From: Paul
Date: Fri, 22 Dec 2000 11:27:57 +0000
Subject: Re: [TSSP] E-Tesla6.11
I wrote: > >This might make more sense now in the context of pn2511. The relevant > >equ is 8.4, although you'll disagree with paragraph which follows, > >since you've found a way to estimate Ces without knowing omega in > >advance! Terry wrote: > Karl Gauss figured that out in the second to the last century, not me. I > just followed his directions... :-))) Gausss' law is one of those things > that always works despite the fine details. Scarry to think of him coming > up with such a leap in electrostatic thought without ever having had a > Tesla coil :-)) Yes, the point is not the use of Gauss, but the fact that you have worked around the statement in the para following 8.4 by your willingness to assume in advance a V profile. This I used to think was doubtful but as said before I've changed my view on that point. This promises to be successful since although the inflection point may shuffle up and down according to h/d, overall I think the normalised V, ie V(x)/Vtop promises to be reliably guessable. Integrating the flux further smooths out any variation, and I'll bet that you can be justified in claiming 1% or so accuracy on Ces estimates once we establish just how normalised V varies with h/d, size, turns, and topload. > >Your Lfac is Les/Ldc and I'm looking forward to generating a database > >of Lfac values when we run the simulator across the range of secondary > >designs. > > Wow! that would be nice!!! E-tesla6 wants them numbers!! This is the other half of the problem. I'll bet that it varies smoothly enough to be a good candidate for interpolation between reference points established by tsim and by measurements. There is also a neat little thing that you can do with E-Tesla6 that will give you Cee for almost no extra effort. Shall I send you the code? Given Ces, Les (from Ldc via Lfac), and Cee, you then have all you need to employ many of the formulae in pn2511. > E-Tesla6 has a lot of hidden things imbedded in it's more basic > functionality of finding Fo for the practical coiler. And you are justified in using them. Such factors in a program like tsim would not be appropriate - they would be 'fiddle' factors and would hinder the purpose of tsim which is to explore the physics. In E-Tesla6 they are essential, and they could better be described as 'calibration' factors. Cheers, -- Paul Nicholson, Manchester, UK. --
Maintainer Paul Nicholson, paul@abelian.demon.co.uk.