On Wed, 10 Feb 2010 21:52:56 -0800, ScottW wrote
(in article <(E-Mail Removed)>):

> On Feb 9, 7:03=A0pm, Audio Empire <audio_emp...@comcast.net> wrote:
>> On Tue, 9 Feb 2010 16:33:01 -0800, m...@stevet.me.uk wrote
>> (in article <7ted5tF9m...@mid.individual.net>):
>>
>>
>>
>>
>>
>>> Have been looking at some really good interconnects for my system;
>>
>>> MF KW550 Amp, DM25 transport and dac and Keff reference 205 system -
>>> Need RCA and XLR - speaker cables are Chord signiture
>>
>>> Looked at Van De Hul, Chord and some stuff on Russ Andrews. Any
>>> opinions?
>>
>>> Lot of stuff about Nordost but the money is quite significant.
>>
>>> It seems that silver is the best so would be looking at that , but not
>>> sure about the science.
>>
>>> PS also found some people call Pure Reference but no heard of them
>>
>>> Thanks StevieT
>>
>> The science is simple: wire is wire at audio frequencies. A piece of
>> coat-hanger or a pair of =A0$4000/meter Nordost Valhalla interconnects, t=
> hey
>> both will sound identical. Cable is audio jewelry; IOW, bling, and nothin=
> g
>> else. Do the math. It will tell you all need to know about the three
>> components of any conductor: DC resistance/foot, capacitive reactance and
>> inductive reactance. The effect of these components on an audio signal (D=
> C -
>> 50KHz) over any practical length? Negligible, unimportant, a non-event.
>>
>
> DC resistance in the ground path is a source of hum. I strapped all
> the chasis together in my system in a star configuration and I believe
> it made for a subtle improvement even though I had no audible hum at
> the listening position.

I too use star configuration grounding. I do it this way: My preamp is the
central ground point. I use only quasi-balanced interconnects between
components. That's where the shield carries no current and is only tied to
the return "barrel" of the RCA connector at one end. Inside the shield are
two conductors: one is the hot and goes to the RCA's pin on both ends and the
other goes to the barrel on both ends. These two conductors are hot and
return and shield is just a static shield. By making sure that all shield
ends terminate at the pre-amp, irrespective of whether the cable is being
used as an input or an output. This means that I have extended the common
chassis ground out the entire length of all shielded interconnects. It
definitely results in a "blacker" background to the music.
>
> This is worth studying.
>
> http://www.jensen-transformers.com/an/an004.pdf
>
> "The 99 dB figure can be improved further by lowering the shield
> resistance of the 5 foot cable which uses a foil shielded cable with
> #24 gauge drain wire (25 mohm per foot). Cable using a #18 gauge
> equivalent braided copper shield (6.5 mohm per foot) will lower hum
> level by 5 dB from 99 dB to 104 dB."
>
> ScottW
>
>

On Thu, 11 Feb 2010 08:09:21 -0800, Frank wrote
(in article <(E-Mail Removed)>):

