Cable Myths Continued

[DrWho]

I'm trying to think of any caveats, but in general I would say yes. Without knowing what's happening up front, I think there may be some variation in language, but overall I think there will be some level of similarity.

I've done this a couple times for my own projects at work so I'm really speaking from my own limited experience. Unfortunately I don't get to explore everything I'm curious about. It is a business afterall.

Ha ha....understood.

Well, if you get a chance sometime to make an experiment like that, it will make good reading. I am, a doubter for now.

What metric would you like to see tested?

Obviously I need to be careful about what I share, but I bet our local AES chapter would be willing to do some experiments and then it's public knowledge...

Knowing where the correlation breaks down would be invaluable information in my opinion.

[garyrc]

It wasn't too long ago that one could buy all sorts of things to try at home and then return without incident or restocking fee---Nowadays not so much. I am AMAZED how many people buy whole speaker systems on the internet and rely on others' opinion--crazy.

I think manufacturers should pool their money and buy listening salons in large cities, where different brands and models could be compared; then we could watch them fight about placement, room treatments, etc.. It is almost impossible to hear any two high end speakers under the same roof -- or any roof -- anymore, let alone hearing speakers that might given the Golden Ear favorites a run for their money, such as Khorns, Jubs, or Palladiums (Palladia?).

A local, cooperative, knowledgeable dealer in our undisclosed small city continued to loan equipment to try out at home right up until he went out of business, a victim of internet "buy without listening" competition.

[mungkiman]

It wasn't too long ago that one could buy all sorts of things to try at home and then return without incident or restocking fee---Nowadays not so much. I am AMAZED how many people buy whole speaker systems on the internet and rely on others' opinion--crazy.

I think manufacturers should pool their money and buy listening salons in large cities, where different brands and models could be compared; then we could watch them fight about placement, room treatments, etc.. It is almost impossible to hear any two high end speakers under the same roof -- or any roof -- anymore, let alone hearing speakers that might given the Golden Ear favorites a run for their money, such as Khorns, Jubs, or Palladiums (Palladia?).

A local, cooperative, knowledgeable dealer in our undisclosed small city continued to loan equipment to try out at home right up until he went out of business, a victim of internet "buy without listening" competition.

Good for the consumer does not always equal good for the manufacturers.

[DrWho]

I think we're talking at this from two sides, so I just want to clarify - does the following sound like a fair summary?

1) My Perspective: If someone identifies a grainy sound, then I believe we can find a measurable attribute of the device that explains the source of the grain. Then if we change the magnitude of the artifact (as observed from the measurement) then we see a corresponding subjective change in the description of the grain's magnitude.

2) Your perspective: I want a set of measurements that identify the presence of grain.

I don't think #2 is feasible, let alone possible because there are too many attributes that might be perceived as grain. is that the point you're trying to hit home? That maybe we're measuring 999 attributes, but that 1000th attribute that we don't know about yet is the source of the grain?

When selecting equipment (or designing it), I think there is value in understanding the 999 known attributes and using that to guide your decision. If nothing else, understand the compromises at play and choose which metrics you feel to be most important to you at the time (and nothing wrong with changing your mind down the road). I just don't understand the need to defend the mysticism of the unknown variable....especially when people go back and attribute differences to the unknown rather than the known.

I've been doing a lot of research on cognitive biases since this thread came up....it's really quite a fascinating subject.

[Deang]

John Curl has some stuff on this, but I can't find it!

[kev313]

I think we're talking at this from two sides, so I just want to clarify - does the following sound like a fair summary?

1) My Perspective: If someone identifies a grainy sound, then I believe we can find a measurable attribute of the device that explains the source of the grain. Then if we change the magnitude of the artifact (as observed from the measurement) then we see a corresponding subjective change in the description of the grain's magnitude.

2) Your perspective: I want a set of measurements that identify the presence of grain.

I don't think #2 is feasible, let alone possible because there are too many attributes that might be perceived as grain. is that the point you're trying to hit home? That maybe we're measuring 999 attributes, but that 1000th attribute that we don't know about yet is the source of the grain?

When selecting equipment (or designing it), I think there is value in understanding the 999 known attributes and using that to guide your decision. If nothing else, understand the compromises at play and choose which metrics you feel to be most important to you at the time (and nothing wrong with changing your mind down the road). I just don't understand the need to defend the mysticism of the unknown variable....especially when people go back and attribute differences to the unknown rather than the known.

