Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!seismo!ll-xn!mit-eddie!apollo!rees
From: rees@apollo.uucp (Jim Rees)
Newsgroups: rec.audio,sci.physics,sci.electronics
Subject: Re: Mercury Filled Speaker Wire
Message-ID: <3746c216.b8ab@apollo.uucp>
Date: Mon, 14-Sep-87 14:03:00 EDT
Article-I.D.: apollo.3746c216.b8ab
Posted: Mon Sep 14 14:03:00 1987
Date-Received: Tue, 15-Sep-87 06:21:14 EDT
References: <3816@watdcsu.waterloo.edu> <578@uthub.toronto.edu> <1700@crash.CTS.COM>
Distribution: rec
Organization: Apollo Computer, Chelmsford, Mass.
Lines: 51
Xref: mnetor rec.audio:3228 sci.physics:2163 sci.electronics:1292


  And from Dick Olsher:

  This is a key point.  When modeling signal transfer in audio cables, it
  is not only necessary to take into account the variable AC resistance of
  the cable, but its variable inductance as well...

I decided to do an experiment.  I connected an audio sig gen to my
ultra-cheap stereo amp, put the output of the amp on the H input of
my scope and through 100 feet of cheap Radio Shack speaker cable,
and put the output of the cable on a 8 ohm dummy load and the V input
of my scope.  This should show me the phase change of the signal
through the cable.

At very high frequencies (100KHz) I could detect some phase shift.
I was using about 100 feet of cable, so the shift should be about
4 degrees at this frequency.  I was unable to determine whether the
shift varied non-linearly with frequency (which seems to be what
the claim is; am I right?)

Olsher also says,

  The shunt capacitance
  of the cable (but not the capacitive reactance) is fixed by the geometry
  of the cable and the dielectric properties of the insulation used.

I was trying to calculate the electrical properties of this particular
speaker cable, and discovered that vinyl apparently has a dielectric
constant that varies over a wide range with frequency.  It has close
to a 10% change over the audible range (which for me is about 20-25K).
This should result in about a 5% change in shunt capacitance.  It seems
to me that this is a greater effect than skin effect, at least at these
frequencies.

No matter what numbers I plug in, I can't get any of these effects to
account for more than a .05 db change in signal level or a .1 db
additional noise due to phase distortion.

Here's an idea for audiophile speaker cable.  Only vacuum has a
dielectric constant completely independent of frequency.  Vacuum
filled speaker cable!  I envision something semi-rigid and metal-
clad, sort of like heliax.

If you worry about these things, I think the practical solution is
to put the amplifier physically close to the speaker.  Then you only
need to get a low level signal, at reasonable impedances, to the
amplifier.  And you don't need a ton of copper to carry those watts
at 8 ohms.  You could use fiber optics to connect the signal source
to the amplifier.

Olsher must have done some experiments.  What were his results?