482 THE BELL SYSTEM TECHNICAL JOURNAL, MARCH 1957 



In order to see what maximum gain G (60) imposes on a typical sys- 

 tem we assume AFjf = 5 X 10^ c.p.sec. Z = 10'l2; Vc ^ IV. Under 

 those conditions we get from (60) that G has to be smaller than approxi- 

 mately 10^ If on the other hand G is very small the signal level at the 

 audio amplifier input is so low that the flicker noise of the detector can 

 still come in. A good practical figure for the IF amplifier gain is around 

 60 db. 



VIII. SATURATION EFFECTS 



In all the previous considerations RF power saturation effects were 

 neglected, i.e., we have assumed that the power absorbed is proportional 

 to Hi, where Hi is the RF magnetic field. When this assumption is no 

 longer satisfied, the question of sensitivity has to be re-examined for 

 different degrees of saturation. However it is difficult from an experi- 

 mental point of view to change the conditions of the experiment for each 

 degree of saturation and therefore an elaborate analysis of this case 

 does not seem to be warranted. However it might be of interest to see 

 the effect on the in phase component of the signal at complete or nearly 

 complete saturation 



The change in output voltage for a reflection cavity is (15) 



and from (4) 



The RF magnetic field in the cavity is given by 



H^ = CQod - r')P,„ (71) 



where C is a constant dependent on the geometr}^ of the cavity the re- 

 flection coefficient. Assuming a simple homogeneous saturation be- 

 haviour we substitute for x" the saturated value of Xs 



//(I) 



xJ' = ^ = ^^^^- - - (72) 



^' 1 + -Yi'Hi^TiT, 1 + yi'TiT,CQ,{l - r^)P,-„ ^ ' 



where X"" is the unsaturated value of the susceptibility and Pin the 

 power from the microwave source. 



At* V 



r 1 -j- yi-liloQoii- — T')Fin 



and the output voltage AV. 



