[sdiy] equations up

[Magnus Danielson]

From: "metasonix at earthlink.net" <metasonix at earthlink.net>
Subject: Re: [sdiy] equations up
Date: Mon, 16 Jun 2003 01:49:23 -0700

> Ian Fritz wrote:
> > OK, show us your space-charge calculations. We know you have 'em.  :-)
> 
> Actually it seems that the 3/2 power law (usually called the
> Langmuir or perveance law) may be incomplete. I know a
> very smart guy who is writing SPICE models for triodes, and he
> claims that you need a whole Taylor series of terms to cover
> even the simplest triode. There are drift effects, electron
> interactions, fringing effects, etc....

Can I assume that he can't use the 3/2 power law directly or? Then he surely
needs a fairly long Taylor expansion.

> But of course people are arguing about this, most physicists
> regard Langmuir as a god, and the 3/2 equation gets you within 1% in
> nearly all cases; and I hate it when smart guys squabble over arcane
> physical rules, so I won't bring it up anymore. Boring.

Besides that I would pass Langmuir as a possible Whiskey (I am not really into
that) it has been known for long that this is not all there is to it.
Secondary emissions and stuff like that adds to the complexity. The 3/2 law is
surely an approximation, but just like the most basic Ebbers-Moll model is just
a very simple model of real life. The point here is, we keep making models. A
model is always a simplification of the real world and the closer the model is
the harder it gets to use it. Thus, you always have to ask yourself just how
closely you depend on your models correctness for your analysis and simulation
to make accurate predictions.

Actually, the digital gate as a model is very succsessfull these days in
predicting the behaviour. However, it still remains just a model and there is
several things which is wrong with it. It's all analog anyway.

Cheers,
Magnus