132 COMMERCIAL ELECTRICAL TESTING 



After the volts have been reduced from the starting 

 point of the curve to normal, three single-phase watt- 

 meter readings, one above, one below and one at normal 

 voltage, should be taken on the two legs to check the 

 results. Check readings should also be taken with a dif- 

 ferent voltmeter and ammeter. 



The single-phase excitation amperes are theoretically 

 1.73 times the three-phase and twice the two-phase 

 values; that is, the kv-a. has equal values for the motor 

 whether single-phase or polyphase. Practically, the 

 single-phase amperes are from 1.6 to 1.7 times the three- 

 phase, instead of 1.73 times. The same. ratio holds for 

 quarter-phase. The watts excitation is the same for 

 polyphase or single-phase, so far as core loss is concerned. 

 The increase in watts single-phase over the watts poly- 

 phase is equal to the polyphase C 2 R. For instance, if 

 the three-phase excitation requires 1000 watts and the 

 C 2 R three-phase is 100 watts, the single-phase excitation 

 will be 1100 watts. 



Before shutting down, a curve should be plotted with 

 volts as abscissae and the algebraic sum of the watts as 

 ordinates. 



Wattmeter work may be somewhat uncertain, and 

 accurate results can only be obtained under good condi- 

 tions. An endless belt on the driving alternator is neces- 

 sary, a laced belt making the wattmeter needle swing with 

 a steady beat corresponding to the striking of lacing 

 on the generator pulley. Any belts running near the 

 table must have their static charges drawn off by a 

 grounded wire and the cases of all transformers should 

 be connected together and grounded. Wattmeters must 

 be carefully handled on high voltages, since all three 

 phases of the alternator are connected on the table and 

 contact between two of the instruments short circuits 

 one of the phases. 



The two important points on an excitation curve are 

 the watts at normal voltage and friction watts. These 

 points determine the percentage core loss for the motor. 

 Several readings, only a few volts apart, should be 

 taken on each side of normal voltage and the volts 

 and amperes in the different phases at two or three 



