In article , Glenn Richards
wrote:
APR wrote:

All you are doing with bi-wiring is placing the terminal that joins
the the LF and HF driver circuit inputs at the amp terminals instead
of the speaker terminals. If the speaker wire is so inadequate that
you are able to measure any significant changes in HF linearity due to
the effect of speaker wire resistance then the wire run is too long
or the wire is too light.

Ok, for those people that still don't "get it", here's an experiment
that will demonstrate visually what's going on:

Take a 12V DC power supply (one of those bench PSUs capable of supplying
several amps will suffice) and a 12V 1W bulb (any type will do, it's
easier to perform the experiment using a MES bulb and batten holder with
screw terminals though).

[snip details]

This is exactly what is happening with your speaker cables. The cable
run acts like a series resistor, and the load (in this case the driver
voice coil) causes a voltage drop across that series resistor.

The point to bear in mind, though, is that the cable resistance for any
sensible speaker cables should be tiny. Hence any effect should
be negligably small. If it isn't simply use thicker/shorter loudspeaker
cables. The following URL may help assess this.

http://www.st-and.demon.co.uk/AudioM.../lscables.html

For cable lengths of a few metres, the cables which cost the order of
50p per foot are likely to be fine, but the choice is up to you. :-)


Bybi-wiring you are avoiding the load from the LF driver causing a
voltage drop to the HF driver. (Again this is a greatly simplified
description of what's going on.)

To assess this, have a look at the other webpages I have been mentioning.
If you look at the simple example it should be clear that this implies that
any such effect in practice should tend to be so small as to be either
inaudible or smaller than moving your head slightly. If it isn't then
simply using a lower resistance cable should deal with the issue - unless
of course you *want* to alter the frequency response by the deliberate
introduction of 'high' cable resistance.

However if you do want this, then it would be easier and more controllable
to use a single cable, but feed the speaker units via some series resistors
at the speaker end of the cable. This method is familiar to speaker
designers. It is called "modifying the crossover/correction networks". :-)

You may find that a couple of low-value resistors are cheaper and more
convenient that more runs of cable. You can also change their values easily
if you want to fiddle about with the frequency response of your loudspeaker
system.

I've specified 13-strand zip wire in the experiment above as it makes
the results more obvious, but the effect will still stand with thicker
cable, it'll just be harder to see. And eventually you'll get to a
thickness of cable that will have a low enough series resistance that
the effect will no longer occur - but in the case of speakers that cable
will either be so expensive or so unwieldy that it makes more sense to
use two thinner runs of cable.

FWIW I tend to prefer the Maplin cables as they give low series resistance
for just a few quid per cable. The above URL uses them for example, so
people can decide for themselves what might suit for the lengths they need.

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