bi-wire theory

Raphael said:

The gist of it is to separate the x-over into it's hi and lo components--essentially making a parallel x-over that's bi-wirable. This--of course--would be different for many x-over designs and requires new component values so should only be accomplished by those who feel comfortable with x-over modifications and knows or can accurately extrapulate his/her x-over points based on speaker/cabinet design. Basically . . I know enough not to try it!

Click to expand...

Then it's no longer a series x-over and all of the component values are going to be very wrong.

You would lose all of the benefits of having a series x-over, plus basically have to re-engineer the speaker design and build a new x-over.

All for bi-wiring that doesn't do a bit of good in the first place.

BobK101 said:

Hi,
I have Vandersteen II 's and they recomend Bi Wiring... Here is a very interesting article on it from their web site... Worth a read..

http://vandersteen.com/pages/Answr7.htm

Bob

Click to expand...

I just lost a lot of my respect for Richard Vandersteen.

Adding an extra set of speaker wires does not change the fact that you're dealing with one linear circuit. One, not two or more.

http://pespmc1.vub.ac.be/ASC/PRINCI_SUPER.html

http://en.wikipedia.org/wiki/Superposition_principle

Pretty much the only thing you might accomplish by bi-wiring is to reduce the resistance of the speaker wire.

Most people already have speaker wire that is more than adequate in terms of gauge. Turning up your volume control a hair will then do the same thing as doubling up on your wire.

If you don't feel your speaker wire is adequate, then go with a larger gauge of wire. That will be far more worthwhile than "bi-wiring" ever will.

Superpositioning is fine.... But, it is not accounting for moving mass of the woofer creating EMF that needs to be dampened by the AMP that is the on the other end of the speaker cable. Also, it is not considering differences in signal reflections due the re-active load the speakers represent. It would be interesting to see the frequency content of the self-induced EMF and what amount of it makes it back past the x-over.

Remember that crossovers have roll off of 6 or 3 db.....and allow for some limited amount of overlap.

I also think some amount of reciprocity is occurring.... What does that mean? A speaker can actually work like a microphone....

What I heard was exactly what the articles speak of... improved clarity and definition in the Mid/Hi ranges.... I would guess that some of the Woofers Self induced EMF is in the higher freq. range...also remembering that a speaker can act like a microphone... Some of which might make it past the x-over then enters the Mid/Hi range driver as distortion.....

By separating the signal path two things happen to the Self-Induced EMF... One, it has MANY TIMES the length of wire to travel to reach the Mid/HI end drivers. Remember in a normal X-over the distant between the WOOFER/MID/HI X-over components is maybe an inch of copper run.... Now on a BI-WIRE system the Self-Induce hi-freq EMF would have to travel 360 inches... My speaker wire is 15 feet long x 2 ( biwired ) x12 inches/foot = 360 inches.

360" vs. 1" that is 360Xs longer....

Lastly, the dampening load of the AMP in now in the middle of the Woofer and the Mid/Hi Freq drivers.... Where as before the Woofer was 1" away from the Mid/Hi freq drivers and the AMP's dampening load 15 feet or 180 inches away. Bi-wire gives the AMP's dampening load a much greater ability to dampen out the Woofer's Self induced EMF before it reaches the Mid/Hi driver.


In a normal X-over the Mid/Hi drivers are 1 inch away from the woofer and the AMP's dampening load is 180 inches away.... The dampening load is 180 times further away than the X-over for the Mid/Hi drivers with single wire pair....

In Bi-Wired the AMP's Dampening load is now closer to the woofer at the mid point of 180 inches with a total 360 inches of speaker wire than the Mid/hi freq drivers ....and the Mid/Hi drivers are another 180 inches further away. This also puts the AMP's power supply in between the Woofer and Mid/Hi drivers....that puts all your Supply Filter caps in the mid of the X-over too....


This example is based on my speaker cables of 15 feet. Just some numbers to look at.... For 14 gauge wire....15 feet is about .04 ohm.... for 1 inch it is .00021 ohm.... for 30 feet of 14 gauge wire is total is .08 ohm.... For an AMP with a Dampening factor of 100, the internal resistance would be .08 ohm for a 8 ohm load.... Just some math.


jk

jk, your description sounds pretty reasonable to me. I suspect there are real, tangible benefits to bi-wiring.

