MICROPHONIC NOISE IN VACUUM TUBES 615 



the discoverer of one of its sources and by several other investigators.^ 

 It is the fundamental noise arising from the circumstance that the elec- 

 tron current is a stream of discrete particles rather than a continuous 

 flow. For ordinary types of low power tubes of good design, over the 

 audio band of frequencies, the root-mean-square amplitude of this 

 noise is equivalent to about 1 microvolt (120 db below" 1 volt) of noise 

 voltage applied to the grid of the tube. G. L. Pearson ,2 in a recent 

 paper, has pointed out that for the best signal-to-noise ratio, the input 

 impedance should be so large that the thermal noise arising in this 

 impedance predominates over the fluctuation noise arising in the tube. 

 In many broad-band or high-frequency systems, however, such an 

 ideal condition is practically unattainable, and fluctuation noise 

 remains as a limiting factor. 



It is with the second type, microphonic noise, that we are particu- 

 larly concerned here, though sputter noise is also of interest and will 

 be dealt with briefly in a later paragraph. Microphonic noise, as the 

 name is usually applied, is the familiar gong-like sound which is always 

 produced when a vacuum tube, followed by a sufficiently high-gain 

 amplifier and a sound reproducer, is subjected to a mechanical shock. 

 Its origin is in the vibrations of the various elements of the tube, which 

 make minute, more or less periodic changes in the spacings of the ele- 

 ments and therefore make corresponding changes in the plate current, 

 whose value at every instant depends on these spacings. Its intensity 

 in a given tube depends on the type and intensity of agitation to which 

 the tube is subjected. For a given agitation, microphonic noise may 

 be reduced either by stiffening and damping the tube structure, thereby 

 reducing the amplitude and duration of vibration of the elements, or 

 by cushioning the tube so that it receives only part of the original 

 agitation. 



In order to treat the problem of noise reduction intelligently, it is 

 necessary to have a measure of the effectiveness of treatments applied. 

 To this end, the properties of microphonic response in vacuum tubes 

 have been studied, and a test set has been designed and built for labora- 

 tory use which affords a quantitative measure of microphonic response 

 in tubes, and of effectiveness of cushioning in cushion sockets. Some 

 of the more important characteristics of microphonic noise will now be 

 considered. 



^"The Schottky Effect in Low Frequency Circuits," J. B. Johnson, Phys. Rev., 

 V. 26, pp. 71-85, July, 1925. 



"A Study of Noise in Vacuum Tul)es and Attached Circuits," F. B. Llewellyn, 

 Proc. I.R.E., V. 18, pp. 243-265, Feb., 1930. 



"Shot Effect in Space Charge Limited Currents," E. W. Thatcher and N. H. 

 Williams, Phys. Rev., v. 39, pp. 474^96, Feb. 1, 1932. 



"Fluctuation Noise in Vacuum Tubes," G. L. Pearson, Physics, v. 5, p. 233, 

 September, 1934. Also published in this issue of Bell Sys. Tech. Jour. 



