FLEXIBLE REPEATER DESIGN 95 



relative humidity under 10 per cent if the water intake were five times 

 this amount. 



After glass seals are silver brazed into the ends of the copper tube of 

 the repeater the helium test is repeated to check the braze and to re- 

 check the seal. For this test the entire repeater must necessarily be sub- 

 merged in high pressure helium. Obviously, in order to sense a possible 

 passage of the gas from the outside to the inside, the leak detector vac- 

 uum system must be connected to the internal volume of the repeater. 

 For this and other reltsons a small diameter tube that by-passes the seal 

 is provided as a feature of the seal design. After the leak integrity of the 

 repeater is established by this means for all but the access tube, this 

 tube is then used as a means of vacuum drying the repeater and then 

 filling it with extremely dry nitrogen. Following this, the tube is closed 

 by welding and brazing. This closure is then the only remaining leak 

 possibility and is checked by a radioisotope leak test. 



Radioisotope Leak Test 



Of various methods of detecting the passage of very small amounts of 

 a liquid or a gas from the outside to the inside of a sealed repeater, a 

 scheme using a gamma-emitting radioisotope appeared to be the most 

 applicable. 



The relatively small region of the welded tube referred to above is 

 surrounded by a solution of a soluble salt of cesium^^^. With the entire 

 repeater in a pressure tank, hydraulic pressure in excess of service pres- 

 sures is applied for about 60 hours. The repeater is removed from the 

 tank, the radioactive solution is removed and the test region is washed 

 by a special process so as to be essentially free from external radioac- 

 tivity. A special geiger counter is applied to the region. If there has been 

 no leak the gamma radiation reads a low value. If an intake has occurred 

 of as much as one milligram of the isotope solution, the radiation count 

 is about four to five times greater than that of the no-leak condition. 

 The rate of leak indicated is an acceptable measure of soundness of the 

 repeater closure. 



The helium and subsequent isotope leak tests are made on a repeater 

 not only when its glass seals are installed but are performed again on 

 each rubber seal after it is brazed in place. 



Electrical Tests 



Prior to assembly into the repeater the various networks are tested 

 under conditions simulating as nearly as is feasible the actual operating 