> Dick Pierce wrote:
>> Frank wrote:
>>> I agree about speaker cables but ... "bad" audio
>>> low signal cables do exist and have high capacity.
>>> Try putting 2 or 4 nF across your rca connector
>>> terminals and hear the difference please.
>>
>> Let's see. 2 to 4 nF of capacitance... Well, assume
>> the load impedance presented at the receiving end of
>> the chain is, oh, 10 kOhms, and assume the sending end
>> of the chain has a source of, let's say, 1 kOhms, the
>> result is an additional 0.23 dB down at 10 kHz with
>> 4 nF shunt capacitance. The response would be 3 dB
>> down at about 43 kHz.
>>
>> Maybe you'd hear it, maybe not. Likely not.
>>
>> Now, I'd bet you could hear it if you hung a 1000 uF cap
>> across the RCA jack as well. So what?
>>
>> It's hard to find coaxial cable with more than about
>> 50 pF/ft of shunt capacitance.
>>
>> That means to achieve a 4nF shunt capacitance , you'd
>> need, let's see, 4nF/50pf/ft ...
>>
>> That's a piece of interconnect cable 80 feet long. With
>> more typical figures in the, oh 25 pF/ft realm, that's
>> a 160 ft long interconnect.
>>
>> The notion that a 4 nF shunt capacitance is at all a
>> reasonable test for demonstrating cable effects is,
>> well, absurd.
>>
>> Let's look, instead, at a 3-foot piece of cable with
>> a shunt capacitance of 30 pF/ft, for a total of 90
>> pF shunt capacitance.
>>
>> You'd end up with a system that isn't 3 dB down until
>> well past 1.5 MHz.
>>
>> You'd have to come up with some pretty wierd cables
>> to end up with the level of concern you express here.
>>
>>> Those 2000 $/meter cables are rubbish and surely sound
>>> different from a good real low capacity double shielded
>>> solid "video" cable.
>>
>> And your reason for this is?
>>
>
> My ears. Young and working. Never been to a disco.

Good for you. I envy you your young ears. Still and all, such tiny losses
(less than a half a dB) at the top extreme of human hearing is unlikely to be
heard in such a way that it would be noticed by even a 12-year old girl.
>
> Would you like to share the formulas you use to obtain that result and
> what happens if loads are higher?

I did that in another post. You should be able to see them on your news
reader now. They've shown up on mine.
>
> I've seen cables with a 350-400 pF capacitance per meter (circa 3 feet?)
> marked low c.

I'm sorry Frank, but that seems pretty unlikely to me. Are you MEASURING
these cables? Can I ask you what your methodology is? About the highest
capacitance that coax gets is 50 pF/foot, but one would be fairly incompetent
to specify such a cable for audio interconnects, when plain-jane average coax
of the type commonly used for audio interconnects is about 30 pF/foot and
relatively inexpensive low capacitance cable is available at 20 pF/ft.

I've used a lot of different cables. I've settled with Blue Jean Audio
for straight up speaker wire and banana plugs, and monoprice for
everything else. The products both companies sell are inexpensive
(relatively) and the build quality is excellent. I would like to know
who their supplier is so I can buy direct from them.

On Wed, 10 Feb 2010 05:15:49 -0800, Frank wrote
(in article <(E-Mail Removed)>):

> Il 10/02/2010 4.03, Audio Empire ha scritto:
>> On Tue, 9 Feb 2010 16:33:01 -0800, (E-Mail Removed) wrote
>> (in article<(E-Mail Removed)>):
>>
>>> Have been looking at some really good interconnects for my system;
>>>
>>> MF KW550 Amp, DM25 transport and dac and Keff reference 205 system -
>>> Need RCA and XLR - speaker cables are Chord signiture
>>>
>>> Looked at Van De Hul, Chord and some stuff on Russ Andrews. Any
>>> opinions?
>>>
>>> Lot of stuff about Nordost but the money is quite significant.
>>>
>>> It seems that silver is the best so would be looking at that , but not
>>> sure about the science.
>>>
>>> PS also found some people call Pure Reference but no heard of them
>>>
>>> Thanks StevieT
>>>
>>
>> The science is simple: wire is wire at audio frequencies. A piece of
>> coat-hanger or a pair of $4000/meter Nordost Valhalla interconnects, they
>> both will sound identical. Cable is audio jewelry; IOW, bling, and nothing
>> else. Do the math. It will tell you all need to know about the three
>> components of any conductor: DC resistance/foot, capacitive reactance and
>> inductive reactance. The effect of these components on an audio signal (DC -
>> 50KHz) over any practical length? Negligible, unimportant, a non-event.
>>
>> Add that to the fact that NO double-blind test (and there have been lots of
>> them) has EVER been able to detect any difference between an expensive
>> speaker cable and a length of zip cord. No double-blind test has EVER been
>> able to detect any difference between the "sound" of a set of Radio Shack
>> throw-away audio interconnects and any expensive interconnect that you care
>> to name. Spend your money on better speakers and forget this cable nonsense.
>>
>
> I agree about speaker cables but ... "bad" audio low signal cables do
> exist and have high capacity.

