Don Pearce wrote in message


It acts as a transmission line at every frequency. The question of how
you choose to model it is quite different.


No, you aren't modelling it the way it is, but as an approximate
abstraction - more than good enough to give you the results you want.
That is of course the nature of all models, but some are further
abstracted than others.


I agree, if we are making a standard model from R, L, C, and/or even
RF transmission lines. There is one model type, however, that is as
accurate as the best test equipment. That would be a two-port,
S-parameter, data file. A speaker cable's S-parameters can be measured
by an HP Network Analyzer, (down to the lowest RF frequencies). Then
below RF, Y-parameters can be used down to the lowest audio
frequencies. (The Y-parameters can then be converted to S-parameters.)
Now, you have an S-parameter, two-port data file, that goes from audio
to RF. A circuit analysis program can use this file to simulate the
speaker cable. The two-port data file is treated, as a *component*.
The output of the program will be exactly the same as the data the
network analyzer measured.

The nice thing about this two-port data component model is: it will
also work in low frequency AC circuit analysis programs, with *any
source and any load impedance*. Using a voltage source and a typical
*loudspeaker load*, it will give the response of that cable into that
load, even though the speaker cable was originally measured using
equal source and load impedance.

Bob Stanton