148 THE BELL SYSTEM TECHNICAL JOURxXAL, JANUARY 1954 



external to the relay winding. Wherever a resistor or power term appears, 

 the sum of all resistances or powers is indicated. 



Maximum Versus Average Operate Time Presentation 



In w^hat follows, load controlled operation data are given in terms of 

 maximum times, whereas mass controlled data are average. Either choice 

 could be used for both. The data for converting to average or maximum 

 respectively will be described. However, the choices made here are con- 

 sistent with the normal use of the data. 



Mass controlled, sometimes called speed, relays operate with rela- 

 tively large power and ordinarily are used in common control circuits 

 where many events occur in succession. The total number of similar 

 common equipments necessary in an office is related to the control circuit 

 holding time. For a system design, the cost of power is balanced against 

 the cost of equipments to arrive at a minimum office cost.^ Now the 

 maximum holding time per call is never the sum of the maximum possible 

 times of each of the several relays operating in succession. It rather is 

 more nearly the sum of the averages, increased by considerably less than 

 the common maximum to average ratio of one of them. For example, if 

 n relays are assumed to have a distribution approaching normal, an 

 analytical expression for the probable maximum is: 



(2) 



where 6/av is an allowance for short time deviations of the manufactured 

 product from the long time average. Thus for mass controlled relays the 

 most directly applicable type of data is in the form of average time, and 

 maximum to average time ratio. 



For load controlled relays just the opposite is true. Here the operating 

 times are relatively long, either because the power drain is to be kept to 

 a minimum, as in a long holding time circuit, or it is an event which takes 

 place while several successive mass controlled e^'ents occur. For either 

 case, it is a single event and only its maximum duration is desired. Here 

 then the most directly applicable type of data is maximum. In addition 

 to these, other data for minimum times are needed for studies of 

 timing when two parallel circuit paths occur. 



Because how far the relay armature has to move is the outstanding 

 variable in mass controlled relay timing, these charts are prepared for 

 each nominal distance which can be chosen. When there is no concern as 



