Why Bi-wiring Makes No Sense.

[jneutron]

I am baffled it took so long to make this point clear. Why argue that bi-wiring has any purpose? It is more cost effective to just get a higher gauge from my understanding of these posts by Mtrycrafts, MDS, Clint, Gene, Jonnythan, or anyone one else that has made contributions to the argument against Bi-wiring.

For the example I provided, the copper required to run mono was three times heavier than biwire. That is efficient use of materials for that application.

While they have argued, they have not provided any factual basis to dispute the analysis..indeed, they still do not fully understand the analysis. I cannot blame them for that however.. Far stronger e/m people have taken much longer to understand.

Higher cost defeats the purpose of bi-wiring, regardless if two 12 gauge cables perform better than just one. If you buy expensive 10 gauge wire and that becomes the argument, let that be your problem. I would sooner purchase Romex cable than high gauge snake oil crap.

Who said biwiring forces higher cost? I showed #2 AWG being replaced with two #10's..that is one third the cost.

I do not buy, nor recommend, high priced cables of any type. In fact, my analysis provides a methodology to circumvent high guage wire as a band aid solution.

To go even further, it is trivially possible to get rid of the issue entirely and run small guage wire to the speaker while eliminating the dissipation modulation. I figured two methods to do that years ago. Alas, both require amplifier modifications. But their darn efficient when it comes to copper..

The thread title is "Why Bi-wiring makes no sense". The way I see things, it still does not offer any advantage over traditional single wire configurations. I will admit that many posts made previously went way over my head, but there have been enough posts that I could understand the jest of it....

The posts you understand do not address nor speak of the analysis. So, you do not understand the gist of the analysis, just the posts where they also do not understand the gist.

, Bi-wiring is not an efficient or effective way to connect speakers to an amplifier! Even in the circumstance that the amplifier doesn't have wiring running from it's output to the terminal. A sensible person would not bi-wire their speakers.....

sigh..a sensible person would use the engineering expertise available to him or her to decide which path optimizes sound and money.

This site is dedicated to such a concept..and as such, I have posted the analysis of biwiring to further allow others to make an informed decision as to biwiring or not.

Cheers, John

 

[emorphien]

Jneutron I wish I had the EE knowledge to keep up with what you're saying, but I find in spirit that what you say is very valuable information. I think it's important to not downplay anything until there are measurable results. Too many comparisons are done here and elsewhere based on theory, weak measurements or faulty techniques.

I don't know the benefits of bi-wire, whether they exist or are worth the effort most times, but I would say that there are other things which are definitely effective and probably worth dealing with first (IE room acoustics).

 

[jneutron]

Jneutron I wish I had the EE knowledge to keep up with what you're saying, but I find in spirit that what you say is very valuable information. I think it's important to not downplay anything until there are measurable results. Too many comparisons are done here and elsewhere based on theory, weak measurements or faulty techniques.

The bulk of the writeups I find on this site are quite good and very informative (even the ones I was not involved in.. ).

Honestly, I come here whenever I need to learn about all that hdmi/dvi/whatever stuff. Recently purchased a 40 inch lcd, and the articles here helped tremendously. When I was ready to purchase, I at least had a good understanding of all the nomenclature..what an alphabet soup it is..almost as bad as at work.. Not knowing better, I would have purchased a "name brand" hdmi cable for 99 bucks, whereas sites listed here, 15 bucks.. So, while Gene and I have our "tiffs" so to speak, I still consider this site rather valuable for info. (on a side note, when I was there, the salesperson was pushing the hdmi with "nitrogen filled dielectric"...I asked him what a "nitrogen filled dielectric" was and why it was good, my significant other told me to "step back", and she told him "not to go there if you know what's good for you"...oh man, I was looking forward to "lunch"..)

As for cables...the bulk of what is written here is totally useful for the vast bulk of the population. Not too many people listen in the sweetspot, not too many worry about imaging, they are a rather select group. Novices coming here receive very valuable information.

I've mentioned here several times, that one should distinguish between high end listeners and the bulk of the population..they have different requirements. Making the distinction allows a better discussion of what is reasonable to do, and what is excessive.

