TRANSMISSION DATA FOR RATING TELEPHONE CIRCUITS 335 



be compared with other circuits under typical operating conditions. 

 In order to meet the last two requirements, it has been necessary to 

 adopt for the present a working reference system which is described 

 in terms of particular instrumentalities instead of in terms of physical 

 measurements. As plant conditions change other working reference 

 systems may be required. If any other reference system is adopted 

 it may be rated in terms of the existing working reference system, in 

 which case the overall rating of any complete circuit which can be 

 rated in terms of both systems will be approximately the same in 

 terms of either reference system. 



The working reference system which has been adopted is shown 

 schematically in Fig. 1. It consists of tw^o representative subscriber 



VARIABLE 

 TRUNK 



CHARACTER- 

 ISTIC 

 RESISTANCE 

 = 600W 

 CUT-OFF 

 = 3000"^ 



R=25^DC AS AN ALLOWANCE FOR 



RELAY, OFFICE WIRING AND HEAT COILS 



LINE NOISE = IOO NOISE UNITS IN RECEIVER 

 "TYPICAL" ROOM NOISE AS SHOWN IN FIGURE 2 



Pig 1— Working reference system for the specification of effective losses. 



sets on three-mile 22-gauge cable loops connected, through repeating 

 coils supplying 24-volt talking battery, to a trunk having a pure re- 

 sistance characteristic impedance and an adjustable attenuation. The 

 present working reference trunk has a very high attenuation above 

 3000 cycles to simulate loaded lines. Below this frequency the atten- 

 uation is independent of frequency and can be varied distortionlessly 

 so that differences in effective ratings can be expressed in terms of 

 differences in trunk attenuation. It has a 600-ohm characteristic 

 impedance. The circuit noise in the receiver of the working reference 

 system is 100 noise units. The room noise associated with this refer- 

 ence system is that distribution of noise which normally will be found 

 in relatively quiet offices and in relatively noisy residences. The 

 average magnitude of noise in such locations relative to other familiar 

 conditions is shown by Fig. 2. 



Any convenient rating may be assigned to the working reference 

 system. The Standard Cable Reference System with a trunk of zero 

 length, w^hich was used for specifying volume losses, was the best 

 circuit commercially available at the time it w^as adopted and was 



