616 BELL SYSTEM TECHNICAL JOURNAL 



Factors Affecting Microphonic Noise Levels 

 In the first place, it may be pointed out that the production of micro- 

 phonic noise in commercial types of vacuum tubes is an extremely 

 complicated phenomenon. Each individual component of the mechan- 

 ical structure is a complete vibrating system having several modes of 

 vibration and natural resonant frequencies, and usually very little 

 damping as compared with electrical circuits. These components, all 

 coupled together mechanically in various ways, form a mechanical 

 network much more complex than the electrical networks encountered 

 in communication engineering practice. 



The complexity of the mechanical vibration is reflected in the com- 

 plex character of the noise itself, and is admirably illustrated by the 

 frequency-response characteristics published by Rockwood and 

 Ferris,^ and by similar characteristics obtained in the course of this 

 work. It is further demonstrated by the experimental fact that when 

 a large group of supposedly identical tubes is tested by applying the 

 same mechanical vibration to each tube in turn, mounted in the same 

 socket, the response levels of individual tubes may differ from each 

 other by as much as 30 db for representative types of tubes. Such a 

 magnitude of variation would not be expected to result from the 

 comparatively small dimensional variations tolerated in manufacture, 

 and must be explained by the exaggerating effect of intercoupled me- 

 chanical resonances in a complicated vibrating system. A curve 

 showing a typical distribution of microphonic response levels in a 

 group of tubes of the same type measured under identical conditions of 

 agitation is shown in Fig. 1. The general shape of this curve is char- 

 acteristic of any function subject to random variations about a mean, 

 and the range of levels included between the quietest and the noisiest 

 tubes, approximately 30 db, is about average for different types of 

 tubes. Tubes of exceptionally firm construction and of fairly wide 

 spacing, may vary over as small a range as 15 or 20 db, while tubes in 

 which there is a possibility of slight looseness of parts, may vary over 

 as great a range as 40 db or even more. 



The nature and intensity of the agitation, the vibrational charac- 

 teristics of the tube mounting, and the type and degree of mechanical 

 coupling between the tube and its mounting also play an important 

 part in the determination of the microphonic response of a particular 

 tube, since the tube and its closely coupled mounting make up a single 

 vibrating system. The reason for considering the coupling apart 

 from the mounting is that it is usually of the pressure-friction variety 

 and is subject to random variation. 



' "Microphonic Improvement in Vacuum Tubes," Alan C. Rockwood and Warren 

 R. Ferris, Proc. I.R.E., v. 17, pp. 1621-1632, Sept., 1929. 