Stupid things are an entirely different category.. most of the articles here are quite good for those.

I don't know the benefits of bi-wire, whether they exist or are worth the effort most times, but I would say that there are other things which are definitely effective and probably worth dealing with first (IE room acoustics).

We concur.

 

[mtrycrafts]

here's the excel in tst form..

Cheers, John

Look at the powerloss comparisons..I'll get back and edit this post to detail..
bummer, I note the columns didn't remain correct..case 2 powerloss totals moved to the right a bit..

If I don't ask questions, it is harder to learn.

That power loss of 1% represents how much dB spl? In what frequency band?That modulation of hf you mentioned, is that the same amount or even less? If less, what is that amount, in dB spl?

And, why are we interested in this instantaneous event beyond pure numbers? Is this something visible on the scope? Can you see these interactions on the scope? Is this something that is audible?

 

[stratman]

Great thread, very informative even for us cave crowd. Bottom line: no need to bi-wire, use good quality "common sense" wiring, make sure you pay attention to room acoustics and speaker placement, make sure all your connections are tight.......apart form the intellectual EE arguments and pleas, to a simple mind such as mine comes the natural conclusion stated above. Common Audioholics sense wouldn't you say?

 

[jneutron]

If I don't ask questions, it is harder to learn.

That power loss of 1% represents how much dB spl? In what frequency band?That modulation of hf you mentioned, is that the same amount or even less? If less, what is that amount, in dB spl?

And, why are we interested in this instantaneous event beyond pure numbers? Is this something visible on the scope? Can you see these interactions on the scope? Is this something that is audible?

Ah, so many questions....

The first thing to note, is that while you are searching for specific numbers to provide a reference level with regard to established literature, what it affects, you will not find..it affects localization, and to spot it requires differential localization testing. This does not exist in the research world as of this date..

That said..

Last Q first..that 2AB product I detailed, is the product of two frequencies, so there are frequencies produced that are not the fundamentals, It's kind of like IM distortion, but the signal itself is a zero power one, not amenable to FFT's. So great, can't measure it that way either...sheesh.

Let's look at the 5% option. If you monowire at 5% line loss, you incur a .2 dB penalty on a per channel basis.. For a two channel program, this means that one channel's bass information has the ability to level shift the highs on that channel by about .2 dB. For a speaker pair spaced 10 feet apart and 10 feet away, that represents a horizontal virtual image shift of FIVE FEET...five frikken feet!!! Biwire stops that dead in it's track.

What does it take to keep the side shift to a foot? about .08dB tracking accuracy..I don't know bout you, but I call that SMALL.

And if you're not trying to discern a virtual image, those numbers are so ridiculously low that even I wouldn't try to claim audibility..

That is why I do not recommend biwiring to the masses..it is hugely inappropriate, as most are not concerned about it.

But as you can see, the effect on soundstage can be rather large.

And Gene, yes, the numbers are based on equal power and the wire having 5% of the load resistance.. Shift the low to high ratio by a factor of 4, you drop to the 6-12 inch differential localization range. Discernable? maybe, maybe not..By me? Why, I don't care at this point in my life. I prefer to listen to the music.

Do not forget, I have not included either the line inductance, nor the driver reactance, which add to the localization errors in a very nefarious way....they change the ITD parameters...time shift of the information based on modulation..The actual analysis is not very easy to do..I've pushed it with just the resistance aspect of the cables.

Cheers, John

 

[jneutron]

I maintain the main benefit of Bi-wiring (if any) is lower cable resistance but higher reactance.

BTW, couldn't let this pass..

The main benefit is not lower resistance, it is the elimination of resistance as a dissipation modulation component.

The reactance..there are two parts.

Capacitive will indeed be higher.

Inductive will be reduced on the whole, as they are paralleled in essence..parallel inductors are lower inductance.

Remember, inductance is the relation between the drive current, and the stored energy within the magnetic field. If you double the current within the wire, as monowiring does with respect to biwiring, you quadruple the stored energy within the wire.

Cheers, John

 

[gene]

John;

I don't argue your math, just your premise.

