RECEIVERS AND MICROPHONES 



573 



wall of a cylindrical drum 30 cm. in diameter and 7 cm. deep. The 

 two openings were spaced 180° with respect to the axis of the drum 

 and on radii of 7.5 cm. Cracks between the microphones and the 

 wall were carefully sealed. The wall thus formed a baffle of the same 

 general character for each microphone. The drum was mounted on a 

 shaft passing through its axis, about which it was rotated at a speed 



40 50 100 500 1000 5000 10,000 



FREQUENCY IN CYCLES PER SECOND 



Fig. 8 — Response of moving coil microphone. 



of 100 r.p.m. Slip rings were provided for making electrical con- 

 nections to the microphones. The drum was placed in a sound field 

 set up by a moving coil loud speaker supplied with current from a 

 vacuum tube oscillator. The voltage generated by each microphone 

 was then measured with an amplifier and thermo-galvanometer. 

 With this arrangement each microphone passed through practically 

 the same sound field. By virtue of the symmetrical character of the 

 drum its rotation has very little influence on any standing wave 

 patterns in the room. A check on the reliability of the measurements 

 was the fact that, if the position of the loud speaker was changed very 

 little dift'erence was observed in the ratio of the voltages generated 

 even at the higher frequencies. Likewise, no change was observed 

 when the electrodynamic microphone was moved a small distance 

 axially in its mounting. The condenser microphone used in these 

 tests had been calibrated by means of a thermophone, but a correction 

 was made for the resonance due to the cavity over the face of the 

 diaphragm, which is not measured in the thermophone calibration. 

 The response of the microphone as determined in this way is shown by 

 the curve A in Fig. 8. 



The disagreement between the observed and computed values at 

 the higher frequencies is believed to be due to resonance oscillations 

 within the air-chamber beneath the diaphragm. In order to reduce 

 the magnitude of these oscillations the chamber was connected through 

 a narrow slit r-, (Fig. 9) to a small cavity formed within the central 



