266 THE BELL SYSTEM TECHNICAL JOURNAL, MARCH 1953 



are two practical limitations on this procedure. First, there is the 

 mechanical difficulty of grinding particles to smaller than the few 

 microns diameter that now represents the commercial limit. Second, as 

 the size of the individual particles is made smaller more air gaps are 

 introduced in the magnetic circuit by the insulating material, and the 

 effective over-all permeability of the core is reduced. If carried too far 

 the advantages in using the magnetic material in the first place are largely 

 lost. 



The need for high resistivity magnetic materials has been recognized 

 for many years, and, in fact, the naturally occurring magnetic iron 

 oxide, Fe304, or lodestone, is such a material. Its permeability, however, 

 is only about 3 or 4 and its use in coil work has been very restricted. 

 The problem of producing an oxide or mixture of oxides having higher 

 permeabilities was investigated by Philips Gloeilampenfabrieken in Hol- 

 land during the late war and they were successful in producing oxide 

 mixtures, or ferrites, having permeabiUties of 1500 and higher.^ 



After the war. Bell Telephone Laboratories initiated a program of 

 ferrite development, and at the present time ferrite cores as well as 

 inductors and transformers which take advantage of their unique proper- 

 ties are in commercial production. 



The most commonly used ferrites consist of solid solutions of oxides 

 of manganese, zinc and iron, or nickel, zinc and iron. They are prepared 

 by mixing the materials, pressing them into the required shape, and heat 

 treating them under carefully controlled conditions. The resulting ferrite 

 parts are hard and brittle but they can be machined with diamond tools. 



Although ferrites belong in the class of so-called semiconductors 

 their conductivities are of the order of a millionth of those of metals 

 and the eddy current losses are proportionately smaller. There are fre- 

 quency limitations in ferrites but these occur at higher frequencies than 

 are ordinarily of concern in magnetic core inductor work. 



Besides offering the coil designer new possibilities because of low eddy 

 current losses the ferrites have a secondary advantage in that they do not 

 have to be subdivided to keep those losses low. The powdered metallic 

 materials are fragile, are difficult to press except in simple shapes such 

 as rings or cylindrical plugs, and are difficult to machine. As a result the 

 coil designer, for virtually every use, is restricted to the toroidal type 

 coil which, although it has some important advantages, is inherently 

 expensive to wind and adjust. With ferrites, on the other hand, a variety 

 of shapes can be produced. They have ample strength and can be ground 

 and machined. 



The combined magnetic and mechanical advantages of ferrites put 



