🚀 go-pugleaf

In article <d1dLzCAjKri2EwAr@jmwa.demon.co.uk>,
	John Woodgate <jmw@jmwa.demon.co.uk> writes:
>
> The enquirer is clearly a beginner, who needs the maximum assistance to
> understand the topics that he will be exposed to. It is very likely
> that, if he is not frustrated by over-rigorous expositions, he will soon
> come across circuit analysis software, such as SPICE, which uses
> controlled-source models. It will be no help to him to have been told
> that the *only* valid controlled-source model is the VCCS, even though
> that corresponds to the physical model of controlled conduction that you
> advocate. 
> 
> The exclusive adoption of the VCCS model for valves, bipolars and FETs
> would actually obscure the essential differences between them, at least
> up to undergraduate level.

This thread started with someone (Nightjar) asking how they worked. 
Out of four replies I followed up on one because the answer was wrong,
the other three were right.  The orignal thread was not asking how
they were modelled.

> I can't see any good reason for not mentioning that the configuration of
> the device when in use affects the optimum choice of controlled-source
> model.

I'm sorry my use of the term VCCS has led to this pointless arguing that
BJTs etc. are not "ideal" VCCS. I didn't say anything about modeling,
I just said that all of valves, FETs and BJTs are physically
device where current is controlled by a voltage, a vccs for short.

Equivalent circuit modelling, use the best for the job. s-parameter,
hybrid pi whatever. I don't personally use the CCCS model too often as
it doesn't hekp in most of the circuits I design (apart from current
mirrors).

Duncan

-- 
________________________________________________________________________
Duncan Barclay          | God smiles upon the little children,
dmlb@ragnet.demon.co.uk | the alcoholics, and the permanently stoned.
________________________________________________________________________