Nevertheless, I would not recommend it when a far superior solution, namely bi-amping, is available.

Today amps are sooo cheap, especially mid to low power amps. When we bi-amp, we really don't need huge amps powering the mids & tweeters. Those high quality mid and low power amps that everyone ignores, work just fine!

Bi-amping gets all the benefits of bi-wiring, PLUS:

1. zero back EMF from the woofer will ever get to the other drivers
2. splitting the xovers reduces the potential for self resonant circuits within these complex xovers
3. reduces the potential for IM distortion in the amps

Regards,
Jerry

Jerry,

I agree.... You would think that with a 4400, I would have enough amps and preamps to Bi-amp my setup.... One of these days I'll get all the 32 year old bugs worked out of it.....

But I was surprised at the Bi-wire improvements....

johnk

Could I get the same effect by using a .04 ohm resistor betweeen my woofer and mid/hi? That would be a lot cheaper than using another 30 feet of wire.

I also think high efficiency speakers make the improvement more noticeable....

Would a high efficiency woofer make more self induced EMF? And then benefit more from being Bi-wire? My speakers are efficient....

jk

Scuzzer said:

Could I get the same effect by using a .04 ohm resistor betweeen my woofer and mid/hi? That would be a lot cheaper than using another 30 feet of wire.

Click to expand...

Don't think that would work....


Think of it like a Y..... The top legs of the Y being a very short copper run in the X-over and the bottom leg is a long speaker wire run.....


What you want to do is lengthen the top legs and shorten the bottom leg as much as possible..... But I had two spare X-overs to add to woofer legs...

Adding resistance in the speaker will not do that....

jk

jstang said:

Don't think that would work....


Think of it like a Y..... The top legs of the Y being a very short copper run in the X-over and the bottom leg is a long speaker wire run.....


What you want to do is lengthen the top legs and shorten the bottom leg as much as possible..... But I had two spare X-overs to add to woofer legs...

Adding resistance in the speaker will not do that....

jk

Click to expand...

Yeah, I understand the concept of bi-wiring. It's the interactive effect that is more difficult to get my head around. I don't have a firm grasp of back EMF and it's ramifications so I ask questions and try not to look too stupid.

From your answer above the wires must be doing more than just adding resistance and that effect is somehow changing the interaction between woofer and tweeter.

Went over board on this, but what the heck.....


Standard wiring..... Woofer to Mid/Hi is .00021 Ohm, Woofer to amp is .08ohm.

Woofer
|
|(Single X-o)__________________________________________180" X 2 = .08 ohm_____________________ AMP internal load of .08 Ohms
|
|
Mid/Hi Driver


Bi-wired

Woofer (Separate X-o) _________________________________180" X 2 = .08 ohm_____________________AMP internal load of .08 Ohms
Mid/hi Driver(Separate X-o)_____________________________ 180" X 2 = .08 ohm______________________/


The Biwired amps internal load would have a better chance of dampening the EMF from the woofer. Also, the wire has a self inductance and capacitive value that has an influence on the signal.


One item I forgot to mention....and calculate is that the lengths are actually twice the calculated lengths and Ohm readings because I forgot to add the return path...

720" vs. 1" that is 720 Xs longer....

This example is based on my speaker cables of 15 feet TIMES 2. Just some numbers to look at.... For 14 gauge wire....15 feet ( ground & signal ) is about .08 ohm.... for 1 inch it is .00021 ohm.... for 30 feet of 14 gauge wire the total is .16 ohm.... For an AMP with a Dampening factor of 100, the internal resistance would be .08 ohm for a 8 ohm load.... Just some math.


I also tried moving the woofer X-overs from the speakers to the output of the AMP.... I did not like the sound as it seemed to kill the Bass..... not sure why that would have of happen. I mover the woofer X-overs back to speaker.

