Audiolabyrinth
Cables are indeed science and DO INDEED make a cable sound in specific ways. The electromagnetic spectrum could care less about price, that's a given. Price is a human emotional endeavor based on marketing and the wants of the consumer. So I don't care about price. The raw material cost of even the most expensive cables is maybe a few hundred dollars. You pay for what they sound like relative to one another, regardless of the material costs. You WANT the SOUND, not the price. The price you pay to GET the sound. Well, most of us do this. A few sure do buy based on a price and stick it in a corner like a statue never to be heard from again.
Now to the DESIGN. Sorry, but there is but ONE best way to move data in a given electromagnetic situation. Anything outside of that is simply less capable and not as linear. Keeping cables as honest as you can based on well-known principals is important.
"My" science (I never was thinking I had science so much as the world we live in has physical science) is as defined as it can get for R, L, C and skin depth. It's all 100% repeatable by anyone. Following specific design guidelines will indeed make a better cable. Most market their differentiation, not good design. Physics can't be differentiated, only misunderstood and used incorrectly.
What isn't as repeatable is the exact interaction of variables as a whole as, like I said before, we have a very nonlinear electromagnetic spectrum audio is stuck in. So the balance of all variables inside the preferred "box" of measured values will shift some depending on what speaker and amplifier you use. But, a well-designed cable has ALWAYS sounded better with any amp or speaker combination I've ever used, or compared cables on. Try C4's with W-8's and then Martin Logan Summits with McCormack amps. Can't go much more different than that. Good cables stayed good. The changes in fundamental transmission characteristics aren't so large as to require a completely different cable design.
I don't listen to cables? From my post, I'd guess all I do is listen to cable, but anyway...I have $5,000.00 NORDOST cables in my system and listened, and still am listening, to speaker cable up and down the price line. If time is somehow important in all this, I've been at it longer than you, I'm 55.
I paid thousands for the Tyr2 NORDOST XLR pre to power interconnects based on their sound (I listened to many other XLR's). They have a superior overall design that puts their overall fingerprint well inside the box of electromagnetic reason (for XLR cables). I do not find their superior overall sound to be surprising as the measurements driven by the design allows them to sound good. There are other vendors that follow strict design guidelines that allow superior performance, but way too many sell FUD and innuendo and / or jump to RF communications principals at audio frequency ranges in which nearly all the variables of influence CHANGE! So WHY ON EARTH go to RF calculation in audio cables?
I'm somewhat confused by your demand that I include price into this. No, I include DESIGN. That's it. You can ONLY buy and understand the physics of a DESIGN. The price is irrelevant to the sound. The more you understand how cables work, the less you need to pay for differentiation that is nothing more than that. Usually, the differentiates tears apart the ability of a cable to follow best in practice designs, and you pay MORE for that privilege? |
Cables sometime raise two kids and let the third one run wild. Low R and L, but high C as an example. To get low capacitance in a design (you have to hold the design exactly the same) inductance will go in the opposite direction. If it goes too wacky, it eats up amplifiers and / or requires Zobel networks to offset capacitance just like power-line power factor correction circuits. Fundamentally superior designs have low R, L and C from the get go. You can't force too high cap to go lower as it pushes inductance up. The design is simply poor. More paper, start over!
With compensation you no longer have a cable per say, but a circuit. One is useless without the other. Many do indeed market circuits. This isn't completely wrong, but it isn't an ideal "cable" either. Take away the Zobel network or "pole articulation filter(s)" and the effort is lost. So, those products are a different breed than cable design, and more expensive as they extend the margins on both fronts.
Some even try to say they "separate" the magnetic fields and electrical fields into different areas of the cable to improve the sound. I'll use Ebm's comment, "Say what??Dude??". This is impossible, as one needs the other to exist. They are tied at the hip like capacitance and inductance. Unless we move to planet ZEN, this will always be so. Not sure about ZEN, actually, has anyone been there except in the game?
The key concept to understand on fields is that the two fields take away energy, and thus the signal. They don't come from anywhere, and they don't go to anywhere instantly. They steal away energy to exist. This is distortion that is required in the real world when we pass current in two wire separated by distance, like it or not. There is no perfect cable, but there are GOOD DESIGNS.
Why do we have to live with imperfect cables? Here we go (those that don't like longer posts should leave now).
Knowing what a perfect conductor is allows you to recognize a good compromise that fits the world we live in. I was going to put this in the initial thread but I think I wore myself out writing it!
A perfect wire is / has / can;
1.0 Has ZERO resistance and thus loses no signal amplitude. This is a benign distortion in that it is linear. So in reality ZERO resistance isn't really critically important UNTIL you include the speakers X-over. Now it's a problem.
2.0 Is infinitely small. There is NO skin effect as ALL as current across ALL frequencies are IDENTICAL. Well, this is a tough one so we have to gauge (pun intended) the wire to our frequency pass band. A wire has to exist in reality, but to what size? The dimensions are limited by the designers ability to manage many wires.
