THE TORONTO-BARRIE TOLL CABLE 



599 



mile. In Fig. 6 is shown the mean of the attenuations of three pairs 

 plotted against frequency. On one pair the measurements were made 

 at frequencies up to 6500 cycles, from which cut-off was determined to 

 take place at 4000 cycles. Assuming a 60-mile circuit, the frequency 

 at which the attenuation is about 10 db greater than 1000 cycles, is 

 about 3500 cycles. 



Before the return loss tests were made, impedance-frequency mea- 

 surements were taken on two of the balancing networks for each of the 

 three adjustments provided, and on a representative number of cable 

 pairs, to determine the optimum network adjustment. The resistance 

 component of the impedance for one of the networks and one cable 



500 600 600 1000 1500 2000 



FREQUENCY IN CYCLES PER SECOND 



4000 5000 



Fig. 6 — Attenuation-frequency characteristic; mean of measurements on three pairs. 



pair is shown in Fig. 7 (the two networks were found to be identical). 

 From these tests the optimum network adjustment was determined to 

 be that corresponding to a cable capacitance of 0.088 mfd per mile, 

 which adjustment was then used for the return loss measurements. 



As it was desired to obtain the singing point to be expected under 

 operating conditions, the return loss measurements were made with 

 the building-out condenser on the return loss set adjusted for optimum 

 return loss at 2600, 2700 and 2800 cycles. The results of these 

 measurements are given in Fig. 8 for comparison with the computed 

 curve '^' mentioned previously. The improvement at the higher 

 frequencies of the actual over the computed values is due almost 

 entirely to the method employed in making the tests, and indicates 

 the advantage to be derived from individual adjustment of each circuit. 



