[Audio Empire]

On Wed, 15 Sep 2010 20:35:41 -0700, Trevor Wilson wrote
(in article ):

"Sebastian Kaliszewski" wrote in message
...
Trevor Wilson wrote:
Audio Empire wrote:
On Sun, 12 Sep 2010 16:41:18 -0700, Trevor Wilson wrote
(in article ):

Audio Empire wrote:

Speakers were a pair of Magnepan MG 3.6s.
**Maggies exhibit a relatively flat impedance curve (close to
resistive) and, therefore, can be expected to be far less demanding
on speaker cable geometry than almost any other speaker. Note the
impedance charateristic (particularly at HF):

http://stereophile.com/floorloudspea...03/index6.html

Here is the impedance curve of a couple of speakers which can be
expected to elicit quite marked differences between different cable
geometries:

http://www.rageaudio.com.au/index.php?p=1_12

IOW: Maggies are almost the worst possible choice for a speaker
cable test (unless, of course, everyone on the planet happens to use
them).



That's irrelevant.

**Incorrect. It is very relevant. Do the math. Use either of the two
curves in my last cite.

But doing the math do it properly, using proper logarithm function (see
below).

**Indeed.



Speaker cable is a CONDUCTOR. It has resistance
(very little) capacitance and inductance which, at audio frequencies,
is of NO consequence for runs less than 50 ft.

**Again. Incorrect. Standard 'zip' type speaker cables (which is 99.999%
of all speaker cables) exhibits inductance figures of approximately
1uH/Metre.

IMHO 0.7 is more realistic.

**My measurements of a dozen of so zip type cables (speaker and mains power)
have revealed figures ranging from around 0.8uH to 1.1uH, with one cable
(really badly designed) measuring 1.5uH. 1uH seems to be a reasonable
average figure IME. Perhaps Australian cables exhibit higher inductance
figures, due to (possibly) larger diameter insulation.


An exotic speaker cable, such as Goertz MI-1 exhibits an inductance of
approximately 0.012uH/Metre. My favourite cable (RG213/U) exhibits an
inductance of around 0.25uH/Metre. Zip type cables (Monster et al) will
be close to the 1uH/Metre, whilst other constructions generally lie
between the extremely low figure of the Goertz and zip cable. One
standout, is Naim cable, which has an extremely high (and undesirable)
inductance figure.

Taking the example of the electrostatic speaker in my second cite, you
will note an impedance of approximately 0.55 Ohms at 16kHz.

Use of broken design speakers noted...

**Nope. Not broken. Just ESLs and inconveniently designed ones, at that.
They were commercially available and quite popular when they were
manufactured.


Those speakers could trip protection circuits in many amplifiers
(especially those with another dip in mid-bass range).

**Sure.



Feeding in your example of a 15 Metre speaker cable run, you will note
(resistive effects will be ignored, though they will be potentially
audible):

Cable L (zip): 15 X 1 X 10^-6 H = 15 X 10^-6 H
Cable L (Goertz): 15 X 0.012 X 10^-6 H = 1.8 X 10^-7 H
Cable L (RG213/U): 15 X 0.25 X 10^-6H = 3.75 X 10^-6 H

Cable XL @ 16kHz (zip): 1.5 Ohms
Cable XL @ 16kHz (Goertz): 0.018 Ohms
Cable XL @ 16kHz (RG213/U): 0.38 Ohms

Plugging those numbers into the above, reveals substantial and
potentially audible effects when using the zip cable and the RG213/U with
the speaker cited.

Attenuation @ 16kHz (zip): -13.1dB
Attenuation @ 16kHz (Goertz): -0.282dB
Attenuation @ 16kHz (RG213/U): -4.6dB

How did you come to those numbers?

**Stupidly, as it happens. My old calculator threw in the towel and I picked
up one from the supermarket. Sadly, I failed to check what I was doing
carefully enough.


My numbers are -5.71dB, -0.14dB & -2.23dB respectively, and for 0.7uH/m
for zip cord (which seems to be more realistic -- e.g. 2 2mm diameter
wires with 3mm insulation between them, thus 5mm axis-axis distance) I
get -4.64dB.

It seems to me you erronously used natural logarithm instead of base 10
logarithm in your calculation (in many colculating programs base 10
logarithm is denoted log10 and natural logaritms is dentoed log which is
contrary to quite popular notation using log and ln respectively).

**What can I say? I miss my old calculator. It served me well for 20 years.



So even with such borderline case of boutique speakers with extremely deep
dips in impedance connected by unusually long for home setup cables (15m)
the effect is on the borderline o being hearable at all (4.5db level diff
at 16KHz is at the edge of ear detedctability for young and healthy ears).

**Not so much. Up until I was in my 30s, I was unable to remain in the same
room as a TV set which had loose core material in the line output
transformer (15,625Hz, in Australia). When in my early 20s, I found myself
in a warehouse and had to leave due to the unbearable noise from the
ultrasonic motion detectors used.


Cut the wire length by half, use speakers from manufacturer who knows how
do design her speakers electrically and the effect is well beyond
detectability.

**Not the point. The claim that I responded to is this:

"Speaker cable is a CONDUCTOR. It has resistance (very
little) capacitance and inductance which, at audio frequencies, is of NO
consequence for runs less than 50 ft."

That is the claim I responded to. I proved that claim to be incorrect. I
alos demonstrated that Maggies are the most speaker cable insensitive
speaker available and, therefore, are an extremely poor choice for speaker
cable testing. THAT is the extent of my claim.



Stereophile, as well as other audio magazines ought to be ashamed of
themselves for perpetuating this speaker cable nonsense.

**Your opinion is duly noted. OTOH, I note that Stereophile, whilst
employing a good deal of 'flowery' language to describe products, go to
quite substantial lengths to publish measured characteristics of products
under test. In fact, I cited one of their tests to show just what a poor
choice Maggies are to delineate differences in speaker cables.

Use any other *properly* designed speakers and effect will be the same.
Even if some typical 4 Ohm speakres have impedance peaks around 20 Ohm or
even 40 Ohm, it doesn't make cables audible, as only exteremely low
impedances could possibly make them so.

**Impedance peaks are of no consequence. It is impedance minima that is
relevant.



I don't care
if they do get advertising money from cable manufacturers. This is a
product area of the business that the audio community can do well
without!

**Actually, it is an area that many people should educate themselves in.
I've just shown that, under certain conditions, there will be very
audible differences in speaker cables.

You didn't show so, as your math has significant bug...

**Acknowledged.


I've also shown why Maggies are a poor choice of loudspeaker to judge
speaker cables with.

They are good choice as any other reasonable speaker as impedance peaks
only diminish cable effect on signal not boost it.

**Again: Impedance minima are the issue.




The bottom line is that speaker cables can be pretty much eliminated as a
source of sound degradation and that for the lion's share of audio systems in
the world, 14 gauge zip cord is audibly indistinguishable from the "boutique"
cables that can cost thousands of dollars for just a a few meters of
elephant-trunk sized wire.

Did you bother to read:

http://www.roger-russel.com/wire/wire.htm

?

Everything that I and others have stated in this thread is bourn-out by Mr.
Russel's impressive missive on speaker cable.