ACCEPTOR AMPLIFIER— POSITIVE FEEDBACK METHOD 



remains to find an expression for the Q. When the gain has fallen to 1/(2)^/^ 

 of its maximum value, it will be 



(2)1/2 (3 _ ^5) 

 A 



{2 (3 - ABffl'^ 



but this is the value of the starred equation when {(coloj^) — (ojjco)} = 

 J^(3 — AB), from which we can deduce the upper and lower values of co 

 (let them be coi and 0^2) ^t which the response is down to 1/(2)^''^ of its maxi- 

 mum. 



For let cOp -f (<3co/2) = (O2, and co^ 



so that 



and 



Oh 



COo 



CO. 



O), 



CO. 



{dcxifl) = o>i, 

 "T 



CO 



to. 



CJO„ 



dot 



T 



CO, 



Then {(mJco^) — (0JJ0J2)}, which is the positive case of ((co/co J — (cojco)} 

 do) . dco I _ 1 



Q 



— , but 



CO. 



0). 



Q 



3~ AB 



Thus, provided A > 3, a. degree of amplification difficult to fail to achieve, 

 any desired selectivity may be had by appropriate choice of B. 



Valve 

 1 



; f(co) 



(^ 



Hl^—U Valve 



^ 



, -ve feedback 



■ve feedback 



63 



+ve feedback 

 Figure 13.5 



Although the minimum amplification required for indefinitely large Q is 

 so low, it is usual to employ two valves in acceptor amplifiers of this kind 

 for three reasons: 



(1) it facilitates the application of the feedback; 

 ~ (2) the spare gain can be lost in negative feedback loops round the valves 

 only — this will stabilize A and thus the stability of the system as a whole 

 {Figure 13.5); 



(3) where considerable gain is required but not a large Q, the amplification 

 has to come from the valves and not from the positive feedback. 



199 



