Analysis of speaker cables

[Art K]

Over the years we have carried out countless 'blind' listening of cables and other items in the hi-fi chain and customers can always hear differences. In fact, blind listening is usually the way we conduct all our comparative demonstrations. All that is, apart from speaker comparisons. If the differences are not apparent, the system must be very inferior or the hearing is impaired.

Of course, differences are often very small but audible and it's up to the individual to determine if the differences are improvements and if they are worth paying for.

Very well said. I have said something very similar many times. There are always differences, sometimes improvements and sometimes not. Sometimes it correlates with expense and sometimes not. It's up to the consumer to determine whether or not the differences are an improvement and whether they are worth paying for. If you don't hear any differences then by all means don't buy them, I do so I will.

[singslingr]

I agree with Art K and hifi_dave. Exactly the same reasoning applies to power conditioners, isolation feet, cones and other such tweaks which lead to subjective improvements in sound.

Ultimately, we all vote with our wallets.

[STHLS5]

Fine. Maybe somebody here can list down the names of the cables, price, design, termination and the so called improvement. Anything is fine by me, if it can really improve my system.

ST

[STHLS5]

Well if you can tell the difference between these two cable then you can always collect the 1 Million dollars from James Randi organization. Something which even the CEO of the cable companies seemed reluctant to do.

ST

[honmanm]

Anything is fine by me, if it can really improve my system. ST

The point is, I think, that to a greater or lesser extent the cable interacts with the other components of the system. And that's the conclusion I took away from the two studies linked to in the first post (however that's just two studies, and it's very likely there is other research that can shed light on this topic).

There is no "silver bullet" that will magically improve everyone's system (that's the kind of snake-oil salesmanship that has turned you and others off, and quite rightly so). Thinking that something will "improve" a system is probably the wrong mind-set - the best one can do is eliminate problems that are degrading the sound quality... and those problems will vary a lot from one system to another.

But it seems just as silly to deny that there may be an audible effect when a component has a measurable effect on audio-band frequency response (and other measurable criteria).

Where objective information isn't available, the best one can do is healthy skepticism as advocated by hifi_dave. If you're being asked to pay a hefty premium over the cost of the materials and labour that went into a product, just say "no thanks".

[A.S.]

Some quarter century ago, as the cable phenomena was gripping audiophilia, I found myself torn between the natural desire to improve the fidelity of my system and incredulity that cables could make the huge claimed differences. As I've mentioned many time here, based solely on observing how I myself react, I believe that if there is more than about one second of silence between comparing A with B the accuracy of reliable, 'scientific' conclusions drawn about A v B is greatly diminished. Maybe even impossible. Sure, there may well be a difference - I'd expect that - but is the difference certainly due to the characteristics of A v B or to the test method itself?

So, I wanted to compare my bog standard QED 79 strand (or similar non-audiophile cable which I still use) against any other cable I could lay my hands on, which to be honest as I wasn't going to buy any, was not a comprehensive market research. The essence of my comparator is that there is a completely silent switchover from A to B. There is no break in the music (ok there is for about a thousandth of a second which is inaudible), and to be absolutely sure that there has been a switchover from A to B, there was an LED which only illuminated after the circuit had indeed changed over. Any 'degradation' of subjective or objective characteristics due to the switching elements (and I don't believe that there was any) would be common to A and B so in my opinion could be utterly ignored.

If you're interested I'll make a sketch and give you a parts lists. The results, which as I've said were not based on a wide ranging selection of cable which may well have given a different outcome, settled my mind on this matter.

[honmanm]

If you're interested I'll make a sketch and give you a parts lists. The results, which as I've said were not based on a wide ranging selection of cable which may well have given a different outcome, settled my mind on this matter.

Yes please!

[hifi_dave]

It beats me that cable sceptics will readily accept that components, such as resistors and capacitors from various manufacturers, can sound different when they have similar specs, yet they don't believe that cables with wildly different construction will sound different.

[STHLS5]

It beats me that cable sceptics will readily accept that components, such as resistors and capacitors from various manufacturers, can sound different when they have similar specs, yet they don't believe that cables with wildly different construction will sound different.

