Cable nonsense -- article #1
John Dunlavy
Having read some of the recent comments on rec.audio.opinion and high
end, concerning "audible" differences between interconnect and
loudspeaker cables, I could not resist adding some thoughts about the
subject as a concerned engineer possessing credible credentials.
To begin, several companies design and manufacture loudspeaker and
interconnect cables which they proudly claim possess optimized
electrical properties for the audiophile applications
intended. However, accurate measurements of several popularly selling
cables reveal significant differences that call into question the
technical goals of their designer. These differences also question
the capability of the companies to perform accurate measurements of
important cable performance properties. For example, any company not
possessing a precision C-L-R bridge, a Vector Impedance Meter, a
Network Analyzer, a precision waveform and impulse generator, wideband
precision oscilloscopes, etc., probably needs to purchase them if they
are truly serious about designing audio cables that provide premium
performance.
The measurable properties of loudspeaker cables that are important to
their performance include characteristic impedance (series inductance
and parallel capacitance per unit length), loss resistance (including
additional resistance due to skin-effect losses versus frequency),
dielectric losses versus frequency (loss tangent, etc.),
velocity-of-propagation factor, overall loss versus frequency into
different impedance loads, etc.
Measurable properties of interconnect cables include all of the above,
with the addition of those properties of the dielectric material that
contribute to "microphonic noise" in the presence of ambient
vibration, noise, etc. (in combination with a "D.C. off-set" created
by a pre-amp output circuit, etc.).
While competent cable manufacturers should be aware of these
measurements and the need to make them during the design of their
cables, the raw truth is that most do not! Proof of this can be found
in the absurd buzzard-salve, snake-oil and meaningless advertising
claims found in almost all magazine ads and product literature for
audiophile cables. Perhaps worse, very few of the expensive, high-tech
appearing cables we have measured appear to have been designed in
accordance with the well-known laws and principles taught by proper
physics and engineering disciplines. (Where are the costly Government
Consumer Protection people who are supposed to protect innocent
members of the public by identifying and policing questionable
performance claims, misleading specifications, etc.?) --- Caveat
Emptor!
For example, claiming that copper wire is "directional", that
slow-moving electrons create distortion as they haphazardly carry the
signal along a wire, that cables store and release energy as signals
propagate along them, that a "final energy component" (improperly
labeled as "Joules") is the measure of the tonality of cables, ad
nauseum, are but a few of the non-entities used in advertisements to
describe "cable performance".
Another pet peeve of mine is the concept of a "special configuration"
included with a loudspeaker cable which is advertised as being able to
"terminate the cable" in a matter intended to deliver more accurate
tonality, better imaging, lower "noise", etc. The real truth is that
this "special configuration" contains nothing more than a simple,
inexpensive network intended to prevent poorly-designed amplifiers,
with a too-high slew-rate (obtained at the expense of instability
caused by too much inverse-feedback) from oscillating when connected
to a loudspeaker through a low-loss, low-impedance cable. When this
"box" appears at the loudspeaker-end of a cable, it seldom contains
nothing more than a "Zobel network", which is usually a "series
resistor-capacitor" network, connector in parallel with the wires of
the cable. If it is at the amplifier-end of the cable, it is probably
either a "parallel resistor-inductor" network, connected in series
with the cable conductors (or a simple cylindrical ferrite sleeve
covering both conductors). But the proper place for such a network,
if it is needed to "insure amplifier stability and prevent
high-frequency oscillations", is within the amplifier - not along the
loudspeaker cable. Hmmm!
Having said all this, are there really any significant "audible"
differences between most cables that can be consistently identified by
experienced listeners? The answer is simple: very seldom! Those who
claim otherwise do not fully grasp the power of the old
"Placebo-Effect" - which is very alive and well among even the most
well-intentioned listeners. The placebo-effect renders "audible
signatures" easy to detect and describe - if the listener knows which
cable is being heard. But, take away this knowledge during blind or
double-blind listening comparisons and the differences either
disappear completely or hover close to the level of random guessing.
Speaking as a competent professional engineer, designer and
manufacturer, nothing would please me and my company's staff more than
being able to design a cable which consistently yielded a positive
score during blind listening comparisons against other cables. But it
only rarely happens - if we wish to be honest!
