798 



Non- Synchronous Operation 



Taking up the non-synchronous dis- 

 charge first, let us imagine that the ro- 

 tating part of either gap is revolved by a 

 direct current motor and that its speed of 



Popular Science Monthly 



Power tronst 



fi9*< 



A diagram of a radio transmitter with a 

 rotary spark gap interposed in the apparatus 



rotation has no particular relation to the 

 alternating current supplied to the power 

 transformer in Fig. 41. As the secondary 

 of the transformer charges the condenser, 

 the spark-gap rotating electrodes R, R, 

 are constantly moving toward or away 

 from the fixed electrodes F, F. If the 

 instant of maximum potential of the trans- 

 former should occur while the spark gap 

 contacts are widely separated, no spark 

 would pass and the condenser would dis- 

 charge back into the transformer second- 

 ary on the next half cycle of applied 

 power. If, on the other hand, the spark 

 gap electrodes were quite near together 

 when maximum potential was reached, a 

 spark would pass and a group of radio 

 frequency oscillating currents would be 

 produced in the primary circuit. When 

 the spark gap rotor is driven inde- 

 pendently of the applied alternating 

 current power, i.e., non-synchronously, it 

 is evident that the time a spark will pass 

 (and, in fact, whether or not a spark will 

 pass at all) depends entirely upon chance. 

 The only way to be sure that a spark will 

 pass for each half cycle of alternating 

 current applied to the power transformer 

 is to increase the speed or number of 

 electrodes of the rotary gap so that at 

 least one opportunity for sparking will 

 exist near the maximum voltage portion 

 of each half-cycle. In commercial prac- 

 tice this has usually been accomplished 

 by running the spark gap at a speed which 

 corresponds to approximately GOO sparks 

 (or, more strictly, "opportunities to. 



spark") per second when the supply cur- 

 rent is of 60 cycles per second frequency. 

 Thus in each half-cycle of secondary volt- 

 age there are five instants at which the 

 condenser might discharge across the 

 spark gap, provided only that at each of 

 these times the condenser voltage is 

 higher than the minimum required to 

 break across the shortest spark gap. 



How the Condenser Discharge Time 

 Is Varied 



How the adjustment of the gap affects 

 the times of sparking may easily be seen 

 by studying Fig. 42. Here the solid 

 curve represents the numerical potential 

 value of the secondary condenser charge, 

 as it is produced by the power transformer 

 and without allowing for effects of with- 

 drawing energy by the spark discharge. 

 The dashed curve above represents the 

 breakdown potentials of the rotating 

 spark gap, and both are drawn for the 

 same successive instants of time. The 

 divisions along the horizontal axis repre- 

 sent time intervals of 1/600 second, and 

 consequently ten of them are contained 

 in two half-cycles or one complete cycle 

 of audio-frequency voltage at 60 cycles 

 per second. The voltage curves are all 

 plotted above the axis, since in this case 

 we are concerned only with the numerical 

 value of the voltage and not at all with 

 its direction — the spark gap will break 



rig. M 



Tima 



Curves showing the oj)eration of the non- 

 synchronous spark gap for a wireless set 



down whenever the potential rises above 

 a certain approximate value, substan- 

 tially without regard to the direction of 

 potential stress. The voltage is assumed 

 to vary from zero to ten thousand. 



Continuing with Fig. 42, the upper 

 dashed curve may be understood by 

 imagining the successive separations be- 



