INSTRUMENTS FOR THE NEW TELEPHONE SETS 345 



preceded by loud speech. If a poorly designed transmitter is held in a 

 position such that the carbon granules tend to fall away from the 

 movable electrode when this test is applied, the non-fluid action of 

 the carbon will prevent the reestablishment of contact with the elec- 

 trode surface with the result that volume losses of as much as 20 db and 

 a serious degradation in quality take place. Furthermore, these losses 

 persist until the transmitter is jarred or moved about sufficiently to 

 change the configuration of the granules. On the other hand, if the 

 effect of the frictional forces within the granular mass has been taken 

 fully into account in the design of the carbon element, these forces will 

 not react in a manner such as to prevent good contact with the electrode 

 following the large amplitude produced by loud speech and uniform 

 volume and good quality will obtain at all times. The new transmitter 

 meets this test with a substantial margin. 



Carbon leakage is prevented in the new instrument without impair- 

 ment of transmission by the resilient silk closure for the carbon 

 container previously mentioned. 



Aging. — Transmitter design has advanced to a stage where heating 

 at the points of contact in the carbon element need no longer be an 

 important source of aging. Therefore, such aging of the granular 

 material as occurs in a well-designed instrument is limited almost 

 entirely to that resulting from changes in the properties of the granules 

 caused by abrasion of their surfaces when the transmitter is subjected 

 to mechanical shocks such as occur when the handset is placed on the 

 mounting. As in the case of the earlier transmitter, the new trans- 

 mitter is machine filled ' with the result that the motion of the granules 

 and the resultant surface abrasion is reduced to a minimum. 



The changes in resistance due to the residual aging have little adverse 

 effect insofar as volume is concerned. In fact, the constants of most 

 of the circuits in which the transmitter is used are such that an increase 

 in resistance adds to rather than decreases the electrical output because 

 of the greater amount of power supplied to the transmitter from the 

 central office battery. 



On the other hand, an increase in resistance, though small, may 

 prove to be important in certain circuits where a critical relationship 

 between transmitter resistance and the performance of associated 

 apparatus exists. Under these conditions variations in transmitter 

 resistance may result in failure of the associated apparatus to perform 

 satisfactorily if certain limiting values of resistance are exceeded. In 

 determining the limits to be placed on these values account must be 

 taken of all the variables in the circuit in which the transmitter is 

 connected. Obviously combinations of variables of this nature 



