TUNED AMPLIFIERS 



With any reasonable amplifier {A -\~ ly will be much greater than 2 



16 



•• \co„ CO I (A + lf 



As we saw in the positive feedback acceptor amplifier, if we take the upper 

 of the 1/(2)^/^ points co^ the n{(f02/wo) — (ojglco2)} is positive and equal to 

 dcolco^. 



doj 4 



• • w~ ^ "^ A + 1 



but 



6co 



Q = 



1 



Some performance curves for acceptor amplifiers of this type are shown in 

 Graph 31. It is clear that, amplifier gain for ampUfier gain, the positive 

 feedback system can give greater selectivity; also only two elements have to 

 be adjustable to vary the frequency in the positive feedback case. On the 

 other hand, where constancy of gain at the resonant frequency is important 

 the negative feedback scheme is probably safer. 



HT+ HT+ 



Cc 





Input 



Ah 



— Output 1 



-*► Output 2 



HT+ 

 Figure 13.10 



In designing circuits of this type it has to be borne in mind that the parallel 

 r equation used here assumes the network to work into an infinite impedance. 

 It ought therefore to feed back straight to a valve grid, itself unconnected to 

 anything else, and the question then arises as to where to feed in the input. 

 An elegant solution is to make use of a long-tailed pair, which has two input 

 grids. The circuit has the appearance o^ Figure 13.10, and has the incidental 

 advantage of providing balanced outputs if required. R^ can be comparable 

 in value with the anode load resistors, and C^Rc should be chosen to produce 



202 



