Bi-amping and Bi-wiring

[kkantor]

1- What the heck are you talking about? There is nothing in Dick's tutorial about "back EMF."

2- I have known Dick for about 20 years. If you contact him, and show him this thread, he can help you understand the conceptual error you are making. If you actually understood the tutorial you posted a link to, you would know that it supports exactly what I said. I quote: "...the electrical model the above suffices quite well for predicting reality."

3- For the past couple of months, my assistant and I have been measuring a variety of AR tweeters and crossovers. The situation is quite interesting, but I will postpone discussing the results until I can publish a complete report, with all the data. That should happen in about a week. For now, I will just say that:

"Removing the pot entirely will yield a large, ugly resonant peak at the low end of the tweeter's response. Even running the pot wide will show this."

... is a correct statement, but only for a certain generation of tweeters. Some show it, some do not. For now, I realize, this is an unsubstantiated assertion. (What? An unstubstantiated assertion? Here???) However, you can argue with the data when it is in front of you. I'm not interesting in proving you wrong, or me right. I am only interested in what is really going on, instead of either of our opinions.

4- Sir, respectfully, this is my field, not my hobby. I know what I am talking about. I don't know everything, and I often make mistakes. But, you could learn a lot if you would only try, instead of arguing. I suggest that that would be a better use of both of our time.

5- Here's a link that pertains to your question about what kind of model I tend to use. This is a very simple one:

http://www.stereophile.com/reference/60/index.html.

-k

www.kenkantor.com

http://kkantor.spaces.live.com

[Carlspeak]

Thanks Ken for your comments. Your position on bi-wiring has become crystal clear over the past couple of days. I was neither advocating nor pooh-poohing the benefits of bi-wiring/bi-amping but just making folks aware (via some references) of perhaps why these mods MAY be beneficial, so folks who were interested in the subject might want to investigate and find out for themselves. After all, isn’t that what the hobby is all about (besides the music)?

I find the whole discussion synonymous to the classic $1000 esoteric speaker cable vs $5 zip cord debate that’s raged for years.

A cursory search of speakers currently being manufactured indicates that at least some offer binding post configurations which allow the option of bi-wire or bi-amping. Vandersteen, Klipsch and Snell are but only a few I found. These are not fly-by-night loudspeaker companies.

If all this is snake oil, as you suggest, your position seems a bit odd coming from someone who’s been a part of this industry for so many years.

I’d be curious to know roughly what % of quality speakers marketed today have this capability. Could you help with that?

It's all about the music

Carl

Carl's Custom Loudspeakers

[soundminded]

A few more thoughts on back EMF. A google search of back emf > loudspeakers will turn up about 152,000 hits so there must be some people who consider it important. There are even some patents to deal with it but I don't know how useful they are. I haven't checked all 152,000 sites yet, so I'll report back when I do. As a rotary motor and generator are actually the same machine only optimized in their configuration for their intended purpose, a linear motor or loudspeaker driver and a dynamic microphone are also the same device. In fact in the bad old days of the 1950s before you could buy a microphone made in China for 3 cents, old "wireless intercoms" which transmitted their signals over house and building wiring used a 4" speaker driver wound with a 72 ohm voice coil to make it more suitable for use as a microphone as well. Della, Perry. Would you come into my office, I want you to take dictation, I need to send a letter to lieutenant Trag.

