The cable conundrum

[ack]

More stuff to ponder which, again, is the basis of MIT's claims:

The question arises, how are these phase angles measured? Recall from the phasor form of Ohm’s law the phase difference of the voltage and current is equal to the phase angle of the impedance: ?z = ?v - ?i. This is an extremely important result. The phase angle of the impedance is a quantity that can be measured directly using an impedance analyzer. Thus, insight can be gained about the relative phase angle of the voltage and current in an audio cable simply by looking at its impedance. [...] The power factor is defined as PF = cos (?v-?i) However, the power factor is usually given as PF = cos (?) where ?, the phase difference between the voltage and current, is known as the power factor angle. Since we can measure the impedance phase angle of cable, we can calculate the power factor of audio cable by simply taking the cosine of our measurement. We can then use the power factor as a tool to characterize how efficiently an audio cable is functioning as a coupling network.

In the patent quoted before, he then uses an impedance analyzer and claims to find capacitive phase angles ~-87 to ~-85 degrees (or ~-3 to -5 off optimal) at various frequencies in the cable under test, which he considers significant...

 

[DonH50]

Using TL terms for audio cables has been a time-honored way to introduce audiophiles to scientific theory, whether it applies or not...

Cobra Cables were developed to limit resistance and inductance by using woven wires insulated from each other. Capacitance was through the roof, of course, and the result made some amps unstable...

Of course you can measure the phase angle and it will not be exactly -90 degrees because there is always a finaite amount of inductance. Whether 3-5 degrees matters I am not sure... I still do not know why they want a purely capacitive phase angle.

How efficiently a cable is working in terms of power transfer only makes sense if the source and load is considered as well, so my 1/2-second look says they are applying metrics to the cable in isolation; in all the systems (power line through microwave) I have analyzed the source and load impedances must be known to facilitate a match. And, in this case, the cable is normally orders of magnitude from the speaker (load) impedance so I am not sure where this is all going.

 

[DonH50]

This assumption is only true at radio frequencies!

At audio frequencies, you have to use the long form formula, which is very messy.

Jim Brown covers this in depth.

Transmission Lines at Audio Frequencies, and a Bit of History
by Jim Brown
Audio Systems Group, Inc.
http://audiosystemsgroup.com

http://www.audiosystemsgroup.com/TransLines-LowFreq.pdf

I quoted the full equation, same as in the paper you linked, not sure what is missing for audio? I said if R is small and G large you can neglect them; the usual catch for audio cables is that R is not all that small, particularly for speaker cables. It gets really messy at microwave frequencies when skin effect means the R is nonlinearly dependent upon frequency, and really, really messy beyond 10 GHz or so when normal T-line theory breaks down entirely because TEM no longer holds.

I usually try to simplify for this audience so most certainly any other engineers out there can poke plenty of holes in all this... I could myself but generally feel it's just not worth it; most audiophiles do not have advanced degrees in EE.

 

[Speedskater]

It's the simplification and audio that don't go together! We see audiophiles over and over again, take electronic engineering formulas, calculations & "good engineering practices" that are for one frequency range or one power level or one type of circuit and grossly misapply them. (just look back a few posts in this thread)

 

[microstrip]

(...) Of course you can measure the phase angle and it will not be exactly -90 degrees because there is always a finaite amount of inductance. Whether 3-5 degrees matters I am not sure... I still do not know why they want a purely capacitive phase angle. (...)

This is the key question. But after reading the whole patent my interpretation is that they suggest it does not matter ...
But IMHO there is not audio science in this patent, just some electrical very basic statements and a few considerations about "changes of energy storage" versus frequency. The patent just covers a particular technique to build audio cables.

Patents are not supposed to be scientific, however patent examiners should reject them if the invention principles are found to violate established science - e.g. if someone invented a passive cable with a power gain of 10 it would not receive a patent.

 

[Groucho]

Haven't read the whole thread, but my approach is to view the subject from the other side of the argument: instead of fabricating pseudo-engineering reasons for the existence of exotic and/or expensive cables, my question is whether anyone here really believes that the manufacturers of said cables have any special knowledge or powers to impart fairy dust into a recording in the gaps between equipment. I don't! People who design amplifiers, or DACs or speakers, then yes, I believe they know some stuff that takes inspiration, intelligence and years of experience to learn. But people who eschew such an easy way to make a living, and instead dedicate themselves to the long hard slog of paying Chinese people to..., sorry running a team of highly-trained scientists to wrestle with that complex maze of two conductors obviously deserve our thanks, and as much money as we can give them. (Not)

 

[rbbert]

Haven't read the whole thread, but my approach is to view the subject from the other side of the argument: instead of fabricating pseudo-engineering reasons for the existence of exotic and/or expensive cables, my question is whether anyone here really believes that the manufacturers of said cables have any special knowledge or powers to impart fairy dust into a recording in the gaps between equipment. I don't! People who design amplifiers, or DACs or speakers, then yes, I believe they know some stuff that takes inspiration, intelligence and years of experience to learn. But people who eschew such an easy way to make a living, and instead dedicate themselves to the long hard slog of paying Chinese people to..., sorry running a team of highly-trained scientists to wrestle with that complex maze of two conductors obviously deserve our thanks, and as much money as we can give them. (Not)

But this is no more or less of a "belief" than that of people who "believe" cables make a big sonic difference. In neither case are we presented with any actual evidence to support either "belief".

