Re: topload questions

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Paul Nicholson by way of Terry Fritz <teslalist-at-qwest-dot-net>" <paul-at-abelian.demon.co.uk>

In the output from fantc, the values shows as (exact => .... pF)
apply to the demo cases, and have been pre-computed to an accuracy
of 5 places after the decimal, using Legendre functions.  This is
intended as a cross-check on fantc arithmetic.

For a 16" x 7" toroid, this gives 19.21259 pF, and for 16" x 0.001"
we get 6.040 pF, which I can only manage to three places.

Godfrey's approximate formula appears to do very well.

Antonio wrote:
 > Fantc takes forever to analyze the 16" x 0.001" case

The generalised minimum residuals algorithm which fantc uses to
obtain the charge distribution is probably struggling to converge
when faced with the dense bundle of surface filaments packed into a
0.001" diameter, which will slow it right down.  Also, the accuracy
will be very poor, because the calcs which generate the potential
matrix are not well suited to this geometry.  If the C of thin wires
is required, we might have to add another component type to the list,
ie: toroids, spheres, discs, cylinders, ... and wires.

  Size "          Fantc pF     Legendre pF
  16 x 7           19.14        19.21259
  16 x 1           13.94        13.9560
  16 x 0.1         10.18         9.875
  16 x 0.001        9.3          6.04

Clearly, fantc is breaking down with conductors less than about 0.1"
diameter.
--
Paul Nicholson
--