Finally an aspect of speaker design for which thermodynamic analysis is useful to tell us something and nobody uses it. Pity. While electrical energy is stored in the inductors of a speaker system as magenetic fields and in the capacitors as electrical fields, it is also stored as potential energy in the compression and expansion of whatever spring it is working against whether it is air in the case of an AS speaker or the mechanical suspension of a non AS driver. This potential energy is at a maximum at the extreme of the cone's excursion on each cycle and zero as it crosses its zero point. By contrast, the kinetic energy of the moving mass (1/2*m*v*v) where v is the velocity of the cone is greatest at the zero crossing point and least at the extremes of excursion where the cone stops and reverses direction. This energy has to be dissipated. It is either dissipated mechanically by performing work against a frictional loss such as pulling and pushing air through the spaces between the stuffing or it must be dissipated electrically. If the amplifier is sold state, its output impedence is usually negligable, far lower than any of the other circuit elements in the output stage. If the wires are of sufficient gage, then their electrical resistance is negligable too. This is how we want this back emf to be dissipated. But, if there is little or no mechanical work because the speaker is undamped at the resonant frequency of an air column it is working into and the amplifier output impedence is high because it is a vacuum tube design with an impedence matching transformer trying to couple to a tube with an output impedence of 5 to 15 thousand ohms and there is no negative feedback and/or the speaker wires are so long and thin they offer significant resistance, then that electrical energy will not only dissipate slowly as the speaker oscillates in a spurious resonance but it just might be dissipated into the voice coils of a midrange or tweeter if say its crossover frequency is sufficiently low and/or it only has a 6 db/octave slope. I don't see how any mathematical model of a speaker can give accurate results unless it takes these factors into account but I haven't looked at any in a long time (since our last discussion about this here probably over a year ago.) Just a thought, but not really of great interest to me. I use exclusively solid state amplifiers and mostly AS speakers and so don't consider the problem of practical signifigance to me.

[onplane]

>1- What the heck are you talking about? There is nothing in

>Dick's tutorial about "back EMF."

Ken, I'm NOT an advocate of back EMF other than its existence begins to explain the complexity of characterizing a speaker with a scattering of passive components.

>2- I have known Dick for about 20 years. If you contact him,

>and show him this thread, he can help you understand the

>conceptual error you are making. If you actually understood

>the tutorial you posted a link to, you would know that it

>supports exactly what I said. I quote: "...the

>electrical model the above suffices quite well for predicting

>reality."

When this issue first came up, I ran to Dick and he explained the complexity of characterising a speaker with linear components (i.e. a single coil and series resistance). Dick's exact words were, "In effect, it's not a single resistor in series with a single inductor, but it's an infinte number of series circuits, each consisting of an infinitesimal resistor in series with an infinitesimal inductor."

>3- For the past couple of months, my assistant and I have been

>measuring a variety of AR tweeters and crossovers. The

>situation is quite interesting, but I will postpone discussing

>the results until I can publish a complete report, with all

>the data. That should happen in about a week. For now, I

>will just say that:

>

>"Removing the pot entirely will yield a large, ugly

>resonant peak at the low end of the tweeter's response. Even

>running the pot wide will show this."

>

>... is a correct statement, but only for a certain generation

>of tweeters. Some show it, some do not. For now, I realize,

>this is an unsubstantiated assertion. (What? An

>unstubstantiated assertion? Here???) However, you can argue

>with the data when it is in front of you. I'm not interesting

>in proving you wrong, or me right. I am only interested in

>what is really going on, instead of either of our opinions.

Ken, this is the area that interests me most and look forward to seeing your data.

>4- Sir, respectfully, this is my field, not my hobby. I know

>what I am talking about. I don't know everything, and I often

>make mistakes. But, you could learn a lot if you would only

>try, instead of arguing. I suggest that that would be a

>better use of both of our time.

Well, Ken, I believe discussion is healthy in any field and never have I made any disparaging remarks about you or your work. Indeed, I respect what you have accomplished. Further, I never accused you of making any mistakes.

I do, however, respectfully question models that fail to predict results that are measurable. (I know first hand the frustration of bread board circuits that fail miserably in performing as models predict.)

Finally, I did try! That is, I measured the tweeter voltage over its useful frequency range and that’s how I know something (and I don’t know what) is wrong with the current model.

>5- Here's a link that pertains to your question about what

>kind of model I tend to use. This is a very simple one:

>

>http://www.stereophile.com/reference/60/index.html.

>

Ken, I've seen this two way model before. (I think I found it on a Google search.)

That model does a good job characterizing the mechanical/electrical peaks and the resulting load on an amplifier. What interests me more is what’s going on between the peaks and can we predict accurately the interaction of the drivers with other passive components.

