240 THE BELL SYSTEM TECHNICAL JOURNAL, JANUARY 1957 



Usefulness of the Loop-Gain and Pulse Methods. 



Both ineth(xls require that all the repeaters between the testmg ter- 

 minal and the fault can he energized. If the fault is in the cable there is a 

 \ery high probability that the center conductor will be exposed to the 

 sea, in which case the power circuit can be maintained on one side of the 

 fault at least, although it may be somewhat noisy. Because the system is 

 short, it is permissible to energize the link full}' from one end onh'. The 

 condition can never arise — ■ as it can in the Oban-Claren\-ille link — that 

 the line current is limited by the maximum permissible terminal voltage. 



The loop-gain test is concerned with the amplifiers in their linear 

 regime and gives no mdication of the overload point ; for this the pulse 

 test must be used. On the other hand, the pulse test does not permit ac- 

 curate measurement of levels, since the pulse level reaching a particu- 

 lar repeater may be restricted by the o^•erload of an earlier repeater in 

 the chain. The pulse test is particularly useful in providing a check that 

 both sides of each amplifier are in operation and in locating a fault of 

 this type. 



Each method depends for its operation on non-linearitj' at a point 

 within each repeater and can only identify a fault as lying between two 

 such consecutive points in the link. It is therefore desirable that these 

 points should be as close as possible to the terminals of the repeater 

 in order to ensure that the faulty unit can be identified. In this respect 

 the loop-gain test has the advantage o^•er the pulse test. 



EXECUTION OF WORK 



Problems due to the remoteness of the site were overcome without 

 undue difficult}' with the co-operation of the other parties concerned in 

 the project, but the present paper would be incomplete without a brief 

 reference to the cable- and repeater-laying operations in Newfoundland 

 and at sea. 



The terrain and conditions in Newfoundland were quite unlike those 

 with which the British Post Office normally has to contend, involving 

 trenching and cabling through bog, rock and ponds in country of which 

 no detailed surve}' or maps were available. Maps were constructed from 

 aerial surve}', and alternative routes were explored on foot before a final 

 choice was made. As much use as possible was made of water sections 

 in the sea, river estuarj^ and ponds; some 22 miles were accounted for in 

 this way, leaving about 41 miles to be trenched by machine or blasted. 

 A contractor Avas engaged for this purpose and to lay the cable in the 

 trench, but all jointing was done by the Post Office. The standards of 

 conductor and core jointing were the same as those in the cable factories 



