On 20/02/2012 11:14, Martin Brown wrote:

It wouldn't surprise me if some power amplifiers reacted to having coax
wire used to connect to speakers by oscillating wildly.

Highly unlikely - the capacitance won't be that much higher than with
any other type of 2-wire line. Most amplifiers are stable with load
capacitances at least a couple of orders of magnitude more than
presented by a typical speaker cable run .

You really want the equivalent of about 1mm^2 cross section of copper
(or more) in each wire of a decent speaker cable intended to preserve
bass response.

Possibly OTT. The only thing that matters is the total (loop)
resistance in the speaker lead, but it doesn't have to be absurdly low,
only low compared to the resistance of the speaker's voice coil - since
that is what dominates in the 'damping circuit'. In practice there will
be no audible or measurable improvement from reducing the cable's
resistance below about 5% of the nominal impedance (and 10% is quite
good enough if you're not super-critical). Hence the heftiness of cable
required is entirely dependent on the length of run.

Basic figure of 8 speaker cable using annealed copper is fine.
You don't need stuff hand spun by mermaids in an oxygen free atmosphere.

We'd all agree on that, I hope:
http://www.douglas-self.com/ampins/f...ailproj.htm#p8

--
Andy