188 



BELL SYSTEM TECHNICAL JOVRSAL 



As a final clieck of furimila ((J) by means of this tuned circuit, liie 

 maximum amplification was measured at 170,000 cycles with different 

 values of extra resistance, Rtxi, inserted in the circuit between c and d 

 in Fig. 3. The results of these measurements agree v'ery well with 

 the formula as will be seen from Table III. For i?„(=160 it was 

 found necessary to connect the plate across the entire coil in order 

 to get maximum amplification and thus a further increase of J?„( 

 beyond KiO ohms will make it impossilile ii> obi.iin inaximuni amplifi- 

 cation with this circuit. 



T.\BLE III 

 /=170,000 cycles, L = 1.63X10-' henries, /?p = 22,OO0 ohms, ai=61. 



The variation with fre<iiieiu\- of the resistance of the coil is shown 

 on Fig. G. These resistance \alues are obtained from the resonance 

 curves in F"ig. 4, and hence indicate also the losses in the variable 

 condenser and the loss due to the input impedance Rg. 



The curve in Fig. 6 gives what may be called the "true" resistance 

 of the circuit, which is to be distinguished from the "apparent" 

 resistance of the circuit as measured for instance !>> ihi- well-known 

 resistance variation method. By this latter meihoil, the resistaiue 

 of the coil is assumed to be equal to such an amoimt of extra resistance, 

 as inserted in the circuit will decrease the resonance current to half 

 its former \alue, but this assumption is only true when the distributed 

 capacity of the coil is negligible as compared with the caijaciix- of the 

 variable condenser ('. or wIumi the resist. nut- is iiii rddiuid in liie 

 center of the coil. 



Il fiilliiws from fiinnula ((il llial for a gi\eii coil the niaxinuim 



amplification is proportional lo ,,,,.m(l the measurements iiuniioiu'd 

 \ /< 



above seem to indicate that the maxiiiiuni of this ratio has alreadx' 



been passed in ilie last case {d, F"ig. 4) when ilu' coil is used sim|)l\- 



