Could someone explain the reasoning behind biwiring? (1 Viewer)

[Donald_Spry]

I have my main system consisting of Paradigm studio series biwired because it was highly recommended. However, I am setting up an audio only, no HT, for my bedroom. I am running the new KEF Q3's powered by NAD.

I had every intentions to biwire, but in an effort to keep cost down the thought of running single wire has crossed my mind. What is the logic behind biwireing? Isn't the same power source powering both any ways? Being that I am not biamping.

Thanks,
DJ

I think I found the answer buried in the search files. Forget it.

 

[BrentPollard]

Go here
http://pachome1.pacific.net.sg/~axiom13/biwire.html

 

[James Zos]

That's a nice link, Brent, thanks!

This in particular caught my eye:

"If you do not wish to bi-wire but your speakers are bi-wire ready, then by all means eschew the supplied bridging posts (which are typically of inferior quality and negate much of what you have spent on your precious cables) and construct your own from mid to superior grade speaker cables (often after stripping off the external coat)."

What do you all think of that idea? Has anyone done this?

As for bi-wiring, in my opinion it is very difficult to screen out the placebo effect. I remember when I first tried it I thought I could hear a difference, but later decided I couldn't and went back to the old way. Now I feel tempted to try it again (sigh).
I would love to conduct an A/B study some time to see if I could actually detect the difference, but my girlfriend thinks I'm crazy enough as it is and would be unlikely to help out.

 

[Chu Gai]

temptation...a dastardly thing although pacino rather liked vanity
it's spelled wrong though...buy-wire.

 

[Bill_D]

I always [rant]preserve[/rant] those bridging posts in the speaker boxes. I made several sets of jumpers from extra Kimber 8TC a while back.

 

[Christopher Lyn]

I lost one of my jumper bars and used Audioquest Type 4 cable that I had left over. Works fine.

 

[Donald_Spry]

Yes, excellent link Brent. I guess I will end up biwiring anyways. It would probably eat at my conscience everyday and I would end up spending the money anyways! Then I be out the cost of the single wire.

DJ

 

[brucek]