Once again, Ken, I look forward to seeing the results of your measurements.

Regards,

Jerry

[Guest robotczar]

This is not the forum to indulge in debates about audio fads and I won't say any more. But, I found over two million hits in goggling "easter bunny" and over 900,000 hits for "flat earth". I am not trying to be snide, but looking at the number of google hits is not exactly evidence. Note also that I did not claim back EMF does not exist but rather that it would be the same for a speaker that is biwired and one that is wired conventionally. I indicated that all the drivers are shorted to each other AT the amp terminal (i.e., before the amp circuit components from the dirvers' perspective), so there really is no difference in a biwired circuit and a conventionally wired on (except more wire).

[soundminded]

>This is not the forum to indulge in debates about audio fads

>and I won't say any more. But, I found over two million hits

>in goggling "easter bunny" and over 900,000 hits for

>"flat earth". I am not trying to be snide, but

>looking at the number of google hits is not exactly evidence.

>Note also that I did not claim back EMF does not exist but

>rather that it would be the same for a speaker that is biwired

>and one that is wired conventionally. I indicated that all

>the drivers are shorted to each other AT the amp terminal

>(i.e., before the amp circuit components from the dirvers'

>perspective), so there really is no difference in a biwired

>circuit and a conventionally wired on (except more wire).

Perhaps you didn't understand my first posting above. I am not debating a fad, I am giving you a circuit configuration to analyze if you care to which you may have overlooked. What you said would be true IF the speaker wires had no impedence. This is not the case. Biwiring clearly places the midrange and tweeter on a separate circuit node from the woofer with the amplifier between them separated by the impedence of the wires to each. The impedence of the output circuit of the amplifier can be at least a full order of

magnitude lower than the speaker wire impedence especially if it is a solid state amplifier. I also said that these are theoretical consdierations for which I have no parctical experience, no measurements, and do not particularly advocate. I don't know why you are making an issue of it with me, I leave it to others to fill in the blanks with hard data if they have it but I will stick with what I said. If you have an objection to my circuit analysis, please state it and we can discuss that. By the way, I hope the Easter bunny brought you plenty of chocolate to eat this year while you listen to your audio system. :-)

[Pete B]

>>1- What the heck are you talking about? There is nothing

>in

>>Dick's tutorial about "back EMF."

>

>

>Ken, I'm NOT an advocate of back EMF other than its existence

>begins to explain the complexity of characterizing a speaker

>with a scattering of passive components.

>

>

>

>>2- I have known Dick for about 20 years. If you contact

>him,

>>and show him this thread, he can help you understand the

>>conceptual error you are making. If you actually

>understood

>>the tutorial you posted a link to, you would know that it

>>supports exactly what I said. I quote: "...the

>>electrical model the above suffices quite well for

>predicting

>>reality."

>

>When this issue first came up, I ran to Dick and he explained

>the complexity of characterising a speaker with linear

>components (i.e. a single coil and series resistance). Dick's

>exact words were, "In effect, it's not a single resistor

>in series with a single inductor, but it's an infinte number

>of series circuits, each consisting of an infinitesimal

>resistor in series with an infinitesimal inductor."

>

>

>>3- For the past couple of months, my assistant and I have

>been

>>measuring a variety of AR tweeters and crossovers. The

>>situation is quite interesting, but I will postpone

>discussing

>>the results until I can publish a complete report, with

>all

>>the data. That should happen in about a week. For now,

>I

>>will just say that:

>>

>>"Removing the pot entirely will yield a large,

>ugly

>>resonant peak at the low end of the tweeter's response.

>Even

>>running the pot wide will show this."

>>

>>... is a correct statement, but only for a certain

>generation

>>of tweeters. Some show it, some do not. For now, I

>realize,

>>this is an unsubstantiated assertion. (What? An

>>unstubstantiated assertion? Here???) However, you can

>argue

>>with the data when it is in front of you. I'm not

>interesting

>>in proving you wrong, or me right. I am only interested

>in

>>what is really going on, instead of either of our

>opinions.

