REPEATER DESIGN — NEWFOUNDLAND-NOVA SCOTIA LINKS 267 



The completed gland assemblies were all subjected to a minute X-ray 

 examination, followed by a pressure test at 5 tons*/square inch for three 

 months. Whilst under pressure, the glands had to withstand a voltage 

 test of 40 kv dc for 1 minute, to show no ionization effects when a volt- 

 age of 3 kv (r.m.s.) at 50 cycles was applied and to have an insula- 

 tion resistance greater than 20 X 10^^ ohms. 



MANUFACTURE 



General 



The manufacture of the electrical units was carried out in accommo- 

 dation specifically designed for submerged-repeater production. Tem- 

 perature was controlled at 68°F and the relative humidity was less than 

 20 per cent in the component shops and 40 per cent in the assembly and 

 test shops. Filtered air forced into the building maintained a slight posi- 

 tive pressure with respect to the outside and eliminated the ingress of 

 dust. With the exception of the tubes and some resistors all components 

 were manufactured in this 'diary' (Fig. 11). Operators are specially 

 selected, and they must change into clean protective clothing in an ante- 

 room before entering the working area, where no smoking or eating is 

 permitted. All operators are particularly encouraged to report or reject 

 any condition which is abnormal or in which they have not complete 

 confidence. Rigorous inspection and testing were carried out by the con- 

 tractor at all stages, and a Post Office team collaborated with 'floor' 

 inspection and the examination and approval of test results. 



The Electrical Unit 



The components, after the most careful examination and testing, 

 which in some cases included an aging test, were assembled into their 

 cans and then subjected to a shock test before undergoing detailed elec- 

 trical characteristics tests. Initial tests on the amplifier were done with a 

 set of 'standard' electron tubes, which were later replaced by the final 

 tubes for the complete tests. The cans were then assembled in a repeater 

 chassis and the electrical tests required before sealing performed — this 

 included a gain response to determine the best settings for the trinmier 

 pads in each transmission band. After fitting and sealing the outer brass 

 case, dry nitrogen was blown through the unit for 24 hours and the gas 

 holes then sealed. The overall electrical characteristics were then taken, 

 the repeater was energized for a two weeks' 'confidence trial' and the 

 characteristics were rechecked. During the 'confidence trial' the gain was 



