|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
98.95.209.195
In Reply to: RE: Skin effect roll of chart from Lavy forum, almost zero!!!!!! posted by Tweaker456 on March 17, 2014 at 12:21:38
As far as I can tell, no one ever claimed it WAS about the amplitude, that is a red herring thrown out by many of the objectivists.
No, the one currently measurable scientific parameter that concerns what we might be hearing with skin effect is phase shift.
Under normal conditions, with a first order roll-off, when the amplitude is at - 3dB, the phase shift is at 45 degrees.
Now, regarding the chart from somewhere, this chart may or may not be accurate. It turns out that the common text book equation for skin effect pertains to a flat sheet of material, and not necessarily a round wire out in space all by itself. (which is also a mis-leading concept, see later)
So, unless the originators of the chart can assure you that it uses the correct formula, then it may not be accurate.
Other charts say other things, see:
http://www.powerstream.com/Wire_Size.htm
Is this one right? I don't know where the column for
"Maximum frequency for 100% skin depth for solid conductor copper"
came from, but it is surely different from the one you are citing.
Practical experience with many DIYers and cable manufacturer's shows that wire gauges above 24-22 tend to cause a smearing and indistinct treble, esp. transients seem dulled or blunted. This is a fairly well known and accepted sonic occurrence, and does not need additional back-up or verification. Typically, only die-hard naysayers insist that it can't be so, just look at the amplitude charts .......
The definition of skin effect involves the depth that the current flow density reaches approx. 37% of that of the current flow density at the surface. Note that this definition does not state the the current flow at the surface at any given frequency is the same as it is at some lower frequency.
So when a given chart (or person), is saying that a wire is good till XX frequency, where the diameter of the wire is equal to or less than the skin depth, it means that the current level in the center is only about 1/3 that of the level at the surface. While this may not be enough to induce significant amplitude loss, or even very much phase shift, it is certainly enough of a difference in current density to drive the phenomenon know as "strand jumping", thus rendering an argument for the use of solid core wire over stranded, even with relatively small wire sizes.
Of course, one can get all wrapped up in the issues of skin effect, or the additional effect of self-inductive roll-off, and the effect of twisting a hot and ground pair together tightly so as to minimize that self-inductive roll-off. That is where the one wire out in space by itself is also not a realistic method of charting the HF effects of a wire carrying an audio signal.
Or, one can trust one's ears, and go with what sounds good.
If you are happy with 12 ga. speaker wires, then enjoy them.
There is no need to justify that by some sort of clutching at a "scientifically based straw", you hear what you hear.
The only reason that any folks are trying to bring up the experience based facts about the large diameter wires, is that ultimately, these size wires ARE acting as a sort of HF filter, softening and smoothing the end result ever so slightly. In your system, with your components, this may be just right to make things turn out for the good.
Later on, if you upgrade another component, or change out another cable, or try a new tweak, the sum total may not sound quite right, it may sound a bit soft, etc. That could then be the large diameter speaker cable wire coming into play (or not, there may be other components that are in play too). It does help to know about this, and keep it handy as a point to consider down the road for future upgrades, tweaks, etc.
In the meantime, enjoy your cables, listen happily and don't worry about a "scientific" justification, it's all good what sounds good.
Jon Risch
Follow Ups:
Fat wire has less inductance.
This is only true for a single wire out in space all by itself.
It is only true in absolute terms as well. That means that as a percentage of the DCR, the inductance does not go down as quickly as the DCR does with the increasing wire diameter.
Then, once you try to pair a larger wire with another larger wire, the spacing that is forced by the larger diameter is greater than the spacing for a smaller wire (wire center to wire center), thus the cancellation of the self-inductance by the mutual inductance coupling is reduced compared to a smaller diameter pair of wires.
Since most all audio wire pairs, especially speaker cable wires, are placed in close proximity to each other, when the spacing is larger, the self-inductance dominates over the mutual inductance, and the net result is a higher inductance for the working pair.
Looking at a single wire in isolation is a red herring, and does not reflect the reality of the real world construction or results.
Jon Risch
The 100% thing came from the same chart you sent. Tweaker
Yes,
I see the charts are identical in content RE the skin effect frequencies.