I've been doing a lot of research on cognitive biases since this thread came up....it's really quite a fascinating subject.

I'm not sure 1 and 2 are mutually exclusive. A few steps away from each other, but not necessarily totally different.

[kaiser SET say]

Not having the time to read through all these posts as I stumbled upon getting ready to leave work.

Did anybody ask one of Marty's original questions about the wire from turntable to phono pre-amp? The signals are very tiny so I think good shielding to keep out interference would be important?

I'll have to go find the links in the other thread DeanG posted.... probably won't understand them.

I used to think bigger was better but the 6 gauge jumper cables are kind of heavy to hook up to my little t amp.

True dat but it makes a great ground wire for your home

[Ski Bum]

This is a fascinating discussion, trying to come up with new metrics, and applying these novel metrics to complex devices.

But, to bring this back on topic, what does it have to do with speaker cable?

Cables are simple lumped parameter components. L, R, C, that's it. Simple conductors do not produce non-linear distortion at all, and they won't effect linearity unless they are poorly chosen for the task. There has yet been any evidence presented showing that these mysterious differences arise from the cable, or that it even exists, yet the appeal to novel metrics in pursuit of audiophile religion persists. I don't get it.

This whole discussion is based on a false premise, one whose persistence and promotion is a detriment to the entire forum. It's unsound, unsupported, bad advice to lead someone to believe that speaker cables are anything more than they are, or that they will somehow transform one's system. They would be better off reading up on room acoustics and tackling that than wasting any time, effort, or money on fancy magic wire.

[Islander]

This discussion is interesting, but I don't think anyone has answered the question I asked many pages back: larger speaker cables have lower resistance, which is a good thing in terms of not compromising amplifier damping factor, and helps high transient peaks be transmitted without being diminished, but is there a point beyond which speaker cable size compromises the sound?

Can a speaker cable be overly large, and do factors other than resistance start to impair the sound they're carrying? At what point does this occur, 6 AWG, 2 AWG, or does it show up at more typical sizes like 10 or 12 AWG?

[Islander]

Thanks for responding.

The cables I use are twisted pair in PVC jacketing, with an outer clear PVC jacket over the twisted pair. The 8 gauge is composed of 805 strands x 2 of 36 gauge tinned OFC copper, and the 10 gauge is 462 x 2 strands of 36 gauge tinned OFC copper.

[Islander]

It is impressively large, due to the outer jacket. That keeps the conductors just a bit further from any other conductors, and thus might reduce EM interference. The 10 gauge is 14mm (9/16") in outside diameter, and the 8 gauge is 17mm (11/16"). The very fine strands make the cable quite flexible and easy to handle.

It's a bit heavy, so I support it by looping it around the tweeter stand and providing a clamp to support the woofer cable, in order to prevent it tugging on the speaker connectors.

As well as sounding good, if cable can be said to have a sound, or at least not impart any unpleasant quality to the music, it's not expensive. The 10 gauge is only $1.09 a foot, and the 8 gauge is only $1.65 a foot. 100-foot lengths cost less per foot.

When I bought the 8 gauge cable, I was thinking that I could relax and listen to the music, secure in the knowledge that at least one aspect of the system (the speaker cable resistance) had definitely not been compromised. Anything that doesn't take away from enjoyment in a way adds to it.

[DrWho]

This discussion is interesting, but I don't think anyone has answered the question I asked many pages back: larger speaker cables have lower resistance, which is a good thing in terms of not compromising amplifier damping factor, and helps high transient peaks be transmitted without being diminished, but is there a point beyond which speaker cable size compromises the sound?

Can a speaker cable be overly large, and do factors other than resistance start to impair the sound they're carrying? At what point does this occur, 6 AWG, 2 AWG, or does it show up at more typical sizes like 10 or 12 AWG?

Here are two easy to compare datasheets between Belden 6000UE (12ga) and Belden 6300UW (18ga):

http://www.belden.com/techdatas/english/6000UE.pdf

http://www.belden.com/techdatas/english/6300UE.pdf

Here's a quick comparison:

Belden 6000UE:

12 gauge (19x25 stranding)

L = 0.14 uH/ft

C = 36 pF/ft

R = 1.6 mohm/ft

Belden 6300UE:

18 gauge (7x26 stranding)

L = 0.15 uH/ft

C = 34 pF/ft

R = 6.5 mohm/ft

Capacitance goes up as you increase the gauge, but resistance goes down. Inductance usually goes down until you hit the skin effect.