3.0 Carries energy in BOTH directions at the exact same time. Umm...this is a short circuit in reality. But, if we could do that, the magnetic / electrical fields would be equal and opposite. We would not waste energy creating fields around the single full duplex audio wire. Just think infinite transient response...with a little help from a vacuum, that's next.
4.0 The dielectric would be a vacuum so we have ZERO capacitance and velocity would be 100%. Great, we just got rid of phase effects caused by band-pass filter characteristics and rise time voltage distortion.
So, we have a perfect audio wire, in our mind anyway. And, that 's a good thing when shopping for cables.
We all would like this impossible to make wire. The next best thing is a ZIP cord, many feel. Well, if you can make a zip cord with 24 AWG wire, keep it short so that the DCR doesn't influence the speakers response, and use a good dielectric they can sound pretty darn good...and they should as they obey all the required tenants of a good cable. Except that they have too high a DCR in any meaningful length and / or speaker efficiency. Once you try to make a ZIP cord bigger, it goes to hell. DCR get lower, but BIG wire loses skin current management, stranded is a bad screen and gets worse over time. This forces many smaller wires... that are hell to manage with all those wires increasing capacitance. A capacitor is a dielectric on each side of a conductor. Make it LONGER or WIDER and the capacitance goes up...a LOT. It's a squared law increase, so many wires is hard to do, not impossible. So here we are with much larger AWG requirements and about 36 wire.
A clue to excellent B and E field management is if the overall cable has lower inductance than a SINGLE "unit" used in the cable. This demonstrates, with out magic, that the design leverages superior field cancellation. You can't fake it.
Another issue to some is RF. Electrical fields leave the cable perpendicular to the surface of the wire, always, and cancel at ninety degrees intersection to other like electric fields. The problem is, you can't go "forward" with one wire stuck at ninety degrees to another. That's an "X"! So, you try to manage RF by as good a cross-field cancellation as you can with passive field cancellation through reasonable wire crossings of LESS THAN ninety degrees. Remember, the magnetic fields are cancelled by the close parallel proximity of each wire where closer is better, and the electric fields cancel by cross-field interaction cancellation. A design has passive RF cancellation or it doesn't.
Overall shields? YIKES! This pretends we ALL HAVE A PROBLEM (we do, just not in our systems) and uses a another less bad, but still bad solution...a shield. This is still bad as it drives up capacitance to unnecessary level UNLESS the second worst problem (a shield) is better than the worst (no shield). If you have no real RF issues, ditch the shield! On a speaker cable the problem is probably egress FROM your speaker cable TO you electronics, not ingress FROM the outside world into your speaker cable.
Another interesting factoid is that a dielectric polarizes by nature (that's WHY it is a dielectric) in about 10 to the minus fourteen seconds! Hey, that's fast. The polarity reversals need little help at audio frequencies to switch polarities. Yes, at HIGH, HIGH and I mean HIGH frequencies in the many giga hertz, the dielectric can lose the ability to keep up. But it’s fastest at 1 Hz and gets slower from there. But at 20 kHz it's not even an issue at all. And, most polarization modes don't even count in audio. Yes, there are about three, and I won't bore you with them all.
So, we never have perfect sound. Every cable will be a fingerprint of its design, more than materials. A good-looking people tend to always look good! The clothes aren't going to fix the major issues with the person’s looks, as an example. Buy the person FIRST. DESIGN, DESIGN, and DESIGN! No, you can't measure all the effects of a system as complex as an audio cable as it lives in the electromagnetic transition zone, and drives an unstable load (speaker) with respect to frequency. BUT, you can segregate good design from bad based on a BALANCE of electricals for truly superior cable. Take those type cables home, and I'm certain that what I've mentioned will perk up to the top of the list when you listen.
Digital cables are more about reflection errors. Lengths are important based on cable bit rate speeds that define the cables worst reflection lengths, called RL or Return Loss. Rise time errors aren't so bad with modern electronics (thank-you 100 Gig Ethernet!) and pretty short leads. If a zero or a one is sent and received, the DAC is the sound as it re assembles the data. Not an easy job, by the way! You never get to hear a zero or a one as they are simply logic "states" that the electronics uses to compile analog data. Digital starts with analog, so it's goal is to not screw that up in the middle between A to D and D to A converters.
Cables I use? I have ONE TYR 2 NORDOST XLR interconnects that is good sounding. I'll use it as a reference when I work on other XLR cables. I've auditioned the NORDOST speaker’s cables. I brought them home based on their good overall DESIGN, NOT the not so good price! But, the fact that the design is solid is weighed out in the sound. The price is emotional. I was looking at DESIGN and the connection so sonic ability. Good cable stays good all the time. You do NOT want to use cables to "tune" your system, especially if you buy them. Make sure they go ANYWHERE with ease. A well-designed neutral cable is what you really want. A perfect cable, for instance, is what we really want, and it does NOTHING to the signal. So if you are really into this for a good DESIGN cable you want an equal hand of justice to weight the attributes just so. I listen for an even tonal balance top to bottom. Trust me, use three or four cables an you'll hear an even handed product pretty quick.
