EFFICIENCY OF TRANSFORMER 



167 



induced in 3 being added to the p.d. which exists across the mains. 

 Half the reading of the voltmeter across S gives the impedance 

 drop in one transformer. 



91. Transformer Efficiency Tests 



Although the efficiency of a transformer might be determined by 

 measuring the useful power across its secondary terminals, and the 

 total power supplied to the primary, and dividing the former by the 

 latter, this method is neither so exact nor so convenient as methods in 

 which the lost power is measured directly. A very simple method of 

 carrying out the efficiency test, originally suggested by Dr. Sumpner, 

 consists in finding, by means of a wattmeter, (1) the core loss and (2) 

 the copper loss at full load. Since the core loss remains sensibly 

 constant at all loads,* its value at full load may be taken to be the 

 same as on open circuit. Hence the core loss is equal to the power 

 supplied to the primary when the secondary is open-circuited.t In 

 order to measure the copper loss, the secondary is short-circuited (as 

 in the regulation test), and the primary supplied at a p.d. sufficient to 

 produce the full-load currents in the windings. The core induction 

 in this case is so small that the hysteresis and eddy-current losses in 

 the core may be neglected, and the power supplied to the primary 

 measured, as in the first test, by means of a wattmeter taken to 

 represent the copper loss. The sum of the core and copper losses 

 being known, the efficiency is 

 easily calculated. 



Another method of determin- 

 ing transformer efficiency, also 

 due to Dr. Sumpner, consists in a 

 modification of the well-known 

 Hopkinson test for dynamos. 

 Two similar transformers are 

 required for this test. Let us 

 suppose the primaries of the 

 transformers connected across the 

 mains, and the secondaries con- 

 nected so that their e.m.f.'s oppose 

 each other, and let, as shown in 

 Fig. 121, a wattmeter Wi be con- 

 nected up to measure the power supplied to the two transformers, 

 r\ denoting the non-inductive resistance in series with the fine-wire 



* See 53. 



t Strictly speaking, this power also includes the (extremely small) copper loss in 

 the primary dae to the magnetizing or no-load current. 



FIG. 121. Measurement of Core Losses of 

 Two Similar Transformers. 