Oh yes, we have heard of golden-eared audiophiles who claim to be able
to consistently identify "huge, audible differences" between cables.
But when these experts have visited our facility and were put to the
test under carefully-controlled conditions, they invariably failed to
yield a score any better than "chance". For example, when led to
believe that three popular cables were being compared, varying in size
from a high-quality 12 AWG ZIP-CORD to a "high-tech looking" cable
with a diameter exceeding an inch, the largest and sexiest looking
cable always scored best - even though the CABLES WERE NEVER CHANGED
and they listened to the ZIP Cord the entire time.
Sorry, but I do not buy the claims of those who say they can always
audibly identify differences between cables, even when the comparisons
are properly controlled to ensure that the identity of the cable being
heard is not known by the listener. We have accomplished too many
"true blind comparisons" with listeners possessing the right
credentials, including impeccable hearing attributes, to know that
"real, audible differences" seldom exist - if the comparisons are
properly implemented to eliminate other causes such as system
interactions with cables, etc.
Indeed, during these "comparisons" (without changing cables), some
listeners were able to describe in great detail the "big differences"
they thought they heard in bass, high-end detail, etc. (Of course, the
participants were never told the "NAUGHTY TRUTH", lest they become an
enemy for life!)
So why does a reputable company like DAL engage in the design and
manufacture of audiophile loudspeaker cables and interconnects? The
answer is simple: Since significant measurable differences do exist
and because well-known and understood transmission line theory defines
optimum relationships between such parameters as cable impedance and
the impedance of the load (loudspeaker), the capacitance of an
interconnect and the input impedance of the following stage, why not
design cables that at least satisfy what theory has to teach? And,
since transmission line theory is universally applied, quite
successfully, in the design of cables intended for TV, microwave,
telephone, and other critical applications requiring peak performance,
etc., why not use it in designing cables intended for critical
audiophile applications? Hmmm! To say, as some do, that there are
factors involved that competent engineers and scientists have yet to
identify is utter nonsense and a cover-up for what should be called
"pure snake oil and buzzard salve" - in short, pure "fraud". If any
cable manufacturer, writer, technician, etc. can identify such an
audible design parameter that cannot be measured using available lab
equipment or be described by known theory, I can guarantee a
nomination for a "Nobel Prize".
Anyway, I just had to share some of my favorite Hmmm's, regarding
cable myths and seemingly fraudulent claims, with audiophiles on the
net who may lack the technical expertise to separate fact from fiction
with regard to cable performance. I also welcome comments from those
who may have other opinions or who may know of something I might have
missed or misunderstood regarding cable design, theory or secret
criteria used by competitors to achieve performance that cannot be
measured or identified by conventional means. Lets all try to get to
the bottom of this mess by open, informed and objective inquiry.
I sincerely believe the time has come for concerned audiophiles, true
engineers, competent physicists, academics, mag editors, etc. to take
a firm stand regarding much of this disturbing new trend in the
blatantly false claims frequently found in cable advertising. If we
fail to do so, reputable designers, engineers, manufacturers, magazine
editors and product reviewers may find their reputation tarnished
beyond repair among those of the audiophile community we are supposed
to serve.
Best Regards,
John Dunlavy
Pekka Aro
Hello John,
To me it is exceedingly obvious that you are referring to MIT cables
and Spectral amplifiers. Am I correct?
I was really surprised to read your opinion on cable issues since I
have visited two of your dealers who demonstrate SC-V's with the new
$11K or so (!) MIT 850 Shotgun EVO speaker cable and claim that
without these cables the full performance potential of the SC-V's
will not be realized. One went as far as to claim that the SC-V with
the $11K MIT cable was "much better" than the SC-VI with the DAL
cable! Furthermore, one of the dealers used Spectral amplifiers to
demo the SC-V's.
I do not doubt the validity of your opinions, but something seems
amiss here. If you are correct, your dealers in this case are either
incompetent hacks or dishonest rip-off artists. In either case, why
would you allow dealers like that represent your product? Perhaps
this is news to you...
In any case, you certainly design fine sounding and reasonably priced
speakers.