 

[Groucho]

But this is no more or less of a "belief" than that of people who "believe" cables make a big sonic difference. In neither case are we presented with any actual evidence to support either "belief".

That's rather like saying that a person who claims that fairies do not exist must provide proof of it, otherwise it is just as irrational a belief as believing in fairies.

There are two claims being mixed up: one that everyone wants to believe is that cables sprinkle fairy dust onto recordings - it's one of the basic tenets of modern audiophilia - but in a confrontation on technical grounds it transmutes into the rational-sounding "Of course, the best cable is no cable" defence.

 

[MylesBAstor]

That's rather like saying that a person who claims that fairies do not exist must provide proof of it, otherwise it is just as irrational a belief as believing in fairies.

Hey Groucho you must be mistaking that most of us really give a rats ass. You don't believe that cables make a difference, fine; no one is twisting your arms and we don't need to be saved.

 

[microstrip]

But this is no more or less of a "belief" than that of people who "believe" cables make a big sonic difference. In neither case are we presented with any actual evidence to support either "belief".

But it is an imaginative and humorous "belief" - although I can not agree that Bruce Brisson looks alike Tinker Bell.

 

[microstrip]

(...) we don't need to be saved.

Sorry Myles, I would love to be saved, top cables are expensive. But they sound so good in my system ...

 

[opus111]

Haven't read the whole thread, but my approach is to view the subject from the other side of the argument: instead of fabricating pseudo-engineering reasons for the existence of exotic and/or expensive cables, my question is whether anyone here really believes that the manufacturers of said cables have any special knowledge or powers to impart fairy dust into a recording in the gaps between equipment. I don't!

You seem to me to be dragging the thread off topic by introducing belief. Belief doesn't belong in science, only religion. If a person hears then that's evidence for them, if they don't then there's no need to purchase.

Do please explain why belief is relevant?

 

[Groucho]

You seem to me to be dragging the thread off topic by introducing belief. Belief doesn't belong in science, only religion. If a person hears then that's evidence for them, if they don't then there's no need to purchase.

Do please explain why belief is relevant?

OK. Strike my use of the word "believes" in the quote above. Replace it with the word "thinks". Problem solved.

 

[opus111]

Not so fast

So now we've moved a step away from religion, which is one step in the right direction in my estimation. But 'thinks'?

The only place that 'thinks' has in science that I can see is in terms of formulating a hypothesis. As in 'I think X causes Y, let's try an experiment to test this.'

In this case your hypothesis (my paraphrase) is that the company (MIT) hasn't got any special knowledge or understanding of cables not already widely understood by engineers generally. I'm ignoring your aside to 'fairy dust' as I don't think anyone knows what this actually is.

In order for this hypothesis to be a scientific one, it has to be testable and be capable of being falsified. So how do you propose to test your hypothesis by experiment? What experimental result would it take for your hypothesis to be shown wrong?

 

[ack]

Using TL terms for audio cables has been a time-honored way to introduce audiophiles to scientific theory, whether it applies or not...

Cobra Cables were developed to limit resistance and inductance by using woven wires insulated from each other. Capacitance was through the roof, of course, and the result made some amps unstable...

Of course you can measure the phase angle and it will not be exactly -90 degrees because there is always a finaite amount of inductance. Whether 3-5 degrees matters I am not sure... I still do not know why they want a purely capacitive phase angle.

How efficiently a cable is working in terms of power transfer only makes sense if the source and load is considered as well, so my 1/2-second look says they are applying metrics to the cable in isolation; in all the systems (power line through microwave) I have analyzed the source and load impedances must be known to facilitate a match. And, in this case, the cable is normally orders of magnitude from the speaker (load) impedance so I am not sure where this is all going.

These are, of course, important points you are making. MIT appear to be looking at the cable's characteristics exclusively, and whether this is good enough or not I don't know. But what I can tell you is that they directly attack the transmission line theory as you described it earlier (the constant aspect of things):

Earlier, in the section on energy storage elements, we showed how capacitive and inductive reactances of ideal components change with frequency. But what about the phase angles of the elements (L and C)do they change with frequency? Many articles have been written about modeling audio cable using the discrete components L, C, and R. But all have one major pitfall, they assume that the impedance phase angles of these components are ideal
and therefore stay constant over the audio frequency range. This is not the case. Our measurements will show that the inductive and capacitive phase angles of an audio cable actually change with frequency, particularly below about 1kHz. This has a major implication: If the impedance phase angle changes with frequency, then the power factor changes with frequency. This means that the cable’s power-handling characteristics are frequency dependent.

