432 KENNELLY, TAYLOR— PROPERTIES OF 



the clamping of the diaphragm is unchanged when the amphtude 

 measurer is appHed. 



In order to investigate the matter further, the diaphragm B of 

 the receiver was mounted in the vibration explorer Fig. lo, de- 

 scribed in a preceding paper.^° It may be seen that the diaphragm 

 is here supported under a clamping ring, which is attached to the 

 main frame by four machine screws. Two of these screws, dia- 

 metrically opposite each other, were removed, leaving the diaphragm 

 clamped tightly under opposite points only, thus simulating the 

 effects of the clamps of the amplitude explorer in Fig. 9. The 

 diaphragm had a central load of 0.975 g"i-> ^^^ the same receiver B 

 was screwed into the explorer to actuate it. The impressed fre- 

 quency was maintained at 584 '—', and with an alternating-current 

 strength of 2.26 milliamperes rms. The amplitude of vibration 

 under these conditions was measured at various angular positions 

 15° apart, at uniform radial distances, 1.93 cm. from the center, the 

 clamping-ring diameter being 5.24 cm. The results obtained are 

 shown geometrically in Fig. 11. Here the axis of the tightly 

 clamped boundary is horizontal, and the axis of the loosely clamped 

 boundary where the screws were removed, is vertical. It will be 

 observed that the average amplitude in the latter axis is about 9 

 microns ; while in the former it is about 7.6 microns ; so that the 

 mean amplitude along the loosely clamped diameter is some 18 per 

 cent, more than that along the tightly clamped diameter, with inter- 

 mediate values in intermediate directions. This indicates the im- 

 portance of securing uniformly tight clamping around the boundary 

 of a telephone diaphragm. 



The motional-impedance diagram for this case is shown in Fig. 

 12. No localized distortion loop is visible ; but there is a general 

 shrinking of impedance, and therefore of diaphragm velocity, over 

 a considerable range of the diagram (559'—' to 618''—'). No distor- 

 tion loop was obtained at any time in the vibration-explorer tests, 

 as though no definite absorption frequency was brought about ; but 

 there was a fairly proportional degree of absorption over a wide 

 range of impressed frequency. 



10 Bibliography No. 9, Kennelly and Taylor. 



