REPEATER DESIGN — NEWFOUNDLAND-NOVA SCOTIA LINKS 269 



Fig. 12 — Completed repeater. 



continuously monitored on a recorder (duplicated to distinguish between 

 test equipment and repeater vaiiations) on which changes in gain of 

 0.01 db were clearly indicated. On satisfactory completion of these tests 

 the unit was ready for housing. 



Assevibly of Electrical Unit in Housing 



The first step was to complete the molded joint between the tail 

 cables from the A-end of the electrical unit and the low-pressure side of 

 the appropriate bulkhead. This joint was X-rayed and proof tested at 

 20 kv dc for 1 minute. The electrical unit was then bolted to the bulkhead, 

 the slack tail cable being correctly coiled into the recess provided, and 

 the whole assembly was lowered into the housing for the first silver- 

 soldering operation. Following this sealing the tail cable joint was made 

 to the B-end bulkhead, which was then lowered into the housing and 

 sealed. 



A leak test was then made by applying an internal air pressure of 

 5 lb/square inch (gauge) and observing the surface when wetted with a 

 solution of a suitable detergent in water. After flushing with dry nitrogen 

 the vents were sealed and the housing was pressure tested. Finally the 

 brass gland covers were fitted and filled with compound, the extension 

 pieces were screwed on, the flexible diaphragms were fitted and the in- 

 ternal space was filled with polyisobutylene. The housing was then ready 

 for further electrical testing. 



Tests on Complete Repeater 



After housing, the repeaters were submerged in a tank of water for a 

 three-month electrical 'confidence trial.' Before and after the trial the 

 complete characteristics were checked and the noise was monitored on 

 both terminals; during the first and last few weeks of the trial the gain 



