594 BELL SYSTEM TECHNICAL JOURNAL 



of the resultant carrier level with distance are given for the point of 

 equal signal strength A and for the points B and C where the field 

 strength ratio is 6 and 10 db, respectively. Any departure from iso- 

 chronism will have the effect of making these points of maxima and 

 minima move along this line in space at the rate of one-half wave-length 

 per cycle difference in frequency. 



Now since the side band frequencies must perforce differ in wave- 

 length from the carrier, they will arrive at any given point out of phase 

 with the carrier, the amount depending on the distance from the trans- 

 mitter and the side band frequency. Thus the side bands will not for 

 the most part be in phase opposition at the same points in space as are 

 the carriers, and distortion will result from the elimination of the 

 carrier while strong side band components are present. The magni- 

 tude of this distortion is primarily a function of the existing field 

 strength ratio between carriers and, while the distortion occurs for only 

 a small proportion of the fading cycle, it is extremely objectionable 

 where the field strengths approach equality. Here the carrier is al- 

 most entirely eliminated momentarily and the resultant program con- 

 sists mainly of second harmonics and other distortion products. 



It is entirely outside the scope of this paper to attempt to present a 

 complete analysis of this problem but an effort has been made to indi- 

 cate the quantitative results that may be expected by selecting a few 

 typical examples. The signal being detected has been assumed in all 

 cases to consist of the ordinary carrier and double side band transmis- 

 sion. The theoretical work which follows has been based upon the use 

 of a square-law detector as being representative of the majority of the 

 existing receivers. In order to avoid undue complexity the curves have 

 been computed for a single frequency audio signal. 



In a square-law detector distortion appears primarily in the form of 

 second harmonics and the ratio of these to the fundamental has been 

 taken as a measure of the distortion present under the varying condi- 

 tions of reception that may exist in the middle area between the sta- 

 tions. There are so many variables concerned in this problem that it is 

 necessary to hold first one and then another fixed while different as- 

 pects of the situation are studied. 



The first set of curves, Figs. 12, 13, and 14, shows the conditions 

 which exist at the point directly between and equidistant from the 

 two stations when the audio signal supplied to the two transmitters is 

 exactly synchronized, i.e., the audio phase angle /3 = 0. With this 

 variable fixed the curves in each successive figure of the series have been 

 plotted for successively decreasing signal strength ratios in order to 

 show the effect of the varying radio phase angle with different degrees 