>

>

>Ken, this is the area that interests me most and look forward

>to seeing your data.

>

>

>>4- Sir, respectfully, this is my field, not my hobby. I

>know

>>what I am talking about. I don't know everything, and I

>often

>>make mistakes. But, you could learn a lot if you would

>only

>>try, instead of arguing. I suggest that that would be a

>>better use of both of our time.

>

>

>Well, Ken, I believe discussion is healthy in any field and

>never have I made any disparaging remarks about you or your

>work. Indeed, I respect what you have accomplished. Further,

>I never accused you of making any mistakes.

>

>I do, however, respectfully question models that fail to

>predict results that are measurable. (I know first hand the

>frustration of bread board circuits that fail miserably in

>performing as models predict.)

>

>Finally, I did try! That is, I measured the tweeter voltage

>over its useful frequency range and that’s how I know

>something (and I don’t know what) is wrong with the current

>model.

>

>

>

>

>

>>5- Here's a link that pertains to your question about

>what

>>kind of model I tend to use. This is a very simple one:

>>

>>http://www.stereophile.com/reference/60/index.html.

>>

>

>

>Ken, I've seen this two way model before. (I think I found it

>on a Google search.)

>

>That model does a good job characterizing the

>mechanical/electrical peaks and the resulting load on an

>amplifier. What interests me more is what’s going on between

>the peaks and can we predict accurately the interaction of the

>drivers with other passive components.

>

>

>Once again, Ken, I look forward to seeing the results of your

>measurements.

>

>Regards,

>Jerry

Jerry, Ken is right and you are reading into what Dick Pierce explained to you.

Back EMF has nothing to do with explaining the difference between a lumped

model and a distributed model which is what Pierce was explaining to you. I wrote

this about back EMF some time ago:

http://www.diyaudio.com/forums/showthread....4559#post554559

Here is another reference:

http://www.diyaudio.com/forums/showthread....3489#post553489

Many people, even some "professionals" don't understand back EMF.

Let me point out that, yes speakers are non-linear and yes there

are times when you might want to use a distributed model for a

speaker (very rarely), however, back EMF has little to do with explaining either of these.

Yes, many are confused about back EMF.

Pete B.

[Pete B]

>Thanks Ken for your comments. Your position on bi-wiring has

>become crystal clear over the past couple of days. I was

>neither advocating nor pooh-poohing the benefits of

>bi-wiring/bi-amping but just making folks aware (via some

>references) of perhaps why these mods MAY be beneficial, so

>folks who were interested in the subject might want to

>investigate and find out for themselves. After all, isn’t that

>what the hobby is all about (besides the music)?

>

>I find the whole discussion synonymous to the classic $1000

>esoteric speaker cable vs $5 zip cord debate that’s raged for

>years.

>

>A cursory search of speakers currently being manufactured

>indicates that at least some offer binding post configurations

>which allow the option of bi-wire or bi-amping. Vandersteen,

>Klipsch and Snell are but only a few I found. These are not

>fly-by-night loudspeaker companies.

>

>If all this is snake oil, as you suggest, your position seems

>a bit odd coming from someone who’s been a part of this

>industry for so many years.

>

>I’d be curious to know roughly what % of quality speakers

>marketed today have this capability. Could you help with

>that?

>

>It's all about the music

>

>Carl

>Carl's Custom Loudspeakers

They offer bi-wiring posts because it helps to sell speakers,

since many customers want them. It is a marketing decision.

Vandersteen speakers are fairly good, I've owned them and reverse

engineered them, however there are many half truths in the marketing

hype from Vandersteen IMO.

[kkantor]

Carl-

1- I never said bi-wiring was "snake oil." I said that it has been widely investigated, and generally found to make very little difference in the vast majority of cases. I did this myself, too, as carefully as I could, with no agenda at the time about what the results would/should be.

2- "Bi-amping" and "bi-wiring" are not at all the same thing. I think the former is much more justifiable, if done with care.

