RELIABILITY OF RADIO TELEPHONE CIRCUITS 



187 



London-New York short-wave telephone circuits during 1930, 1932 

 and 1934. These three curves have very nearly the same mid-range 

 slope, showing that the general character of the distribution was the 

 same over the wide range of transmission conditions experienced 

 within this interval. To those familiar with transatlantic short-wave 

 transmission during 1930, the year 1934 would rate as comparatively 

 undisturbed, and yet these curves indicate that for an equal percentage 

 of measurements or, as will be apparent later, for equal lost circuit 

 time during these two years the difference in required transmission 

 effectiveness would be only 13 or 14 db. 



Fig. 3 (b) shows that there was relatively small db separation be- 

 tween the 6-A noise distributions for 1930, 1932 and 1934 on the low 

 latitude South American circuits but that the position of the curves 

 is reversed, 1930 being better than 1934. An examination of field 

 intensit>' data for these years indicates that this is due to a change in 

 noise rather than to a change in signal transmission. 



Fig. 4 (a) compares the distributions for the circuits Buenos Aires- 



, ALL LONDON TO 

 NEW YORK CIRCUITS 



o— « ALL BUENOS AIRES TO 

 NEW YORK CIRCUITS 



o OALL HONOLULU TO 



SAN FRANCISCO CIRCUITS 



LOW 

 NOISE 



HIGH 

 NOISE 



\ 



10 20 -30 -20 



6-A NOISE IN DECIBELS 



Fig. 4 — Percentage distribution curves of 6-A noise measurements made on (a) 

 all London to New York, and Ruenos Aires to New York short-wave circuits during 

 1930 and on (/;) all London to New York, Buenos Aires to New York, and Honolulu 

 to San Francisco short-wave circuits during 1932 and 1934. 