Follow me on these points please:
1. You will NEVER have the same current draw on the tweet as you will the woof. More likely it will be a 1:10 (tweet:woof) ratio. Thus your assumptions about equal current to both wires is leading to a wrong conclusion for a real world scenario
2. Our brains are NOT good at deciphering instantaneous changes like this, please review our articles on Human Hearing for more elaboration
3. The power loss from the wire 1 watt vs 100 watt into speaker represents a 20dB SPL difference - again not audible instantaneously
4. Your assuming correlated power losses between each cable which again isn't real world since a speaker is NOT a resistor
5. The change in voice coil resistance (at rest vs under power) is so profoundly GREATER than the cable loss you discuss that the cable becomes moot again - measure a driver DCR at rest vs the DCR after loaded on an impedance analyzer to demonstrate this to yourself

You did basic math on an oversimplified model which lead to overcomplicated results.

Now I suggest you do some real measurements in a lab with real equipment and stop simulating resistive losses in cables attached to resistive loads running identical current magnitudes

• Use a current probe to measure current for HP and LP sections comparing monowire and biwire under the power levels of your simulations. Be warned however most tweeters will instantly fry if you feed them more than 20-50 watts of power.

• Measure the DCR of voice coil of the drivers before and after loading and factor that into you losses.

Then redo your sims. If you need, I have a PSPICE model for a real world speaker somewhere which you can use to build a more accurate simulation.

 

[gene]

BTW, couldn't let this pass..

The main benefit is not lower resistance, it is the elimination of resistance as a dissipation modulation component.

The reactance..there are two parts.

Capacitive will indeed be higher.

Inductive will be reduced on the whole, as they are paralleled in essence..parallel inductors are lower inductance.

Remember, inductance is the relation between the drive current, and the stored energy within the magnetic field. If you double the current within the wire, as monowiring does with respect to biwiring, you quadruple the stored energy within the wire.

Cheers, John

Yes using identical cables for biwiring:
capacitance will be 2 x
Inductance will be 1/2 to 1/4 depending on coupling
resistance will be 1/2

So now you run the risk of running a highly capacitive cable when biwired and causing an amp instability

 

[jneutron]

John;

I don't argue your math, just your premise.

Follow me on these points please:
1. You will NEVER have the same current draw on the tweet as you will the woof. More likely it will be a 1:10 (tweet:woof) ratio. Thus your assumptions about equal current to both wires is leading to a wrong conclusion for a real world scenario

You still do not understand conceptually what I am talking about.

Higher woofer currents are not better. The effect approaches maximum when the woofer current equals the tweeter current. The fact that in reality the lows are much stronger does not support your statement.. When you understand the analysis, you will see that.

I could have used DC as the lows excitation if I wished.

2. Our brains are NOT good at deciphering instantaneous changes like this, please review our articles on Human Hearing for more elaboration

Where in your article do you discuss localization parameters?

3. The power loss from the wire 1 watt vs 100 watt into speaker represents a 20dB SPL difference - again not audible instantaneously

Gene, you should re-state that. As written, it has no meaning. (I proof my posts multiple times to find my silly errors..)

4. Your assuming correlated losses between each cable which again isn't real world since a speaker is NOT a resistor

I have no idea where you got that from. In point of fact, the analysis I provided assumes orthogonal signals, one to each driver. So, they are in fact, already uncorrelated..again, you should be asking questions rather than making incorrect statements.

5. The change in voice coil resistance is so profoundly GREATER than the cable loss you discuss that the cable becomes moot again - measure a driver DCR at rest vs the DCR after loaded on an impedance analyzer to demonstrate this to yourself

Again, you must understand the analysis and the effects. The 4% per degree C change in copper resistivity which corresponds to power compression is a small effect with respect to the wire losses...remember, if the vc ups 10%, the diss in the wire becomes 10% less.

You MUST understand the analysis and effect before you can make relevant statements..please ask, either here or via e-mail.

You did basic math on an oversimplified model which lead to overcomplicated results.

Actually, that is quite wrong. I have presented only the aspects which I believe people here will be able to understand. So far, that assumption has not born fruit.

