622 



BELL SYSTEM TECHNICAL JOURNAL 



points lie, within the limits of experimental error, on parallel straight 

 lines -so drawn that along them the microphonic noise level expressed 

 in volts is proportional to the initial velocity of the tube mounting. 

 Some such relation as this would be expected to hold as long as the 

 response of the system is linear. The departure from this law at higher 

 values of momentum, then, probably indicates non-elastic motion 

 either of elements of the tube with respect to one another or of the tube 

 with respect to the socket. It may be noted in passing that the No. 

 264B Tube is exceptionally rigid in structure and that in more loosely 

 constructed tubes, the straight line part of the response curve ends 

 at much lower intensities of agitation. 



Since the noise energy is spread over a band of frequencies, the 

 microphonic response observed in any given reproducing system de- 

 pends also on its frequency-response characteristic. In the usual type 

 of volume indicator, the response is substantially uniform over the 

 audio range of frequencies, but where the final auditory sensation is 

 being considered, the overall characteristic is modified by that of the 

 ear of the listener.^ The effect of changing the overall response 

 characteristic is illustrated for one particular case in Fig. 6 and Table I. 



0-I5 



Z 



^-20 

 O 



01-25 



500 1000 5000 



FREQUENCY IN CYCLES PER SECOND 



Fig. 6 — Microphonic noise amplifier frequency characteristics. 



10,000 



Here, two sets of measurements have been made on each of three types 

 of tubes, one set using an amplifier having a fairly uniform gain char- 

 acteristic, curve (a), Fig. 6, and the other set using a weighted ampli- 

 fier, weighted as in curve (b) in the figure. The same agitator and 



^ "Speech and Hearing," H. Fletcher. D. Van Nostrand Co., 1929. 



