Skin effect (was:Primary and copper)

From:  Jim Lux [SMTP:jimlux-at-earthlink-dot-net]
Sent:  Friday, April 03, 1998 1:01 PM
To:  Tesla List
Subject:  Re: Skin effect (was:Primary and copper)

Tesla List wrote:
> ----------
> From:  Antonio C. M. de Queiroz [SMTP:acmq-at-compuland-dot-com.br]
> Sent:  Thursday, April 02, 1998 10:35 PM
> To:  Tesla List
> Subject:  Re: Skin effect (was:Primary and copper)
> Jim Lux wrote:
> > For a close wound coil, I suspect that you could, at some level, treat it
> > as a long cylinder of copper with the current flowing around the cylinder.
> > In this case, the adjacent turns "shields" the conductor of interest,
> > reducing the effective cross sectional area.
> This makes sense. Probably with an "effective angle" in the conducting
> tube at the wire surface where there is current. This angle would vary,
> being large when the skin depth is small and small when the skin depth
> is large. I will verify with an electromagnetic simulator, if I find one
> that can compute this (I have used the Qfield simulator for some problems,
> with reasonable results).

I just downloaded the "student" demo version yesterday, but haven't started running it yet.

> > FWIW, has any one tried modelling a tesla coil using something like NEC to
> > do a complete EM simulation?
> The complete simulation of a coil with hundreds of turns is something
> rather complex.

That is, generating the model would be complex. I was thinking about writing a little program that would 
generate the wire segments to approximate the helical turns as polygons. A highly repetitive system like 
this might cause numerical problems in the simulation code. I just found my copy of NEC2 to try this with. 
As soon as I finish with the manual and figuring out how to define the model, I'll start it.

> The FDTD simulator in the APLAC circuit simulator has a
> command to define a thin helical conductor that may be usable in Tesla
> coil simulations (I haven't tried yet, and expect problems with a large
> coil).Does it have analytical models for helices built in, or does it generate some approximation?

Surely someone has done some simulation work with the helices on TWT's, for instance. However, a typical 
helical antenna doesn't have anywhere near the number of turns we are talking about.