Speaker cables? Oh I have some. You can't buy them. |
No, didn't forget directionality it at all, it doesn't matter. Audio signals are not truly symmetrical or balanced (plus to minus signal don't subtract to zero) as impulse noises are more positive amplitude than the negative decay. Look at a piano or drum strike, for instance. Yep, more stuff on the positive side than the negative side. But this is no reason to think the "positive side" needs help.
The waveform is alternating current and travels BOTH directions with respect to polarity. What half of the waveform do you want to be biased with the copper "grain" structure? If this was a DC signal, rubbing against the copper grain structure the wrong way may be a fun FUD argument. But since audio cables aren't a cat and get pissed when rubbed the wrong way and are AC in nature, this concept of current "direction" when it has reversal changes to even play music is rather odd.
I already discussed polarity reversal timing (fast as the dickens) in the dielectric relative to single directional changes. How on earth could AC signals even be transmitted if the dielectric were "slow"? Coaxial cables transmit in the 6 Gig frequency range all day with out directionality just fine, thank you. And, their dielectrics don't need help...
DC dielectric bias systems are also perplexing. DC bias systems further tell you that the signal could care less about directionality and polarity as a DC bias system theoretically makes it HARDER to polarize in the opposite direction of the bias. All good dielectrics polarize so fast that a DC impediment doesn't audibly affect the sound, just your pocket book. Why so much concern for the BIAS in the ONE direction only? And no, a dielectric doesn't take seconds, minutes, hours or days to polarize. It takes 10 to the minus fourteen seconds. Do they wander around after power is removed? Yes. Do they snap back into control IMMEDIATELY when a signal is applied? Yes. If you listen to music at the 10 to the tenth power in frequency...be worried, very worried, as the polarization can now not keep up. So how does a DC battery bias system add speed to an audio signal dielectric and to both polarities, not just one?
But if you must, mark your cable and add an arrow to make you feel good along with a battery and heavens don’t put the battery in the wrong direction! :)
I'm not against REAL design metrics that is repeatable and you can design around. To jump to RF wave-guide signals and force an issue at audio is simply misguided. Audio cables are hard enough without using physics the wrong way.
All the attributes I discussed are measurable at audio, and the improvements to them rational in a good design. Diectionality, DC bias and many others are not. But if you want to add nonesense to cables, why not add it to a cable that at least is designed right up to the arrows and battery installation steps? Are we tossing out what we can measure and replacing it with what we can't? Yep, we see pages of FUD "data" (differentiation)with no mention of the compliance to the physical basics. If a cable meets the basics, at least the FUD stuff didn't hurt the sound, just your pocket book. |
...The high end has known about wire directionality for at least 10 years; isn't it about time for everyone else to catch up? ...
In my opinion, no it isn't. There isn't any proof to catch-up to. Is the "audio" community the only place where sinusoidal information is transmitted? Why has no other scientific discipline, with far more fragile signals than audio noticed this phenomenon (well, it would be a phenomena if there was any true evidence it existed) and hasn't taken advantage of it?
Yep, we can take a picture of copper grains in drawn wire and say, see...the grains say to go THAT way! We can add dielectric polarization and batteries to also say we go THAT way. Until, Kickoffs current rule and polarization principals say no, we go BOTH ways and the circuit can't work. Sorry, but a picture isn't a sound. It isn't a proof, it's just the grain structure of copper, nothing more. It has to be directly applied to a signal attribute.
Alternating signals aren't "directional" they are BOTH directions. Which "direction" or polarity is it that you refer to? When there is a single shred of ANY evidence that can be put to the physical, let me see it. Don't tell me it's too complex to explain. EACH individual attribute needs to stand, with evidence, on it's own. This is how the math works. Everything can be taken apart to its constituent components. Nature is built in steps. Sometime a step is hinged. Capacitance and inductance are tied together. Magnetic fields and electric fields are tied together. One creates the other at the same instant. So they have to be looked at as a step in two directions.
Yes, the superposition of all the individual elements can be unpredictable in sound, but each one is readily analyzed on it's own. There is no, "we believe" (maybe because we paid for it?) in my world.
"When you control the mail you control information." - Newman
Ya, that's cute. Who's controlling what? I put fourth my opinion (it is mine) and we try to shut it down with the hi-end audio community acceptance of totally unproven electrical phenomena? The use of the words "hi-end audio" is interesting as it remains complete as anyone not believing it (it can't be proven) is thrown out of the fold so the community remains 100% right? Ummm,I'm pretty hi-end, and no, I don't "believe" it. Not only that, I have never heard it to accept that a proof exists to explain it.
Things like this are generally let alone as it doesn't hurt anyone except their pocket book. The FDA could care less about drugs that are placebo's if you want to buy them. You can change your oil every mile if you want to. But to try to pretend something is real based on blind faith and tell me ignoring it won't make it go away like it existed with any proof in the first place? No, I'm more scientific than that. I don't buy funny pills, I change my oil on proven service intervals, and make sure the cable I buy follow very real design principles. And, I'm sure glad the brilliant minds exist that do indeed continue to PROVE the physical world in repeatable ways. Just believe it? Are we living in the dark ages?
