Bi-Wiring

Andy W — Fri, 31 Aug 2007 20:46:00 GMT

Nothing gets people’s blood boiling like a good “conversation” about speaker cables. I’ll just wrap it in with bi-wiring.

Bi-wiring is the practice of running two pairs of wire to each speaker, one pair for the woofer, and one pair for the tweeter. The theory is that only the electrical current that pertains to the woofer will flow in the woofer wires, and likewise, only the current that pertains to the tweeter will flow in the tweeter wire, and that this will improve the sound coming out of the speaker.

The proponents of bi-wiring claim better performance, and there are a myriad of wire types, all with claims as to why they are better for the given application. The proponents of bi-wiring can be found in the form of audio forum gurus, magazine writers, and even loudspeaker companies. Not only that, many times the type of wire recommended changes depending on whether it’s for the woofer or the tweeter.

For a number of years now, the Klipsch Reference line of speakers has been equipped with binding posts that accept bi-wiring. I’m not sure exactly how much this costs in terms of material and labor, but I have an idea, and it’s comparatively small, and it varies depending on the quality and materials of the binding posts. So the logic is, if we include it, the people that don’t bi-wire won’t care that much (since the cost is comparatively small), but if we don’t include it, the people that do bi-wire won’t buy the speakers. And since we are the performance leader in loudspeakers, we have to have bi-wire binding posts. After all we don’t want to be perceived as anything less than the best speaker in the showroom, even if it doesn’t matter one iota to the quality of the sound.

Does that also mean a three-way would need three pairs of binding posts? Apparently it does.

At any rate, let’s say that 95% of people don’t bi-wire… Is this money that could better be spent someplace else that mattered more? Like improving the tweeter? Or the crossover network? Someplace that would make a distinctly audible improvement to a larger set of people? And therefore making it a better value in Hi-Fi sound for your hard-earned dollar?

From an electrical engineering standpoint, there are three things I can measure in any given speaker cable: resistance, inductance, and capacitance. The three things combined give you the cable impedance, which when connected to a particular speaker will be lumped into the speaker’s impedance. The cable impedance will be distributed through the length of the cable, and generally speaking there will be an added resistance at the termination, in fact most of the resistance will exist at the terminations and connections. The cable can also have a particular construction, which includes the wire gauge, the gauge and number of strands, the insulation size and type, and the type of termination(s) used.

The resistance of a cable is determined by the wire gauge and the length of the cable. The inductance is also affected by the gauge and length. The capacitance is affected by the length and spacing to the other conductor(s) and perhaps to a certain extent, by the wire gauge/geometry/insulator. The impedance of the cable could possibly affect two things: the amplitude response and/or the phase response. You may notice I did not say the cable could affect distortion. The distortion of a speaker cable is utterly unmeasureable. If you believe you can hear cable distortion, well, I can’t help you; go read someone else’s blog.

Those three measurements, resistance, inductance, and capacitance, will in every case determine the performance of the cable. Period. There are no magic bullets, no magic connectors, no special copper crystal structures, or cable geometries that cannot be measured in resistance, inductance, and capacitance that can in any way affect the way a speaker cable “sounds”. If I give you two completely different speaker cables that have exactly the same impedance (resistance, inductance, and capacitance) then they will “sound” exactly the same to your ears. If anyone tries to sell you anything different, then I say it’s snake oil. If they can prove it in a double-blind ABX test to a statistically significant difference then I will personally give them $100. I would say $1,000, but I don’t have $1,000, and then my offer would be a lie, but I do probably have $100 at any given time. (Note to would-be muggers: I don’t keep that kind of cash on me.)

Even if the two speaker cables don’t measure exactly the same, I truly doubt that any person living can, in a double blind ABX test, tell the difference between any two reasonably constructed speaker cables of reasonable length, including generic lamp cord at 20¢ a foot, unless they are specifically designed to alter the sound rather than simply pass it.

Now, why would I say that? Let’s look at what it would take to hear an audible difference in terms of frequency response. Can you hear differences of 1dB? Probably. 0.5dB? Maybe. 0.1dB? No. So let’s pretend we have an 8 Ohm speaker and a perfect speaker cable. By that I mean zero resistance, inductance, and capacitance (no such cable exists, but it gives us a place to start). To change the response 1dB due to resistance I would need a resistance of 0.976 Ohms, or about 386 feet of 14 gauge wire. Well, I really only need 10 feet to hook up my speakers, and that would give me a change of -0.0274dB. Not at all audible in any way shape or form. And that’s compared to a magic speaker cable that can not exist in the real world. Compared to another real cable, like the one you would use, the difference is even less.

OK, so what about inductance and capacitance? Just for fun I pulled out the cheapest speaker wire I could find; it’s 12 feet of 18AWG zip cord. Capacitance: 166pF. Inductance: 3.90µH. Let’s take them one at a time… first the capacitance. If you hang a capacitor on the output of your amplifier (don’t do this for real, you might make it unstable) it will form a low-pass filter with the output impedance of the amplifier itself. The -3dB frequency for this filter, even if the output impedance of the amplifier were very, very high, say 8 Ohms, would be 120 MHz. At those frequencies, other things happen that become more important, but let’s just say that there is no effect at audio frequencies.

How about the inductance? 3.9µH must make some difference with an 8 Ohm load right? If I put an inductor in front of a speaker, again I get a low-pass filter. If I have the aforementioned cable the -3dB frequency would be 326 kHz. Not even a bat can hear that.

What about the self resonance of the cable? Sorry.

Skin effect? Ah, now this one gets some people to believe, but the physics for skin effect don’t begin to take effect until 100 kHz or so.

Speaker wire critics have been known to use the example of 100 yards of cable and 10kHz square waves to demonstrate differences in speaker wire. Well…

I want someone to show me a) the listening room with 100 yards of speaker wire; b) the musical instrument that generates 10kHz square waves; c) the studio microphone that can convert it to an electrical signal; d) the studio recording equipment that can record it; e) the storage medium that can archive it; f) the playback device that can play it; g) the preamp/amplifier that can amplify it; h) the loudspeaker that can reproduce it, and i) drum roll… the human that can hear the difference between an 10kHz square wave and a 10kHz sine wave. (A square wave is the sum of the fundamental and the odd-order harmonics. The lowest odd order harmonic of 10kHz is 30kHz, which is not audible to humans.)

No, the speaker wire is not in any way imaginable the weak link in the recording/playback chain. Does that mean I won’t spend my hard earned money on good speaker cable? Not at all, but I would spend it where it gives me some value, like in the quality and usefulness of the connectors and the cable’s manageability, or to meet code when run inside walls.

Now, remember this: If you can hear the difference between two different speaker cables, you have a bad connection somewhere.

By the way… all copper wire is “oxygen free” and any claims that some type of wire will reduce the “grit” or “grain” in the midrange, or tighten the bass or improve the imaging are a bunch of hooey. But feel free to disagree with me.

Oh, and if you do have a professional installation and you are running 100 yards of speaker cable, you might have to increase your wire gauge accordingly because inductance can start to be a factor. But that could get rather expensive. And every console I’ve ever seen had a treble knob on it. So you can easily save your money and turn up the treble a touch if you need to. As far as resistance, well, again every console I’ve seen had a volume knob.

In the end, it doesn’t matter what I think of speaker cable of bi-wiring from an engineering perspective, because there are people (including our own people) who are convinced that it does matter. And those people buy speakers. And we want them to buy our speakers.

Oh, and yes, I will submit to a double blind speaker wire test. Just don’t tell me I’m listening to speaker wires, I may be biased.

Andy's Blog : speaker cable

Bi-Wiring