Geez Donald, I think you should be listening to James Zos girlfriend, who thinks he's crazy for buying into such nonsense. Believe me, anyone with an electronics engineering background knows this is clever marketing to convince you to purchase more wire product. As you admit, it plays on your conscience.
As far as bi-wiring is concerned, there is simply no scientific support (other than anecdotal evidence). I'll bet I could produce a believable treatise to convince you either way, but as it is, I'm going to let you in on a secret.
Bi-wiring is about as effective as trying to slow down your car by putting your hand out the window. I gotta think there are better ways to spend your money.
Mmmmmmmmmmmmmm....
Donald thinks.............But my speakers are bi-wire ready, with two sets of terminals and the manufacturer recommends it for increased clarity...........
Well, of course the speaker manufacturer is recommending it. If everyone else in town was recommending it, wouldn't you? That's good business.
Donald thinks............But bi-wiring separates the LF (low frequency) signal currents and the HF (high frequency) signal currents and simply must make my speakers sound better, because now these signals won't interfere with each other, as happens when I use a single wire. Bi-wiring, allows the amplifier to "see" each driver as a separate unit, and certainly this claim has a good gut feel to it. To audiophiles, "isolation" has always been a central philosophy which relates to good sound!
Well, sorry, the reasoning is flawed. This is the pillar that bi-wire theories are built on and it just ain't so. You have indeed separated the HF and LF currents. Seems logical and everyone gets on board claiming how much clearer everything is now. But unfortunately, anyone who's even taken a first year engineering course in electronics knows that the claim is meaningless when you apply the superposition theorem and know that this is a linear system and all those signals can easily co-exist on a single wire quite nicely without interfering or modulating each other. Superposition theorum states this, and it is so. I'm afraid there is no disputing it. The signals create a complex waveform, which could be fourier transformed into the original signals. This is a linear system, there is no modulation occurring. I can go into why, but I won't bother - it's boring.
Donald thinks......... Biwiring helps because there's a possibility that one of the drivers in my speaker may distort, producing harmonic products, so the current flowing to the driver may possess a small distortion component. That component, flowing through the speaker cable, causes a distortion voltage that appears in series with the output of the amplifier. Now, if I don't bi-wire, and if I assume that the cable has a finite impedance, those harmonic components in the current will cause a small harmonic voltage between the amplifier's output and the speaker's input. So the tweeter will also see this small distortion voltage. When I biwire, the other driver will not see that distortion voltage, while on the contrary, if I use only one pair of cables, then that distortion voltage is presented to the other driver through the strap.
Well, your idea has merit, but there are a couple of problems with this statement.
The rubric of this statement hangs on the calculated difference in the impedance between the speaker's straps and the impedance of one pair of the bi-wire cable. The assumption is that the speaker cable presents a non-zero impedance (which it does), and that the distortion can indeed reach and affect the other driver. Not likely. Consider the fact that a speakers common terminal "straps" are extremely close in impedance to the near zero impedance path through the bi-wires to the other driver. This is the difference you're considering. Yes, the currents to each driver are different in bi-wiring, but the voltage is the same because the amplifier is a near perfect voltage source. The distorted harmonics generated by each driver will be short-circuited by the low impedance output of the amplifier in either case. Also consider that harmonics are second and third order effects, and consider that the harmonic would have to get through the other driver's crossover, which would be in a high impedance state to that harmonic frequency. Again remember that these calculations are to be performed on the 'difference' between the impedance of the speaker's straps and the impedance of one pair of the bi-wire cable. You better have your calculator set to many, many decimal places......
I'm afraid as far as bi-wiring is concerned (if you already are using it), you'd probably be better off using bi-wire cables with your speaker straps reinstalled. At least then you'd take advantage of the lower gauge realized by combining two cables to each speaker. It would certainly help the LF drivers damping path to ground through the amplifier output.
If you like bi-wire, then use it, no harm done......It's ineffective, others will disagree. As I said, it's about as effective as trying to slow down your car by putting your hand out the window. I suspect you could make the calculations on the effect, but why bother..................
One fact that is probably more important, is that the impedance of the wire that connects from the amplifier to the drivers should be as low as possible. This can be accomplished by using cable with a low inductance per foot value and also using a heavy gauge wire and keeping the distance short. Besides frequency distortion, damping factor is important. Both can be put into a more acceptable range by using a heavy gauge wire with a low inductance per foot rating.
With regard to the gauge of the cable, if you are using two cables in a bi-wire fashion to your speakers, you can lower the resistance of the connection (for example) between the LF driver and the amplifier by putting the straps back in as I've suggested and the LF driver will now see a less resistive path back to the amplifier. This will increase system damping. Let me explain the benefit.
Damping generally effects the "tightness" of bass frequencies. The higher the number the better.
It's the amplifiers ability to control or damp the speakers voice coil oscillations. When a signal is sent to a speaker and then stopped, the speaker cone continues to move, and creates or presents a back voltage to the amplifier. If the output impedance of the amplifier combined with the impedance of the interconnecting speaker wires is very low (read a short) then this oscillation is damped and the bass sounds very tight. The higher this combined impedance, the more the speakers' voice coil will continue to move and effect the sound.
This of course is where the bass mushy-ness of tube amps comes from. Tube amps generally have an attribute unlike solid state amps that their output impedance is usually unacceptably high and is non-linear and rises as the frequency increases. You can see how this non-linearity, besides being a source of distortion (read "warm" sound) in the damping formula makes a tube amp pretty undesirable if you're looking for accuracy. But, some people like the sound of distortion, that's their business.
Anyway, damping factor is a ratio of the load impedance (generally considered a flat 8 ohms) to the output impedance of the amplifier. Very nice, but generally in practice though, you have to take into consideration the impedance of the wire connecting the two devices and the fact that speakers with passive crossovers are an extremely non-linear device. Again, the short answer is that the higher the number the better.
The long answer is that in most solid state amplifiers the output impedance is fairly constant (read linear) and very, very low over the entire audio spectrum and as such can almost be ignored in the formula because the system impedance is overwhelmed by the non-linearity of the speakers crossover plus the interconnecting wire resistance. This is the reason to ignore small differences in amplifier specs of damping. It's usually insignificant in relation to the effect of the speaker wire. I'll show you why shorter or heavier gauge speaker wires are better.
Look at the formula for damping using a figure of 300 for damping. This yields about 0.026 ohms output impedance:
(speaker impedance) / (amplifier output impedance)
= 8.0 / (0.026 ohms)
= 300 : this is a very acceptable amplifier damping factor.
Now add in a typical connecting wire of 12 feet of 12 gauge zip cord.
12 gauge at 12 feet yields about 0.039 ohms of pure resistance. Capacitive reactance can be ignored as insignificant, as can inductive reactance at the lower frequencies and this short length. I calculate at 100 Hz that inductance would only add about .0015 ohms to 12 feet, so I'll ignore it here.
(speaker impedance) / [(amplifier output impedance) + (wire impedance)]
= 8.0 / (0.026 + 0.039)
= 123
Anyway, you can see how the system damping factor dropped quite significantly from 300 to 123 even when using very heavy gauge wire. It shows the importance of using a low gauge interconnect and the importance of using short speaker wires.
Donald, forget about bi-wiring. Get a heavy gauge speaker cable and keep it as short as possible.........
brucek

 

[KeithH]

I tried biwiring my two stereo systems with mid-priced Monster biwire cables. At the time, I had been using bulk Monster wire with Radio Shack banana plugs. I did not hear a benefit from biwiring. Perhaps the biwire cables I used were not that good, or perhaps biwiring just doesn't do anything. Biamping is another story. I biamped my main stereo system last year, and that made a significant improvement.

