Economics of Telephone Relay 

 Applications 



By H. N. WAGAR 



(Manuscript received July 6, 1953) 



Today's telephone central office is largely built around the telephone 

 relay. As each office may use some 60,000 of them, their performance 

 characteristics, as well as their first cost, have a very important influence 

 on the cost of the office. By properly halancing such factors economically, 

 the loicest cost office can be planned. 



This paper shows how performance features and design factors may all 

 be expressed as ''equivalent first cost," and so be related to manufacturing 

 cost itself. The influence of lot-size on manufacturing cost is considered 

 including a determination of optimum lot -sizes, to aid the relay designer 

 in deciding on standardized models. 



With a basic performance language formulated — equivalent first cost — 

 optimum relations among the variables may be established. Such results 

 are given for (a) optimum coil-plus-power costs, (b) optimum power-plus- 

 speed costs, (c) optimum number of relay codes, and other related problems. 

 Methods are also given to evaluate how serious the effect may be when opti- 

 mum conditions are not satisfied. 



Were it not for the application of these methods, central office costs would 

 be much higher. 



INTRODUCTION 



The telephone central office of today is in part an enormous computing- 

 machine which, upon instructions from the customer, gi\'en by his dial, 

 calculates how to find the called party; then the remaining mechanism 

 completes the call. This machine is composed largely of interconnected 

 relays, used in enormous quantities. Every dial call in a large city in- 

 voh'es around 1000 relays. A typical large office contains more than 

 60,000 of them; in fact, they are used so extensively that the Western 

 Electric Company manufactures about ten of them for e\'er3^ new sub- 

 scriber, and their output is figured in tens of millions per year. It is not 



218 



