[Prev][Next][Index][Thread]
Re: STUPID SIMULATIONS
From: Geoffrey Schecht[SMTP:geoffs-at-onr-dot-com]
Sent: Thursday, September 18, 1997 9:07 PM
To: Tesla List
Subject: Re: STUPID SIMULATIONS
Fr. Tom:
I can agree with you on the fact that Electronic Workbench is little more
than an amusing and expensive simulation toy which has schematic capture
capability. PSPICE is somewhat better IMO but any successful simulation
depends upon the quality of the models that you have.
PSPICE handles lossy transmission lines satisfactorally and lumped passives
beautifully but a Tesla coil is a lot closer to a short, loaded antenna or
a form of coupled distributed delay-line system; both of which PSPICE is
pretty clueless about.
I have a Win95/NT compilation of NEC2, which works fine for modelling
general, coupled electromagnetic radiating systems so it would seem that
it should be extensible to Tesla coils. It's incredibly cumbersome to use
and create data input files for, however. Someone would be providing a real
service by writing a piece of code that's needed to generate the large file
needed to describe a Tesla coil.
I've noticed that there are some List members (such as Malcolm and Mark
Rzeszotarski) who are working on the modelling and simulation of TC's; I'd
like to know more about the specifics of the
theoretical treatments, myself. Then again, in many cases you're possibly
just about as well-off with a scope, signal generator, grid-dip meter and a
lot of experience with what everyone's found that works to get the really
big sparks.
Geoff
----------
> From: Tesla List <tesla-at-pupman-dot-com>
> To: 'Tesla List' <tesla-at-poodle.pupman-dot-com>
> Subject: STUPID SIMULATIONS
> Date: Thursday, September 18, 1997 4:07 PM
>
>
> From: Thomas McGahee[SMTP:tom_mcgahee-at-sigmais-dot-com]
> Sent: Thursday, September 18, 1997 11:48 AM
> To: Tesla List
> Subject: STUPID SIMULATIONS
>
>
> >
> >
> > From: richard hull[SMTP:rhull-at-richmond.infi-dot-net]
> > Sent: Thursday, September 18, 1997 3:04 AM
> > To: Tesla List
> > Subject: Re: Potential Transformer
> >
> > At 11:35 PM 9/17/97 -0500, you wrote:
> > >
> > >From: Greg Leyh[SMTP:lod-at-pacbell-dot-net]
> > >Sent: Wednesday, September 17, 1997 5:10 AM
> > >To: Tesla List
> > >Subject: Re: Potential Transformer
> > >
<snipus majoris>
>
> > Well spoken Greg! Kim Goins and I are really hacked at this drift
> away
> > from hands-on in physics and the sciences. Still, it seems the
> wave of the
> > future to simulate instelf of stimulate.
> >
> > Richard Hull, TCBOR
> > >
> > >
>
> Let's see if Richard Hull finds the subject header on this post
> interesting enough to get him to read it! My major problem with many
> so-called simulations is that they DON'T simululate at all. If
> someone with a half-baked understanding of something whips up a
> so-called simulation and then runs this on the computer, it may make
> him feel good because it gives him the answers he wants, but what
> else was it supposed to do, GIVEN THE DATA IT WAS FED???
>
> The true utility of a simulation is that *IF* the model is fairly
> complete and accurate, then running the simulation should give you
> results that closely mimic the real-world behaviour of the system
> upon which the simulation is modelled. This correlation between fact
> and theory is paramount. If your mathematical model gives results
> that do not correspond closely with what we find in the real world,
> then your model is probably flawed. A flawed model cannot produce
> very useful results.
>
> Let's say someone has been experimenting with various models that try
> to mimic the behaviour of a Tesla coil. From the behaviour of actual
> Tesla coils he has formulated a model. If he plugs in the original
> data and gets out the original results, that only means that his
> model works for that particular case. If he now feeds in different
> data and gets results that are verifieable on a wide variety of
> actual Tesla coil systems, then that is a good indication that his
> model is good. If he then uses the model to predict something like,
> say, the optimum H/D ratio for a 6" diameter coil, and then VERIFIES
> that through actual experimentation, then he has a DARN good model.
> But there will be things that his model doesn't predict properly,
> simply because any model begins out simply and then becomes complex
> as the author of the simulation tries to factor in more and more
> things.
>
> I use a simulator for electronic circuits called the Electronics
> Workbench. I can build simulations of circuits that I absolutely know
> work in the real world, and the simulator will give me results that
> are ridiculous. Like oscillators that don't oscillate. The
> Electronics Workbench is nice in that the students can't destroy
> expensive components, but give me a genuine protoboard circuit and
> REAL components ANY day!!!
>
> Fr. Tom McGahee
>
>
>