Roy Lewallen wrote:
John Popelish wrote:

No, you're misinterpreting what you're seeing. Imagine an LC L network
with theoretically lumped series L and shunt C.

Okay, I am imagining an idealized, network made of perfect, impossible
components that is simple to analyze. Got it.

If you look at the
currents at the input and output of the perfect inductor, you'll find
that they're exactly the same.

Right.

If, however, you look at the currents in
and out of the *network* you'll see that they're different, because of
current going to ground through the C.

Got it. Same for any pi, T, or more complicated LC network.

And, as I said before, you can
even pretend it's a transmission line and measure forward and reverse
traveling waves and a standing wave ratio.

Yes. Under some specific conditions.

But with zero length, there
can be no standing waves inside the inductor.

Yes. There are no waves in a single ideal lumped component, so there
can be no waves inside any of them, only a phase shift between the
voltage across them and the current through them. But a network made
of them can mimic lots of processes that internally involve
propagation of waves, including the phase shift between voltages
across the terminals and current into the terminals, and even group
delay, but only over narrow frequency range. It is a model with this
severe limitation.

Yet the terminal
characteristics of the network are the same as a transmission line. You
don't need to imagine standing waves residing inside the inductor in the
LC circuit, and you don't need to imagine them inside the inductor in
Cecil's model, either.
(snip)

Whether or not we need to imagine them to picture what is happening at
the terminals is not the question at hand. The question in my mind is
what is the actual mechanism, inside the device in question that is
causing the effects we see at the terminals. I am not interested in
the full range of models that predict the effect, but in the actual
cause. I accept that my motivation is not necessarily the same as yours.