There will also be a resonant tank circuit created by the cable, but it interacts with the output stage of the amplifier too. The amount of damping you get is related to the resistance of the cable, and the output impedance of the amplifier. This is different from the damping the amplifier provides to the loudspeaker driver.

You can divide the cable affects into two categories:

1) Amplifier Output Stage

2) Reduction in coupling between loudspeaker and amplifier

A good amplifier output stage is unaffected by normal cables: it should be stable into a small capacitive load. An amplifier with a damping ratio of 100 into 10 ohms has a 100mohm output impedance: At some point your cable no longer dominates that behavior.

Output impedance is such a simple concept, and yet I think it gets way too under-stressed when it comes to amplifier design....it's one of the reasons I like the idea of a Class D amplifier with no output inductor strapped directly to the voice coil of a speaker (which then becomes the inductor). You have short cables and you can get an insane damping factor: now all of your frequency shaping can be well controlled and won't shift based on the input signal....

[jacksonbart]

My favorite analogy to interconnects so far has been the comparison of modern HD TVs. Perhaps beyond this thread. That was a hoot and apples to oranges comparison. The actual comparison would be interconnects to HDMI cables.

[Islander]

This discussion is interesting, but I don't think anyone has answered the question I asked many pages back: larger speaker cables have lower resistance, which is a good thing in terms of not compromising amplifier damping factor, and helps high transient peaks be transmitted without being diminished, but is there a point beyond which speaker cable size compromises the sound?

Can a speaker cable be overly large, and do factors other than resistance start to impair the sound they're carrying? At what point does this occur, 6 AWG, 2 AWG, or does it show up at more typical sizes like 10 or 12 AWG?

A good amplifier output stage is unaffected by normal cables: it should be stable into a small capacitive load. An amplifier with a damping ratio of 100 into 10 ohms has a 100mohm output impedance: At some point your cable no longer dominates that behavior.

Output impedance is such a simple concept, and yet I think it gets way too under-stressed when it comes to amplifier design....it's one of the reasons I like the idea of a Class D amplifier with no output inductor strapped directly to the voice coil of a speaker (which then becomes the inductor). You have short cables and you can get an insane damping factor: now all of your frequency shaping can be well controlled and won't shift based on the input signal....

Thanks for the detailed response, Mike! Since the dual-mono amps are Class D, with a damping factor of 200+, and each channel is connected directly to its respective driver (nothing but large cable between the amps and the drivers), it sounds like the situation is near ideal.

[DrWho]

This is a fascinating discussion, trying to come up with new metrics, and applying these novel metrics to complex devices.

But, to bring this back on topic, what does it have to do with speaker cable?

Cables are simple lumped parameter components. L, R, C, that's it. Simple conductors do not produce non-linear distortion at all, and they won't effect linearity unless they are poorly chosen for the task. There has yet been any evidence presented showing that these mysterious differences arise from the cable, or that it even exists, yet the appeal to novel metrics in pursuit of audiophile religion persists. I don't get it.

This whole discussion is based on a false premise, one whose persistence and promotion is a detriment to the entire forum. It's unsound, unsupported, bad advice to lead someone to believe that speaker cables are anything more than they are, or that they will somehow transform one's system. They would be better off reading up on room acoustics and tackling that than wasting any time, effort, or money on fancy magic wire.

Just to be complete, the impedance of a cable will change with temperature - and the signal flowing through it will cause heating due to the resistive losses (the reactive components don't create heat). This will generate a form of hysteric distortion as the rate of heating will be faster than the rate of cooling, and the music source has a high crest factor. What is the magnitude? Ultimately it depends on the size of the wire. Something like those light-bulb fuses (a fuse is an undersized wire) that (try to) protect tweeters sound horrid at almost all levels - even when not hot enough to glow. The same effects will be happening in any conductor, just to a much much much smaller degree (bad pun intended).

