How to buy cables-

[Warren Peck]

This is NOT a debate about whether cable A sounds better than cable B, just a (mostly) scientific approach as to what to look for in interconnects and speaker cables.

Any connectors should be tight fitting, with non-corroding or oxidizing contact surfaces, which is why gold is frequently used, though in varying degrees of purity, thickness, and method of application. The conductor in the cable should be attached to the connector through a high-quality solder joint. High quality means performed at the optimum temperature for the correct length of time, and with the correct amount of solder to allow a new alloy to form that mixes molecules of the base metals of the connector and conductor while using as little of the semi-conductive or non-conductive solder material as possible. Silver content solders have a different plastic state and melting temperature then common lead/tin/nickle/flux core solders, and require different application techniques, but can result in a better joint with less inter-element capacitance.

In an environment with a lot of RF noise, a cable should have shielding running the length of the cable and attached to the connector at ONE END ONLY, preferably the pre-amp/reciever end. This is to avoid a ground path that would negate the shielding. Twisted pair for the signal and return/ground are adequate at audio frequencies, though coaxial is prefered at higher (digital or video) frequencies. XLR is preferable to RCA if possible. There is a misconception that high frequencies travel on the outer skin of the cable, and that silver coating of an analog interconnect or speaker cable enables higher frequencies to travel better. Though it is true that higher frequencies do travel towards the outside of the conductor, these are MICROWAVE frequencies FAR ABOVE audio range.

Conductors should be made of a highly conductive metal, such as copper or silver, and should allow flexibility without breaking, and should be of a gauge large enough to pass the low signal amplitudes with as little attenuation as possible. Cables should have a good quality dielectric that will not change significantly in capacitance with age, light, heat, or current. The jacket should also be of a material that will not degrade significantly under the same stresses. All cable lengths should be kept as short a possible for a myriad of reasons, electrical as well as asthetic. Remember though, the farther from the source and the more demanding the load, the heavier the gauge should be.

Connectors should be WELL MARKED for polarity, especially when they are configured as PAIRS. RED and WHITE are preferable to RED and BLACK, as they are easier to distinguish from each other in a low light environment. Avoid running interconnect cables with power cords that may induce an audible hum. If they must cross each other, it is preferable to do so at an angle. Also, be aware of where power is in walls, especially if you are running any audio cables inside the walls as well. Though it may be easy to use the same conduit or path the electrical wiring uses, don't be tempted! Besides, it is against building codes.

So, there are reasons to spend some money on cables, or to build your own, but don't be lured in by the exotic or wild claims for the reasons a particular brand "sounds better" than another. Any that meets the criteria above should perform exceptionally in any system, from $1,000 to $100,000. Quality of materials and construction should be the factors you use in deciding which cables to use in your system, not exaggerated marketing claims, and you shouldn't have to spend thousands of dollars to connect all your components, regardless of what components you have.

[Smokey]

Although I agree with most of what has been said, there are couple of items might need discussing.

Twisted pair for the signal and return/ground are adequate at audio frequencies, though coaxial is prefered at higher (digital or video) frequencies.

Unless we are talking about balance connections, twisted pair is not recommended in today's HT environment. For audio, the best cable are ones that have double shield (mesh and foil) for low and high frequency protection.

There is a misconception that high frequencies travel on the outer skin of the cable, and that silver coating of an analog interconnect or speaker cable enables higher frequencies to travel better. Though it is true that higher frequencies do travel towards the outside of the conductor, these are MICROWAVE frequencies FAR ABOVE audio range.

Not true. Even at audio range (20-20k hz), higher frequency tend to move away from center of conducter and travel closer to outer surface. The higher the frequency, the less skin depth (distance from outer surface to the center) it will have

[markw]

Not true. Even at audio range (20-20k hz), higher frequency tend to move away from center of conducter and travel closer to outer surface. The higher the frequency, the less skin depth (distance from outer surface to the center) it will have

While this may be true in theory, in practice the effect is inaudiable. Take this from a guy who worked with microwaves in the USAF. There it makes a difference. At audio frequencies? Nah!

Can you say "wave guide, loops and probes and big-ass 50 kw klystrons", kiddies?

[Smokey]

While this may be true in theory, in practice the effect is inaudiable. Take this from a guy who worked with microwaves in the USAF. There it makes a difference. At audio frequencies? Nah!

That is probably true for shorter runs. But since skin effect causes the effective resistance of the conductor to increase with the frequency of the currerent, it might have an effect on longer runs.

To minimize skin effect, a 20 gauge cable is recommended. But as you can see longer length will become a problem.

For longer length, one solution might be to use [lower gauge] silver coated copper cables or go with more expensive Litz design cables

[markw]

That might be acceptable for very short runs but you know that's not really a good idea for longer runs, such as those to rear or remote speakers. The internal resistance becomes more of an issue than any skin effect.

