Dave wrote:
. . .
The problem with the use of the Thevenin or Norton equivalents is that you
have to exactly respect the limitations in order to use them properly...
that is the part of the circuit being replaced with the 'black box'
equivelent must be linear and time invariant, and the analysis is only valid
for sinusoidal steady state. This last one is what gets everyone, it
eliminates all the transients and makes it impossible to use to figure out
what happens when that first reflection physically happens... you have to
ignore all that stuff and only consider the steady state solution.
. . .

That's not true. A Thevenin or Norton equivalent generator can produce a
voltage or current, respectively, which is any function of time. This of
course includes single pulses and pulsed sinusoids, as well as an
infinite number of others.

Roy Lewallen, W7EL