All the best,
Pekka L. Aro
John Dunlavy wrote:
[quoted text deleted -- deb]
> Another pet peeve of mine is the concept of a "special configuration"
> included with a loudspeaker cable which is advertised as being able to
> "terminate the cable" in a matter intended to deliver more accurate
> tonality, better imaging, lower "noise", etc. The real truth is that
> this "special configuration" contains nothing more than a simple,
> inexpensive network intended to prevent poorly-designed amplifiers,
> with a too-high slew-rate (obtained at the expense of instability
> caused by too much inverse-feedback) from oscillating when connected
> to a loudspeaker through a low-loss, low-impedance cable. When this
> "box" appears at the loudspeaker-end of a cable, it seldom contains
> nothing more than a "Zobel network", which is usually a "series
> resistor-capacitor" network, connector in parallel with the wires of
> the cable. If it is at the amplifier-end of the cable, it is probably
> either a "parallel resistor-inductor" network, connected in series
> with the cable conductors (or a simple cylindrical ferrite sleeve
> covering both conductors). But the proper place for such a network,
> if it is needed to "insure amplifier stability and prevent
> high-frequency oscillations", is within the amplifier - not along the
> loudspeaker cable. Hmmm!
------------------------------------------------------------------------------
From ra...@graphics.cornell.edu Sat Nov 15 13:28:43 1997
Posted-Date: Sat, 15 Nov 1997 20:28:22 +0100 (MET)
To: rec-audio...@uunet.uu.net
Path: news
From: Jan Didden <jan-d...@pi.net>
Newsgroups: rec.audio.high-end
Subject: Re: DACs - apparent lack of information
Date: Sat, 15 Nov 1997 21:19:54 +0100
Organization: Sound Advice
Lines: 41
Message-Id: <346E03...@pi.net>
References: <64fueb$h...@eyrie.graphics.cornell.edu> <64g4sc$1...@merckx.graphics.cornell.edu> <64i460$j...@eyrie.graphics.cornell.edu> <64iisd$l...@merckx.graphics.cornell.edu>
Nntp-Posting-Host: 145.220.202.22
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 2.01E-PI-32 (Win95; I)
X-Lines: 40
Status: RO
Content-Type: text/plain; charset="us-ascii"
Content-Length: 2101
Trey Jackson wrote:
>
> Stewart Pinkerton wrote:
> > co...@iofr.com (uh OH! ) writes:
> > >[Trey wrote:]
> > >>[snip]
> Regarding connecting a DAC into a cheap CD player, I asked a salesman
> at the local high-end audio store about that. He responded with
> something about the wow and flutter, which didn't quite make sense to
> me. In my mind, a bit stream is a bit stream is a bit stream.
>
> Do not CD players with digital out just produce a bit stream?
> Isn't the bit stream coming from my $200 player just as good as the
> bitstream coming from a $1000 player? It's hard for me to imagine a
> $200 CD player getting that many errors in the bitstream.
>
> TJ
Well, a bit stream is a bit stream is a bit stream except that:
- depending on the error correction in the
laser/reading/decoding system, your bit stream may have missing
chunks (muted) or sometimes repeating previous chunks, but that has
nothing to do with the 'quality' of the bit stream when most of the
time there are no errors or the error-correction takes care of it
correctly (no pun intended). This system may be better in a more
expensive deck, but I doubt that it makes a significant difference.
How often do you hear muted sound during CD replay or other gross
errors even in a cheap portable player? And if you don't notice it,
it means that the error-correction has taken care of it.
- the bit stream may have jitter, which is something you can
take care off either in the transport or in the (outboard) DAC. You
must do it in the outboard DAC if the transmission from deck to DAC
introduces jitter.
In fact, there is only one technical reason to separate deck and DAC,
and that is to avoid possible noise from the deck electronics to
enter the DAC circuits. Sensible layout and design standards should
take care of that, making the combo less expensive than a 2-box set.
Unfortunately, that's not how Madison Avenue works.
Jan Didden
'Minds are like parachutes. They work best when open'
[excessive sig deleted -- deb]
Pekka Aro
Hello John,
To me it is exceedingly obvious that you are referring to MIT cables
and Spectral amplifiers. Am I correct?