My viewpoint doesn't exclude those that want to address their superstitions (many are bullied into "believing" as those on this very site ridicule them to do so). I use the term superstition as simply that, a belief that an outcome is expected without ant real proof, not as a put-down.
Me, I'm pretty thick skinned and failing to "believe" isn't one of the things that keep me up at night. You better be glad those that design the equipment that we use everyday don't believe faith, as you can't trick the science in those circuits. Mark your cables, add batteries, do what ever, but DO NOT ignore what is proven and pay for those that attributes that are faith based until every real facet of design is well accounted for.
With modern DSP electronics, you could indeed pass a white noise (equal amplitude at all frequencies) through a know system and do a balanced circuit remainder function to "pull-out" the difference signature of various cables. This would be magnitude only. Phase is a more challenging measurement. You can sweep a cable for phase with respect to frequency and probably do a similar analysis. As different as cable do sound good equipment should be able to quantify some, not all, of what we hear.
Count me in when we try to prove what we all can really hear. Have you ever cut your 10K cable in two and pulled all the conductors out of and inch of your cable, and had the grain structure analyzed? We're they really in the same draw direction? OK, the box said they were.
With so called single grain copper it should not be an issue as the copper grain boundaries are gone, and the grains lay parallel throughout the wire till the ends are reached (limited length based on AWG size). You can pay a lot of money to guarantee you're right about directionality by theoretically avoiding it? Well, that's a way to handle it. Just do it in a good DESIGN.
"Maybe if we ignore wire directionality it'll just go away"
Well, what if I ignore something that doesn't exist verses something that does, will it appear? Maybe I'm on your side after all.
I will be VERY careful to not ignore R, L and C while playing my stereo, though, as the consequences could be catastrophic if they go away.
Better yet, see how many cables you can find that adhere to good design practices (ya, even the ones that add snake oil over a good design). The physics say that the proper management of the alternating signal will yield a nice sounding cable. Make a list to go shopping by. Buy more of the real deal in design.
|
These are my general observations from the physics of how all this works. I like repeatable conclusion backed by calculation as that's what physics is, the real world reduced to numbers (with some guesswork even in those!). But, It's at least a common ground for advancement. I just have a hard time with guessing about stuff. I’m open to repeatable, measurable calculations. Since cables are complex superposition of materials and relationships, with each selected or calculated on its own, the final sonic results that are derived are somewhat a black art, no argument there.
Gregm - The skin depth is one of those, when do you stop, arguments. The smaller the wire, the more consistent the current in that wire at all frequencies. It can, in theory, never be the same except at DC. But, lets get real, it can get darn near the same (can't hear it anymore) with numerous small wires. practical reasons usually mean 23-24 AWG as it gets terribly hard to manage more wire than that and not destroy the capacitance or inductance balance. All those wires are trying to hurt the design EXCEPT for skin depth. This is a somewhat an evil business. Every variable is a cancer to the other! A speaker cable is a balance of many attributes, but the overall balance should fall into an optimal compromise. No, they all won't be the same. If you change any one element, the overall relationships all have to move. The physical relationships of the wire and materials result in the blended finished electricals. The compromises made are EVIDENT and KNOWN by the designer, though. This isn't guessing.
So the optimal gauge is going to be at LEAST no bigger than the radius of the wire at the highest frequency of interest. 20 K is about 18-mils. This is just the definition of skin depth, though (37% current differential at that frequency). How small (reduce the current differential even more) you go is going to be locked in by DCR requirement that will rise as you try to optimize one attribute, (forced to use more small wires!) and how far you think your ears can perceive the skin depth issues. Don't rob Peter too much to pay Paul! BALANCE is the key for an overall nice cable. Nothing I have posted here is a secret other than physics we don't understand to put into numbers. But, we have plenty we do understand that can leverage to make better, more logical, cables.
Audiolabyrinth - Buy the sound, yes, not the price. OK, they might be expensive, they might not be. It all depends on what that "differentiation" cost you! Be careful to avoid "trust me" engineering as what you hear, is more than likely the true basics being correct and not the snake oil applied after the fact. Nothing wrong with snake oil, I suppose, as long as the fundamentals are correct. But, many a cable is JUST snake oil and expensive…and sound poorly.
I look at good sounding cables this way, if you have even 10K in a system a $2,500.00 cable that really sounds good to you is about as cheap an upgrade as you get in audio now-a-days. Yes, it's expensive, but can you beat the sound improvements for the price paid? Here is the BUT part, LISTEN to the cables and try to ignore the price and "packaging / marketing". If the fundamentals are there, the cable should sound pretty nice, as the real world physics will allow it to. You can't cheat Mother Nature. The same relationships are in EVERY cable.
|
Almarg,
Yes, you're input is spot-on. One is a goof, the other is more accurate to the current gradient effect based on skin depth, and the last is I can't count zeros! It would be nice to be able to EDIT as the author to improve accuracy, and make it easier on the reader to get the right info fast. Well, as fast as my stuff will allow (ya, I hear you...shut-up already!).
