THE ROTARY CONVERTER. 333 



voltage given out, for various outputs of the converter, 

 the field being adjusted throughout so as to give a 

 minimum current. Under these conditions the power 

 factor was found to be unity, except when the converter 

 was working at no load, when the power factor was '99. 



The curve marked " excitation " shows the excitation 

 supplied in order to maintain these conditions. 



The dotted curves marked C and F 2 show the primary 

 current and secondary voltage for the same loads as 

 before, but with a constant excitation of '35 amps. A 

 comparison of these curves shows a most important 

 property of rotary converters, which will be further 

 alluded to in connection with the compounding of con- 

 verters. The upper curve F 2 shows that if the power 

 factor of the supply circuit remains constant there is a 

 considerable drop of voltage due to armature reactions as 

 the load increases. Also, in order to maintain the power 

 factor constant, a decrease of excitation is required.* 

 Further, contrary to what might have been expected, 

 with a constant exciting current, although the current 

 supplied to the armature is much larger, the variation in 

 voltage is much smaller. It is thus found that although 

 the efficiency is greater with the excitation adjusted so 

 as to maintain constant voltage, and the voltage generated 

 is higher, the effect of thus decreasing the excitation is to 

 increase the variation of voltage. 



Since a diminution in excitation with increase of load 

 produces an increased drop in voltage on the direct-current 

 side, it appears at once to be likely that if the excitation 

 were increased with the load, a decreased variation in 

 voltage might result. This reasoning is supported by facts, 

 and the idea is frequently carried out by adding a series 

 magnet winding in the direct-current circuit, which auto- 

 matically increases the ampere-turns on the field as the 

 load increases. 



Effect of Variation of Excitation. In the case of a rotary 

 converter driven from the alternating-current side, it is 

 evident that no change in speed can follow from an 

 alteration in exciting current, since the machine must 

 continue to run synchronously at a speed depending only 



* The percentage variation in voltage and excitation shown in the curves 

 greater than would be obtained in practice in a large commercial machine. 



