the cable should have been spliced on a 

 random basis, without regard to capacitance unbalance in individual 

 lengths, the r.m.s. average residual unbalance per loading section would 

 have been about four times as great as that actually obtained. 



On the whole, crosstalk results, including the effects of the loading coil 

 unbalances, were considered fairly satisfactory for an initial pioneering effort, 

 but not good enough as a standard of excellence to work to in subsequent 

 projects which were to be of an entirely different order of magnitude and 

 importance in the scheme of nation-wide telephony. 



Since the cost of the capacitance unbalance test splicing was small rela- 

 tive to the total cost, the general technique used on the Boston-Neponset 

 cable was subsequently standardized for general use in installing quadded 

 cable. Eventually, substantial reductions in the amount of the test splicing 

 resulted from improvements in cable manufacture. 



Loading Details: Except for its phantom working feature, the side circuit 

 loading for the Boston-Neponset cable was similar to the old standard 

 medium- weight loading, originally developed for non-quadded cable. 175 

 mh coils were installed at intervals averaging about 8520 feet. A total of 72 

 side circuits were loaded, using two cases, each containing 36 coils at each 

 load point. Eighteen phantom circuits were loaded with 106 mh coils at 

 an average spacing of about 8450 feet. The differences in average spacing 

 resulted from manhole space limitations which led to the phantom coils 

 being installed systematically at manholes next in line to the associated 

 side circuit loading points, at distances ranging from 214 to 490 feet. This 

 layout would have simplified the removal of the phantom loading, if the 

 phantom-to-side crosstalk should have been large enough to make phantom 

 working impracticable. The phantom loading just described conformed to 

 the established cut-off frequency standard for cable loading (approx. 2300 

 cycles) and gave an attenuation loss reduction of the same order as that 

 provided by the side circuit loading. The absolute attenuation in the loaded 

 phantom was approximately 20% below that in the associated side circuits. 

 The nominal impedances of the cable phantom and side circuit loading were 

 about 800 and 1300 ohms, respectively. 



The new types of loading coils used on the Boston-Neponset cable were 

 generally similar in basic design features to the side circuit coils and phan- 

 tom coils used in the open-wide trial installation of phantom loading, but 

 were much smaller in dimensions. Mainly because of size differences, the 

 unbalances in the cable coils were much smaller than those in the open-wire 

 coils. 



Exploitation of the New Developments 



The pioneering developments just described were so basically important 

 in extending the limits of long distance telephony, and reducing the cost of 



