182 



THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1954 



20 



LU 6 

 1- 5 



2 



1.5 



1 1.5 2 3 4 5 6 7 8 9 10 15 20 30 40 50 60 80 IOC 



N^C X I0"6 (tURN2 X MICROFARADS X 10"^) 



Fig. 15 — ■ Release time as a function of N^C 



Fig. 5 are used. This permits exact simulation for anj^ winding and net- 

 work from a time standpoint. For presentation of data, an arbitrary 

 separated product form for the more important variables is used. The 

 particular factors chosen are: 



(1) A^'^C, a measure of frequency, with A^/o (release ni) as a param- 

 eter, 



(2) /2C, a damping term, with A^/o as a parameter. 



(3) Soak A^/ with A^ /jR as a parameter. 



Essentially what is assumed is that the release time can be represented 

 well enough as a function of the form: 



h = MN'C) X MRC) X MNI). (38) 



Then by holding two factors fixed and varying the third, each factor in 

 turn is evaluated. Generally the accuracy is better than 10 per cent even 

 with this elementary approach. For a few cases the error approaches 15 

 per cent. Typical charts for these three factors are shown in Figs. 15, 

 16 and 17. 



Release Time for Series Relays and an RC Shunt 



With two series relaj^ windings it is usual to provide a single RC net- 

 work, as shown in Fig. 18. This results in a double infinity of possible 

 combinations if handled directly. However, it can be broken into two 



