[sdiy] "Time Winding" in Audio Cables ???

[Magnus Danielson]

From: Glen <mclilith at charter.net>
Subject: [sdiy] "Time Winding" in Audio Cables ???
Date: Mon, 11 Jul 2005 10:51:25 -0400
Message-ID: <4.3.2.7.2.20050711104216.03f37b60 at mail.charter.net>

Glen,

> On a totally separate forum, I have a person who has been to all the 
> Monster Cable(tm) sales seminars telling me that low frequencies travel 
> through an audio cable more slowly than the high frequencies. Is this 
> remotely true?

Yes, ofcourse they travel at different speeds! The question is how big is the
difference and do we care. To the two later questions I say no and no.

> I had just mentioned on this forum that I wouldn't be buying any Monster 
> Cables, because I make my own cables instead, and that I have had good luck 
> with mine so far. Here is part of his response:
> 
> >well the truth is a grew up with a soldering iron in my hand and I have no 
> >doubt either one of us could make a cable that we could swing from. But, 
> >there is a difference, it's called time winding. That's something we can't 
> >do. Think about it, lower frequencies travel slower then higher 
> >frequencies. They are wound so they compensate for that...I can't do that.
> 
> Is this fellow just brainwashed by the Monster seminars he attended? Is he 
> perhaps financially dependent on Monster sales, and wants to tow the "party 
> line"? Or does he actually have a valid point?
> 
> I wouldn't expect any appreciable difference in the propagation speed of 
> audio signals of different frequencies, assuming those frequencies are 
> confined to the same cable. So who is correct here? If there is some sort 
> of difference in propagation speed, I'd like to know more about it. I don't 
> remember hearing of this issue before.

OK, I haven't measured speaker cables in ages, but a few years back these guys
where talking about stuff like skin-effect and how that would effect the top
end of the audio spectrum. Skin effect is a real effect, however, it takes a
whole lot of higher frequency to make any major inpact. So, I built two coax
cables with BNC in one end and XLR in the other. Tying them to my rather
accurate (amplitude, phase, group delay and frequency) HP network analyser I
was ready to make some measure. Normalizing so the effect of the adaptercables
go away and then I dug around in the drawer and came up with the first
microphone cable I could find, an inherited 3 m standard Mogami cable with
Neutrik XLR connectors, nothing fancy but what you can find when digging around
in a drawer. It was dead flat. It *did* loose some in the top end, but we are
talking below 0.1 dB here (I have the real numbers hidden away somewhere) and
the -3 dB point where at.... 46 MHz! Actually, that whole slope was mainly due
to the impedence errors I made when I jumped from 75 Ohm coax to the 110 Ohm
mic cable and back. That creates a standing wave behaviour and a first cancel
just above the -3 dB point.

Also, althought I did not measure the phase explicitly (I can do that
experiment again), the flatness would indicate a rather flat group-delay as
well in my experience.

As to your question about delay, sure there is difference in delay for
different signals, but there are a few things that help to confuse the issue.
In a cable, or indeed any other medium we toss electromagnetic waves thru, you
have a distributed system of inductance, capacitance and losses. These will all
contribute to the speed of light and losses to the frequency dependent delay.
There is math to all that, nothing new, rather just an old school thing.
Further, since the characteristic impedance of the system is not matched in
either end, we *will* have waves going in different directions and contributing
to the resulting phase shift. Infact, the cables is so short that we can
approximate its effects very well with a lumped parameter view (serially
connected inductance and serial loss transistor, shunting capacitance and
parallel loss). Notice that we must include the source impedance of the amp
and the terminating impedance of the speaker to make *any* sense out of this.

Then, just to help making this a complex issue, there is *two* different delays
going on. One is the phase-delay, which is really the phase shift for a
frequence divided by 2 Pi that frequency to get the phase delay in seconds.
This is the delay experienced by a single sine. However, for a modulated sine
or a group of frequencies (the kind of signal you really is going to use) you
will experience another delay, which is the group-delay. To get the group
delay you diffrentiate the phase with respect to the frequency and divide by
2 pi for correct group delay. Thus, this is shift in phase which is of
interest.

I recommend you to look at the phase-plot of normal speakers. They are all over
the place being inductive, resistive and capacitive depending on frequency.
Trying to match that is a hell. Getting a *different* responce by changing the
cable or amplifier is trivial. Getting a *better* responce is another issue
totally. Usually the ear experience *different* as more interesting and thus
it is not hard to accept it as better when it infact is only different. In this
case different can be both better, worse or just differently bad.

Personally I beleive that there is alot of BS going on, and that is part of the
buissness. It's just another variant of the horse or car salesman. It's a
buissness and they need to keep the wheels turning. If people think they get a
better system and is willing to pay for it, who am I to say they are wrong?
However, don't expect me to be overly entusiastic about the "progress". From my
experience in building large PA systems I just get amused/borred/annoyed
(depending on mood) with what they sell stuff with which arguments. I don't get
upset anymore, except when someone just *badly* wants to be right when they
arn't and there is just too much fluff for one to care to clear it up.

Cheers,
Magnus