SLOW RELKASK KKLAV DKSKiN 



ABAMPERE TURNS, vF/aTT 



J L 



ATT 



Fig. 1 — Decreasing magnetization of an electromagnet. 



ing values of <p, as in Fig. 3, or against the corresponding values of 

 -V/, as in Fig. 4. In comparing two models, corresponding to curves 

 1 and 2 in these figures, the curves of t/G versus ^p in Fig. 3 indi- 

 cate the comparati\'e release times for the same spring load, as the pull 

 versus ^p relation is the same for similar structures. Thus in Fig. 3 the 

 times for the pull to drop to a common \'alue for curves 1 and 2 cor- 

 respond to the points c and 6, for example, and are measured b}^ the 

 areas o-c-c' and o-e-h' respectively. Hence cur\'e 1 gi^'es longer times in 

 this comparison than cur\'e 2, as illustrated in Fig. 3. 



If, on the other hand, the spring loads for the two models are adjusted 

 so that they release on the same ^'alue of A^/, the release times to be 

 compared correspond to decay to the points a and 6, for example. The 

 corresponding \'alues of t/G are given by the areas o-d-a! and o-e-b'. 

 These differ little from each other, and consecjuently curves 1 and 2 in 

 Fig. 4 are similar. It is therefore possible to adjust individual relays to 

 give nearly the same release time by adjusting the spring load so that 

 they release on the same \-alue of Ni. Advantage is taken of this fact in 

 the adjustment practices employed with slow release relays, as discussed 

 in a latter section. 



This graphical method for the prediction of release times, which was 

 developed by H. X. Wagar," is described here l)ecause of its utility in 

 demonstrating the relation between the timing relations and the char- 

 acter of the demagnetization curN-es. It provides a means for accurately 

 predicting the release times in specific cases, but does not afford as 



