THE TRANSFORMER. 151 



carried out for the sake of an experiment rather than to 

 obtain information about a special transformer, this course 

 may be recommended to the student. 



The curve of voltage has been added in Fig. 68, and is 

 necessarily a straight line passing through zero, since the 

 magnetic flux is maintained constant. In carrying out the 

 experiment it is convenient to obtain this voltage curve first, 

 after ascertaining the range of speeds available, but before 

 taking readings. At any given speed the voltage of the 

 transformer must afterwards be regulated to agree with the 

 corresponding point on the voltage line. 



The effect of eddy currents becomes more marked at 

 frequencies higher than those shown on the curve, and the 

 losses become very serious at frequencies above 100 if not 

 specially guarded against by very thin lamination and 

 satisfactory insulation between the core plates. 



This method of determining the iron losses may be applied 

 to samples of iron in order to find the loss per cubic centi- 

 metre, and thus to judge of the quality. For this purpose samples 

 are made up into a ring, or other form of magnetic circuit, 

 and a number of turns of wire are wound on the circuit. 

 The test is then carried out exactly as above described. 



If the losses in the resistance of the winding employed in 

 this experiment are appreciable, they must be calculated, and 

 subtracted from the wattmeter reading, as explained in 

 Experiment XXV. 



Separate Determination of Copper Losses The method of 

 calculating the voltage drop due to the resistance of the 

 windings of a transformer has already been given. The 

 power lost in the windings can be simply calculated from the 

 voltage drop by multiplying by the output current. Thus, 

 if v is found to be the drop in volts due to the- resistance of the 

 windings at any load current C.,, then the watts spent in 

 heating the windings are equal to v x C.,. 



The following experiment is a method for rapidly deter- 

 mining the copper losses by direct measurement, and depends 

 on the fact that the iron losses in a transformer are very small 

 when the magnetic flux in the core is low, since both eddy 

 current and hysteresis losses decrease rapidly with decreased 

 magnetic densities. 