I am not so sure that different capacitors really alters the sound. After spending close to the price of the CD player itself I could hardly tell if the blackgate and Auricaps helped. My preamp designer told me not to waste time changing the stock capacitors. He said the design puts the limit to the parts used.


Anyway, back to cables. This is the picture of the wire used inside an amplifier.



How do you think a 10 feet cable going to improve this? Wouldn't it be the best to connect these wires directly to the speakers' terminals?

ST

[A.S.]

Thanks for the picture. You've highlighted the standard cable used to connect the circuit board to the amplifier's output terminals. But you can take a number of steps backwards from those cables. Have a look at the copper foil tracks on the amplifier's PCB, barely a mm or two wide and wafer thin. What about them? Shouldn't they be rewired with fancy or at least thick cable? And another step backwards .... to the semiconductor junctions inside the amp's power devices .... you need to see them under the microscope. Shouldn't they be fattened up somehow?

Here is a PDF of an electron microscope photo of a semiconductor device's output wire. The picture is not scaled, but we're looking at a joint that is perhaps a thousandth of a mm across. That joint carries the current from the working part of the transistor - the pn junction - to the metal pins of the transistor that we can see from the outside, so it is by far the weakest link in the current carrying chain to your speakers. Here is an example of the external transistor case that we see; the working junction and its tiny wires are buried inside the case.

Although those visible external transistor connection pins are fat, the actual working part of the device is only atoms thick .... surely we should bear that in mind when becoming over anxious about cable? Is it not the weakest part of the chain which defines the performance of the entire signal chain?

[honmanm]

How do you think a 10 feet cable going to improve this? Wouldn't it be the best to connect these wires directly to the speakers' terminals?

If the amplifier has no zobel networks on its outputs (thinking of classic Naim gear here), that would be a bad idea! Most amplifier designs need some inductance on their outputs to damp HF oscillations (Naim specified a minimum length of a particular type of cable).

If you amp does have Zobel networks - which is very probable! - then the less cable the better... and the smaller the effect that the cable will have.

Now looking at the transistor complement of that amp (and those power resistors), it probably has a nice high damping factor. Going back to the first study, the frequency response of their test system when using such an amplifier was not significantly affected by the cable used in the test.

However if you were to use a high-capacitance cable (litz or similar) it might affect the performance of the amplifier as described in Nelson Pass' study.

[A.S.]

I illustrated the semiconductor junction and its tiny lead-out wires previously. If we now take a razor blade to the PCB track of the crossover (which is much thicker and broader than the PCB track on the amplifier) you can see just how thin it is. Thin, but entirely adequate for the current that it will carry.

So does the speaker cable need to be much thicker i.e. have more copper per metre than this? Well yes, thicker would be good considering the length the current must travel from the amp to speaker (and hence to minimise loss as heat) but as thick as a hose? No.

Note the fat wires from the adjacent coil, itself much thicker than the track foil but flattened out, probably about the same number of copper atoms per unit volume. Do bear in mind that this crossover track is likely to be several times thicker than the track inside the amplifier. Anyone have a close-up of an amplifier PCB tracking adjacent to a ruler for scale?

[Euler]

Over the years we have carried out countless 'blind' listening of cables and other items in the hi-fi chain and customers can always hear differences.

Simple logic tells us there are two and only two possibilities here: either (a) by "blind listening" you mean what I do, a rigorous and careful AB or ABX comparison, where the volume remains constant and you and the customer do not know which cable is being used at any given time, the kind of test described by Alan in various posts, or (b) you mean something different by "blind listening." In case (a), you and your customers should immediately contact James Randi, who, as ST pointed out, will hand you 1 million dollars. But I suspect that (b) is the case. And that being the case, it would be more accurate for you to say, not "customers can always hear differences," but rather "customers always _believe_ they hear differences."

But "belief" doesn't prove anything; evidence does. To restate a simple point: if "golden ear" folks or just regular folks could indeed reliably detect (in a careful ABX test) sonic differences between pairs of cables, wouldn't some such tests have confirmed this? Again, think about how much money Monster, Pear Anjou, or Opus cable companies would make if they could prove that their cables "sound better" with such rigorous testing. Yet no one has published such test results. That's a proof by contradiction: if folks really could detect sonic differences, then some rigorous tests would have confirmed this. But no such test has been published. Q.E.D.