I'll admit the effects of skin depth are "real" based on calculations but the audible nature is hard to pin down EXCEPT that I have used IDENTICAL R, L and C cable on purpose and listened to designs with little regards (four wires) to no regards (two big wires) for skin depth. Those with more exceptional regards to skin depth at 12-20 wires in each polarity had improvements that were immediate. So I have to look at the overall design, and try to figure out what is going on as you use multiple wires. What is really skin depth management and what is something else? In theory, it can't be PHASE as the capacitance is VERY low (less than 30 pf/foot)in both designs.
Calculation say that capacitive roll-off first order filter PHASE shouldn't be audible, either, with reasonable capacitance. But, a superior design can simply bypass this effect to make it a none issue and still get sub 0.150 uH/foot inductance.
I agree, and point out, that speaker cables between a speaker / amp combination will sound more or less better moving from system to system. But, I also point out that "I" have not had an exceptional cable fall in my rankings switching systems, from dynamic drivers to electrostatic panels.
As to wire coatings and wire...has anyone heard tough pitch copper (~1500 grains per foot) verses OFC copper (~300 grains per foot) verses functionally perfect copper (~30 grains per inch) and single crystal copper in the exact same design? Go ahead and add verses silver plated to any type. I'm hot on the trail of trying to do just this between at least two copper types if I can.
Yes, I added copper quality, but tend to throw it down the list as good design eclipses what the copper supposedly brings to the table. Copper seems to be a knife in a gunfight. Yes, it can get some of the job done if the heavy lifting is already out of the way.
This is facinating stuff, in that it's like an unsupervised free for all. Like the light out in the cafeteria...seriously fun, but when the light come on, the mess has to be accounted for and blame placed on the right suspects. Let's get the light on, shall we? |
Jneutron
...First however, one clarification on a comment you made earlier, that a vacuum dielectric would have no capacitance. Totally wrong. Capacitance is proportional to epsilon free space times epsilon relative. For a vacuum, epsilon relative is 1, and epsilon free space is 8.854 times 10e-12 farads/meter. I DO NOT UNDERSTAND HOW AN ENGINEER COULD MAKE THIS MISTAKE....
Good, this is correct. Sometimes my head is in a vacuum...sometime not. That number is much higher than I would have expected, too.
...""I've stated that as well If you wish a general feel, the Belden website illustrates this for a general 75 ohm cable. For audio use, I recommend a value 2 to 6 times the hf value""..
I'd rather know what they are at audio, but they are of little use in a speaker cable circuit load which do not act like true transmission lines in the manner we use them. On one hand we want precision, and the other we don’t, it seems. I do agree this can get really hard to characterize, and sooner or later it is indeed precise enough.
...You need better sites. Just because it's on the internet doesn't mean it is correct...
Yes, the methods commonly used are correct, and verified in three places. But, you feel at lower frequencies, like impedance, differing assumptions are made. So there may be different methods, but the object is to determine exactly why, and to what degree the various methods are limiting performance. If the equations overlap, and there is no ideal wire small enough to be "perfect", than knowing approximately where to be on wire size is OK. Since we can’t make a perfect skin effect wire, at what point is one “good enough”? You comments, please.
""If this question is intended for me, you are barking up the wrong tree. What I speak of here is a very small subset of what I do for a living.""
Nope, I need not speak of people, I speak of principals only...so I did not intend the statement to be derogatory to anyone so much as getting the right information. Some feel God like in their presentations, some don't. Your posts aren't really about information they seem to be about you. I’ve said this before, the “push” (fill in your awesome ness) then “tell” (your story or throw your arrows) is not a sign of maturity or constructive advancement of the subject matter. I’m here for the subject matter and advancing the understanding for all involved. I haven’t figure out your purpose yet.
...Be nice, get nice. Be arrogant, get same...
Yes, I agree with that. But, I think some of this deservedly stuck to me with a little bouncing off me and sticking to you. True class keeps its class at all times. Yea, it's tough, I know. We do our best. Me, I'm fine with good information, being corrected and moving on. I don't use my knowledge to go hunting for those that don't without any real regards for the actual topic, just the hunt. Grow some civility with your posts and more will follow. You seem to be aware of this, so why not change it? Witness...
“…ps. are all posts on this site moderator approved, or is this a trial period for bad eggs like me?…
No, we love having you. What you say is good enough. We’ll make good use of your input and being respected for that should be a better feather in your cap than simply using the push tell prose (you guys stop me when I do it to…yes, you too Mapman!)to mash people.
OK, enough of that lets do something.
Your comment here Jneutron, …cable impedance changes at all frequencies based on the amount of real to imaginary components in the complex impedance. The RF return loss will vary significantly with each frequency point. The higher the imaginary component, the higher the reflection signal and the worse the signal transfer to the resistive load. The cable can be 75-ohms at all frequencies even (not likely!), but have differing reflection coefficients. Throw that into a speaker resistive load, and it gets messy for sure. Low frequency signal is a whole different world.
As low as the resistance is in 10 AWG speaker cable, arriving at an eight ohm load requires a good amount of "something" imaginary (most cables end up most capacitive) to create a complex impedance that high relative to the cable's very low resistive component. I don't see realistically consistently matching the speaker cable to the load by adding capacitance. True, some cable can be made to a specific length to “add” capacitance. Has anyone played with this? Not that it would be easy to do (cut up expensive cables) and or buy various lengths of cable).