Yet, the cheapest Radio Shack cables do not have significantly high
capacitance, and even if they (or anyone else's interconnects, for that
matter) DID have high capacitance, it would have to be impossibly high for it
to have any affect whatsoever on an audio signal at the typical half and one
meter lengths used in most audio systems to connect between components.

> Try putting 2 or 4 nF across your rca connector terminals and hear the
> difference please.

4 NANOFARADS (10 to the minus 9)? Are you serious? You've actually SEEN
cables with that high of a capacitance? Are you aware that this would give an
X sub C of almost 2000 Ohms at 20 Khz (assuming I didn't get any decimal
points in the wrong place)? I question whether it's possible to make a
reasonable length of coaxial cable with that high of a capacitance. Now,
maybe a cable 100 meters long could have that kind of capacitance, but 1/2
and 1 meter interconnects? I'm dubious. What brands might these be?

The highest capacitance per foot coax that I've ever heard of is 50 pF/ft (50
X10 to the minus 15) with most coaxial cable types being in the neighborhood
of 30 pF/ft and real low capacitance being around 20 pF/ft. Average coax
would make a 1-meter run approximately 100 pF total. That's several orders
of magnitude less than your 2-4 nF!

> Those 2000 $/meter cables are rubbish and surely sound different from a
> good real low capacity double shielded solid "video" cable.

You mean that they surely sound NO different from a good, real low capacity
etc.?

Robert Peirce wrote:
> The longer the run the more likely you may encounter resistance,
> capacitance and reluctance issues, but it would take a really long run.

The only reluctance you will find in long cables is the
reluctance I and other hasve in believing that reluctance
is a relevant property of interconnects. :-)

Reluctance is a property of magnetic materials and circuits.
It's not a relevant factor in the conduction of eletrical
signals in cables.

That being said, I assume you meant "inductance" and not
"reluctance."

--
+--------------------------------+
+ Dick Pierce |
+ Professional Audio Development |
+--------------------------------+

Frank wrote:
> Dick Pierce wrote:
>>> Try putting 2 or 4 nF across your rca connector
>>> terminals and hear the difference please.
>>
>> Let's see. 2 to 4 nF of capacitance...
>>
>> It's hard to find coaxial cable with more than about
>> 50 pF/ft of shunt capacitance.
>>
>> That means to achieve a 4nF shunt capacitance , you'd
>> need, let's see, 4nF/50pf/ft ...
>>
>>That's a piece of interconnect cable 80 feet long. With
>>more typical figures in the, oh 25 pF/ft realm, that's
>>a 160 ft long interconnect.
>>
>>The notion that a 4 nF shunt capacitance is at all a
>>reasonable test for demonstrating cable effects is,
>>well, absurd.
>>
>> ...
>>
>>You'd have to come up with some pretty wierd cables
>>to end up with the level of concern you express here.
>>
>>> Those 2000 $/meter cables are rubbish and surely sound
>>> different from a good real low capacity double shielded
>>> solid "video" cable.
>>
>>And your reason for this is?
>>
>
> My ears. Young and working. Never been to a disco.

I'm sure you detected what you perceived to be a difference.
However, considering the vanishgly small difference in
actual response, that such differences are smaller than
small positional changes in listening AND that unless
you've gone through extraordinary measures to allow direct
and time-proximal comparisons to overcome the inevitable
failings in fine-detail auditory memory, then the
best we can do is suggest the conclusions are anecdotal
and unverified.

