110 BELL SYSTEM TECHNICAL JOURNAL 



Conclusions 



As was previously pointed out, the effect of the shielding action 

 of the iron armor of a submarine cable is to diminish the electro- 

 magnetic field which is propagated through the sea water, and which 

 gives rise to the return current. Combined with this effect is the 

 shielding action of the sea water adjacent to the cable, upon the dis- 

 tant portions. The total shielding effect increases with the fre- 

 quency until a point is reached where practically the whole of the 

 return current is carried by the armor wires. 



Several remedies have been suggested for diminishing the damping 

 effect of the armor wires. It can be proved, for example, that for a 

 given size of core and .weight of armor, the number and size of armor 

 wires can be chosen so as to give a minimum value of return im- 

 pedance. A proper choice of the electrical constants of the ma- 

 terial of which the armor is constructed would also be of advantage, 

 since the return impedance is somewhat larger for iron than it is for 

 material of higher or lower conductivity. 



Another method of diminishing the return impedance, which has 

 been used in practice, is to wrap the cable core with a number of con- 

 centric layers of conducting tape before it is covered with jute. The 

 return current, as it crowds in toward the core with increasing fre- 

 quency, will then have a path of comparatively low impedance, and 

 at the higher frequencies only a small portion of the current will be 

 carried by the armor wires and the sea water. The impedance of 

 the return path can be calculated for this case by the methods given 

 in the preceding pages. The following table compares the values 

 of the resistance of the return conductor calculated by three different 

 methods, and determined experimentally, for a cable provided with 

 a brass tape 5 mils in thickness. 



Resistance of Return Conductor — OHMS per Statute Mile 



Frequency Approx. Method* 



Cycles Approximate Corrected by Exact Experimental 



per Sec. Method „ ^ 2 Method 



V actor — 



3,000 4.00 3.15 2.87 2.92 



10,000 4.90 4.25 4.45 4.60 



The experimental values are the results of a series of measure- 

 ments made by the Department of Development and Research of 



* This is an empirical formula which has been found to be fairly close in most 

 cases. The correction factor suggested itself in that it takes care of the increased 

 surface of the armor wires, as compared with the corresponding continuous sheath. 



