196 



BELL SYSTEM TECHNICAL JOURNAL 



words, in a multiple-stiige amplifier is it possible to get the total 

 amplification curve from the curve for the amplification per stage by 

 multipKing them together? 



The answer to this question is that the total ampliticaiion of a 

 multi-stage amplifier will, in general, be lower than the value obtained 

 by multiplying the amplification values per stage, and the reason 

 for this is to be found in the input impedance of the tubes. So far, 

 we have assumed the input impedance of the tube after the amplifier 

 to be high as compared with the impedance of the tuned circuit (or 

 transformer) and this is correct for a plate cur\ature detector, in 

 which the impedance of the load in the plate circuit is negligible 

 at the frequency of the amplified current but if the next tube is another 

 amplifier it is only true at lower frequencies. It has been shown ^ 

 that the input impedance of a vacuum tube can readily be calculated 

 by means of the constants of the tube and the output impedance. 



For the tubes used in the foregoing e.xperiments we have the fol- 

 lowing approximate constants: 



Cg_p = Gr\A to plate capacity =3X10~'' farad 

 Cg =Grid to filament capacity = 5X10"'^ farad 

 Cp = Plate filament capacity =5X10 '^ farad 

 Rp = Plate impedance =20,000 ohms 



/i = Amplification constant =6. 



The outinit inii)edancc including the plate-filament cajiarity will 

 be assumed to be a resistance equal to the plate impedance. 



If the input impedance is represented by an apparent resistance 

 R'g in parallel with an apparent caparit\- Cg, we get for R' g and 

 C'g the values given in Table I\'. 



T.MILE IV' 



R' = 



o,'R,Cl^ (m+2) 



C'. = C,-f2C,_, 



M+2 



u^RlCI-, +i 



" H. W. Nichols, Phys. Rei'., Vol. 13, p. 405, 1919. John M. Miller, Bureau of 

 Standards— Scien. Pap. No. 351, 1919. 



