REGENERATION THEORY AND EXPERIMENT 



691 



tential is supplied to those plates of the oscillograph which produce a 

 horizontal deflection. To the common branch of the other wattmeter 

 is applied a potential equal in amplitude to the input voltage but lag- 

 ging behind it by 90 degrees. The rectified output of this wattmeter is 

 proportional to the product of input and output voltages multiplied 

 by the cosine of the transfer factor phase angle minus 90 degrees, or in 

 other words proportional to the sine of the transfer factor phase angle. 

 This voltage is supplied to those plates of the oscillograph which pro- 

 duce a vertical deflection. We have then across one pair of plates of 

 the oscillograph a steady potential proportional to the real component 



Fig. 7 — Schematic diagram of the circuit used to plot the transfer factor diagram on 

 the screen of a cathode ray oscillograph. 



of the transfer factor, and across the other pair of plates we have im- 

 pressed a steady potential proportional to the imaginary component 

 of the transfer factor. These two components act upon the beam of the 

 oscillograph to produce a deflection which in amplitude and in phase is 

 the resultant of the two component deflections and so corresponds to 

 the transfer factor. 



It will be observed that the above procedure requires a two-phase 

 source of constant amplitude, the frequency of which is variable over 

 the range necessary to establish the properties of the amplifier. In the 

 present instance the frequency range extends from 0.5 to 30 kilocycles, 

 and the accuracy required is of the order of five per cent. 



A schematic of the two-phase oscillator used is shown in Fig. 8. 

 This oscillator is of the heterodyne type. Two independent sources 

 are used, one of constant frequency (100 kilocycles), the other variable 

 in frequency and practically constant in amplitude over the range of 

 100 to 130 kilocycles. As indicated in the figure, the variable fre- 



