422 



BELL SYSTEM TECHNICAL JOURNAL 



9.7 Severed Helix With Drift Space 



In actually putting concentrated loss in a helix, the loss cannot be con- 

 centrated in a section of zero length for two reasons. In the first place, 

 this is physically diilficult if not impossible; in the second place it is desirable 

 that the two halves of the helix l)e terminated in a refiectionless manner at 

 the cut, and it is easiest to do this by tapering the loss. For instance, if the 

 loss is put in by spraying aquadag (graphite in water) on ceramic rods sup- 

 porting the helix, it is desirable to taper the loss coating at the ends of the 

 lossy section. 



Perhaps the best reasonably simple approximation we can make to such a 

 lossy section is one in which the section starts far enough from the input 



0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 



CN 



Fig. 9.7 — Suppose that the circuit is severed and an increasing wave only is incident 

 at the tjreak. How does the signal build up beyond the break? The solid curve shows 

 (6 = QC = 0). dl) is the level of the incident increasing wave. 



so that at the beginning of the lossy section only an increasing wave is 

 present. In the lossy section CA" long we will consider that the loss com- 

 pletely shorts out the circuit, so that (8.28) holds. Thus, in the lossy section 

 we will have onlv two values of 5, whicli we will call hi and hu . 



8i = jk 

 5// = —jk 



k = iVc 



(9.21) 



(9.22) 

 (9.23) 



Let Vj and l'// be the voltages of the waves corresponding to 5/ and 6,i 

 at the beginning of the lossy section. Let 5| , 62 , 5:i be the values of 8 to the 

 left and right of the lossy section. Let I'l be the amplitude of the increasing 



