256 BELL SYSTEM TECHNICAL JOURNAL 



cause of the more severe service conditions, the predominant factor in 

 the aging of the handset transmitter is mechanical. 



Laboratory aging tests equivalent to about four years of handset 

 service under severe conditions, indicate an increase in carbon noise of 

 only 6 to 8 db above the initial value. Neither before nor after aging 

 is the noise produced by the handset transmitter of practical impor- 

 tance. 



The output of the handset transmitter is decreased somewhat and 

 the average resistance increased to about double its initial value as the 

 result of aging. The relatively small change of resistance with posi- 

 tion in this design, however, effectively prevents the frequent occur- 

 rence of resistances sufficiently high to interfere with the operation of 

 signalling apparatus in circuit with the transmitter. In this respect 

 the new design represents a notable improvement over earlier types of 

 instruments. 



The method adopted for filling the handset transmitter contributes 

 materially to keeping the aging rate low. In most of the transmitters 

 which preceded the present instrument, the carbon occupied only 

 about three-quarters of the volume of the carbon chamber. This 

 allowed the granules to move freely and caused rapid aging. Ob- 

 viously, the movement of the granules would be reduced to a minimum 

 if the chamber were filled full of carbon. Inasmuch as the space oc- 

 cupied by a given weight of granular material is dependent upon the 

 configuration of the granules, it is evident that a full filling cannot be 

 obtained merely by pouring carbon into the chamber. Definite 

 means, therefore, must be provided for bringing the carbon into a 

 minimum volume state if a full filling is to be obtained. 



A machine has been designed for this purpose which vibrates the 

 transmitter during the filling operation. This settles the carbon in 

 much the same way as if the transmitter were held in the hand and 

 tapped. The carbon chamber cannot only be filled full of carbon in 

 this way, but the effect of differences in the volume of the chamber, 

 due to the commercial variations in the size of the parts, is eliminated. 



It has been found that the loss in sensitivity which results from filling 

 a transmitter in this manner is relatively small. There is, however, 

 one feature of a full filling, often referred to as "mechanical packing," 

 which is objectionable and has led to a filling slightly less than full. 

 If the diaphragm of a transmitter which has been filled full by the 

 method described, is displaced sufficiently to permit the granules to 

 change their configuration, the volume of the filling will increase and 

 the diaphragm will not return to its original position. A marked in- 

 crease in contact pressure and lowered sensiti\'ity takes place. When 