In practice it is very easy to mitigate those effects, but I still think there is value to understanding the behaviors. This website spells it out pretty well for anyone that wants to calculate actual values:

http://www.engineeringtoolbox.com/resistivity-conductivity-d_418.html

Copper is shown as having a temperature coefficient of 4.29 x 10^-3, which will yield a 0.4% change per degree C. And then the variation isn't even linear either - which gets exaggerated by impurities and more so at colder temperatures.

http://www.copper.org/resources/properties/cryogenic/

Now all that said, it is also very easy to ensure that these effects are small - especially in comparison to the behavior of the loudspeaker. Klippel has done plenty of research showing the effects of voice coil heating, and I've found those effects to be quite audible in my applications. The balance between my direct radiating woofer and horn loaded tweeter changes based on how loud I listen....or maybe subjectively I would describe it as the woofer can't keep up with that tweeter. (There's a lot more going on, but heat is definitely one of the factors).

http://www.klippel.de/measurements/power-handling-heat-dissipation/voice-coil-temperature.html

http://www.klippel.de/uploads/media/Nonlinear_Modeling_of_Heat_Transfer_01.pdf

So which wires are we talking about? There's definitely more than just LRC going on, but it depends on the application when it matters...

Edited February 26, 2014 by DrWho

[DrWho]

As you know, designers continue to go to very elaborate ends trying to improve the sound of amplifiers that measure nearly perfect in every regard we can measure.

I think that is a false premise. While a few attributes may measure nearly perfect, I've never seen a design where compromises weren't made to achieve those results. In fact, there is very little in the audio world that doesn't have some opposing metric forcing a compromise. Even something apparently as trivial as a resistor value forces a lot of compromise, and that's even assuming a perfect resistor. The compromise tree just compounds further with the active circuits.

So to provide a counter-view: whenever I hear someone says "it measures perfect, but....", I interpret that to mean the person is unaware of the compromises at play. Nothing measures perfect.

Or maybe to put it another way, could you provide an example of something that measures even near perfect in all measurable attributes?

[Islander]

So to provide a counter-view: whenever I hear someone says "it measures perfect, but....", I interpret that to mean the person is unaware of the compromises at play. Nothing measures perfect.

Or maybe to put it another way, could you provide an example of something that measures even near perfect in all measurable attributes?

Some things, after a great deal of time and effort, can approach perfection, but even then it's possible to get it wrong, like the mirror of the Hubble Space Telescope. It was one of the most precise/perfect man-made items ever built, but it was built to the wrong specs and had to have corrective lenses installed after it was in orbit.

[Don Richard]

So to provide a counter-view: whenever I hear someone says "it measures perfect, but....", I interpret that to mean the person is unaware of the compromises at play. Nothing measures perfect. Or maybe to put it another way, could you provide an example of something that measures even near perfect in all measurable attributes?

If I provided a suitable picture of her I'm afraid I would get too many warning points.

[Deang]

"Or maybe to put it another way, could you provide an example of something that measures even near perfect in all measurable attributes?"

Mark mentioned amplifiers. I'm sure he was thinking of the solid state variety. Read the claims by companies like QSC and Bryston. measurably speaking, what's not "perfect"?

On the other, a combination of power compression and the equal loudness curves are responsible for why the highs run away from the lows when you play loud - I know you know that already - I don't think wire plays much of a factor in that.

Edited February 27, 2014 by DeanG

[DrWho]

"Or maybe to put it another way, could you provide an example of something that measures even near perfect in all measurable attributes?"

Mark mentioned amplifiers. I'm sure he was thinking of the solid state variety. Read the claims by companies like QSC and Bryston. measurably speaking, what's not "perfect"?

On the other, a combination of power compression and the equal loudness curves are responsible for why the highs run away from the lows when you play loud - I know you know that already - I don't think wire plays much of a factor in that.

Are we talking marketing specs or actual real measurements? Does your QSC have an audible noise floor? What's the distortion like? How much power supply rejection you getting? Crosstalk? As crazy as this sounds, you know the distortion from your circuit before ever needing to measure it....and you're trading it for lower noise, or lower power, or cheaper cost, or trying to mask artifacts by pushing them around in level, etc...

As far as the balance difference due to voice coil heating.....it's a measurable phenomenon. I've seen the frequency response change by as much as 1-3 dB depending on the speaker if you push it loud enough. Port compression is even worse. It really has nothing to do with equal loudness perception in this scenario. In fact, turning it up should yield more bass perception, but I still think that's a bad interpretation of Fletcher Munson. Our listening should be calibrated to the real world.