274 THE BELL SYSTEM TECHNICAL JOURNAL, MARCH 1953 



This ac loss is due to a combination of skin effect and eddy currents 

 in the conductor. It is proportional to the sixth power of the diameter 

 of each of the strands of wire that go to make up the conductor, and the 

 square of the number of strands . If the size of an inductor is increased, 

 therefore, the ac loss may rapidly assume important proportions. Some 

 help can be obtained by finer stranding of the wire and this has, in some 

 cases, been carried to the extent of using 810 separately insulated strands 

 in a single conductor. Even if fine stranding is sufficient to reduce the ac 

 loss to a tolerable value, a penalty is paid in the form of higher dc re- 

 sistance, due to the space occupied by the separate insulation on the 

 strands. This amounts to a decrease in the winding efficiency, ky, , in (6). 



The effect of distributed capacitance on the dissipation factor of a coil 

 is a function not only of the volume, directly, but of the absolute value 

 of the dissipation factor. If the dissipation factor is low compared to 

 unity it is given, approximately, by 



(Do + d)^ 



Dc = -^ (24) 



1 - — 



where Do = D — Deis the dissipation factor that would obtain were it 

 not for capacitance 



d = dissipation factor of the distributed capacitance 



C = distributed capacitance 



Cg — resonating capacitance of the inductor. 



If (24) is written in an alternative form 



d C 



(25) 



2) _ ^ A C, 



Co 



it becomes evident that a coil with an inherently high Q is more seriously 

 affected, proportionately, by distributed capacitance than is a low Q 

 coil. Fig. 3 shows this information graphically, using the value 0.01 

 for d, which measurements on model inductors indicate to be of an 

 appropriate magnitude. 



Each of the curves in Fig. 3 is based on a constant value for C/Cg . 

 In practice, however, as volume is increased it becomes more difficult to 