Some URLs that speak to what is happening in general....these two items below are helped by more dampening, Biamping and Biwire... I think it helps keep the mid ranges near the crossover point clean of Back EMF/Microphone distortions from the woofers.....


http://www.sweetwater.com/shop/live-sound/power-amplifiers/glossary.php

"Back-Emf
Literally,, back-voltage, is a phenomena found in all moving-coil electromagnetic systems, but for audio is most often used with respect to loudspeaker operation. This term describes the action where, after the signal stops, the speaker cone continues moving (due to inertia), causing the voice coil to move through the magnetic field (now acting as a microphone), creating a new voltage that tries to drive the cable back to the power amplifier's output. If the loudspeaker does too much of this, the cone flops around unpleasantly. It is not pleasant-sounding. To stop back-emf, the loudspeaker must "see" zero ohms looking backward (a dead short), or as close to it as possible from the output of the amplifier. "

http://www.meta-gizmo.com/Tri/june/Darkness.html
"EVERY SPEAKER IS A MICROPHONE:
When this was demonstrated to me, thirty years ago, it blew my mind and it is very easy for you to discover it if you have an oscilloscope. Attach one speaker to a tube amplifier that is on, but no making any music. Connect your oscilloscope across the B+rails. With your other speaker and amplifier turn up the music and watch the oscilloscope trace bounce up and down, as the loudspeaker, acting as a microphone, is producing an AC signal that is surging back into the power supply…as a form of distortion. By this way this phenomenon will not affect your amplifier if you are using actively regulated power supplies.

While your speakers are creating sound waves in your rooms, those sound waves are bouncing back at some time delay and striking the driver, which generates an AC signal. The bigger the loudspeaker driver the more efficient it is as a microphone.

The use of an electronic crossover also diminishes this effect."




jk

stuartk said:

Then it's no longer a series x-over and all of the component values are going to be very wrong.

You would lose all of the benefits of having a series x-over, plus basically have to re-engineer the speaker design and build a new x-over.

All for bi-wiring that doesn't do a bit of good in the first place.

Click to expand...

Stuark . . you just repeated everything I said--but in a negative light. I said "essentially a parallel design" and "require new component values."

As for Bi-wiring being a waste . . reflect on your post.:yes:

Raphael said:

Yes it does "spoil the point" but isn't that what you are trying to do? Get away from the dependence the drivers have on each other in formulating the spectrum of sounds and allow them to independently focus on the ranges they were designed for given a well engineered crossover point (no freq gap or overlapping and filtered to within the easy driving range of each driver).

From previous reading--which I am only now remembering--Nakdoc hits the nail on the head regarding the benefits of Bi-wiring--namely higher DF resulting in faster/cleaner sound (however significant or slight).

Click to expand...

You speak as if drivers are ideal things that don't alter when a signal is fed into them. If you're driving them with a bit of power, then the voice-coils heat up a bit, altering the inductance and changing the crossover. If you've got a parallel crossover, what that means is that now you have a bump/dip in frequency response where before it was flat, or flattening the curve where before there was a bump/dip. If you've got a series crossover, then the crossover is affected as a whole and the response ends up the same.

But, if you don't drive your speakers hard, then I can understand that this would not occur.

I have an amp for each speaker and active crossovers, so it's all academic to me anyway. DIY rules !

Thanks for the reply.

I understand what bi-wiring is and I understand what back EMF is. What I don't understand is how the length of wire between the woofer and tweeter can modify the back EMF. You talk about the resistance of the wire but then say that a simple resistor can't do the same thing. My question is: What other property of the wire besides resistance is helping to reduce back emf?

Once again, thanks for helping me think this through and providing a sounding board for my poorly formed questions.

I can't say for sure....but the topology of the wiring changes ( where things are placed in the circuit ) of being biwired, plus the self inductance/capacitance of the wire may be a part of the re-active load. Wire has an inductive and capacitive load that it presents to the signal.

Where things are place in a circuit can affect the signal. Remember I mentioned, in biwired, I moved the crossover to the amp side of the woofer wire and that changed the "voice" of the speakers. Some how that seemed to lower the bass....why, I am not sure. I put it back to being at the speaker and never thought about it again....

jk

A series crossover can allow back EMF from the woofer into the tweeter or mid and apparently cause IM issues etc. Bi Wiring cannot help a series crossover- it could only make its function worse. I just ran a few impedance plots on normal versus biwired speakers and guess what- exactly the same.