Now I suggest you do some real measurements in a lab with real equipment and stop simulating resistive losses in cables attached to resistive loads running identical current magnitudes

Silly you...you make the assumption that I have not???? And that I have not used some "real equipment" at my disposal..amusing..

Use a current probe to measure current for HP and LP sections comparing monowire and biwire under the power levels of your simulations. Be warned however most tweeters will instantly fry if you feed them more than 20-50 watts of power.

Why in gods name would I use such a piece of garbage to measure high slew rate currents?? I strive for accuracy, and you talk about current probes???puuullllleeeeeese.. seriously, I strive for Ghz level components for these types of measurements..that way I have no B dot errors in the CVR's. My typical load resistor has 60 picohenries of inductance, and no measureable current slew rate error. If I wished to be so sloppy, I'd trust a current probe.

Measure the DCR of voice coil of the drivers before and after loading and factor that into you losses.

As I stated, that is of no concern to the analysis.

Your still counting beans there Gene. It's not about the overall efficiency of the system, but rather, how that power is lost (or gained) with respect to a monowire, and how that loss impacts localization.

Gene, you really need to sit back and understand what I've been saying. It's clear you do not, and if you do not try, this discussion will not work very well. I'm off for now, relax, cool down, and re-visit what I've shown here when you're in a better mindset.

BTW, here's a graph showing the dissipative modulation when the lows are presented as DC. A DC signal is still orthogonal to a sine, therefore, non correlated..

Cheers, John

 

[jneutron]

Yes using identical cables for biwiring:
capacitance will be 2 x
Inductance will be 1/2 to 1/4 depending on coupling
resistance will be 1/2

So now you run the risk of running a highly capacitive cable when biwired and causing an amp instability

Only if the cap sum exceeds a stability level. If you examine your own writeups, you will see that the majority of the cables you measured had rather normal capacitances..so your point is moot. Even if you double the cable lengths in the mono situation, which has the same effect, it's not an issue.

If one tried to do up a whole mess of cat 5's, or some other ridiculous cable scheme, it'd be rather easy to go into oscillation.

It's not a normal risk, so don't worry bout it..

Have a good night, we can pick this up tomorrow if you wish..

Cheers, John

 

[dave1490]

Only if the cap sum exceeds a stability level. If you examine your own writeups, you will see that the majority of the cables you measured had rather normal capacitances..so your point is moot. Even if you double the cable lengths in the mono situation, which has the same effect, it's not an issue.

If one tried to do up a whole mess of cat 5's, or some other ridiculous cable scheme, it'd be rather easy to go into oscillation.

It's not a normal risk, so don't worry bout it..

Have a good night, we can pick this up tomorrow if you wish..

Cheers, John

the last few post are great keep argueing.

 

[Audioholics]

"I argue very well. Ask any of my remaining friends. I can win an argument on any topic, against any opponent. People know this, and steer clear of me at parties. Often, as a sign of their great respect, they don't even invite me."

- Dave Barry​

 

[mtrycrafts]

Ah, so many questions....

Yep, that is me, until I have an understanding that doesn't need more Q
What can I say. Wasn't born an Einstein or some other greats

Last Q first..that 2AB product I detailed, is the product of two frequencies, so there are frequencies produced that are not the fundamentals, It's kind of like IM distortion, but the signal itself is a zero power one, not amenable to FFT's. So great, can't measure it that way either...sheesh.

Let's look at the 5% option. If you monowire at 5% line loss, you incur a .2 dB penalty on a per channel basis.. For a two channel program, this means that one channel's bass information has the ability to level shift the highs on that channel by about .2 dB. For a speaker pair spaced 10 feet apart and 10 feet away, that represents a horizontal virtual image shift of FIVE FEET...five frikken feet!!! Biwire stops that dead in it's track.

Now we are getting to my questions that I can at least relate to. So, this effect affects horizontal image shift and level shift. Level change is rather small, especially with music, so good luck hearing that with speakers.

Now, it happens for a very short time period, right? And, only under those specific conditions when the signals, low and high in the same speaker are exactly the same level, right? And, when signals are other than this, these effects are much smaller, right?