With such LONG wavelengths, it seems odd to suggest reflections are from true wavelength load reflections. Yes, I hear you on measuring the reflections, but the load has to be taken into account, too. RF uses a resistor equal to the characteristic impedance. I’m still not convinced of how reflections allows us to convey this as a transmission line.
For power transfer, the speakers get mighty hot, the cables not so much. I’m not clear on exactly where you suggest all the energy is dissipated any why. Resistors absorb power, and the cable is not a good resistor. Audio is still a pretty slow AC signal not too far removed from DC. We want power on the load.
When the waves begins to approach the length of the cable used it is plausible to consider "reflections" as a transmission line. We aren't really close to that in audio. The cable isn’t alone in all this, the back EMF from the speaker is a hell of a circuit in itself.
When you “freeze” the circuit, it’s a static situation. The current and voltage can be examined in turn. AC voltage signals in a transmission line are “pulsing” in the line so to speak, at each frequency if the ideal transmission line acts like it should. But a speaker cable with LONG wavelengths into a speaker with low output impedance from an amplifier?
Trelja
No, I have one NORDOST cable as a reference. Trust me, it was much better than the alternatives. And, to "end up with" is a certainty no matter what we use, you included. Are you to be a poorer judge of cables if one is deemed better than yours is by someone? Many feel differently about NORDOST speaker cables (I don't have those). But enough on that, you'll have Jneutron on your case in no time. Tough love, but he is knowledgable.
Mapman
"I like Jneutron's statement earlier that it is design that matters, not cost. A good mantra for any decision making process involving technology"
No, that was I, go to the top of the thread..."DESIGN is first...." This was the purpose of the thread, to put design ahead of myth that is paid for.
And, my point was, and still is, to get audio cable out of the closet. Many are simply afraid of conflict, as that seems to be the general direction that "knowledge" tends to be going all of a sudden, and NOT to the real subject matter. WHAT are the major issues with audio cable design that can indeed be characterized by known principals?
I stated what is somewhat relevant, but the issue is to drill that down to a finer level so we can all better examine cable constructions for neutrality and possible performance.
I'd rather have half the knowledge and work towards a solution than all the knowledge and sit there with it. Jneutron, where are "our solutions" as you see them? What attributes do you look for?
This group should use each member’s input to help move the topics forward, not throw arrows or use indefensible arguments that can’t be analyzed when this thread is about just that. Believing is fine, but this is for the other half.
|
Go here;
http://www.audioholics.com/education/cables/debunking-the-myth-of-speaker-cable-resonance
In reality cables DO NOT resonate at all! The model represented here is single RLC lumped circuit for simplicity and is only accurate at audio frequencies for circuit analysis. A speaker cable is actually a distributed element and should be represented as infinite number of lumped RLC models. As an infinite number of lumped RLC circuits are modeled becoming its true distributed form factor, we see the resonance frequency go to infinity.
In addition, once we approach much higher frequencies such as in the RF region we must re-evaluate the cable as a transmission line. In that respect the characteristic impedance becomes the SQRT (L/C) =SQRT(8.8*10^-6/700*10^-12) = 112 ohms. So if our source and load terminations at transmission line frequencies (1/6th the wavelength) do not match, we see reflections in the line, which can appear like a resonance behavior, but in reality are simply reflections or power loss down the line.
Also note that when an exotic cable vendor claims Inductance, Capacitance and Resistance dramatically varies within the audio band, that this is more total and utter nonsense as can be seen in the following real world measurements...
As always, we welcome any cable vendor to furnish us proof of their claims, and cable samples for us to conduct our own testing for verification purposes. I agree, the FTC should be involved in this business, as it is a consumer product based on engineering truths that must not be ignored. - edited to remove vendors.
END
I don't know you everyone.
We look at 20KHz signal (lower than this is even more far fetched) that are far, far too long to properly conduct as a transmission line, add-in the fact that the load is not matched to the cable, and varies with frequency as does the cable too and you have a line model that is closer to your 110-volt wall outlet than your CATV outlet.
END
http://sound.westhost.com/cable-z.htm
In order to obtain a low characteristic impedance, it is necessary to have very low inductance and relatively high capacitance, and the high capacitance may impose serious constraints on the amplifier. Indeed, many amplifiers will become unstable if there is sufficient capacitance connected directly to the output, causing oscillation which may damage the amplifier. As described above, regardless of anything else, the cable does not act as a true transmission line at audio frequencies, and claims to the contrary are fallacious.
Matched impedances ensure maximum power transfer from source to load, and this is obviously very important for RF transmitters and telephony applications. It is completely irrelevant for a solid state audio power amplifier however, since the drive principle (known as voltage drive, or constant voltage) does not rely on maximum power transfer, but relies instead on the amplifier maintaining a low output impedance with respect to the load.
