43G 



KENNELLY, TAYLOR— PROPERTIES OF 



approximately the natural frequency of the loaded diaphragm ob- 

 tained from the undistorted circle OCAD of Fig. 14. The curve of 

 Fig. 15 has two sharp resonance peaks, whereas the inferred am- 

 plitude curve of the diaphragm, without abnormality, has, as usual, 

 only one, ABCDE. It was noticed during the test represented in 





X ■» J I i i - 



"TEL-EPHONE Receiveif? B 



NATURAL FI»EQUENCY. 7iO^. 



Spring L Odd at Center. 



Natural Tretjueney if 

 Load, abou^ T50fv. 



Fig. 15. Relative Amplitudes of Diaphragm Vibration witli Spring Attached. 



Fig. 14, that the receiver gave practically no sound at 753 '— ' ; but 

 gave loud sounds for frequencies slightly removed on either side 

 of this. This peculiar property of a spring-loaded diaphragm to be 

 silent at a certain selected frequency, but to sound loudly at a small 

 departure therefrom on either side, may have practical applications. 



The natural frequency of the spring load was then altered to 

 about 913 '—', leaving the natural frequency of the diaphragm at 

 712-—', a little below its preceding value. The effect of this change 

 on the motional-impedance diagram is shown in Fig. 16. Here 

 the distortion loop is reduced in diameter nearly one half, and is 

 located much nearer to the origin on the left-hand side of the im- 

 pedance circle. 



By lowering slightly the nattu^al frequency of the diaphragm to 

 about 698 '-', leaving the natural frequency of the spring load un- 

 altered, the distortion loop in Fig. 17 is brought still nearer to the 

 origin O , and its dimensions further diminished. 