3- Originally, many speaker companies put dual terminals on their speakers to allow bi-amping, including AR. In the 70's, active crossovers like the Pioneer were quite popular with audiophiles. Bi-wiring gradually took advantage of this situation. It started off as simply a way to conveniently attach a heavier wire gauge to the small binding posts of the day. It wasn't until somewhat later that a few of the highend mags started hearing other benefits to it.

4- It costs almost nothing for a manufacturer to add a second set of inputs. Well under a dollar. The only products that don't use it are the very cheapest ones. It's a sales advantage and, once in a great while, a technical advantage, that comes almost for free.

www.aural.org/audio/misc_info/The_Audio_Critic_interview_01a.pdf

www.aural.org/audio/misc_info/The_Audio_Critic_NHT33_review.pdf

-k

[kkantor]

Thanks. I thought your linked post was very articulate.

Stored energy is fully acounted for in the complex input impedance, including mechanically-stored energy. I don't know why this is hard for people to grasp?

http://www.aural.org/audio/articles/speakers_by_design/

-k

www.kenkantor.com

http://kkantor.spaces.live.com

[onplane]

>Back EMF has nothing to do with explaining the difference

>between a lumped

>model and a distributed model which is what Pierce was

>explaining to you. I wrote

>this about back EMF some time ago:

>

>

>Many people, even some "professionals" don't

>understand back EMF.

>

>Let me point out that, yes speakers are non-linear and yes

>there

>are times when you might want to use a distributed model for

>a

>speaker (very rarely), however, back EMF has little to do with

>explaining either of these.

>Yes, many are confused about back EMF.

>

>Pete B.

Hi, Pete!!

You are correct! When Dick explained the problem of frequency related inductance AND resistance, he was referring specifically to pole piece eddy currents. I only used back EMF as one example of how a speaker voice coil differs from a standard stationary coil. Like you mentioned, Pete, the difference is the storing of mechanical as well as … electrical energy.

Pete, I tend to agree with your arguments that distributed models for speakers is probably unnecessary and “overkill” when attempting to calculate amplifier loads.

Nevertheless, when attempting to predict interactions between raw speakers and other “more linear” passive components, is where the simple model starts to give us trouble. In short, SPICE will find resonance, but the real result can never be similar to a simple tuned circuit.

As you well know, Pete, resonance with linear devices occurs when the capacitive and inductive reactance cancel each other. Then above the resonant frequency the inductive reactance dominates. Well, if the inductance of a tweeter DECREASES with frequency, this “domination” may never kick in or it may be much, much more gradual than SPICE would predict. Net actual result would be a plateauing or flattening of the response curve.

Pete, does this make sense??

Regards,

Jerry

[kkantor]

Onplane,

Since you are seriously interested in loudspeaker modeling and have a technical background, I've posted a few references on the topic in the Mods/Tweaks forum. I couldn't find open-source links to a number of publications but these, at least, highlight some recent concepts. There is much more sophisticated stuff around than my "Stereophile" box!

http://www.classicspeakerpages.net/dc/dcbo...109&mesg_id=109

-k

www.kenkantor.com

http://kkantor.spaces.live.com

[soundminded]

>Stored energy is fully acounted for in the complex input

>impedance, including mechanically-stored energy. I don't know

>why this is hard for people to grasp?

>www.kenkantor.com

Maybe it's because most models are so oversimplified that they are a joke and any accurate model is so complicated as to be useless to all but the most determined scientists and engineers making it more of an esoteric laboratory curiousity than a practical tool.

