Limits to Amplification * 



By J. B. JOHNSON and F. B. LLEWELLYN 



The amplification obtainable in a vacuum tube amplifier is limited by 

 the noise in the circuit. Of the various sources of noise the most funda- 

 mental and inevitable is thermal agitation of electricity. Other sources 

 are the influence of ions and of shot elTect and flicker effect on the current 

 in vacuum tubes, poor contacts, mechanical vibration, and hum from a.-c 

 cathode heating. These noises and their effects in limiting amplification 

 are discussed in this paper. Although the natural noise level of an amplifier 

 is exceedingly low, modern amplifiers have reached such a stage of perfection 

 that their noise levels often are practically at the natural limit. 



NOISE in amplifiers is now a familiar term. Any one who has had 

 his favorite radio hour ruined by static knows the effect of an 

 incoherent background of noise. Although static was one of the first 

 noises observed in radio amplifiers, its origin is really outside the 

 circuits. At one time it seemed that other sources of noise of purely 

 local origin, such as poor batteries, loose contacts, gassy tubes and 

 induction from power lines, might be eliminated entirely so that the 

 circuits would be capable of amplifying any signal, no matter how 

 small. It was found, however, that the noise level cannot be lowered 

 indefinitely; that there are limits below which, in the nature of things, 

 noise cannot be reduced. 



Of the sources of noise, the most fundamental and inevitable is 

 thermal agitation of electricity. In a perfect amplifier all other 

 noises would be reduced to a level below that of thermal agitation. 

 Next in order comes the influence of ions and of shot effect and flicker 

 effect on the current in the vacuum tubes. Under control to a greater 

 extent, but nevertheless of a malignant nature, are the effects of poor 

 contacts, mechanical vibration, and hum from a.-c. cathode heating. 

 In dealing with these disturbances, the circuit and vacuum tube of 

 the first stage of the amplifier are the most important, for here the 

 signal being amplified is at its lowest level. When the signal is so 

 faint that it is masked by the noise remaining as the natural limit of 

 the circuit, then the only possible remedy is to raise the signal level. 



The natural noise level is exceedingly low, yet modern amplifiers 

 have reached such a stage of perfection that their noise levels often are 

 practically at the natural limit. This is true not only of special 

 amplifiers built for experimental purposes, but of many amplifiers 



* Published in November 193-4 issue of Electrical Engineering. Scheduled for pres. 

 entation at Winter Convention of ^. /. E. E., New York City, January 22-25, 1935 



85 



