INTERTOLL TRUNK NET LOSS MAINTENANCE 



963 



Table II 



Number of IntertoU Trunks in the Connection . 



Distribution Grade in db 



Per cent of Connections Departing 

 from Average 



±2 db or more 



±4 db or more 



±8 db or more 



2.5 



42 

 11 

 0. 



4.4 



65 

 36 



7 



5.0 



69 

 42 

 11 



8* 



5.6 



73 

 47 

 15 



Includes two trunks over 500 miles long. 



noise or low volume. Furthermore, undesirable contrast may be encoun- 

 tered on successive calls between the same two telephones. 



The results of the study as well as experience with the beginning of 

 automatic alternate routing show that the performance of the existing 

 trunk plant must be improved. Three immediate objectives have been 

 set: 



1. Reduction of distribution grades to about }^ of the values men- 

 tioned above, i.e., about 1.0 db. 



2. Maintenance of office bias within ±0.25 db. 



3. Removal from ser\'ice of individual trunks differing widely from 

 their design losses (in the order of 4 or 5 db). 



To achieve these objecti^'es requires effort along four lines. First, 

 systems should be designed to have sufficient stability once they are 

 adjusted. This involves the inclusion of stable circuit elements and the 

 provision of automatic regulating devices to compensate for unavoidable 

 transmission variations arising from natural causes. These features have 

 been applied to existing systems within limits imposed by economic con- 

 siderations and the state of the art. Further extension of these features 

 will be required in the future in order to meet the above objectives. 



Second, before a trunk is placed in service, each of its component parts 

 and the over-all trunk should be adjusted to give the correct loss. From 

 the transmission maintenance point of view, it is extremely important for 

 each trunk to start out with all of its adjustments correctly made. 



Third, existing and incipient troubles, and deterioration or maladjust- 

 ment of components, must be detected and corrected by routine mainte- 

 nance of indi\'idual systems used in making up trunks. Such activity 

 must make up for the inability to design systems to have the desired 

 stal)ility. 



Fourth, significant departures from trunk design losses must be de- 

 tected by over-all transmission measurements, and must be corrected be- 