 

[Andrew Walbert]

Bi-wiring isn't going to gain you anything but a lighter wallet. However, if you see yourself eventually moving to a bi-amp setup, then by all means go ahead and run bi-wire connections if you're building an HT from scratch. But otherwise, there's little reason for doing so.

 

[John Royster]

when I bi-wired my speakers there was too much bass. overpowering even.

So I am back to a single pair of wire with small jumpers made of the same wire.

 

[Donald_Spry]

Donald thinks......... He just got schooled on Intro to EE. Brucek, thanks for the thorough post. That is exactly the scientific explanation I have been wanting to find. Everything I have read prior to this was peoples loose interpretations on why biwireing is better. People can theorize all they want, but until I have scientific proof then it it all bunk. You just saved me a decent amount of cash. Thanks!

By the way, where did you pick up such a strong EE background?

DJ

 

[Donald_Spry]

Brucek,
One other question. Is their any specific wire type/brand I should look for. Monster? Canare? And when you say heavy, what do you consider heavy- 14? 12? Heavier?

 

[KeithH]

Andrew said:

However, if you see yourself eventually moving to a bi-amp setup, then by all means go ahead and run bi-wire connections if you're building an HT from scratch.

Don't buy pre-made biwire cables if you plan on biamping down the road. If you plan on biamping later on, biwire with two separate pairs of cables rather than a pre-made pair (i.e., a pair with two terminals on one end and four on the other). You can biwire with two pairs of cables. Buy one pair with banana plugs and one with spades. You should be able to fit both a banana plug and a spade on one binding post on the amp side. Then you will have your two separate pairs of cables when you are ready to biamp.

 

[Andrew Walbert]

That's actually what I meant -- two seperate pairs of cable. If it's not four unique cables, then it's not really bi-wire, if you ask me.

As for what cable brand to go with, everybody will say something different. Your best bet is probably to grab a Canare cable and a Monster cable and compare them. Maybe even an Ixos cable for the fun of it. If you can hear a difference, then go for the best one. If not, then get the cheaper one. 14awg is usually plenty thick enough for speaker wire, unless you're going to be pumping a lot of power (200W or more) to the speakers. Most Monster and Canare cables are 14awg, I believe.

 

[KeithH]

Duplicate post.

 

[KeithH]

Andrew said:

That's actually what I meant -- two seperate pairs of cable. If it's not four unique cables, then it's not really bi-wire, if you ask me.

Got it. I've read that it is much better to biwire with separate pairs of cables, as interference can result between the high pass and low pass sections on a single biwire cable.

If I were to biwire, I'd do so with two separate pairs of cables since they are not as limiting as a single biwire pair. Given the expense that one is likely to incur in biwire cables, it makes sense to buy separate pairs. One could get more use out of separate pairs down the road (biamping, different systems, etc.).

 

[brucek]

Donald,

One other question. Is their any specific wire type/brand I should look for. Monster? Canare? And when you say heavy, what do you consider heavy- 14? 12? Heavier?