I was really surprised to read your opinion on cable issues since I
have visited two of your dealers who demonstrate SC-V's with the new
$11K or so (!) MIT 850 Shotgun EVO speaker cable and claim that
without these cables the full performance potential of the SC-V's
will not be realized. One went as far as to claim that the SC-V with
the $11K MIT cable was "much better" than the SC-VI with the DAL
cable! Furthermore, one of the dealers used Spectral amplifiers to
demo the SC-V's.
I do not doubt the validity of your opinions, but something seems
amiss here. If you are correct, your dealers in this case are either
incompetent hacks or dishonest rip-off artists. In either case, why
would you allow dealers like that represent your product? Perhaps
this is news to you...
In any case, you certainly design fine sounding and reasonably priced
speakers.
All the best,
Pekka L. Aro
John Dunlavy wrote:
[quoted text deleted -- deb]
> Another pet peeve of mine is the concept of a "special configuration"
ShLampen
In article <64a986$3...@eyrie.graphics.cornell.edu>, John Dunlavy
<10236...@compuserve.com> writes:
>The measurable properties of loudspeaker cables that are important to
>their performance include characteristic impedance (series inductance
>and parallel capacitance per unit length), loss resistance (including
>additional resistance due to skin-effect losses versus frequency),
>dielectric losses versus frequency (loss tangent, etc.),
>velocity-of-propagation factor, overall loss versus frequency into
>different impedance loads, etc.
Since the "characteristic impedance" of any cable does not stabilize
before 1 MHz, and sometimes higher, how can this be of any effect at
audio frequencies? And, since impedance is a wavelength-dependant
effect, how can impedance be of any effect at audio wavelengths? The
wavelength of 20 kHz is 15,000 meters. Your speaker cable or
interconnect would have to be at least one-quarter wavelength before
the impedance had any effect on signal propagation. ( 1/4 of 15,000
meters = 3750 meters = 2.33 miles).
The skin effect of a 20 AWG conductor at 100 kHz is THE ENTIRE
CONDUCTOR, so how could "skin effect" have any effect? At 20 kHz, the
skin effect of a 10 AWG speaker cable (for example) is greater than
the width of the conductor.
Velocity of propagation (Vp) of a rea bad PVC is around 50%. Top of
the line gas-injected foam polyethylene or Teflon is around 84%. At
20 kHz, what difference does this make in signal propagation? Much
less than 1 degree of phase difference regardless of choice of
plastic.
Of MEASURABLE effect, there is no doubt that resistance and
capacitance can make measurable and audible differences. Inductance
(a tiny effect in cable) might be considered a borderline effect.
>Since significant measurable differences do exist and because
>well-known and understood transmission line theory defines optimum
>relationships between such parameters as cable impedance and the
>impedance of the load (loudspeaker), the capacitance of an
>interconnect and the input impedance of the following stage, why not
>design cables that at least satisfy what theory has to teach? And,
>since transmission line theory is universally applied, quite
>successfully, in the design of cables intended for TV, microwave,
>telephone, and other critical applications requiring peak
>performance, etc., why not use it in designing cables intended for
>critical audiophile applications? Hmmm!
I would humbly suggest that we at Belden have been doing precisely
this since 1902!
--
Steve Lampen Technology Development Manager Belden Wire & Cable Co.
www.belden.com
My new book, "Wire, Cable, and Fiber Optics for Video and Audio Engineers"
has just been published by McGraw-Hill.
Stewart Pinkerton
Pekka Aro <pek...@ix.netcom.com> writes:
>Hello John,
>To me it is exceedingly obvious that you are referring to MIT cables
>and Spectral amplifiers. Am I correct?
>I was really surprised to read your opinion on cable issues since I
>have visited two of your dealers who demonstrate SC-V's with the new
>$11K or so (!) MIT 850 Shotgun EVO speaker cable and claim that
>without these cables the full performance potential of the SC-V's
>will not be realized. One went as far as to claim that the SC-V with
>the $11K MIT cable was "much better" than the SC-VI with the DAL
>cable! Furthermore, one of the dealers used Spectral amplifiers to
>demo the SC-V's.
Perhaps this is indeed the case if the obviously dubious Spectral amps
are used. A clear solution would be the purchase of stable Krell
amplification, and Rat Shack cable......................
