[Prev][Next][Index][Thread]

Re: Using skin effect



Subject:       Re: Using skin effect
       Date:   Mon, 28 Apr 1997 06:59:01 -0700
       From:   Bert Hickman <bert.hickman-at-aquila-dot-com>
Organization:  Stoneridge Engineering
         To:   Tesla List <tesla-at-pupman-dot-com>
 References: 
            1


Tesla List wrote:
> 
> Subject:  Re: Using skin effect
>   Date:   Sun, 27 Apr 1997 15:48:15 +0500
>   From:   "Alfred A. Skrocki" <alfred.skrocki-at-cybernetworking-dot-com>
>     To:   Tesla List <tesla-at-pupman-dot-com>
> 
> On Tue, 22 Apr 1997 08:03:26 +1200 Malcolm Watts
> 
> > HI Alfred, all,
> >                 One skin depth in copper is 66/SQRT(f) mm.
> 
> I had checked my notes and found the equation 2.6/SQRT{f} inches.
> Which when multiplied by 2.54 to convert to mm. yields 66.04/SQRT{f}
> mm. which compares favorable with your equation.
> 
> > The definition of a skin depth is that distance inwards to the point
> > where current has dropped to 1/e of its total value. The equation
> > above has resistivity factored in so will be different for silver.
> 
> Hmmm, in my notes the equation is listed as being independent of
> materials or resistance.

> 
> > In fact, the skin depth on silver is not as deep due to the higher
> > conductivity. If a silver film is not thick enough I have heard
> > things could actually be worse than bare copper.
> 

Alfred,

It turns out it is a function of the resistivity AND the magnetic
permeability (u) of the conductor. The latter is usually ignored, since
u for silver or copper are approximately unity. A somewhat more exact
form of this equation is:

   x = Pi*D*SQRT(2*u*F/p)
 
where:  D = Diameter in cm
        u = permeability of the conductor
        F = Frequency (Hertz)
        p = specific resistivity of conductor, ohm-cm^-3 
        x = penetration depth (cm)

Plugging in 1.724 X10^-6 ohm-cm^-3 for Cu and changing the units results
in the shorthand form of the equation that Malcolm presented and you
confirmed. Since silver has has greater p than copper, x is less. 

<SNIP>

-- Bert --