The question is of course, "How much performance are you willing to lose?" - because, no matter what, you'll always lose performance with any speaker wire. Increase the length, or decrease the size, - the worse it gets.
I think 12 gauge should cover most situations of reasonable lengths. Anything much over 50 feet you might like to use 10 gauge. Personally I don't think the brand matters too much, as long as the specs are good. Many will disagree, but unless I'm missing some new primary constants that can be applied to a cable, the electrical characteristics of cables such as this are well understood in the audio spectrum. Specifications are usually better in the more expensive cables. Whether you hear the difference is up to you. The performance can easily be calculated given the cables specifications.
It's easy to calculate the impedance of your speaker wire run and then using the characteristic impedance of the speaker that you're using determine what percentage of power you'll lose in the cable and what your damping factor in the system will be.
There's an old rule of thumb that says you shouldn't lose more than 10% of your power in your speaker cable and that a damping factor of greater than 10 is OK. Well, this is valid for some people, but it wouldn't satisfy any audiophile that I know. I personally go quite a bit beyond that and use 9 gauge cable - kinda silly. The resulting impedance and damping factor is very good. Does that mean it would be mathematically better if I went further and used 8 gauge. Yes, but I feel I wouldn't hear any difference and so I have reached the point where I am comfortable with the cable I'm using. Really, anything beyond 10 gauge and it becomes an art form.
Basically, you want the cable's impedance to be as small a factor in relation to the speakers impedance as possible. The speakers impedance is not linear with frequency and if the wires impedance is a large enough proportion of that impedance you will indeed change the response of the output.
Anyway, 12 gauge speaker cable is usually an excellent choice and is available from Home Depot or Radio Shack....If you decide to step up to 10 gauge, you'll not likely hear the difference. Here's a quick comparison of standard 12 gauge and 10 gauge speaker wire.
The 12 gauge cable will exhibit a simple (frequency independent) resistance of about 1.62 ohms per thousand feet. So for example, with a typical 10 foot speaker cable, your cable's simple resistance will be about 0.0324 ohms. In relation to the 8 ohm speaker you're likely using, this can be considered statistically insignificant, and effectively lossless.
The 10 gauge will exhibit a simple resistance of about 1.02 ohms per thousand feet. So, for a typical 10 foot speaker cable, your cable's simple resistance will be about 0.0204 ohms. In relation to the 8 ohm speaker you're likely using, this again can be considered statistically insignificant, and effectively lossless.
If you like percentages, the 12 gauge in this example with respect to its simple resistance would exhibit a loss of about 0.4% loss of power in the wire in the form of heat and the 10 gauge would exhibit a loss of about 0.25 % loss of power in the cable. Not much difference there.
Both cables will also exhibit some series inductive reactance, which increases with frequency and length. In a typical zip parallel cable, a standard ball park value is about 0.20 microhenries per foot. The interesting thing, is that the inductance changes very little with changes in wire diameter. Yes, the higher the gauge (smaller wire), the higher the inductance, but it's basically insignificant.
So, at 1000 Hz your frequency dependant series resistance (inductive reactance) would be 0.00125 ohms. Again, completely insignificant. At 20000 Hz (worst case), then your frequency dependant resistance (inductive reactance) would be about 0.2512 ohms. In relation to your 8 ohm speaker, this would be a 0.13dB loss in the form of heat at worst case 20Khz. Can you hear a 0.13dB loss at 20 Khz, or even hear 20 Khz? Either way, both the 12 gauge and 10 gauge will suffer about the same fairly insignificant roll off from this reactance factor. There are expensive speaker cables that possess better inductance specs per foot than standard zip, and as such will mathematically result in better performance. Whether you hear the difference is up to you to decide......
The loss from parallel capacitive reactance in a low impedance interface isn't even worth deriving because it is so insignificant. You see, we don't generally get concerned about capacitance in speaker cables because this is a low impedance interface. I'll explain. The output impedance of a typical amplifier is around .02 ohms and it's attached to a load of about 8 ohms. This is a very low impedance circuit. Unlike inductance, capacitance provides us with a "parallel" reactance. This frequency dependant resistance is "across" or in parallel with the load. The load is 8 ohms. A typical parallel load that the capacitance of zip cable provides is many orders higher than 8 ohms. It's in the thousands of ohms. If you want to do the math - go ahead. It's not significant enough to even talk about. If we were discussing interconnects, of course the opposite is true because it is a high impedance connection.
So what does all this crap mean? It means there isn't much difference between 12 gauge and 10 gauge. For most applications 12 gauge is a good choice. If the cost is the same, I would choose the 10 gauge.
brucek

 

[John Royster]

Bruce,

Thanks for the info. How does inductance and capacitance relate to a speaker that is 4 ohms nominal and dips to 1 ohm in the higher frequencies?

 

[brucek]

John,

Well, it's a simple voltage divider between your load driver in question and the cable you're using to supply it.

If I take an extreme example where you're using this speaker 50 feet from an amplifier, and I'm using 12 gauge parallel zip cable with a standard inductance spec of 0.20 microhenries/foot and a standard resistance of 1.1619 ohms/1000feet and an assumed spec of 1 ohm at 20KHz for your speaker (a guess from your post).

This results in a speaker cable DCR of 0.1619 ohms, and at 20 KHz an inductive reactance of 1.26 ohms. This realizes a total series (frequency dependant) speaker cable resistance of 1.42 ohms at 20KHz.

Consider that the amplifier is a perfect voltage source, you have a fairly awful voltage divider setup between the 1.42 ohm impedance of your cable and the 1 ohm impedance of your speaker in question. You are losing about 58% of your signal at 20KHz in your cable (in this example).

To say that you would have high frequency roll-off in this situation would be an understatement...........

It would be nice if speaker cable impedance could be zero and given that an amplifier was indeed a perfect voltage source, then we wouldn't be that concerned about the fluctuations in impedance of drivers or crossovers, but it just ain't so.
So, you attempt to achieve a speaker system impedance curve that doesn't do nasty things (like drop to 1 ohm), and you try and keep your cables as short as possible with the best possible specs.

The short answer.
Inductance of a cable will have a large effect on that speaker. Capacitance can be ignored.....

brucek