'54 TRANSFORMERS. [Exp. 



determined at about half and full voltage and note how closely 

 a straight line drawn through them coincides with the curve 

 through the working range. (It will be found that this straight 

 line construction, based on two readings, can not be used when 

 the transformer is worked at a very high flux density, as in the 

 29-cycle curve of Fig. 2.) 



7. Take from each curve the value of 7 for normal voltage 

 (48). Resolve the exciting current, 7 , into two components: 

 the in-phase power component /H (which supplies the core losses 

 due to hysteresis and eddy currents) ; and the quadrature mag- 

 netizing* component, /M. These are determined by the following 

 relations : 



The value of W \ is taken from curves, Fig. 3 and Fig. 5, 

 described in the next paragraph. At no load, power factor 



= /H-f-/ . 



If the transformer under test had a core made of improved 

 steel, with less core loss, the component /H would be somewhat 

 less than indicated in Fig. 2. On account, however, of the very 

 much greater value of the component /M (due to the higher flux 

 density common with such iron), the total exciting current 7 

 would be greater than shown in Fig. 2. Furthermore, the point 

 for normal voltage would be near the knee of the curve, so that 

 the straight line construction would not be accurate, as has 

 already been pointed out. 



Compute, as in Fig. 2, the values of 7 , /H and /M as per cent. 

 of the normal full-load current of the coil on which the test 

 is made; thus, the full-load current for a loo-volt coil of a 

 2 K.W. transformer, Fig. 2, is 20 amperes. Expressed as per 

 cent., the results will apply to any coil. 



* Usage is not fixed in regard to the terms " exciting " and " magnet- 

 izing" currents, these terms being not infrequently interchanged. 



