one for the theorist

[Haible Juergen]

	>H field is never apart from E field, unles it is DC (well, all this
	>audio stuff is actually DC anyway...). Change in H-field makes
	>E-field and vice versa - this way we have radio waves.
	>E and H are more or less like voltage and current in Ohm law,
	>and air impedance (vacum too) is about 377 ohms, belive it or not.

Far field. Another approximation that can be very useful (;->).
The 377 Ohms can lead to remarkable errors when applied to the near field.

	>Another point - wire without return path
	>Well, for me it is antenna. Such wire has certain impedance with
	>relevance to ground, and it depends on frequency. Might be real
	>(resistance) or complex, inductive or capacitive depending on ratio
	>of frequency (or actually a wave length) to wire length.

Sure. And my point was that these are two different reigns that must
not be mixed up. 

Either (1) you're way below the wavelength with your
structures: Then the concept of "inductance" makes sense, and
you also need a return path (as Martin pointed out). Depending on the shape
of your structures (open or closed loop) you will have a dominating
capacitive
effect (E-field) or a dominating inductive effect (H-field). Sure you will
have a H field caused by a current, caused by a voltage or E-Field (if there
is
some conductivity in your circuit). But the *direct* gereration of H-field
by the mere *change* of E-field (even without conductivity) can be
neglected in that case.

Or (2) your structures are large enough for the coupling of E and H field
not to be neglected anymore. Then you have "HF radiation", and you're in the
reign of antennas, and not isolated capacitors and inductors anymore.
This is where you don't need something like a "return path" anymore,
and, as you said, a piece of wire, or a pair of wires, can behave as
"inductance"
or "capacitance" depending on frequency.

In a practical sense, the whole concept of "inductance" is that you have
a more or less constant factor that describes the frequency dependency
of some impedance: It's more convenient to write "L = 1mH" rather
than "Z = j*2*pi *f * mOhm/Hz", just because the word inductance implies
a linear dependency from f. If you have slight aberrations from this
linear law, it can still be convenient to draw a courve for L rather than Z,
but you see where this will have a natural end.

In the "HF domain" ((2) from above) you can still talk about "inductive"
or "capacitive" *at a certain frequency*, but this is just another way to
express that your phase is positive or negative at that frequency. Doesn't
have much to do with the behaviour of an inductance at lof frequencies.

JH.