Even though most power amplifiers are limited to at most a few hundred kHz or so, there can still be some energy at higher frequencies - typically noise. What often happens is that an amp can be quite stable with a capacitive load and no signal, but as soon as it is driven it "excites" the whole system, and it then bursts into sustained oscillation.
At audio frequencies, speaker cables are not transmission lines. They are merely cables, with inductance, capacitance and resistance. Despite popular belief, they are bereft of any magical properties, only physics.
It is worth noting that a cable will never act as a true transmission line with a defined (and maintained) Zo unless its source and load impedances are equal to the line impedance. This means that no audio cable will ever be a transmission line, (almost) regardless of length, unless the amplifier output impedance, cable impedance and load impedance are all equal at all frequencies within the desired range. No known amplifier or loudspeaker system can meet these criteria. Alternatively, the cable may be infinitely long, however this is usually impractical in a domestic environment.
END
The above is pretty much what I've said all along. And will continue to say. Keep capacitance lower is better, and the cable is NOT a transmission line.
I do not agree that wire is wire to the extent that audioholics goes to. Make a cable with two large stranded conductors and one with multiple solid AWG strands of the same AWG (or just a different design) and the differences are definitely there.
I'd would indeed like to visit audioholics with the two types of cable and set-down and measure the cables and have them formulate the impact of the design on the sound through measurements. I haven't seen this done, so you can't deny that it could not be done. This would be tremendously informative.
I'm not going to hide behind "my" hearing and say XYZ exists (transmission-line effects) or any other "invisible" attribute. This is to properly define a good audio cable with realistic attributes everyone can enjoy.
The wealth of evidence is not in the favor of audio as a transmission line. |
Jneutron
You seem to be a little too easy to get upset about some things so it's going to be hard to please you. I'll do my best.
Zero capacitance? Sure, my perfect conductor example was just that. You can't do anything in that cables example let alone the capacitance. Not sure why you got so excited about it. Yes, a dielectric constant of 1.0 is a give away to the fact that capacitance can't be zero in THIS world, but I wasn't limiting the perfection to the real world but the "perfect" world where things do go infinitely fast, have no resistance, ETC. I see no sense in making a "perfect" wire half-in one world and then the other. So, I put it all in ONE world.
And yes, you're right, I do not have a source for very low frequency skin effect. Nor does it seem anyone else! As long as the numbers are "close" and you use a wire size that is smaller than the calculation, and use the number of wire to meet DCR things should get into the reasonable range.
As far as transmission line effects well, the source impedance sure is small at less than 0.05-ohms on average of an amplifier's output stage. The cable impedance is small (looking at the real component of the cable) compared to the speaker's varying input impedance so it is hard to imagine the speaker cable as a classic transmission line (matched source, line and load impedance’s). Not to mention the wavelength are WAY long to even begin to couple between the source and the load. A reflection bridge certainly can show reflections but to say this is a transmission line?
If I shove my amplifier up to the speaker terminals where is all the power dissipation going? the speaker. If I stick a cable in there, the power is still going almost all into the speaker and not the cable. The L and C energy eventually goes into the load but is lagging one way or the other. Probably not a good thing.
True, the "impedance" of the cable can be complex in nature and a higher vector magnitude per low frequency equations (mostly capacitive, as the resistive load value is so small in the cable). But at such a low frequency it's really hard to see this as a transmission line, or impedance values near the speakers input impedance.
How do we negate the effects of the LONG wavelengths relative to the line length at audio? Even worse, go lower than 20KHz.
I'm all ears on your low frequency transmission line model (pun intended)and impedance matching. The Z=SQRT(L/C) is only good above 1 MHz.
The ability to design to metrics that are repeatable would be a great benefit to designers who want to use the best of what's really capable in design without "faith" based engineering. But, I'd rather try to do the right thing with a few bumps in the road than do the wrong thing perfectly. Too little risk can limit the outcome as bad as anything else. Yes, we lose a few that insist on only what they perceive and hear. That's fine, but for right now I want to concentrate on the "knowns" (yes, even the one's I don't know!) to look for in a nice, and reasonable, cable. Design elements that are always beneficial to sound.
Bridges go up and fall down, space shuttles go up and explode... all these things have "experts" at the wheel and still failed. People may be experts, but the ones that do their jobs can overreach even their understanding(s).
So, as far as being the second one, I sure didn't know I was competing with the first. Sorry about that. If I only stick with what I know today, what benefits do I achieve tomorrow?
Looking at cable, you can't see a common design thread across the lot of them that indicates forward thinking to a known set of conditions. I can even see a few, not hundreds, of designs for a specific set of amp (SS or VALVE)/speaker (dynamic or electrostatic) combinations.
I listen to cable with the same R, L and C and am amazed at how different they sound. I am not even close to shutting down my ears. That’s what got me in this mess! How do you recognize correctly made audio speaker cables.
I did look back at some of your post so as NOT to pester you, and yes, the main things my ears hear is much improved openness, imaging stability, precise location of the image and debth in good speaker cords.
I can't say I grasp speaker cable impedance at such low frequencies as they "rise" as the frequency drops, making consistent low impedances at audio seem implausible, at least to my way of thinking about the measurements. My guess is if you look at the cable like a T-line, the impedance is the same at any cable length. Of course, the low pass nature of the cable changes too.