Jon Risch
A roll off of 6db or more per octive starting at 4150 for 12awg seems to me to likely be a problem, thats more than 12db down in amplitude at 20k. Harmonics and some fundamentals would be effected, probably audible. Now if the roll off once again started at 4150 and signal was down at (point) .14 db at 20k, what audioholics is saying than that would likely be little problem. He is also saying there is no phase shift. The chart you sent me was where the roll off starts, not measured decrease in dbs down, amplitude. I don't see, I can't find any evidence for a 6 db roll off. Not saying it's not there, but I can't find it. I don't understand the down 37% current density thing. I know what current is but I don't know what current density is,yet. Isn't this fun. Tweaker
If you only look at one dimension among many, then you can ignore all the other dimensions completely. Bliss.
You have to realize that places like Audioholics are NOT neutral in their presentation of material about audio. They have a definite agenda, and the gist of it is to remove anything but their one dimensional view of the audio world from the complete picture.
Early in their existence, while they were struggling to make a name for themselves, I came across a web sight where they were profiled as a new online business. Their avowed business purpose: to sell audio equipment and make a profit from the sheer traffic to the site on the advertising they have there, via a variety of means, which included stoking controversy in the audio industry and providing discussion forums to facilitate this goal. These goals and objectives, and the means to reach them, were all being discussed fairly openly on this site, in the manner of a businesses interview.
Once I realized that they weren't genuine, their actions and so-called scientific articles made a lot more sense. They weren't really interested in the truth, only in presenting their version of it, and stirring up as much controversy with subjectivist's as they possibly could.
Interestingly, this web site where I saw this industry profile and interview was later blocked from open viewing, it require a subscription and registration to view the pages, and archived articles were only opened at the discretion of the web site manager/editor.
Do you know how many e-mails I was sent saying they were going to do an actual listening test to high end cables? They were even going to test my DIY CC89259 speaker cables. Instead, after many promises and delays, they simply performed some measurements using some very one dimensional lab equipment, and postulated ALL of their conclusions from their own special interpretation of simple and basic LCR parameters.
BTW, they didn't even bother to measure a sample of the CC89259, but instead, used LCR figures I had measured and published. There was no opportunity to cross-check their measurements, as the cables they did actually measure were most all fairly new brand names of the oversized zip cord variety. It turned out that virtually all of the brands "tested" were offered for sale later on at the site as well. How convenient.
Cables were not looked at as a series of extended networks, or transmission lines, or any other quite acceptable models currently considered correct and acceptable in the scientific world, they looked strictly at simple numbers based on lumped parameter models. This was "explained" as relevant based on their own interpretation of what was relevant. Any other way of looking at cable measurements, cable measurement data, or other models or simulations or network type complex measurements were ignored or deemed irrelevant, based strictly on their POV, and not necessarily on science or current signal transmission theories or practice.
If you lock in on a one dimensional parameter POV, then everything can then be reduced to a single parameter number which can then be declared (erroneously or not) a figure of merit.
The single most relevant parameter they chose to hang their hat on was resistance, and therefore by inference, simple amplitude response as a figure of merit.
Quite simply, if amplitude was all it took to do audio right, then all anyone would need as a really decent equalizer, and voila! audio Nirvana is yours.
Guess what, it doesn't work. You can equalize till you are blue in the face, and you won't achieve anything even remotely satisfying sonically.
A $3,000 system can not be equalized to sound like a $30,000 system, nor can a $30,000 system be equalized to sound like a $300,000 system.
You need to realize, that if all of the BS that objectivist's spout has any basis in truth, that according to their own belief system, that it should be possible to make the $30,000 system sound as good or better than the $300,000 system does without benefit of a separate EQ.
On a $30,000 system, the noise should be low enough, the distortion low enough, the various parameters that they deem required to reach audio perfection are all going to be present, and if their rhetoric is to be believed, the simple act of EQ'ing the system should do the trick.
But it doesn't work that way. The "intangibles" of clarity, soundstage depth, pinpoint imaging, jump factor, PRAT and all the other high-end audio system benefits are present or not in varying degrees. A properly and well set-up $300,000 system will usually have those intangibles in spades compared to the $30,000 system, and no equalizer now known to man can "fix" the $30,000 system to get it to the next level.
If it could be done, someone would have done it by now, and sold it to the masses. Since it has in fact, not happened, I rest my case.
So if you want to be blinded by "science", go to Audioholics and read all of their "technical" articles, the end result will insure that you can only look at audio one confusing and constricted dimension at a time. After all of that reading, a Sony receiver, a simple HT cube and mini-sub speaker system, a $99 Blu-ray player, and you are set for life. As long as it is all hooked up with 10 Ga. wire of course.