>I do not doubt the validity of your opinions, but something seems
>amiss here. If you are correct, your dealers in this case are either
>incompetent hacks or dishonest rip-off artists. In either case, why
>would you allow dealers like that represent your product? Perhaps
>this is news to you...
If this is not a troll, why would you not be aware that the typical
high-end dealer is indeed an incompetent hack and rip-off artist. Do
you not read r.a.o?
--
Stewart Pinkerton | Music is art, audio is engineering
ASP Consulting |
(44) 1509 880112 |
AWRigby
From: Pekka Aro <pek...@ix.netcom.com>:
Thank you for your response to John's posting, and your kind words regarding
our products.
<<I was really surprised to read your opinion on cable issues since I
have visited two of your dealers who demonstrate SC-V's with the new
$11K or so (!) MIT 850 Shotgun EVO speaker cable and claim that
without these cables the full performance potential of the SC-V's
will not be realized. One went as far as to claim that the SC-V with
the $11K MIT cable was "much better" than the SC-VI with the DAL
cable! Furthermore, one of the dealers used Spectral amplifiers to
demo the SC-V's.>>
One must consider that regardless of the results of the numerous
tests and measurements performed at DAL, the final determining
factor in judging sound quality is the individual listener.
While we personally feel that the SC-V is the second best loudspeaker
in the world, we find it hard to imagine that any type of
loudspeaker cable (regardless of price) could immediately elevate
its overall sonic performance to better that provided by our
flagship SC-VI. However, the dealer may have honestly felt, in
his/her own opinion, that this was indeed the case.
It is very difficult to argue or find fault with a persons concept of
"good sound" or "bettersound", no matter how much objective data is
supplied with a particular product. (And this is why we perform
both critical listening evaluations as well as extensive bench tests
for our products.)
With regards to the Spectral amplification, I believe (and someone
correct me if this is inaccurate) that in order to validate the
warranty on their electronics, MIT cabling must be employed.
We definitely don't have a problem with dealers demonstrating
particular products in conjunction with our equipment, but we do not
endorse or encourage the use of hyperbole and pseudo-scientific
terminology that some manufacturers (and dealers) employ in their
advertising or sales tatics that could mislead consumers regarding
the performance of that specific product in relation to other
similar products.
Sincerely,
Andrew Rigby
DAL
William Hunt
Pekka Aro wrote:
> To me it is exceedingly obvious that you are referring to MIT cables
> and Spectral amplifiers. Am I correct?
>
> I was really surprised to read your opinion on cable issues since I
> have visited two of your dealers who demonstrate SC-V's with the new
> $11K or so (!) MIT 850 Shotgun EVO speaker cable and claim that
> without these cables the full performance potential of the SC-V's
> will not be realized. One went as far as to claim that the SC-V with
> the $11K MIT cable was "much better" than the SC-VI with the DAL
> cable! Furthermore, one of the dealers used Spectral amplifiers to
> demo the SC-V's.
>
> I do not doubt the validity of your opinions, but something seems
> amiss here. If you are correct, your dealers in this case are either
> incompetent hacks or dishonest rip-off artists. In either case, why
> would you allow dealers like that represent your product? Perhaps
> this is news to you...
>
> In any case, you certainly design fine sounding and reasonably priced
> speakers.
[ quoted text deleted -- jwd ]
My wife and I visited a high end dealer and listened at length to the
Dunlavy SC-IV speakers. MIT cables and a Spectral amplifier were
used. The sound had some obvious faults -- I mean car radio level
problems. Changing to a different amplifier and different cables
fixed the problems.
We have listened to other speakers with MIT cables in several stores
and found the same kind of problems. It is often difficult to get the
salesman to understand what we are saying and to agree to use other
cables.
My thanks to DAL for talking sense.
Bill Hunt
AWRigby
From John Dunlavy:
With respect to some of the recent posts regarding audiophile cables
and their performance properties, a lot of questionable technical
information is still being touted as gospel on the NET.
