LOAD CARRYING CAPACITY OF AMPLIFIERS 577 



analyzer the resonant circuits of the ampHfier detector and the auxil- 

 iary oscillator were tuned to the harmonic to be measured. The 

 input currents and the attenuators Li and Lo were adjusted so that 

 the readings of meters Mi and ,1/2 did not change when the four-pole, 

 double-throw switch was thrown from one position to the other. It 

 will be seen that the difference between the settings Li and L-i then 

 gives the proportion of harmonic to total r.m.s. value of the out- 

 put wave of the amplifier. The proportion of harmonic necessary to 

 cause a 1% change in the r.m.s. value of a wave is approximately 

 14% and for the values obtained in the experiment the r.m.s. value 

 of the output could be taken as equal to the fundamental in the out- 

 put. The difference between Li and L2 therefore gives the proportion 

 of harmonic to fundamental. In the present work the difference 

 between the frequencies to be separated was large enough to avoid 

 any difficulty in obtaining sufficient resolution by the use of simple 

 resonant circuits. 



It is, of course, necessary that for measurements of this kind the 

 current supplied to the amplifier under test should be a single fre- 

 quency. A vacuum tube oscillator which was known to give a very 

 pure wave was used as the source of current. To obtain sufficient 

 energy for all the measurements made it was necessary to amplify the 

 output current from this oscillator and subsequently filter it to remove 

 the harmonics introduced by the amplifier. Final analysis of the wave 

 applied to the amplifier under test showed in most cases less than 

 0.2% and in all cases less than 0.5% of third harmonic and less than 

 .1% of all other harmonics. Greater purity could have been obtained 

 at the expense of more time and trouble but this was considered 

 sufficient for the purposes in view. Where necessary a small correc- 

 tion for the harmonic content of the input wave has been applied 

 to the results. 



The voltages to operate the cathode ray oscillograph were ob- 

 tained by a step-up transformer for the input and directly off a resist- 

 ance potentiometer for the output. The use of a step-up transformer 

 for the output wave is in general undesirable because it introduces 

 phase and amplitude changes which differ for the component fre- 

 quencies of the wave and thus the transformer itself introduces a 

 distortion which renders the interpretation of the figure as applied to 

 the amplifier distortion more difficult. This limits the method to 

 cases where a minimum of 10 volts is available in the output. For 

 the amplifiers dealt with here this voltage was available and the 

 limitation was not felt. On the input side the step-up transformer 

 has to transmit one frequency only so that the same difficulty does 



