708 THE BELL SYSTEM TECHNICAL JOURNAL, MAY 1956 



offices bear a generic resemblance to digital computers and it is therefoi-e 

 not surprising that the magnetic drum has engaged the attention of tele- 

 phone engineers, since the speed and flexibility of such a device offei-s 

 much promise in connection with forward-looking telephone office de- 

 sign. One system has already been described^' ^ involving the use of mag- 

 netic drums for telephone switching control applications in an entirely 

 new form of telephone office; it is the purpose of this article to describe 

 another application of less complexity which could function in coopera- 

 tion with equipment in existing telephone offices. 



The standards of reliability and ruggedness which must be met by any 

 equipment proposed for Bell System use are in some respects a good deal 

 higher than those imposed on other commercial sytems such as digital 

 computers. Thus when a new type of apparatus such as a magnetic drum 

 and its associated electronic components is considered for a telephone 

 job, it is necessary to determine whether the apparatus is capable of being 

 designed to meet these stringent requirements. This was judged to be the 

 most important objective of the undertaking about to be described, and 

 it strongly influenced the choice of experimental application for the drum. 



The program which the designers set for themselves to determine the 

 possible suitability of the magnetic drum type of equipment might be 

 summarized as follows: 



(1) Choose an existing telephone application in which a magnetic drum 

 system can receive a satisfactory work-out without disordering the sys- 

 tem. 



(2) Design a magnetic drum system to work cooperatively with exist- 

 ing office equipment, using existing power facilities. Assume that the de- 

 sign is aimed at practical application so that due regard is given to operat- 

 ing economies, and protection against power failures. 



(3) Construct a full-scale model following the design, and test the 

 model in the chosen environment long enough to determine the failure: 

 rate and the reasons for each failure. 



(4) Evaluate the results in order to determine the sphere of useful- 

 ness, and the proper design philosophy for applying magnetic drum sys- 

 tems of any kind in existing telephone offices. 



One telephone switching application which meets the qualifications of 

 (1) above exists in the new No. 4 A toll switching offices. Here, due to the 

 demands of nationwide dialing, a large-scale translation function is re- 

 (]uired to convert destination codes into information which will properly 

 loute each call. The volume of information which nuist be stored foi' 

 1 laiislation purposes, and the relatively rapid access desired, fall close to 

 the optinnim parameter values of magnetic drum systems. The action 



