🚀 go-pugleaf

In article <fuNEGVAumei2EwR5@jmwa.demon.co.uk>,
	John Woodgate <jmw@jmwa.demon.co.uk> writes:
> <76b8i0$rqm$1@computer.my.domain>, Duncan Barclay
> <dmlb@ragnet.demon.co.uk> wrote:
>>Sorry, you are wrong twice. Both valves and transistors (BJTs and FETs)
>>are voltage controlled current sources.
> 
> This is no area for dogmatic statements. Not even, perhaps, for that
> one!

But it is true irrespective of network "model".

The fundemental operation of valves, BJTs and FETs is to modulate a
current carrying channel by one (or more) electrical fields.  Seems to
be a VCCS to me.

> A true voltage-controlled device would have infinite input impedance -
> approximately true for valves and FETs at low frequencies.
> 
> A true current-controlled device would have zero imput impedance - not
> true for BJTs.
> 
> A true voltage source would have zero output impedance - only very
> approximately true for earthed-cathode triodes: closer for
> cathode/emitter/source-followers (earthed anode/collector/drain).
> 
> A true current-source would have infinite output impedance -
> approximately true for earthed-cathode/emitter/source pentodes, BJTs and
> FETs.
> 
> Controlled source models of active devices are useful in circuit
> analysis because, although they are approximations, the models can be
> made much less approximate by the addition of a few passive components.
> 
> You can use *any* controlled-source model if you attach the appropriate
> passives for correcting the approximations, but some choices of type
> controlled source are clearly more appropriate than others in a given
> case.

I said nothing about equivalent circuit models and don't have any
problem with the above but it's not relevant.

The collector current of a transistor is govered by the base emitter
voltage, look it any decent text book or reference any SPICE manual,
Gummel-Poon or Ebers-Moll will help altavista.

Ic = Is * (exp (Vbe / (n * Vt)) -1)

Where Vbe is the base emitter voltage, Is is the saturation current of
the device, n is a constant, Vt is kT/q about 26mV at room
temperature.

The base current is due to the reverse leakage current from the
collector-base junction, it is a secondary effect.

For FETs the equation is often quoted as
Id = beta (Vgs-Vt)^2,
this is just an approximation and not true apart from JFETs with a
channel length of more than about 1um in Si or GaAs.

I take it you are actually talking about common-emitter/source
_circuits_ and not transistors.

I do know what I am talking about, Chapter 2 of my D.  Phil thesis
describes in detail GaAs MESFET modelling and I have spent the last
three years designing BJT RFICs.

Transistors and valves are VCCS end of story.

Duncan

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