> Would you like to share the formulas you use to obtain
> that result

The results were obtained using a straightforward
SPICE simulation. The model for your 4 nF case looks like:

*
Vin 1 0 AC SIN 1.0 0.0
RSrc 1 2 1K
Ccable 3 0 4NF
RLoad 3 0 10K
.AC OCT 3 20 2MEG
.PRINT AC VDB(2,0)
.END

Running this, you'll find that there's a 0.8 dB loss
at low frequencies, due to the attenuative effect of the
series source and load resistance. Now for the more
realistic case:

*
Vin 1 0 AC SIN 1.0 0.0
RSrc 1 2 1K
Ccable 3 0 90PF
RLoad 3 0 10K
.AC OCT 3 20 2MEG
.PRINT AC VDB(2,0)
.END

Now, here's a side-by-side comparison of the results,
where they are of any real significance, i.e at audible
frequencies. What happens above that is of no significance

Frequency 90pF 4000pf diff
--------- ------ ------ ------
1000Hz -0.8dB -0.8dB 0.0dB
1260 -0.8 -0.8 0.0
1590 -0.8 -0.8 0.0
2000 -0.8 -0.8 0.0
2500 -0.8 -0.8 0.0
3150 -0.8 -0.9 0.0
4000 -0.8 -0.8 0.0
5000 -0.8 -0.8 -0.1
6300 -0.8 -0.9 -0.1
8000 -0.8 -1.0 -0.2
10000 -0.8 -1.0 -0.2
12600 -0.8 -1.2 -0.4
15900 -0.8 -1.3 -0.5
20000 -0.8 -1.7 -0.9
25000 -0.8 -2.1 -1.3
31500 -0.8 -2.7 -0.9
40000 -0.8 -3.5 -2.7
50000 -0.8 -4.5 -3.7
63000 -0.8 -5.8 -5.0
80000 -0.8 -7.3 -6.5
100000 -0.8 -8.9 -8.1

Now, you might argue that the difference of 0.9 dB
would be audible, except there are two good reasons
why I would bet good money they are not:

1. In a fair comparison, with trained listeners in
carefully controlled time-prximal maximum-sensitivity
comparison, these sorts of differences are tough,
at best, to discern reliably (I don't care HOW old
you are and what you have or have not listened to),

and, more importantly:

2. Your 4nF capacitance figure is, at its very best,
utterly unrealistic, thus the represent an absurd
condition. Why is an absurd, unrealistic condition
a reaosnable point of comaprison

> what happens if loads are higher?

What happens if you use more realistic figures for cable
capacitance? After you deal with that, then you get to
ask about load impedance variations.

> I've seen cables with a 350-400 pF capacitance per meter
> (circa 3 feet?)

But that's a LOT different, by a full order of magnitude,
then the rather absurd claim of 4 nF. In such a case, the
differences are WELL below 0.1 dB out beyond 64 kHz.

> marked low c.

I would challenge you to actually name any cable intended
for use as interconnects between audio components that
exhibits a 4nF shunt capacitance.

--
+--------------------------------+
+ Dick Pierce |
+ Professional Audio Development |
+--------------------------------+

In article <(E-Mail Removed)>,
Dick Pierce <(E-Mail Removed)> wrote:

> Robert Peirce wrote:
> > The longer the run the more likely you may encounter resistance,
> > capacitance and reluctance issues, but it would take a really long run.
>
> The only reluctance you will find in long cables is the
> reluctance I and other hasve in believing that reluctance
> is a relevant property of interconnects. :-)
>
> Reluctance is a property of magnetic materials and circuits.
> It's not a relevant factor in the conduction of eletrical
> signals in cables.
>
> That being said, I assume you meant "inductance" and not
> "reluctance."

You are correct.

Audio Empire wrote:
> The highest capacitance per foot coax that I've
> ever heard of is 50 pF/ft (50x10 to the minus 15)

Uh, no. pico is 10^-12, NOT 10^-15. BIG difference.
The prefix for 10^15 is femto.

