[Chris Hornbeck]

On Thu, 12 Oct 2006 14:51:57 +0200, "Ruud Broens"
wrote:

: This is actually the approximation for a diode, but the
: form for a triode is similar.

actually, it's more general than that, it is applicable to all tubes
under space charge conditions, subject to modification when there are
more influences (such as grids). in a penthode, for instance, the diode
curve comes back for both G1 and G2 voltage vs Ia.

other interesting stuff snipped for bandwidth

Another thing folks who take Terman's introductory college
texts as Gospel is that Child actually is more interesting
in the original (_Phys. Rev._, 32, p.498, 1911 [!} ).

Child includes a term for distance between electrodes that
we always just roll into k because it's outside our control.
But he also specifies certain assumptions:

1) that the electrodes are parallel infinite planes

2) that the plate current is limited only by space charge
(not by emission saturation)

3) that there is no gas within the tube (or at least not
enough to cause retardation of the electrons due to collisions
with gas molecules)

4) that equipotential electrodes are used

5) that the electrons are emitted with zero velocity

6) that the contact potential at the plate is negigible.

Since most of these assumptions are never wholly realized
in actual tubes, the exponent (*always* given without
caveats) generally differs somewhat from 3/2, being
somewhere between 1 and 5/2.

There's a very good description in Eastman _Fundamentals
of Vacuum Tubes_ Third Edition, edited by Terman, 1949,
probably still pretty available, from his famous MIT series.


My thoughts are that interesting real-triode deviations are
largely geometric, ei. from the first assumption above, and
that a realistic triode model must begin in a geometric
rather than an algebraic form.

Much thanks, as always,

Chris Hornbeck