The last sentence is, of course, what I have been discussing in this thread; and ponder to what extent does all this matter... I ask this in the larger context of the question asked earlier: can some cables make a huge difference. What I can tell you is that the latest Matrix cables I am trying make a staggering difference for the better all across the spectrum, something that others here in the forum also attest to, in our private messages. My quest is to understand why this is...

 

[opus111]

Earlier, in the section on energy storage elements, we showed how capacitive and inductive reactances of ideal components change with frequency. But what about the phase angles of the elements (L and C)do they change with frequency?

No, they don't change with frequency - they're ideal L and C in the models.

Many articles have been written about modeling audio cable using the discrete components L, C, and R. But all have one major pitfall, they assume that the impedance phase angles of these components are ideal and therefore stay constant over the audio frequency range.

The TL theory works with just L and C however no real-world TL is truly lossless. Once we introduce R then of course the phase angles change with frequency because we have a loss component. A real capacitor and a real inductor both have loss and therefore never exhibit the perfect 90 degree phase angle. The better models realize that R is itself a function of frequency - in the case of inductors because they exhibit skin and proximity effects. It gets rather tricky modelling these frequency dependent losses - SPICE doesn't support this directly, you have to faff around with external networks which exhibit the necessary 3dB/octave slopes.

This is not the case. Our measurements will show that the inductive and capacitive phase angles of an audio cable actually change with frequency, particularly below about 1kHz. This has a major implication:

I for one am skeptical of this claim, can we see the measurement set-up somewhere? Below 1kHz its vanishingly unlikely that inductive losses are coming into play so I rather suspect measurement error.

If the impedance phase angle changes with frequency, then the power factor changes with frequency. This means that the cable’s power-handling characteristics are frequency dependent.

They're using inappropriate terminology - power factor is where we are concerned with power, not signals. Makes me all the more skeptical of their claims.

What I can tell you is that the latest Matrix cables I am trying make a staggering difference for the better all across the spectrum, something that others here in the forum also attest to, in our private messages. My quest is to understand why this is...

To make a start on this question please go ahead and describe the sound with the Matrix cables in place in detail. Details include specifics of the music played and what aspects of the sound were different and in what ways compared to the previous cables. Then we might be able to make a start in formulating hypotheses to explain the observations. But science begins with observations and 'makes a staggering difference' isn't one.

 

[NorthStar]

-- I am reading E.V.E.R.Y.T.H.I.N.G., the entire thread from A to Z. :b

* Buy the best quality cables (interconnects, speaker wires, power cables, audio/video, analog/digital) that you can afford.

But do your research first, and audition them with your own set of ears and connected to your own gear and soul. :b

 

[DonH50]

I don't know what MIT is doing. I tend to agree with opus111's comments. I am reminded how hard it was to show someone why the "hideous" 5000 ppm/degC (or whatever, I think that may be the number for gold instead of copper) tempco of his speaker wires did not in fact cause terrible thermal distortion he was sure he could hear...

ack -- Are you speaking of interconnect or speaker cables? Either way, power factor depends upon the load, and power loss in speaker cables is in the mud compared to speakers. What I can easily believe is that a network added to a cable (interconnect or speaker) can change the sound. The change will depend upon the source and load, of course. I am not sure PF enters into it, though any network would also change the effective PF as well. It sounds like they are addressing a real effect, one that can be measured, but one that does not seem to make a practical difference.

Look back at the T-line equations and you will see that the phase angle depends upon resistance (R) and I suspect that is causing the phase angle to change. I wonder what superconducting speaker cables would cost, along with the nitrogen or hydrogen dewar...

Bottom line is I am not sure I can contribute any more to this discussion. I will follow as it is interesting, but most of these threads end with irreconcilable differences and I prefer to part friends.

However, if you can nail down some specifics about what you hear, I would be glad to work with opus111 to gen up a model that might explain it. I am curious. Amp, speakers, and what you hear as differences would be welcome.

 

[opus111]

Bottom line is I am not sure I can contribute any more to this discussion. I will follow as it is interesting, but most of these threads end with irreconcilable differences and I prefer to part friends.

Reminds me of one of my favourite aphorisms which I first heard from Lotfi Zadeh but perhaps didn't originate with him - 'Friends come and go, but enemies accumulate.'

However, if you can nail down some specifics about what you hear, I would be glad to work with opus111 to gen up a model that might explain it. I am curious. Amp, speakers, and what you hear as differences would be welcome.

I'm up for that

 

[NorthStar]

-- You know the adage (adagious);
spend 10-15% of your total sound system's budget on the connections (cabling).

If your system cost one million bucks, that's $100,000-150,000 right there.

And if your total system cost only hundred bucks, that's $10-15 right there. :b

The less money you have the more you save.

- And if you feel like it, you can increase that figure to 20%.