More completely:

deci 10^-1
centi 10^-2
milli 10^-3
micro 10^-6
nano 10^-9
pico 10^-12
femto 10^-15
atto 10^-18
zepto 10^-21
yocto 10^-24

--
+--------------------------------+
+ Dick Pierce |
+ Professional Audio Development |
+--------------------------------+

On Thu, 11 Feb 2010 17:20:31 -0800, Walt wrote
(in article <(E-Mail Removed)>):

> Audio Empire wrote:
>> The bottom line is that if you can hear, using controlled double-blind
>> testing, any difference between two interconnects or speaker cables it's
>> because one of the manufacturers has purposely designed the cable to
>> attenuate (that is, take away from) some portion of the audio spectrum and
>> this is NOT anything that one should want.
>
> I mostly agree with what you have said, and am no fan of "magic cables".
> But one thing is worth correcting:
>
> For speaker cables, long runs of high-gauge cables can be audible. Say,
> 100 feet of 20 gauge speaker wire. This gives a series resistance of
> about 2 ohms, which is close enough to the load impedance (nominally 8
> ohms, but many speakers have a dip at certain frequencies, sometimes as
> low as 2 ohms) to create audible effects.

Well of course. As I've stated several times here over the last several days,
long runs will start to have deleterious effects on high frequency signals.

> This, of course, depends on the impedance plot of the speaker.
>
> Given the capatance per foot and the resistance per foot of the wire,
> the output impedance of the amp, and the impedance plot of the speaker
> one can calculate the frequency response of the circuit. For most
> reasonable wire at reasonable lengths, cable effects are negligable.
> And for line-level signals that are normally terminated with a high
> impedance (10k ohms is typical) the response at audio frequencies
> doesn't change until you reach multiple hundreds of feet.
>
> So, yes, you can hear the difference between cables, but you need a
> *lot* of cable!

AS I and several other have pointed out several times.
>
> //Walt

Walt wrote:
> I mostly agree with what you have said, and am no fan of "magic cables".
> But one thing is worth correcting:
>
> For speaker cables, long runs of high-gauge cables can be audible. Say,
> 100 feet of 20 gauge speaker wire. This gives a series resistance of
> about 2 ohms, which is close enough to the load impedance (nominally 8
> ohms, but many speakers have a dip at certain frequencies, sometimes as
> low as 2 ohms) to create audible effects.

This is a high-end audio discussion group. I surmise this
because the group is called "rec.audio.high-end."

Assuming my assumption is correct, let me ask a question:

Can you imagine ANY "high-end" audio system using 100 feet
of 20 gauge wire for connecting a speaker system to an
amplifier?

Anyone?

Okay, since no reasonable response jjustifying this scenario
is likely to be forthcoming, then why not carry it to its
extreme absurdity.

Let's, instead of using 100 feet of 20 gauge wire, let's
use 100,000 feet of 36 gauge wire. Gee, that ends up with
a resistance of 82,800 ohms! Surely that would have a
noticeable audible effect, woudln't it?

And so what?

What on earth is the relevance in actual scenarios of
hundreds of feet of 20 gauge wire or 4 nanofarad capacitors
on actual situations?

> Given the capatance per foot and the resistance per foot of the wire,
> the output impedance of the amp, and the impedance plot of the speaker
> one can calculate the frequency response of the circuit. For most
> reasonable wire at reasonable lengths, cable effects are negligable.

Yes, so what does going to UNreasonable and even absurd
conditions have to do with anything?

> And
> for line-level signals that are normally terminated with a high
> impedance (10k ohms is typical) the response at audio frequencies
> doesn't change until you reach multiple hundreds of feet.
>
> So, yes, you can hear the difference between cables, but you need a
> *lot* of cable!

Yes, as would putting your speakers on the mmon and hooking them
to your amp on Jupiter would very likely demonstrate.

--
+--------------------------------+
+ Dick Pierce |
+ Professional Audio Development |
+--------------------------------+