Re: Magi initial design

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

Original poster: Bart Anderson <classi6-at-classictesla-dot-com> 

Hi Antonio,

Tesla list wrote:

>Original poster: "Antonio Carlos M. de Queiroz" <acmq-at-compuland-dot-com.br>
>Tesla list wrote:
> >
> > Original poster: Bart Anderson <classi6-at-classictesla-dot-com>
>
> > The programs don't assume any particular percentage, they
> > actually compute the effective capacitance. In Javatc, the topload
> > capacitance identified is what it is. The value of Ces is where the topload
> > and everything else is pulled in and the work is performed to compute the
> > effective system capacitance.
>
>I imagine that the calculation departs from the energy stored in the
>capacitance of the combination. But this depends on the voltage
>profile along the coil. Does the program calculate the profile, or
>assume a certain profile?
>
>Antonio Carlos M. de Queiroz

The charge distribution along the coil is computed, however, the profile is 
fixed. The profiles listed below as pulled from Paul's write up for Geotc:
http://www.abelian.demon.co.uk/tssp/geotc/

G.PROFILE_LOADED. Use a voltage profile typical of a loaded coil.
G.PROFILE_BARE. Use a voltage profile typical of an unloaded coil.
G.PROFILE_DC. Use a uniform voltage along the coil.
G.PROFILE_LINEAR. Use a linearly rising voltage along the coil.

With Fantc, these profiles are selected by the user.
With Javatc, these profiles are determined by  the hd and if loaded or not. 
For example, short h/d coils (< 1.5 ) would take on a linear profile.

It should also be noted that there are instances where a user could install 
a very small sphere or topload object in comparison to the coil. In this 
case, accuracy would fall as the coil would be run under a loaded profile 
because the topload is installed, but would have been better to be run 
under a bare profile. I have thought about letting the program make a 
determination in that area, but have yet to approach on what criteria would 
determine this.

Take care,
Bart