206 ALTERNATORS. 



This value will vary with the armature current and with the 

 field excitation on account of the varying permeability of the 

 iron. 



It thus appears that the voltage drop due to armature 

 impedance is about 20 x 1-317 26.34 volts at normal 

 excitation in the present case. At first, this appears to be 

 contrary to the indications of the .curves shown in Fig. 94 

 where the difference between the volts at the terminals of 

 the armature at no-load and full-load is shown to be only 

 100 - 93.7 = 6-3 volts. The reason for the apparent difference 

 between the two results is that the voltage lost in armature 

 impedance is nearly all due to armature self-induction,and is 

 consequently nearly ^-period out of phase with the terminal 

 volts, which are only slightly affected. The diagram, Fig. 95, 

 shows the relation between the voltages of the example 

 worked above. 



The value of the self-induction obtained by this method 

 will always be slightly too high, since the weakening of the 

 field caused by the armature magnetic reaction is neglected. 



For many practical purposes it is not important to distin- 

 guish between loss of voltage due to weakening of the field, 

 and to self-induction. Neither cause produces loss of power ; 

 both diminish the available voltage. 



A further point to be noticed is that as the resistance of 

 the external circuit becomes lower, a greater proportion of 

 the total voltage generated is spent in overcoming the 

 armature impedance, and for this reason the distance between 

 the curves becomes greater as this voltage becomes a more 

 and more important factor of the total voltage and differs 

 less and less in phase from it. Thus, even if the volts lost 

 were the same for all excitations, this would cause the 

 difference between total and terminal volts to increase at 

 lower excitations. When the machine is short-circuited 

 the voltage overcoming impedance is the total voltage, and 

 its full value would therefore be seen on the curve. These 

 changes should be followed out in connection with the changes 

 occurring in the shape of the triangle, Fig. 95, when drawn 

 for various current outputs. 



The point where the load magnetisation curve cuts the 

 horizontal axis is given by the value of the excitation required 

 to send the load current through the armature when short- 

 circuited, and where the terminal voltage is consequently 