Any accurate model of even the simplest loudspeaker, a theoretical acoustic suspension type would have to take into account the mechanism by which mechanical energy is stored and then converted back to electrical energy. It's easy to just write down a resonant circuit in shunt with the rest of the model but since most electrical engineers don't even understand Newton's second law of motion, how will they determine the parameters for a given case? How will they determine how much energy is dissipated as mechanically generated heat and how much as electrical energy working into a resistive load? They also have to know the mechanical/electrical conversion efficiency which is based on the number of turns of wire in the voice coil and the number of magnetic lines of flux cut made all the more complicated if the magnetic field is not completely uniform over the travel of the VC. In a real AS speaker, you have to add the non linear properties of the mechanical energy stored in the driver's suspension. When you get to a ported system, you go into another universe of complexity because the mechanical restoring force or springiness of the air resistance is not only a function of frequency but expresses itself as an exponentially decreasing period function of frequency due to resonance modes at each integral multiple of the fundimental mechanical resonant frequency and an anti resonance mode at every ingegral multiple half octave in between. Once the model is accurate, you have to have a way to verify it by experimental data for various combinations of frequency and amplitude for a particular design and a way to predict and measure them for any proposed design. This seems like an awful lot of trouble to go through considering that the entire phenomenon is easily damped out by nearly any solid state amplifier on the market.

Have you got an a single complete model which takes all these factors into account and has been verified as an accurate predictor of back emf?

[Pete B]

>>Stored energy is fully acounted for in the complex input

>>impedance, including mechanically-stored energy. I don't

>know

>>why this is hard for people to grasp?

>

>>www.kenkantor.com

>

>Maybe it's because most models are so oversimplified that they

>are a joke and any accurate model is so complicated as to be

>useless to all but the most determined scientists and

>engineers making it more of an esoteric laboratory curiousity

>than a practical tool.

>

>Any accurate model of even the simplest loudspeaker, a

>theoretical acoustic suspension type would have to take into

>account the mechanism by which mechanical energy is stored and

>then converted back to electrical energy. It's easy to just

>write down a resonant circuit in shunt with the rest of the

>model but since most electrical engineers don't even

>understand Newton's second law of motion, how will they

>determine the parameters for a given case? How will they

>determine how much energy is dissipated as mechanically

>generated heat and how much as electrical energy working into

>a resistive load? They also have to know the

>mechanical/electrical conversion efficiency which is based on

>the number of turns of wire in the voice coil and the number

>of magnetic lines of flux cut made all the more complicated if

>the magnetic field is not completely uniform over the travel

>of the VC. In a real AS speaker, you have to add the non

>linear properties of the mechanical energy stored in the

>driver's suspension. When you get to a ported system, you go

>into another universe of complexity because the mechanical

>restoring force or springiness of the air resistance is not

>only a function of frequency but expresses itself as an

>exponentially decreasing period function of frequency due to

>resonance modes at each integral multiple of the fundimental

>mechanical resonant frequency and an anti resonance mode at

>every ingegral multiple half octave in between. Once the

>model is accurate, you have to have a way to verify it by

>experimental data for various combinations of frequency and

>amplitude for a particular design and a way to predict and

>measure them for any proposed design. This seems like an

>awful lot of trouble to go through considering that the entire

>phenomenon is easily damped out by nearly any solid state

>amplifier on the market.

>

>Have you got an a single complete model which takes all these

>factors into account and has been verified as an accurate

>predictor of back emf?

You really are a troll, with your deliberately arrogant and offensive view of what engineers do and do not know or understand, fishing for confrontation with competent engineers. If you were competent in the area of speaker design, you'd be able to answer your own challenge, instead you demonstrate your own ignorance. I'll let you dig yourself deeper, not going to comment further and educate you.

[Pete B]

>

>

>>Back EMF has nothing to do with explaining the difference

>>between a lumped

>>model and a distributed model which is what Pierce was

>>explaining to you. I wrote

>>this about back EMF some time ago:

>>

>>

>>Many people, even some "professionals" don't

>>understand back EMF.

>>

>>Let me point out that, yes speakers are non-linear and

>yes

>>there

>>are times when you might want to use a distributed model

>for

>>a

>>speaker (very rarely), however, back EMF has little to do

>with

>>explaining either of these.

>>Yes, many are confused about back EMF.

>>

>>Pete B.

>

>

>Hi, Pete!!

>

>You are correct! When Dick explained the problem of

>frequency related inductance AND resistance, he was referring

>specifically to pole piece eddy currents. I only used back

>EMF as one example of how a speaker voice coil differs from a

>standard stationary coil. Like you mentioned, Pete, the

>difference is the storing of mechanical as well as …

>electrical energy.

