360 BELL SYSTEM TECHNICAL JOURNAL 



than if non-quadded pairs and non-phantom type loading coils should have 

 been used. 



Cable Details: The multiple-twin quad was chosen in preference to the 

 spiral-four quad on the basis of practical features involved in manufacture 

 and installation and the resulting arrangement, in case phantom working 

 should not prove successful. That is to say the twisted-pair side circuits 

 would better serve the customer's plant flexibility needs than would the side 

 circuits of spiral-four quads. Also the early experiments had indicated a 

 high probability of superior balance among the very important within- 

 quad coupUngs. 



Additional factors that resulted in the subsequent standardization of 

 multiple- twin quads were: 



1. Their phantom capacitance is about 60 per cent higher than that of 

 the side circuits, whereas in spiral-four quads the phantom capacitance 

 is upward of three times as high as that of the side circuits. Conse- 

 quently, when the phantom and sides are loaded for equal cut-off 

 frequency and at the same spacing, the resulting impedances are such 

 that the attenuation of the multiple-twin phantoms is considerably 

 lower than that of their associated side circuits, whereas the attenua- 

 tion of the spiral-four phantoms is inherently much higher than that 

 of their side circuits. Since the phantom loading for the open-wire 

 lines had provided phantom circuits which were preferable to their 

 own (loaded) side circuits, due to their lower attenuation, it also 

 seemed desirable that the loaded cable phantom should be better 

 than the loaded side circuits. 



2. The ratio of phantom circuit to side circuit capacitance in multiple- 

 twin quads is close to that in the open-wire lines. In consequence, 

 the loaded entrance cable impedances are better related to the open- 

 wire impedances than would be practicable with simple loading on 

 spiral-four quads. These comparisons again assume the phantom and 

 side circuits to be loaded at the same spacing, with coil inductances 

 giving equal cut-off frequencies. 



In the design of the multiple-twin quad cable, the "staggered- twist" 

 principle which had been found necessary from the crosstalk standpoint for 

 use in loaded non-quadded pair cables was applied to the individual cir- 

 cuits of a quad. The lengths of twist were different for the two side cir- 

 cuits and a still different length of twist was used in the phantom. Adja- 

 cent quads had different combinations of pair and quad twists, and adjacent 

 layers were stranded in opposite directions. 



Notwithstanding these basic design precautions a great many manu- 

 facturing difficulties were encountered in preventing the phantom-to-side 

 capacitance unbalances from being very much too large. In the first 