Well, if you want to get really picky, don't use solid wires. Use stranded cables in the first place.

[jneutron]

..... while using as little of the semi-conductive or non-conductive solder material as possible....

????? semiconductive or non conductive solder constituents??

.....
. Silver content solders ........ lead/tin/nickle/flux core solders,

?????? Nickel??

..... ......but can result in a better joint with less inter-element capacitance.

????? Did you say capacitance??

..... Though it is true that higher frequencies do travel towards the outside of the conductor, these are MICROWAVE frequencies FAR ABOVE audio range.

hmmm..even the approximation equation isn't so black and white, nevermind the bessels..

..... . Any that meets the criteria above should perform exceptionally in any system, from $1,000 to $100,000.

hmmm...no worry bout nuttin else???

Cheers, John

[E-Stat]

hmmm...no worry bout nuttin else???

Of course not. Basic theory handles everything.

rw

[Smokey]

That might be acceptable for very short runs but you know that's not really a good idea for longer runs, such as those to rear or remote speakers. The internal resistance becomes more of an issue than any skin effect.

Canare Star quad design (such as Canare 4s8) might be a good solution for longer runs where instead of running one large gauge per leg (13 gauge), it used two smaller gauge wires (16 gauge) per leg. A sort of trade off.

Of course not. Basic theory handles everything

There is more thruth to that statement than meet the eye. After all, we're only talking about a piece of wire

[Warren Peck]

That is probably true for shorter runs. But since skin effect causes the effective resistance of the conductor to increase with the frequency of the currerent, it might have an effect on longer runs.

To minimize skin effect, a 20 gauge cable is recommended. But as you can see longer length will become a problem.

For longer length, one solution might be to use [lower gauge] silver coated copper cables or go with more expensive Litz design cables

Just out of curiosity, isn't this is a primarily a HOME audio website, not STADIUM audio. My imagination may be limited, but I cannot see even in a very large home (10,000 sq ft) that you would have any interconnect or speaker cable runs longer that 200 feet, and I would wager that it would be a very rare case indeed where an interconnect was over 20 feet or a speaker cable was over 50, even running in walls. I just can't concieve of a situation in a HOME environment where cable length would be so great that skin effect would come into play. Speaker cables are basically transmission lines, much like those that deliver power to your home, only a broader frequency range but lower power requirements. Basic transmission line science still applies, and whatever bells and whistles you may attribute to a particular brand of cable, the science is the same. Electrons (or holes, if you prefer) will flow in the way they do, as they don't care which manufacturer's name is on the jacket.

[markw]

Canare Star quad design (such as Canare 4s8) might be a good solution for longer runs where instead of running one large gauge per leg (13 gauge), it used two smaller gauge wires (16 gauge) per leg. A sort of trade off.

What do you consider a "longer run" where skin effect would become audiable?

[E-Stat]

. After all, we're only talking about a piece of wire

After all, tires are just a piece of rubber. What possible difference can there be?

ps: my cables must be different from yours. Mine are sealed in a dielectric, have shielding and connectors.

rw

[Smokey]

Warren Peck and Markw,

I agree with your statements that in typical home run, Skin Effect will not be an urgent issue to deal with-as it probably will not be audible.

My argument was that it does exist and if anybody want to tackle this issue (or get picky), there are solutions that can address it. Like dealing with worst case scenerio.

Don't know, but may be I'm flip-floping on this issue

[Warren Peck]

After all, tires are just a piece of rubber. What possible difference can there be?

rw

Sorry, but tires are an extremely poor analogy for several reasons. The function of an audio cable is to pass musical information from point 'A' to point 'B'. Whether the audio system cost $200 or $200,000, the function of the cables is the same. Tires are designed for specific types of terrain, weight bearing loads, size of wheels, etc., and frequently a specific tire or limited range of tires can be applied to a particular vehicle, or it will not function as intended, even if the tires "fit". Off-road vehicles and Formula race cars both require tires, but the function of the 2 is vastly different. I suppose you could say the basic function of the 2 is the same- to drive from point A to point B, but then you would have to say that you require different cables for different types of music, and that just doesn't fly.

[E-Stat]

Whether the audio system cost $200 or $200,000, the function of the cables is the same.

Yet, the execution and outcome can be quite different depending upon the system and environment. If you don't detect any differences, then fine by me.

rw

[jneutron]

Sorry, but tires are an extremely poor analogy for several reasons. The function of an audio cable is to pass musical information from point 'A' to point 'B'.

Actually, the function of an audio cable is to pass musical information ACCURATELY.

Power loss is but one parameter that needs to be controlled. Aren't you in the least bit worried about temporal and amplitude issues??

Skinning has several effects.

