I5 2 TRANSFORMERS. [Exp. 



to reduce the exciting current measured on a coil of one voltage 

 to its value for a coil of another voltage. The watts core loss 

 is the same measured on one coil as on another, for the same 

 magnetization. 



CAUTION. If two coils are to be connected in parallel or series, 

 to avoid a burn-out it is necessary to first make sure of their polarity, 

 as described in 6 and 28, Exp. 5-A. 



Be careful of the high-potential terminals in this test. It shoufd be 

 made impossible for loose wires, or for persons making measure- 

 ments, to come in contact with these terminals. Although the testing 

 current and instrument are all of low voltage and although the high- 

 potential coil is open and has no current in it, the potential is there 

 and must be respected. 



4. Data. At normal frequency, say 60 cycles, vary the volt- 

 age (45) from say J to ij normal and determine the core 

 loss W Q , and exciting current / for various voltages. Note the 

 frequency at which the test is made. It is desirable that the 

 frequency be maintained constant, and that the voltage be of 

 sine wave-form. 



Take very accurate readings at two points (within, say, 5 or 

 10 per cent, of half and full voltage) by taking at each of these 

 points a series of five readings and averaging. This two-voltage 

 method is very convenient, since normal and half voltage (as 

 55/1 10 or 110/220) are often available or their equivalent can 

 be obtained by series and parallel connections, as described in 

 46. As will be seen later, Figs. 2 and 5, it is very accurate for 

 transformers built of ordinary iron, at normal and higher fre- 

 quencies, but not at frequencies far below normal. For trans- 

 formers with improved iron, the two-voltage method is not cor- 

 rect unless one observed point is taken very near full voltage, 

 little error being then introduced by obtaining values for full 

 voltage from the curves. 



5. If possible, repeat the data at a second frequency. If 



