SPARK GAP SWITCHES FOR RADAR 593 



where A is the intercept of the extrapolated solid straight line through the 

 value of lA and AC. There are reasons for believing that this is due primar- 

 ily to the higher electrode temperature, but is doubtless aided by residual 

 ionization left over by the high-energy sparks now passing. As a conse- 

 quence the lirst gap always breaks down at voltages intermediate between 

 A and .4+ AA. Gap 2 always breaks down at voltages between C and C 

 -f AC, below the re-ignition voltage of gap 1, and gap 1 always re-ignites 

 at voltages between D and D + AD allowing the main pulse of current to 

 current to pass at a voltage E. However, if the switch voltage is decreased, 

 C + AC will occasionally cross the re-ignition voltage characteristic R of 

 gap 1. Gap 1 can then re-ignite and thus the gaps will not fire on the 

 application of that trigger pulse. A second way in which the gaps can miss 

 is by failure of gap 1 to re-ignite at D. Even though either of one of these 

 events occurs only once in many thousands of pulses, a minimum operating 

 observed points and where B is the slope of this line. The shaded blocks I 

 and II of Fig. 19 (c) were obtained from values of the two terms A and Bt, 

 respectively, showing graphically the agreement between the calorimetric 

 and the oscillographic methods. 



As a result of calorimetric measurements on a wide variety of gaps having 

 either aluminum or mercury cathodes and operated under a wide variety of 

 pulsing conditions, we have been able to establish an empirical formula for 

 the dissipation D in joules per pulse per gap in terms of these gap parameters 

 and pulsing conditions as follows: 



D = 5.7(10)-7/p5 + [40 + 3.9(10)-2 p "i^] Ipt. (2) 



Here Ip is the peak current in amperes, S the gap spacing in mils, p the gas 

 pressure of hydrogen-argon in inches of mercur}^, and / is the duration in 

 seconds of an idealized square-top wave equivalent in ampere-seconds to 

 the actual current wave. This formula holds for either aluminum or mer- 

 cury- cathodes and is independent of gap design. It is moditied only slightly 

 when pure hydrogen is substituted for the hydrogen-argon mixture, the 

 constant 3.9(10)~2 becoming 3.1(10)"-. It is based on many measure- 

 ments in which the parameters covered the following ranges: 



Parameter Range 



S 40-350 mils 



p 28-50" Hg. 



t 1-6 X 10"^ seconds 



Ip 45-1070 amperes 



After calculating the value of D from Equation (2) the dissipation in watts 

 per gap for any project is obtained by multiplying by the pulse repetition rate. 

 This equation does not include the trigger energy dissipated which usually 



