MINIATURE CONDENSER MICROPHONE SYSTEM 455 



pressure remains constant. The diameter can not be decreased 

 indefinitely, however, because of the limitations of amplifiers with 

 which the condenser microphone must of necessity be closely associ- 

 ated. The resistor for feeding the polarizing voltage and the grid 

 biasing resistor must be increased as the diameter of the diaphragm 

 (and consequently the active capacity) is reduced, in order that, at 

 low frequencies, the same proportion of generated voltage may get 

 to the grid of the amplifier tube. They must not be made too large, 

 however, for then the voltages due to thermal agitation ^ in the 

 resistors will become comparable to the signal voltage. Similarly, 

 the capacity of the input leads, vacuum tube and resistors will become 

 comparable to the lowered capacity of the microphone, with a conse- 

 quent reduction in signal voltage at the grid, over the whole frequency 

 range. The actual limit in reducing microphone size can not be 

 defined accurately; but it is definitely greater than one-tenth of an 

 inch. 



Further consideration of the diffraction problem, however, shows 

 that such an extremely small size is not really necessary in order to 

 practically eliminate sound-field distortion. When sound is picked 

 up indoors at some distance from the source, the directly incident 

 sound contributes much less to the microphone output than does the 

 reflected sound arriving from other angles of incidence.^ The greater 

 part of the effective actuating pressure comes then at polar angles in 

 the vicinity of 90°, where the diffraction effect is much less pronounced. 

 If 90° be taken as an effective average angle, it is seen from Fig. 1 

 that the diameter of a microphone need only be reduced to about 

 six-tenths of an inch in order to make this type of distortion negligibly 

 small. 



If sound is picked up out-of-doors, or indoors near the source, 

 the directly incident waves predominate over the reflected waves. 

 For this case it will suffice to place the microphone so that the sound 

 arrives at a polar angle of 90°; and a six-tenths inch diameter is still 

 sufficiently small. 



Inspection of Fig. 2 shows that the phase-difference loss is about 

 1 db at 10,000 c.p.s. for this diameter; so that six-tenths of an inch 

 can safely be chosen as an acceptable design value. 



The Amplifier 



Consideration has now been given to the problem of reducing the 



size of a microphone to such an extent that it would not appreciably 



disturb the sound-field. But all such labor is in vain if the size of 



^ J. B. Johnson, Phys. Rev., July, 1928, pp. 97-109. 

 ^ Loc. cit. 