Note that the above represent my own opinions after my own personal experiences with Audioholics, and do not reflect the views of anyone else at AA including the Bored or Rod.
Jon Risch
First of all I'm not just looking at one thing. I am listening to my interconnect wire experiments, and up to 16awg I feel the sound gets better, fuller bass, more natural timbers, sibilance has a more realistic timber,cymbols sound fatter, no perception of damaging rolloff. I am in my 60's. This once again is a simple parallel mag wire design. I remain incredulous that people love thin wire. RG59 is 18awg, so I read. Not exactly thin. NOW to my point. Jon you said that the roll off is 6db per octive from skin effect and more from phase distortion. Can you please tell me where this info comes from. Seems like you and Gene are not the best of buddies. The Lavy site info shows essentially no roll off. This should be easy. I no longer have test equiptment. Jaeger
Quite simply, skin effect is shown to have a 6 dB per octave roll-off per the laws of physics. So is self-inductance, when NOT countered with an near equal dose of mutual coupling inductance. This is shown by basic physics equations, and is available in many a Physics textbook for perusal.
However, what ever modifies the skin effect influences, and anything that changes the surface area versus the cross-sectional volume will modify the skin effect numbers, effectively changing the roll-off point hinge frequency, or more likely in the real world, modifying it with frequency, so that the corner frequency starts out at the classic text book spot, but the skin effect is reduced at higher frequencies than the hinge point due to other factors.
It turns out that if looked at in isolation, skin effect and self-inductance add up. However, the bizarre truth is that if there is a large amount of self-inductance, this actually reduces the skin effect impact on the signal to some extant. How much depends on how aggressive the self-inductance is, and how rapidly it engages.
That last part may sound odd, but in the case of a steel cored wire, the self-inductance can skyrocket above some certain frequency, and skin effect may fall off quite a bit until it re-establishes dominance of the situation at some higher than typical frequency. Extremely low inductance cables can actually have more of a percentage of skin effect occurring than a high inductance cable.
The biggest problem is that these effects interact AND are frequency dependent, which also tends to cause them to interact yet again with one another.
Trying to model it or to measure the effects of both effects is troublesome and difficult at best. Again, most of the charts are textbook calculations based on the skin effect in a flat planar stretch of metal with a single wire suspended over it. Measurements made with relatively simple LCR meters or Wheatstone bridges, can not and will not show all the actual goings on of the situation with a pair of wires carrying current in opposite directions while in near proximity.
RE your positive experiences with larger diameter wire, as I said in my earlier post:
one can trust one's ears, and go with what sounds good.
If you are happy with 12 ga. speaker wires, then enjoy them.
The input you have received from other regular members of the Cable Asylum regarding what the larger diameter wires tend to do is valid and good info, backed up by years of experience, and basically the same thing that a great many different folks have discovered by trial and error.
None of that practical real-world experience requires any validation via skin effect, self-inductance, or whatever.
If your experience is different, then go with what you hear on your system.
Just realize that it may not always sound that way, it may change as time goes on, or when you change or upgrade your equipment, etc.
Happy listening.
Jon Risch
I'm up to 16awg for interconnect and find it to sound very good. No problems in the treble or soundstage that I can hear. Aggressive peaks are toned down, maybe like you described thick wire can or does do. The 6 db rolloff completely contradicts the science I can find and what your friend Gene Dell Salla says he measured. There are others who like certain interconnects, if I remember a silver transcendance interconnect that are 16awg. I think I'll stop at 16. Thanks for the responce. Tweaker
I think that you are confusing the point on the charts where they say the skin effect "starts" with the -3 dB point, They are NOT the same point!
The center of the wire has to be many skin depths below the surface before you reach the - 3 dB point, where the asymptote of the 6 dB (first order) roll-off is said to begin.
So when the chart says that 12 ga. has a skin depth corner of 4150 Hz, that is NOT the -3 dB point. The -3 dB point DUE TO SKIN EFFECT ALONE is much higher up, usually self-inductance causes a roll-off before the skin effect does.
A skin depth corner of 4150 Hz is the point where the skin effect starts to screw with the self-inductance, and the phase shift begins, and by many accounts, where the wire no longer faithfully passes musical _transients_ accurately.
