286 ELECTRICAL EQUIPMENT 



In the position shown in Fig. 155 the e.m.f. generated in the 

 coil has a value somewhere between zero and maximum, zero cor- 

 responding to a coil position midway between the field poles. 

 The armature current, which, in this case, is lagging somewhere 

 between and 90, can be considered as made up of two com- 

 ponents, an in-phase component having a cross-magnetizing 

 effect, and a 90 lagging component having a demagnetizing 

 effect. At zero power factor the wattless armature current, 

 lagging 90, has a maximum value, and consequently the greatest 

 demagnetizing effect. 



In Fig. 156 the current is leading and its effect is just opposite 

 to that when the current was lagging. It is thus seen that, in a 

 generator, the field is weakened by a lagging current and strength- 

 ened by a leading current. 



The armature reaction in polyphase generators is materially 

 different from that in single-phase machines. In the former its 

 total effect combines that of the several phases and has a constant 

 value provided the load is balanced. If unbalanced it will be of a 

 more or less pulsating nature of double frequency, as is always the 

 case in single-phase generators. 



The armature self-induction is caused by the leakage flux 

 which is set up by the armature current and which does not inter- 

 link with the field flux. Since the armature current is alternating, 

 the local or leakage flux, which does not become linked with the 

 main field, will be continually altering in magnitude and direction, 

 so that there is set up a self -induced e.m.f. proportional to the 

 leakage flux of each phase and lagging 90 behind the current. 

 The armature leakage is usually local, and thus a distributed 

 winding with many slots will have a smaller leakage inductance, 

 since the flux generated by each unit of current will be linked with 

 a smaller number of ampere turns. 



The exact value of the self-induction of an armature winding is 

 somewhat difficult to determine, its magnitude depending upon 

 the reluctance of the paths taken by the leakage flux. There are, 

 however, several methods in use which give results which agree 

 very closely with those afterwards obtained experimentally. 



If L is the self-induction, expressed in henrys, and / the fre- 

 quency, the inductive reactance is equal to 2-jrfL. It is of the same 

 nature as resistance and is expressed in reactance ohms. The 

 counter e.m.f. caused by it is lagging 90 behind the current, and 