Why would anyone believe that they can't hear the difference between cables?

Well, for example, an engineer who understands the physics of current, wires, etc, someone like Alan, say, might well have solid, scientific reasons for expecting that any differences would not be detectable by the human ear. In my case, not being an engineer, I have no expectations one way or the other, so I depend on evidence. And in the absence of any convincing evidence that differences can be detected, but in the presence of scads of cases where tests have failed to demonstrate that such differences can be detected, I remain a skeptic.

Bruce

[honmanm]

Here a photo of both sides of an amplifier board. The board is about 90mm wide, and the output-stage current path is highlighted. When the components are fitted the path length is 35mm from the decoupling capacitor to the output terminal - the current goes through one transistor, one 0.22 ohm resistor, and a 0.5uH inductor.

Other than that there are wires from the +ve power supply to the board, from board to +ve binding post, and from -ve binding post back to the transformer centre tap.

Tracks are 4mm wide, BTW.

[STHLS5]

If the amplifier has no zobel networks on its outputs (thinking of classic Naim gear here), that would be a bad idea! Most amplifier designs need some inductance on their outputs to damp HF oscillations (Naim specified a minimum length of a particular type of cable).

If you amp does have Zobel networks - which is very probable! - then the less cable the better... and the smaller the effect that the cable will have.

Now looking at the transistor complement of that amp (and those power resistors), it probably has a nice high damping factor. Going back to the first study, the frequency response of their test system when using such an amplifier was not significantly affected by the cable used in the test.

However if you were to use a high-capacitance cable (litz or similar) it might affect the performance of the amplifier as described in Nelson Pass' study.

The picture wasn't mine. I got it off the net. But you have raised an interesting observation regarding high-capacitance cables. I will try similar cables. I am still keeping my mind open, as the debate was going on I was changing my players power cord. Alternating between a reference, solid core Belden and the original cable that came with player. Yes, I detected some difference. I have done this comparison many times before and it is always the same when we first change the cable. We detect some difference.

As days go by, you will be asking again what happened to the highs or deep bass that you first heard with the cable change. And we start all over again with different cables and tweaks.

ST

[steveinaz]

Cables. Where to begin? You have Science and perceived real world results. I think the truth lies somewhere in between. I think it's most important to first set your ego and bias' aside for a moment, and research both arguments in a honest effort to get at the truth. That is what we are after, right?

Then we must factor in the variables; listener preference, the physical shape of ones ears, the equipment, the setup. Ear shape is interesting topic that often goes without mention. Next time your listening to music, take your index fingers and push the center of your outer ears forward about 1/16"...notice something? Pretty darn significant huh? How might this single variable effect ones choice of speaker? How might it effect the sensitivity to subtle change? Sometimes people phrase arguments in such a way, that it seems they think the human ear is a device, without error and/or variation. I'd submit that people with "wide" ears are especially sensitive to treble energy.

While I think it's silly to be deeply entrenched in either camp (believer/non-believer) I personally exercise some common sense with my choices. I have experienced differences in some cables. I've often found that multi-strand cables tend to sound out of focus/hashy when compared to solid copper conductor type cables. I also on occasion chose the "cheaper" cable over the more expensive one because it simply sounded better. Audioquest Diamondback vs Copperhead is an example. I have always preferred the cheaper Copperhead.

Personally, I'd never spend more than I have already on cables. My speaker cables (Kimber 8TC v.2) were $456 for an 8ft/pr; and the Kimber Hero IC's at about $285 for 1 mtr set. Having said that, the much cheaper Kimber PBJ is excellent as well; and with hindsight I would have kept it. Digital cable is different, in my opinion. I think a digital cables ability to be as close to the 75ohm spec as possible, far out-weighs any material/topology considerations that you have with analog cables. I love the Belden 1694A, and have had digital cables 6x the price that didn't perform as well, and neutral as the Belden.

[Audentity]

Thanks for the nice, sane post Steve. People get too excited about cables.