7-11] BANDWIDTH AND DYNAMIC RESPONSE 375 



7-10 GAIN VARIATION AND GAIN SETTING 



Gain of an amplifier stage which does not incorporate feedback is equal to 

 the product of the transconductance of the vacuum tube and the transfer 

 impedance of the network which the tube drives. Instability of these 

 parameters results in gain variation. The effective signal transconductance 

 of a tube is proportional to the d-c current through the tube. Gain can 

 therefore be stabilized by operating the tubes so that the d-c plate current 

 is stabilized. This can be accomplished by means of large cathode resistors 

 or by operating a number of stages in series d-c connection. However, the 

 first method introduces transient recovery problems and the second method 

 reduces dynamic range. Application of conventional feedback stabilizing 

 techniques may be employed, but in high-frequency IF amplifiers it is 

 usually limited to a small amount of signal-current feedback which is 

 employed to compensate for input admittance variations of the tube. 



Network transfer impedance variations are on the order of ±0.5 db; and 

 when a small amount of d-c current stabilization is employed with the 

 vacuum tube, the variation in signal transconductance is on the order of 

 ±1.0 db. 



Stage gains are limited by bandwidth requirements in the case of wide- 

 band stages, and stability requirements in the case of narrow-band stages. 

 In addition, restrictions are usually encountered in gain distribution 

 through the receiver as a result of dynamic range requirements. Typical 

 average stage gains in a receiver are between 6 and 20 db incident to these 

 limitations. An amplifier providing 100 db gain might therefore require 

 about 10 tubes. Since the variation in gain of each stage is on the order of 

 ±1.5 db, 15 db reserve gain is required in the design, and provisions for con- 

 trolling the maximum gain of the amplifier over a 30-db range is required. 

 These gain-control variations do not include the gain control that is 

 required to accommodate target signal variations. The gain setting may 

 take the form of a noise AGC loop which controls the current of several 

 tubes or a manual adjustment which is periodically set. 



7-11 BANDWIDTH AND DYNAMIC RESPONSE 



A criterion sometimes employed for best signal-to-noise ratio is that 

 signal plus noise should be filtered by a network which maximizes the peak 

 signal-to-rms noise power. The network which will accomplish this result 

 was determined in Paragraph 5-10 to be simply the conjugate of the signal 

 spectrum; that is, the receiver filter should be "matched" to the signal. In 

 the case of the noncoherent pulse radar, each pulse must be considered as a 

 separate entity; therefore the optimum predetection filter is a bandpass 

 filter shaped like the RF pulse spectrum envelope. The IF characteristics 

 usually employed for maximum detection in thermal noise are reasonable 