Yes, that 5 ft in shift is huge, by the calculations you presented. This is testable as it is not a small shift by any means. Now, can we react to such a very short time period, with music? after all, a shift from left to center is huge. I cannot ever recall such an event, even though my memory is zilch
So, I am trying to get a handle on this huge shift being audible.

But as you can see, the effect on soundstage can be rather large.

Yes, it is huge, and am amazed that I haven't seen such a shift in location, whether from the center to 5 ft left or right, or, from any other intermediate location moving 5 ft. That effect is such huge that one could not forget such a movement. Perhaps the time interval is so short that the brain will not process it, with music?

Do not forget, I have not included either the line inductance, nor the driver reactance, which add to the localization errors in a very nefarious way....they change the ITD parameters...time shift of the information based on modulation..The actual analysis is not very easy to do..I've pushed it with just the resistance aspect of the cables.

Cheers, John

Maybe, perhaps, with all this in, you add music, and is no wonder we don't see 5 ft of shifting in the middle of music?

 

[mtrycrafts]

Common Audioholics sense wouldn't you say?

I would, absolutely

 

[gene]

Let's look at the 5% option. If you monowire at 5% line loss, you incur a .2 dB penalty on a per channel basis.. For a two channel program, this means that one channel's bass information has the ability to level shift the highs on that channel by about .2 dB. For a speaker pair spaced 10 feet apart and 10 feet away, that represents a horizontal virtual image shift of FIVE FEET...five frikken feet!!! Biwire stops that dead in it's track.

Um sticking a chair in the room will have a more profound effect on system frequency and amplitude response

Seriously when any installer worth his/her salt calibrates a system, they do so on a per channel basis first with an accurate SPL meter, and then with an RTA. Any good processor has time alignment adjustments in .1ft increments and .5dB level steps. Also an installer should never set up a system for a single sweet spot IMO. Its impractical and a waste of not being able to share an equally good experience with friends/family.

I've never seen a .2dB variance between speakers b/c of cabling. Most rooms experience +-20dB variances in bass response. Small room acoustics is very difficult below the transition region of 300Hz.

Of course I biamp my mains, run 4 subs in mono, and use spatially average active equalization to achieve bass response in my room +-5dB from 17Hz to 100Hz averaged across the entire listening area so my setup is an exception and far from typical among even the high end audio crowd.

So, while Gene and I have our "tiffs" so to speak

This is far from one my friend. I enjoy this dialog and regret I cannot stay active with all of it with everything else going on. If you send me a formal writeup on "Biwiring" with measurements, graphs, maths, etc, via email, I'd like to sit down, print it out and go over it. Perhaps I can add a more practical flare to it and turn it into a useful article as well so you can be even more proud of the content on this site

Its hard to gather bits and pieces on a forum. Perhaps down the road I can assist with a DBT test on my system

 

[Nick250]

The obvious was always there. Why did it take you 150 posts to chime in?

Is the common inductance problem also so obvious to you?

Cheers, John

I do not understand all the mumbo jumbo and calculations. Then after all that and at 150 posts, we got to the part where it was proclaimed that a single run of fatter speaker wire takes care takes care of any of bi-wire sonic issues, real or imagined which has been written about and discussed a zillion times before. No offense intended, it was just bizarre.

Nick

 

[Tom Andry]

I do not understand all the mumbo jumbo and calculations. Then after all that and at 150 posts, we got to the part where it was proclaimed that a single run of fatter speaker wire takes care takes care of any of bi-wire sonic issues, real or imagined which has been written about and discussed a zillion times before. No offense intended, it was just bizarre.

Nick

You obviously don't hang out with many engineers!

 

[mtrycrafts]

Um sticking a chair in the room will have a more profound effect on system frequency and amplitude response

Seriously when any installer worth his/her salt calibrates a system, they do so on a per channel basis first with an accurate SPL meter, and then with an RTA. Any good processor has time alignment adjustments in .1ft increments and .5dB level steps. Also an installer should never set up a system for a single sweet spot IMO. Its impractical and a waste of not being able to share an equally good experience with friends/family.

I've never seen a .2dB variance between speakers b/c of cabling. Most rooms experience +-20dB variances in bass response. Small room acoustics is very difficult below the transition region of 300Hz.

