DESIGN OF SMALL IRON-CORED CHOKES AND TRANSFORMERS 



diameter, b the length and c the thickness of winding, all dimensions in inches, 

 then the self-inductance in microhenries is approximately given by^ 



0-2aV 

 ^ ~ 3a + 96 + 10c 



DESIGN OF SMALL IRON-CORED CHOKES 

 AND TRANSFORMERS 



This is a highly elaborate subject, and a specialist field in itself. The ability 

 to design efficient small transformers for power, signal and pulse applications, 

 comes to most people only after many years of experience ; readers who feel 

 drawn to the subject may care to look at an excellent book by Macfadyen^. 

 A complete account of the subject cannot possibly be included here; never- 

 theless the author proposes to outhne a procedure, trial-and-error rather 

 than synthetic, by which the reader should be able to make himself small 

 transformers and chokes which, though perhaps not the best that can be 

 done, should prove sufficient for purposes such as LC oscillators and filters. 

 The method is to use an interesting magnetic material 'Ferroxcube', made 

 by the MuUard Company. Unlike most transformer cores, which are metals 

 and are alloys of nickel and iron, this material is a crystalline substance of 

 the form MFe204 where M is a divalent metal. Here we are interested in a 

 grade of Ferroxcube called A4 having the formula MnFe204. The advantages 

 of this material are: 



(1) Extremely low losses permit use up to frequencies of at least half a 

 megacycle, much higher than are feasible with the nickel irons. Ferroxcube 

 is less efficient than the latter at very low frequencies because the saturation 

 flux density is lower; nevertheless, one can reckon to use this single material 

 successfully at any frequency of electrobiological interest, so one is relieved 

 of the need for selecting one's magnetic medium. 



(2) The manufacturers make a range of 1 1 moulded E and 3 / members 

 in Ferroxcube A4 with which it is possible to make up 14 different sizes of 

 core, either by using 2E 's or an E and an /; one of these is likely to suit 

 one's requirements. They also can supply winding bobbins to go with them. 

 Assembly is therefore merely a question of winding the bobbin, slipping the 

 core round it and cementing the whole together. This saves the labour of 

 making ones own bobbin and assembUng a core from laminations, as is 

 usually necessary when using nickel iron cores. {Plate 22.1.) 



(3) The information provided by Mullard Ltd on these cores and the fact 

 that 'iron losses' are usually negligible make designs very straightforward. 



We have to distinguish two cases, according to whether or not the trans- 

 former or choke cores operate with d.c. polarization; we consider the 

 unpolarized case first. 



Ferroxcube transformers and chokes, without d.c. polarization — Chokes 

 {Inductors) — It is required to produce a choke of inductance L which will 

 have a resistance not greater than /?max and at a minimum frequency F^m 

 will operate satisfactorily with terminal alternating R.M.S. voltages up to 



First, choose a core: at first one simply has to make a guess here. It is 

 not likely that this is the core one will finally use, but no matter — with 



309 



