568 BELL SYSTEM TECHNICAL JOURNAL 



axes and passing through the region of overlap. The shaded areas are for 

 the sparking position as indicated and the location of the corona point is 

 shown to scale. Experiment shows that when the moving electrode has 

 reached the position corresponding to 400 X 10~^ seconds after the previous 

 spark, corona is established on the point. Thus the cathode is irradiated for 

 200 X 10~^ seconds prior to breakdown. 



No serious erosion problem was encountered when these gaps were oper- 

 ated for many hundreds of hours in air. No deterioration of the points was 

 observed when their locations were properly adjusted so as to avoid sparking 

 over to them. The cathode erosion rate is so low that appreciable flats were 

 produced only after a hundred hours of operation. The anode erosion was 

 estimated to be less than one-tenth of that of the cathode, and was doubtless 

 associated with a small amount of reverse current shown to be present. The 

 magnitude of the cathode erosion rate for tungsten in air is about twenty- 

 five fold less than that for tungsten in hydrogen under the same conditions 

 which indicates that oxygen plays an important and somewhat unexpected 

 role in making practical the operation of these gaps. 



There was, however, a serious corrosion problem when these gaps were 

 adapted to airborne radar because of the necessity for sealing the modulator 

 unit in a container capable of maintaining atmospheric pressure at high 

 altitudes. Spark discharges in air are attended by the formation of both 

 ozone and oxides of nitrogen, the latter combining with moisture to form 

 nitrous and nitric acids. These reached such concentrations under continu- 

 ous operation in the container that they were damaging to all enclosed equip- 

 ment because of their corrosive action. A solution for this was arrived at 

 after considerable study on the part of the Chemical Department. This 

 consisted of the use of a copper impregnated activated carbon as an absorb- 

 ent. With this absorbent a life of 500 hours was shown to be possible. 



Over 10,000 rotary gap switches of this type were manufactured and used 

 successfully in both ship and airborne radars. However, under the urge to 

 reduce the weight of all possible components used in airborne radar and even 

 to eliminate the necessity for pressurizing, the development of glass-enclosed 

 fixed gaps as switches was diligently pursued. 



The authors would like to acknowledge the cooperation of Mr. N. I. 

 Hall of the Whippany Laboratories whose responsibility it was to engineer 

 and develop these rotary gap switches for manufacture. 



^ II. Fixed Gaps 



Preliminary experiment indicated that a series of fixed gaps could be made 

 to operate satisfactorily as a modulator switch. A study was therefore 

 made to determine the most suitable gas atmosphere, electrode material and 

 gap design for use in sealed gaps. This led to the development of a unit 