Of course I biamp my mains, run 4 subs in mono, and use spatially average active equalization to achieve bass response in my room +-5dB from 17Hz to 100Hz averaged across the entire listening area so my setup is an exception and far from typical among even the high end audio crowd.


This is far from one my friend. I enjoy this dialog and regret I cannot stay active with all of it with everything else going on. If you send me a formal writeup on "Biwiring" with measurements, graphs, maths, etc, via email, I'd like to sit down, print it out and go over it. Perhaps I can add a more practical flare to it and turn it into a useful article as well so you can be even more proud of the content on this site

Its hard to gather bits and pieces on a forum. Perhaps down the road I can assist with a DBT test on my system

I think this might be his case here:
For a two channel program, this means that one channel's bass information has the ability to level shift the highs on that channel by about .2 dB. For a speaker pair spaced 10 feet apart and 10 feet away, that represents a horizontal virtual image shift of FIVE FEET...five frikken feet!!! Biwire stops that dead in it's track.

Not that a fixed .2 dB spl will cause a fixed skew of the horizontal virtual image but that it will be shifting as this effect comes and goes, depending on frequency content and level of the two signals, lows and highs, at the very same instant. With his example, these signals being equal voltages.

5 ft of image shift is huge and a cinch to detect, especially as it comes and goes, like a ghost. But, I have never heard such an effect nor have anyone else reported such an effect before that I know of. So, I am grappling with this issue if this condition would ever occur. And, if it is that rare, what is there to worry about?

 

[rode]

Many of the points made by jneutron are perfectly valid at first glance, but I think that some may not stand up to scrutiny. This depends entirely upon what we look at in the circuit. I shall elaborate on that a little more shortly, but need to cover the general bi-wiring topic first.

Anyone who has visited my site knows that I don't like passive crossovers, although they can make a convenient door-stop in an emergency. An active system has so many advantages that I wouldn't even consider bi-wiring, so to me the point is moot. For those who buy (rather than build) their gear, active systems can become expensive, so passives will be around for a few more weeks yet

Simulations of bi-wiring do show that there are some benefits, and although easily simulated and (somewhat less easily) measured, audibility is not proven. Some people do have much more sensitive hearing than others, and they may be able to hear a difference, where most cannot. I doubt that this is especially relevant, as proper double-blind tests have not been done anywhere to my knowledge. Sighted tests are pointless at best, and warrant no further comment

I ran a simulation based on jneutron's (perfectly valid) claim that DC can be used for one signal in the cable. As predicted (and amply demonstrated), this does indeed cause distortion of the power waveform of the resistive component of the cable. Zero argument there - this much should be quite obvious to anyone reasonably technical ... as hinted at by jneutron.

What is not intuitively obvious is that even though the power waveform of cable dissipation is distorted, this does not affect the voltage applied to the load - the sinewave (or complex waveform) component remains at the same level and distortion with or without the DC (or low frequency AC), and level is affected only by the cable's series resistance. Further effects (such as instantaneous compression) can occur if the power is sufficiently great to cause the cable to become warm or hot, but this will only happen if the wire is insanely thin for the expected power (such as that in a voicecoil ... another topic altogether).

Having only become aware of this thread a couple of hours ago, I may have missed something - if so please let me know (yeah, like you wouldn't )

In the meantime, yes, having both signals flowing in the same cable will cause the power dissipation waveform of the cable to become distorted, but it does not affect the voltage waveform applied to the load one iota. This assumes pure resistance only - including inductance and capacitance will have other effects, but they tend to be linear (as in not introducing additional harmonics as expected from 'non-linear' distortion).

Where there may be audible effects from bi-wiring (and that remains a big maybe), the (alleged) problems are caused largely by the cable's resistance and inductance interacting with the crossover network. These interactions will be small but measurable with most speakers, but potentially large and maybe audible with others.

If the xover network is not designed to accommodate real world source impedances, and/or has significant impedance variation at the xover frequency, then undesirable (i.e. 'bad') things can happen. Bi-wiring may alleviate some of the problems and thus sound better (or at least different).

Does this help at all?

Cheers, Rod

sound.au.com