MICROPHONIC NOISE IN VACUUM TUBES 625 



small, fixed output voltage which is measured and set at a predeter- 

 mined value with the aid of another thermocouple galvanometer. 

 The amplifier gain may be adjusted, by means of a small range, con- 

 tinuously variable potentiometer until the indicator gives the proper 

 reading to correspond with the known level of the applied input. 

 With reasonably steady battery voltages, this calibration is necessary 

 only two or three times in the course of a day's testing. The range of 

 noise levels for which the amplifier is calibrated extends from 10 db 

 above 1 volt to 65 db below 1 volt. This range has been found to 

 include practically all tubes which it has been desired to test with the 

 standard agitator. 



The flat amplifier characteristic, which has been described, is nor- 

 mally used for general testing in connection with vacuum tube design 

 work since it gives the highest microphonic level readings and there- 

 fore the most conservative picture of the performance of the tube from 

 the standpoint of the designer. Provision is made, however, for 

 switching in a specially designed weighted amplifier such as is used in 

 making routine noise measurements in telephone speech circuits.® 

 The frequency characteristic including this unit has already been 

 shown in Fig. 6, curve (h), and is designed to compensate for the in- 

 terfering effect of each component of noise on the average ear plus the 

 effect of the frequency characteristic of the telephone subset. A similar 

 weighting network compensating for the non-uniform frequency re- 

 sponse of the ear alone would also be useful, but has not yet been 

 provided. 



Nature and Measurement of Sputter Noise 



By making a slight modification of the amplifier circuit, this test 

 set may also be used to measure sputter noise. Sputter noise is a 

 descriptive name applied to a class of noises characterized by a harsh 

 crackling or sputtering sound easily distinguished from the gong-like 

 quality of microphonic noise or the steady roar of electron noise. It 

 may occur either with or without agitation and is the result of dis- 

 continuous changes in electrode potential such as may be produced by 

 imperfect contact between conducting members in a tube or by inter- 

 mittent electrical leaks across insulation. 



Sputter noise due to agitation is always accompanied by micro- 

 phonic noise, and though it often contains instantaneous peaks of high 

 intensity which constitute a very disagreeable and annoying type of 

 interference, its total energy content is usually so small that it contri- 



^ "Methods for Measuring Interfering Noises," Lloyd Espcnschied, Proc. I.R.E., 

 V. 19, pp. 1951-54, Nov., 1931. 



