THE TRANSFORMER. 161 



If a complete set of experimental curves were required 

 for inductive loads, they might be obtained in a manner 

 similar to that described in connection with Experiment 

 XXXVI. for the characteristic of an alternator. 



As already stated, transformers are seldom tested on 

 inductive loads, except to obtain the secondary voltage at 

 certain loads to fulfil the conditions of a specification. 



Indirect Methods of Measuring Efficiency. It is frequently 

 more convenient to measure separately the losses occurring 

 in a transformer, and then to state the efficiency as being 

 power given out _ useful output 

 power supplied output + losses. 



This system of measurement has two important advan- 

 tages, namely, that the power required to carry out the test 

 is small, and that the losses are measured directly and con- 

 sequently with greater accuracy than when they are derived 

 from measurements made upon the total input and output of 

 the transformer. 



Method I. The efficiency may be calculated directly 

 from the measurements made in Experiment XXV., page 141, 

 together with separate measurements of the resistances of 

 the windings. 



Let W i = watts spent in iron losses determined as in 



Experiment XXV. 

 RI> ^2 = resistances of primary and secondary 



windings. 



GI, C = primary and secondary currents. 

 W = output. 



Then total losses = W t + C?R l + O, 2 R, 



W 



and efficiency = = ^ n . 2 p n 2 p 



"t i ^i" -n-i + I/a Ky 



W 



This may be written /R l \ where k is the 



W + W j + C 2 " I -jnj + -K-2 I 



ratio of transformation (see page 130). 



The value of W h the watts spent in iron losses, is the value 

 of the watts taken by the transformer on no-load at normal 

 voltage as read on a wattmeter connected in the primary 

 circuit. The iron losses are constant at all loads, and the 

 copper losses are calculated for each load for which the 

 efficiency is required. The efficiency is thus determined by 

 calculation from a single measurement made at no-load. 