1. Alteration of current distribution within the conductor. This impacts both resistance and inductance. The equation normally used for skin depth is WRONG!!!! At 20 Khz, it's off by about a factor of 4. Skinning is not as bad as the approximation equation predicts...simply because the equation is for a planar conductor of infinite size, with a TEM wave impinging on it... Wires are not like that.

2. The internal inductance of a cylindrical wire is 15 nH per foot. Skinning will reduce that. In the hf limit, the 15 nH per foot dissipears.

3. The characteristic impedance of a cable will change with frequency if skinning can reduce the inductance. A cable made with tefzel or teflon will have a dielectric coefficient of about 2.7, and to be 8 ohms, it must have about 10 nH per foot inductance, and 280 pf per foot capacitance. Clearly, skinning of a line which requires 10 nH per foot cannot be tolerated, the structure of the line must be such that the internal inductance of the wire is zero. This is typical of a coax outer shield, as well as flat striplines.

btw...soldering technology is about two thousand years old. I was surprised you provided some erronous information. Both copper-tin intermetallic layers are not insulating nor semiconducting, they are quite conductive. Tis only the flux which is insulating.

Also, I've not really seen a solder with a nickel contituent. Nickel specific fluxes, sure, but nickel as part of the solder alloy? I've never used any.

Course, that's only in my limited experience..

The tire analogy taint actually that bad..depends on how anal one desires to be, whether to agree or disagree on that.

Cheers, John

Yo Estat...how's it shakin? Tain't been around this sleepy hollow much..

[E-Stat]

The tire analogy taint actually that bad..depends on how anal one desires to be, whether to agree or disagree on that.

The point I was trying to make was to illustrate how simplistic the "after all it's just wire" comment was. Even when focused on a single segment like passenger car tires, the difference in performance ("its just rubber") can be significant.

Yo Estat...how's it shakin? Tain't been around this sleepy hollow much.

Stayin' busy and enjoying the stats! I get to visit your neck of the woods at the end of the month. Have business around LaGuardia which means I'll hang out in Sea Cliff that weekend. Speaking of wires, I'll get to (not) hear the Nordost Odins there. Speaking of which, were you ever able to duplicate the Valhalla's characteristics with your DIY formula?

rw

[jneutron]

Stayin' busy and enjoying the stats! I get to visit your neck of the woods at the end of the month. Have business around LaGuardia which means I'll hang out in Sea Cliff that weekend. Speaking of wires, I'll get to (not) hear the Nordost Odins there. Speaking of which, were you ever able to duplicate the Valhalla's characteristics with your DIY formula?

rw

Um,,,yah.. It was simple enough. I duped it using 1/2 wide 3 mil thick copper foil...so the inductance and capacitance were equivalent, but the resistance was a tad higher, up near #14awg. Gave it to some guy in canada to try, but he's done absolutely nothing with them. A shame.. So far, of the several people I've shipped free of charge wire samples to try, only Ted Smith actually tried and provided feedback.

I have some #8awg flat copper braid here (least I think it's #8), I was planning on making a stripline set with a cable Z of 8 ohms for someone to try (maybe even me..). Working on what dielectric to use, it's gonna be some foam thingy, roughly 5 mils thick give or take. Wanna get the prop velocity as near to C as possible..

Cheers, John

[audio amateur]

damn I feel like i'm in physics class all of a sudden... long time no been

[emorphien]

After all, tires are just a piece of rubber. What possible difference can there be?

Well, aside from the fact that there are measurable and quantified differences in tires that have direct impact on performance and can be tested on a track... and that similar scientifically valid measurements do not generally show significant (to the point of being audible) differences between proper gauge cables for a given application; and that proper psycho-acoustic testing has rarely been performed even correctly to try to back up the claims of fancy cables, I'd say that the argument for fancy cables is on pretty shaky ground!

I've had the pleasure to try to prove the science wrong with various expensive cables ranging from midrange pair of speakers to nice car on systems ranging from inexpensive car to very nice car, and there's just not much to be said for spending money on cables. Working on room acoustics or equalization for your favorite pair of speaker will provide better results, cheaper.

[E-Stat]

Well, aside from the fact that there are measurable and quantified differences in tires that have direct impact on performance and can be tested on a track...

Just curious. Which metric(s) quantify the degree of over/understeer tendency exhibited by the tire? Or, which metric(s) quantify the breakaway characteristics at the limit of adhesion?

Working on room acoustics or equalization for your favorite pair of speaker will provide better results, cheaper.

I'll be the first to agree that room treatments (and careful speaker placement) should come first. I've used various approaches for well over twenty years. With my main system, I use a forest of bass traps to flatten the bass (measured quite flat from 25 hz to 200 hz) and a mix of acoustical panels and ficus trees to control the reflection points and back wave. There are a couple of pics in my gallery if you're interested.

I didn't stop there.

rw