From: Bert Hickman
Date: Tue, 19 Dec 2000 02:10:23 -0600
Subject: Re: [TSSP] Proximity effect and Terman
Paul, Thanks for the kind words! Skin effect is a result of the changing magnetic fields that are created when an AC current flows through an isolated conductor. The resulting EM field changes the current distribution in the wire in a direction that always tends to minimize the total inductance of the current path. In the case of an isolated conductor, this means that current is forced to flow preferentially through an outer layer, called the skin depth, of the conductor. Skin depth is only a function of frequency and material properties of the conducting medium. Proximity effect is similar to skin effect in that it's also a magnetic effect. However, with proximity effect, the redistribution of current flow in a given conductor is further constrained by external magnetic fields coming from OTHER conductors that are in close proximity. Skin effect and proximity effect may be of approximately the same order of magnitude in single layer solenoids. Paradoxically, increasing conductor diameter may actually worsen the ratio of Rac/Rdc through proximity effect. And, losses in multilayer coils will tend to be predominated by proximity effect, obviously making proximity effect a major concern for high frequency transformer designers. Proximity effect increases with frequency, and of particular concern for pulse or PWM applications. Terman's analysis is similar to those of Medhurst and Grover - a series of tables are used to provide the appropriate "fudge factors" for the particular coil geometry, wire diameter, and spacing. For single and multi-layer coils, this is covered in Terman, "Radio Engineers Handbook, 1943, pp73 - 83. Much of the information used by Terman was actually based upon earlier work by S. Butterworth reported in 1925 and 1926 (Wireless Eng. and Exp. Wireless, and Proceedings of Royal Society). If there's enough interest, I could scan the portion of Terman's book and make these pages available to the List. I don't have Butterworth's original material. There are also some well written modern discussions of proximity effect. One of the best ones is on a site for a CAD tool (called PROXY): http://kosys.home.mindspring.com/articleproxy/articleproxy.html The above web site references the work of some modern day workers, P. L. Dowell and Bruce Carsten. Dowell developed some closed form equations, subject to constraints which conveniently appear to apply to single layer resonators(!). It would be valuable to obtain copies of these papers: 1. P. L. Dowell, "Effects of Eddy Currents in Transformer Windings," Proc. IEE, Vol 113 No. 8, August, 1966 2. Bruce Carsten, "High Frequency Conductor Losses in Switchmode Magnetics", High Frequency Power Conversion Conference Proceedings, CA, May 1986 Hope this helped! -- Bert -- -- Bert Hickman Stoneridge Engineering Email: bert.hickman@aquila.net Web Site: http://www.teslamania.com Paul wrote: > > In a recent thread on the main tesla list concerning base > input impedance, Bert Hickman commented: > > (Referring to Terman's "Radio Engineers Handbook, McGraw-Hill, 1943.) > > > ... > > Terman's proximity effect analysis assumes lumped behavior - that > > the same current flows through all portions of the resonator. We also > > know that the resonator base current is ultimately balanced by the > > displacement currents "flowing" through the distributed capacitances > > of the resonator and topload - and that these currents are not the > > same in all portions of the resonator. This makes Terman's approach > > of estimating proximity effect of limited use in base-driven > > resonators. > > We are fortunate that Bert recently joined this project list. > > Bert, > > Welcome to the tssp list! From reading your comments in the tesla list, > it sounds like you might be able to shed some light on what is for a me > a mystery - the origin of the proximity effect and how to estimate it. > I'm not familiar with Terman's tome, so maybe you could advise whether > the analysis given there for the case of uniform current could be made > to apply to the non-uniform case, given that we can now calculate the > current distribution of the secondary with reasonable confidence? > > Any comments you can make which might help me understand the mechanism > involved in the proximity effect would be most appreciated. While I > have several books which go into skin effect, none do more than mention > the existence of proximity effects in closewound inductors. > > At present, the tssp simulator makes use of Medhurst's tables to > estimate the effective series resistance, and the resulting input > impedances and Q factors compare poorly with measurement, often > being 50% in error, sometimes more. It would be very nice to find > a more reliable approximation. > > Cheers, > -- > Paul Nicholson, > Manchester, UK. > -- -- Bert Hickman Stoneridge Engineering Email: bert.hickman@aquila.net Web Site: http://www.teslamania.com
Maintainer Paul Nicholson, paul@abelian.demon.co.uk.