In article , Mike Gilmour
wrote:

"Spanner Dangle" wrote in message
...

Going back to my college days I was taught that electrical current
travelled on the outside of a conductor. I assume that in multi strand
cables the current travels along the outside of each strand
(conductor). Exactly how I don't know as thay all touch each other.

So I have always assumed that if you solder a multi strand cable then
at that point the strands are physically bonded and hence you would
get a higher resistance.


A function of frequency though. Its called skin effect and if I also
remember my college theory correctly its caused by the self-inductance
of the conductor which causes an increase in the inductive reactance
with increasing frequencies thus making electrons move towards the
surface. I'm sure someone will correct me if I've got it wrong ;-) Mike

I'd say you'd got it about spot-on. :-) The currents act as if they tend
to prefer to flow near the metal surfaces. Terms like "internal impedance"
are sometimes used to describe the effect. Can be seen as a consequence of
the velocity of EM fields inside a metal being low, and frequency
dependent. In effect, the metal has a very large 'refractive index' which
is frequency dependent. Also quite lossy (i.e. resistance) but since most
of the field is outside the metal, this may not mean a high overall loss.

The situation with multiple strands in contact is complicated by the way
they touch or are near each other, but the basic idea is the same. The
effect of soldering bundles together is more complex still for various
reasons - e.g. below.

Curiously, if a metal has a lower bulk conductivity (i.e. more resistive)
the skin depth at a given frequency is greater. Solder tends to have a
difference conductivity to copper.

The result of all this is twofold.

1) A *precise* analysis of something as simply as normal wiring can be a
nightmare to get right. As soon as the system lacks simply symmetry, etc,
the simple approaches tend to break down if you look at them too hard.

2) In practice (1) rarely matters as the effects at audio frequencies are
usually so low as to not be worth worrying about when you come the the
actual performance. ;-

Hence these things are of interest to crazy academic who like to
underdstand (and explain) things in mind-numbing detail, but usually come
around to "Having spent ages and made my head bleed understanding it, the
result is, you find things generally work fine".

My own impression from various fields of engineering and science is that it
seems to often be the case that systems and devices work best when their
working can be modelled reasonably well by the simpler methods. The more
complex the model you require, the more that implies there are extra ways
things can go pear-shaped. :-)

Slainte,

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.st-and.demon.co.uk/AudioMisc/index.html
Armstrong Audio http://www.st-and.demon.co.uk/Audio/armstrong.html
Barbirolli Soc. http://www.st-and.demon.co.uk/JBSoc/JBSoc.html