GAS-DISCHARGE TRANSMIT-RECEIVE SWITCH 71 



provide a certain amount of self-protection against the spike. It has been 

 estimated that this self-protection seldom exceeds 2 db in practice. 



Leakage Power Measurement. The c-w method of measuring the flat 

 power has already been mentioned. Spike energy and direct coupling 

 must, of course, be measured under normal high level operating conditions. 

 Relative measurements of the spike can be made with an oscilloscope, acting 

 ballistically, and the factors which affect the spikes can be studied in this 

 manner. A correlation between the relative spike energies and the degree of 

 cr>'stal protection can be obtained by trial and from this correktion the 

 operating conditions for adequate protection can be determined. Most of 

 the early studies were made in this way. It is possible to deduce absolute 

 values for spike energ>', flat power, and direct coupling from measurements 

 made when all three are present because of the different ways in which these 

 parameters vary with the recurrence rate, pulse length and transmitter 

 power. The method of doing this is outlined in appendix D. 



A more precise method of measuring the spike energy involves the can- 

 cellation of the flat power by a signal of adjustable phase and amplitude 

 obtained from the high-level transmission line. The average spike power 

 is then measured directly and energy per spike computed. Most of the 

 spike data quoted earUer were obtained in this fashion. 



High-level Loss. The power dissipated in the TR box as a result of the 

 gas discharge is not ordinarily a large enough fraction of the total transmitter 

 power to be of any concern. Using the Po values previously quoted, it is 

 possible to compute the gas discharge power by the use of Fig. 16. At a 

 line power of 100 kw and a low-level loss of 1 db the gas power in the 721A 

 tube is 63 watts. The corresponding figure for a 1.5 db box using the 724B 

 tube is 47 watts. For these cases the high-level loss is therefore less than 

 0.005 db. Low as this fraction is in db it still may be high enough to affect 

 the life of the TR tube, as discussed in a later section. No trouble of this 

 sort is ordinarily encountered with the 724B or 721A tubes. The chief 

 cause of failure of the 1B23 is from loss of Q and this in turn is caused by the 

 sputtering action of the high-frequency discharge. 



Recovery Time. As mentioned earlier a TR box must recover its low- 

 level properties at the end of the transmitting pulse in a very short period of 

 time. The actual "recover)' time" is in fact several orders of magnitude 

 smaller than the deionization times of the usual gas discharge so that a 

 quite different mechanism must be involved. While an exact theory of the 

 recovery is beyond the scope of the present paper, a qualitative picture of 

 the recovery process may be of interest. 



During the transmitting period the free electrons provide almost all of the 

 discharge current, and are replenished by electron-molecule collisions. At 



