258 



MEASUREMENTS 



A system for measuring the response of a loud speaker employing a 

 thermal noise generator is shown in Fig. 11.9. A diode may be used as 

 a source of thermal noise. The output is amplified, filtered and fed to a 

 loud speaker. The frequency distribution of the energy fed to the loud 

 speaker is shown in Fig. 11. 9y^. The output of the loud speaker is picked 

 up by the microphone, amplified and passed through a narrow band pass 

 filter. The response characteristic of the filter is shown in Fig. 11.95. 



LOUD SPEAKER MICROPHONE 



^=m co 



A INPU T TO LOUD SPEAKER 



o o 

 AMPLIFIER 



B F ILTER CHARACTERISTIC 



1 



r. L OUD SPEAKER RESPONSE 



THERMAL 



NOISE 



GENERATOR 



(^^yzz 



FREQUENCY IN KC 



Fig. 11.9. Schematic arrangement of the apparatus employing a thermal noise generator 

 and a band pass filter for obtaining the response frequency characteristic of a loud speaker 

 j4. Input to the loud speaker. B. The response frequency characteristic of the band pass 

 filter. C. Response frequency characteristic of the loud speaker. 



The band width of the filter should be independent of the frequency. The 

 position of the filter pass band is varied with respect to frequency. The 

 output of the filter is detected and measured by means of a meter. 

 The response characteristic of a loud speaker is shown in Fig. 11. 9C. 



Apparatus employing thermal noise for obtaining response character- 

 istics has not been developed to the stage where it may be used with the 

 facility of other methods. It appears, however, that this type of mea- 

 surement will become very important for all types of acoustic measurements 

 when suitable apparatus has been developed. 



3. Calibration of the Sound Measuring Equipment '^^. — The microphone 

 should be calibrated in terms of a free progressive sound wave. The micro- 

 phone, amplifier and detector should have a combined characteristic which 

 is substantially independent of the frequency over the frequency range 

 under consideration. If it is not substantially constant over the frequency 

 range the data must be adjusted for known variations. 



A general schematic circuit arrangement showing one specific way to 



2* Standards on Electroacoustics, Institute of Radio Engineers, 1938. 