Please note that steady state sine wave measurements are NOT going to provide the same info as a transient pulse in terms of where the problems start to crop up. This is another thing that Audioholics sorta ignores since it throws a wrench in their neat little world of single metric numbers and figures of merit.
I think that you are fixating on skin effect, when it is the sum total of all the cables materials, physics parameters, etc., that makes up the sonic signature of any given cable or pair of wires.
Jon Risch
Deleted.
Edits: 03/27/14 03/31/14
I think that you need to re-read what I wrote, rather than jumping to conclusions and setting up a straw-man based on said mis-understanding.
Jon Risch
Deleted.
Edits: 03/29/14 03/31/14
You apparently still did not re-read my post.
Aside from that, regarding transients, a toneburst is one type of transient signal, but a very specialized one, and it is not necessarily what I was referring to.
In the world of measurement signals (which doesn't always correlate directly or strongly to the world of listening to music), a single sample pulse (or dirac function), commonly used for FFT analysis, would be a type of transient more representative of musical transients, esp. cymbal crashes, electric guitar fuzztone spikes, etc.
You don't mention if the toneburst you used was shaped or gated, this will have a significant impact on just how "transient" it is in terms of a test signal, even so, the primary toneburst frequency would still tend to mask or overshadow the true transient portion of the toneburst, if it were not a shaped type.
In terms of showing energy storage, yes, a tone burst can do that for the loudspeaker, primarily AT THE GIVEN FREQUENCY of the toneburst. Away from that frequency, the measurement equipment will be unable to discern the transient behavior of the DUT, unless it is truly terrible and of a large amount of deviation from linearity.
Finally, there is a significant difference between an energy storage resonance (which is what toneburst's measure with a loudspeaker transducer), and other effects that tend to blunt or soften a transient, which is more of a filtering effect. Signals that work well with the former, will not be very effective with detecting the latter.
Previous posts referred to the likely effects more specifically, they did not refer to energy storage issues.
So the bottom line is, your jumping in here with comments about measuring tonebursts is a complete red herring, and really bears no relationship to what was being discussed.
With that in mind, I strongly suggest you give it a rest, and beat a strategic retreat.
Jon Risch
IMHO, the way audible skin effect alters transients due to phase shift may be even more pronounced than high frequency roll-off.
Some folks may consider a softening of the leading edges to be somewhat euphonic or "smoother" while obtaining a bolder sounding sonic signature, especially in an otherwise edgy sounding system.
just my 2 pennies
.
Another chart I looked at was clear taht many variables depend on both the gauge of the wire AND the length.
One reason a small fuse does not matter. It is really short.
So the 6dB down?? is it even with a one mm length of wire? or does it have to be ten kilometers to achieve 6dB down??
You are at it again Lizzy. Just when I thought you were comming around. Fuse type and size can make or break the sound of a system. They make a huge difference as probably millions of AA members can attest to. I think I kept this pretty clean (Jon). Tweaker
I find that ever so slightly hyperbolic. %^)
No disrespect intended for or to the AA, its Mods, or its maven-asylees.
axolotl
MEMBERS.
.
For shorter runs the whole spectrum is good. But when the quad weave of Kimber GQ 26g wire was run XLR (two 26g each leg (+) (-) the sound was missing some midrange and bass in the six meter run.
But I never heard treble like it before. amazingly good.
I am trying to do a dual pair so the 26 quad is paired up with a 19g quad Kimber.(the 26g is over $500 at 6m ) the 19g quad KImber is only $350...)Going to play with running the 19 one each to (+) (-) and two plus an extra 19g as ground, and add in the 26g wires.. (the basic Hero uses a cheaper insulation wire with one each to (+) (-) and two 19g to ground. And that sounded decent before I sold it.
Trying to keep some of the amazing treble and still get some good midrange out of a six meter XLR set.
Right now I have pretty good sound with the Kimber quad weave at 19g pairs to (+) (-) and big old thick 12 gauge silver milspec Teflon coated for the ground. But the treble is only good. not spectacular. But all in alll a well balanced cable using the Kimber GQTCSS
I got really burned out screwing around with that Cardas Parsec 7 meter balanced cable. It SUCKED, ((and everyone was like "let it break in",)) Well three weeks of sucking and absolutely no respite was killing me. So i sent it back. All in all it was a PITA.
Though the one meter Parsec RCA are pretty nice sounding. But they do pick up hum pretty easy.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: