[Ian Iveson]

Chris Hornbeck wrote

...so Ia=k*Vak^3/2, with k some geometrically determined
: constant.
:

better formulated as: so, by induction, we know Ia is proportional
to Vak^3/2,
the multiplier k being dependent on geometry and cathode emission

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

Please note that these are approximations. Put another way,
k is not constant, and not linear. It's the k part
that's interesting.


Perveance is largely dependent on heater power, for a given type of
cathode, AFAIR from discussions long ago.

From a practical point of view, I have asked a few times about how I
might take valve ageing into account in my (Duncan Munro's, as nearly
all are) models. One might assume that any decent model should need to
know what kind of life the valve has suffered?

For the moment, I just vary perveance.

It would be nice to sensibly discuss Duncan's most recent model, and
especially to transfer its grid model to his pentode. I wonder to what
extent reducing heater current has the same effect as ageing?

Unfortunately Duncan has gone very quiet, and has withdrawn the
documents describing the derivation of his models.

No-one has even tried, AFAIK, to capture the lumpiness of typical real
valve characteristics...the unevenness that, for any particular valve,
may result in sweet spots. I haven't seen a tetrode kink either,
AFAIR.

All the extra bits added to Child-Langmuir by the more sophisticated
SPICE models are just curve manipulation and bodging. I can't find
much relating them to the physical reality, except approximately
similar behaviour.

As for delay, I suspect that the kind of periods involved in
propogating electron flows in a valve are too short for SPICE to
contemplate. Very few amateurs have access to an alternative modeling
environment, so if it's not SPICE its not much use.

Where we are now, for all practical purposes, is likely to be the end
of the story.

cheers, Ian