For example, in my post of 11 Nov., I cited several examples -
including magazine advertising using graphs for comparing the
performance of different cables on the basis of Joules Vs. Frequency
and efficiency Vs Frequency. In my view, rating the relative
performance of cables using Joules and/or efficiency conveys no
relevant meaning regarding their real-world performance within an
audiophile system. To support this view, I mentioned that the
standard definition of a Joule is a watt-second - a measurement of
energy storage which has no relevance to cables. Neither do curves of
efficiency Vs frequency, unless efficiency is defined in the normal
manner - which it was not.
Since a Joule is a measure of energy or unit of work Vs time, I fail
to understand its relevance for rating cable performance or for using
it to compare one cable to another. This view seems further justified
when cables are advertised as exhibiting performance in the range of
300 to 600 Joules, while all other cables earn a rating of only about
0.3 to 1 Joule. Hmmm!
Indeed, every one of the many textbooks in my office defines a Joule
as a "watt-second" (in MKS units). In the C.G.S. system, 1 Joule = 10
to the 7th-power ergs/second. Webster's Collegiate Dictionary
defines a Joule as, "the absolute meters-kilogram-second unit of
work" or energy equal to 10>7th-power ergs, or approximately 0.7375
foot-pounds. (So what does this have to do with loudspeaker cables,
pray tell?)
On Nov. 13, 1997, Martin Leese (mle...@hudson.cs.unb.ca), perhaps
misinterpreting my intent, wrote:
>> More relevant definitions are:
>> Energy in a capacitor = (C*V*V)/2
>> Energy in an inductor = (L*I*I)/2
where V = voltage
I = current
C = capacitance
L = inductance
>> I cannot believe you did not know this so must assume that when you
>> chose to define energy in terms of mechanical work done you were
>> being mischievous. This was unfortunate as it damaged the
>> credibility of an otherwise excellent set of posts.
>> Of course with AC signals, the average energy is zero and these
>> energy stores simply present themselves as reactance.
With regard to Martin Leese's assertion that his (C*V*V)/2, etc.
represent more relevant definitions than using Joules, which convey
no relevance to cable performance, I would agree. If Martin is
inferring that I do not understand the meaning of a Joule or that I
do not properly grasp the way cables perform, perhaps he can explain
his viewpoints on the subject in more detail! Anyway, lets keep the
dialog active - with the discovery of truth being our main objective.
With regard to accuracy, the purpose of an audiophile loudspeaker
cable is to transport complex, rapidly-varying A.C. signals between
the output of a power-amp and the input of the loudspeaker with as
few measurable/audible alterations as possible. This means vanishing
levels of frequency-domain distortion and time-domain distortion,
along with the lowest possible signal loss.
The physics and engineering principles associated with the
propagation of even the most complex electrical signals along a cable
have been exhaustively studied and are well-know. Despite this, many
audiophile cable companies continue to advertise that they have
discovered fundamentally new design principles which permit their
cables to achieve levels of performance far beyond that obtainable by
their competitors. And they use bogus specs, misleading explanations,
and meaningless graphs to tout the superiority of their products.
With respect to performance, the important and relevant properties of
loudspeaker cables are: 1) total loss resistance of conductors, 2)
inductance per unit length, 3) capacitance per unit length, 4) total
signal loss incurred Vs frequency when terminated by the input
impedance of loudspeaker, 5) velocity of propagation factor, 6)
maximum safe voltage, 7) maximum safe current, 8) stability of
properties with respect to temperature extremes, 9) dielectric
absorption, 10) loss tangent of dielectric, 11) dispersive properties
in the time and frequency domains, etc.
While the above measurable properties represent most or all of the
important considerations that serious designers of loudspeaker cables
must take into account, the average audiophile need only be concerned
with: 1) power loss Vs frequency, with the cable terminated by the
input impedance of a typical loudspeaker, 2) impulse/step response
(for disclosing any frequency/time dispersive properties), and 3)
quality of the connectors (preferably gold plated and with a large
contact area).
Another property that may prove worthwhile for those wishing to
achieve the most accurate performance possible from a loudspeaker
cable is its characteristic impedance. Although of questionable merit
at audio frequencies, it still represents good engineering practice,
despite the fact that even a 20-30 ft length of cable represents an
insignificant fraction of a wavelength at 20 kHz. However, since
trained ears are often able to discern almost vanishing levels of
distortion, and it is not difficult to design and manufacture a
loudspeaker cable with a characteristic impedance in the range of 6
to 8 Ohms, a case might be made for doing so!