With your respone Jneutron, I close comments in this post. I've bugged you all enough. |
Jneutron -How are you calculating impedance at audio to such low levels? The impedance rises rapidly at audio frequencies and is tremendously non linear. A cable can very easily be 600-ohms at 100 Hz, and drop to 50-ohms at 20 Khz with the open short method.
In don't see any liberties being taken to reduce an engineer's work load when it isn't even working as a transmission line. Oh it's "transmitting" all right, but not voltage.
As short as these cables are, open - short method is used to derive "impedance" even though there is no real impedance as the cable are far too short to manage such LONG audio wavelength. To be a factor, the cable length has to be at least 10X or more the quarter wave length of the frequency of interest. This relates to the fact that a voltage change has to happen BEFORE it gets to the end of the cable and audio speaker cables transit times are too fast for this to happen.
There is NOT true impedance matching reflections in audio cables caused by reactive impedance values verses resistive. The back EMF from your amplifier is many times more severe than so called "reflections" of a hi-current signal in a speaker cable. True, no cable has a 100% transfer to the load (pure resistor), but I think it is somewhat a mistatement to convey it's an "impedance" as it falls well outside what is known under a transmission line situation.
Speaker cable deal in current / power transfer where transmission lines deal in just voltage transfer. With POWER transfer you want the LOAD to be a MUCH HIGHER resistance than the amplifier output or cable so the "power" is dropped in the speaker and not on the amp outputs or the cable. You want the cable to be a pure resistor, too, just NOT a very big one.
So let's say you arrive at your "complex" impedance by adding the vector sum of the real and imaginary (capacitance and inductance) parts. That would be HUGELY capacitive to get to an 8-ohms value with such low real component resistance and inductance. Most of the magnitude is a CAPACITOR! Why on earth would you want to load the circuit with all that capacitance when POWER or VOLTAGE is NOT dropped across imaginary values but only the resistive one? Capacitors and Inductors store voltage and current, only to release it later on (minus their internal resistance, anyway). Add a bunch of imaginary capacitive component to your speaker leads and you create a messy situation even at RF. Talk about phase shift and imaging issues, there is no transfer of energy, just storage and release of energy at in opportune times. It doesn't sit around forever. The higher the capacitance, the worse it gets. We aren't storing nuts for the winter, we want to eat them as the come down the line.
Power is current squared time resistance, and I sure don't want as much POWER dropped on my cable as the speaker (same "impedance")! Of course this doesn't happen since the cable is NOT high resistance so POWER can not be dropped on the speaker cable. But, highly capacitive leads DO NOT aid the transfer of power to the load, either. I also don't want much dropped inside the amps output stage (usually less than 0.05-OHMS). I see nothing here that says you would want to, or can, match a cable to a speaker.
The skin effect calculation is "wrong"? Well, All I see is you have a different opinion right now. Multiple credible sites use the most common methods and all arrive at about 18-mils at 20 KHz. Where is your documentation on your method? I agree that "approximations" can boil stuff down too far. Saying so is one thing, showing us is another. We're all tired of sayings.
For delay, Velocity of Propogation is one over the square root of the dielectric constant. All good dielectrics are stable from 1 KHz to well into the GHz range, Teflon changes less than +/- 0.05 and is 2.15 nominal from 1KHz to 10 Gig and more. The delay is JUST the effects from the dielectric material group delay (some have more than ONE dielectric), and nothing more. It is design agnostic, zip cords or otherwise. You can measure the delay at 20 Hz, but it isn't going to make a huge difference in the arrival time at the end of a ten foot cable.
I still see the yearning to be like the RF guys. Why?
As for others, where's the beef in your audible beliefs? In God we trust, all else bring data.
Directionality - rice puffs.
Cryogenic copper - rice puffs.
We need to start digging out provable facts and using them, not just "hearing" them, to make better and practically priced audio cables. Are people being taken advantage of? Well, what do YOU think? Information is power. Get some.
I have thick skin so getting some good heads to knock me around is actually fun. Learning is NOT a spectator sport. |
Some have asked about ZOBEL networks as we've moved along. Here is a good analysis of what they can do for you;
http://sound.westhost.com/cable-z.htm
Granted, we don't have a true transmission line with a speaker cable since output, cable and load impedances aren't ever matched and the line is too short, but as Jneutron pointed out, there are termination reflections based on cable length. A Zobel network can remove them WHEN you know the "impedance" of your speaker cable. The article doesn't state the frequency of the cable impedance calculation, which is varying with frequency pretty badly. The objective is more for amplifier stability than cable "sound" per say. Still, some amps may benefit from Zobel networks. Those that want to try one, here is a good place to start.
You need to use the right quality / type parts, and depending on your amplifier, you could get tertiary benefits. Long cable low cap or short higher cap cables are more likely to see enhanced your amplifiers performance with a speaker end Zobel network.
Jneutron, where should user's of Zobel networks calculate the impedance? What do you use when you suggest cable "impedance"? |
