RECORDING AND REPRODUCING OF MUSIC AND SPEECH 517 



Equations (12) and (13) which form the basis of the design work con- 

 tain four variables, /c, c, m and 2o. It is, therefore, necessary to deter- 

 mine two of them by the physical requirements of the problem after 

 which the other two are determined. The upper cut-ofif frequency /< 

 was arbitrarily chosen at 5000 pps. as a compromise between the high- 

 est frequency occurring on the record and the increase in surface noise 



=j-^Wi5^-r-'Tnr^ 



Fig. 16 — Electric equivalent of the system shown in Fig. 15 



as the cut-off is raised. The choice of the other arbitrarily set variable 

 came after considerable preliminary experimenting and was fixed by 

 the difficulty of obtaining a diaphragm which is light enough and has a 

 large enough area. Hence the effective mass of the diaphragm mz, 

 (Figs. 15-16) was fixed at 0.186 grams which value can be obtained by 

 careful design. The effective area can be made as large as 13 square 

 centimeters. For convenience let the arbitrary value chosen for/c = fc 

 and the value of m = m^. 



Solving Equations (12) and (13) for c and Zo, we get 



1 



also 



c= — - — , 

 T-fc-nis 



1 



TCfc 



(14) 



(15) 

 (16) 



In order to obtain the low value of mass mentioned, with a large 

 enough area, it was necessary to make the diaphragm of a very stiff 

 light material. An aluminum alloy sheet 0.0017 in. thick was chosen 

 and concentrically corrugated as shown in Figs. 17 and 18. These cor- 

 rugations are spaced sufficiently close so that the natural periods of the 



