136 THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1954 



range of capacity values. For large values of C, the discharge is damped 

 by the coil resistance, and for small values of C it is damped by the cur- 

 rents in the secondary. 



The latter case is essentially the condition applying for the small 

 capacity values used for contact protection. Initially, the flux decay is 

 opposed by the magnetomotive force resulting from both the eddy cur- 

 rents and the coil current. The latter discharges the condenser and 

 develops a charge of opposite polarity. The initial rate of flux decay is 

 therefore similar to that for exponential decay with a conductance equal 

 to Ge + N'^/R, where R includes the external resistor as well as the coil 

 resistance. The effect of the condenser charge is equivalent to a con- 

 tinuously increasing coil resistance, so that the decay approaches the 

 rate that would apply for exponential decay with G = Ge , and the decay 

 rate is then further increased by the reverse magnetomotive force result- 

 ing from the coil current reversal in the subsequent discharge of the 

 condenser. 



If CR is less than Ie , the effect of the coil circuit is to change the charac- 

 ter of the flux versus time relation, without materially changing the time 

 scale of the discharge. The initial decay is delayed, while the later 

 decay is accelerated. Hence the release time is increased for a heavy load, 

 corresponding to a high value of tp for release, while it may be slightly 

 decreased for a light load, corresponding to a small value of <p. The utility 

 of the coil-condenser circuit for contact protection results from the low 

 initial rate of flux decay, which holds the induced voltage across the 

 contact to a relatively low value in the initial stage of contact opening, 

 when the contact separation is small. 



Except initially, the predominant magnetomotive force is that of 

 the eddy currents, and this results in a changing distribution of field 

 intensity, as in the case of simple release. Hence the analysis of Fig. 8, 

 as described above, is not applicable quantitatively, and is therefore 

 not given here in detail. Qualitatively, the relations are similar, the pro- 

 tected release has a flux time relation similar in time scale to that for 

 simple release, with a lower initial rate of decay, and a higher rate for 

 the later stage. An illustration of this effect is included in the article by 

 M. A. Logan* cited above. No analytical treatment is available for de- 

 termining the differences in release time between simple and protected 

 release, at least for values of CR of the same order at Ie , as in the pro- 

 tection networks commonly employed. 



8 Release Motion 



It was stated in Section 7 that the waiting time in release is usually 

 larger than the motion time. This, hoAvever, is not necessarily or in- 