But some might argue that a cable with a characteristic impedance of
6 or 8 Ohms would exhibit a much higher parallel capacitance than
typical 30 to 100 Ohm cables, causing some high slew-rate power-amps
with a poor stability factor to oscillate (usually at ultra-sonic
frequencies). Such oscillations can, of course, destroy some tweeters
and or damage transistors and other components within the amplifier.
The solution for solving the stability problem is simple and is the
reason for the costly, beautifully-machined, high-tech looking box at
the end of some expensive cables, which houses what some
manufacturers refer to as a unique terminating network. Unique? No!
Most often, the network is nothing more than a simple, inexpensive
resistor/capacitor network, at the speaker end of the cable, in
parallel with the cable conductors. Often referred to as a Zobel
Network, it typically consists of a 6-10 Ohm non-inductive resistor
(5-10 watt rating) in series with a 0.33-0.47 uF, 400-600 Volt
capacitor (preferably with a polypropylene, Teflon, etc. high-quality
dielectric). Such a terminating network compensates for the usual
rising input impedance of many loudspeakers at frequencies above the
audio range, thus providing a more-or-less constant resistive load,
independent of frequency, for the amplifier.
But the proper location for such an R-C network is either inside the
loudspeaker or a conjugate version at the output of the amp. Indeed,
many well-designed audiophile power amps use a small inductor wound
around, and connected in parallel with, a 1-2 watt carbon resistor,
connected in series with the output terminals.
Let's now briefly consider what meaningful properties
well-designed audiophile coaxial inter-connect cables should
possess. Perhaps, the most important consideration is a cable's
ability to prevent hum from being induced by nearby 60 Hz hum-fields
surrounding some power transformers, A.C. mains wires, etc. The best
means for minimizing such hum induction is an outer shield with the
highest possible conductivity (lowest series resistance, from
end-to-end, between the connectors). Most of the so called high-tech
solutions being offered in the marketplace, such as those using a
second center conductor to carry the ground connection, and
connecting the shield braid to ground at only one end of the cable,
does not work well for two reasons: 1) it significantly increases the
capacitance per foot and 2) it greatly reduces the shielding effect
(while increasing ground-loop transfer) of the outer braid in
preventing nearby A.C. fields from inducing hum voltages into the
center conductor.
In answer to Doug Plumb's query concerning the values for the
Zobel Network, the values for the resistor and capacito are not
critical - resistances in the range of about 5 to 10 Ohms and
capacitances in the range of about 0.3 to 0.6 being acceptable for
use under most conditions.
Must stop writing and return to the old grind. Thanks for caring!
All comments are welcome. Special thanks to Charles Biller, Howard
Ferstler, Allan Freeman, Vandit Kalia, Colin, Tom Jacobs, Steve
Lampen, Mike Guyote and GRL for their comments and observations. I
will attempt to answer more questions about cable design and
performance in the near future.
Best regards,
John Dunlavy
Richard Dell
In article <64pkk2$m...@eyrie.graphics.cornell.edu>,
ShLampen <shla...@aol.com> wrote:
>The skin effect of a 20 AWG conductor at 100 kHz is THE ENTIRE
>CONDUCTOR, so how could "skin effect" have any effect? At 20 kHz, the
>skin effect of a 10 AWG speaker cable (for example) is greater than
>the width of the conductor.
Actually, the skin depth is the decay rate of the charge carriers
with depth from the surface (ie -- an exponential distribution)
so that there would be measurable, not necessarily significant,
effects on a wires resistance even for 20 gauge wire at 20 kHz,
where the skin effect would increase resistance by 33 percent over
its value at low frequencies.
For 20 kHz, the 'skin depth' for hard drawn copper is .593 mm,
while the diameter of 10 gauge wire is 2.588 mm, or a radius of
1.4 mm, certainly greater than the skin depth. At 100 kHz, the
'skin depth' is .265 mm, certainly less than the radius of 20 gauge
at .406 mm.
Again, with wire chosen for low enough resistance anyways (ie --
negligible in comparison to speaker voice coil and amplifier out
put impedance), this is likely to be insignificant, as the resistance
of the cable should be insignificant.
--
Richard Dell