>

>Pete, I tend to agree with your arguments that distributed

>models for speakers is probably unnecessary and “overkill”

>when attempting to calculate amplifier loads.

>

>Nevertheless, when attempting to predict interactions between

>raw speakers and other “more linear” passive components, is

>where the simple model starts to give us trouble. In short,

>SPICE will find resonance, but the real result can never be

>similar to a simple tuned circuit.

>

>As you well know, Pete, resonance with linear devices occurs

>when the capacitive and inductive reactance cancel each other.

> Then above the resonant frequency the inductive reactance

>dominates. Well, if the inductance of a tweeter DECREASES

>with frequency, this “domination” may never kick in or it may

>be much, much more gradual than SPICE would predict. Net

>actual result would be a plateauing or flattening of the

>response curve.

>

>Pete, does this make sense??

>

>Regards,

>Jerry

>

We are not even close to being on the same page, sorry. You are mixing small signal, and large signal analysis and in the process not treating either correctly. Distributed systems, I know a lot about distributed systems and I did not mean in any way what you describe above. It would take a lot to explain and I just don't have the time. SPICE, who doing serious speaker design uses SPICE for speaker simulations? There are much better simulators for speaker design that address exactly what you seem to think is so difficult to model. It's in the literature, probably as far back as 20 years ago. The only place I'd use a distributed model is in the analysis of transmission line loaded speakers, and in box mode resonance analysis. Sorry, this is not a productive discussion, I'm going to bow out.

[Pete B]

Just let me add a comment that there is and advantage to bi-wiring when a speaker is poorly designed. Resistance in the bass wiring can be used to tune the Qtc through Qes of the fundamental bass resonance, it can also tame peaking in the low pass crossover section. Wiring resistance in the mid/tweeter leads can be used to tame peaking in these crossover sections.

Better to design the system to have the correct Q with low resistance wiring.

[soundminded]

It is clear you have nothing of technical merit to add to this discussion which only leaves one question; since you all you can do is insult me, why do you bother to reply to my postings at all? The answer is obvious. Funny how this seems to happen every time I bring up the subject of Newton's second law of motion. What happened, did you have a life altering experience with your physics teacher in high school? Did he embarrass you in front of the whole class or something?

[Pete B]

>It is clear you have nothing of technical merit to add to

>this discussion which only leaves one question; since you all

>you can do is insult me, why do you bother to reply to my

>postings at all? The answer is obvious. Funny how this seems

>to happen every time I bring up the subject of Newton's second

>law of motion. What happened, did you have a life altering

>experience with your physics teacher in high school? Did he

>embarrass you in front of the whole class or something?

Once again you leap to arrogant assumptions about others, when

in fact you do not know the facts. I notice that you do this about

broad categories of people, consumers, engineers, etc., and you are

most often wrong.

Physics was in fact one of my best subjects, I studied it to the

level of relativistic physics at the college level. Come on,

Newtonian physics is taught at the junior high level where I come

from.

I actually find it amusing that you toss out sweeping claims about

Newton's second law and think people are going to sit back and

believe you. Let me give you a clue, you need second order

differential equations to model the bass roll off of an acoustic

suspension speaker, and 4 th order equations for a vented system.

Partial differential equations are required for the analysis of

transmission line speakers. Your claims about Newton's second law

are absurd. You are a troll!

Now I'm really done here, sorry I stopped in to comment.

[Carlspeak]

>Thanks. I thought your linked post was very articulate.

>

>Stored energy is fully acounted for in the complex input

>impedance, including mechanically-stored energy. I don't know

>why this is hard for people to grasp?

>

>http://www.aural.org/audio/articles/speakers_by_design/

>

>-k

>

>www.kenkantor.com

>

>http://kkantor.spaces.live.com

Here's another reference to a less technical article wherein the author takes a slightly different view of the model by looking at the circuit from the driver's perspective. Amplifier damping factor is also raised as a variable in the back EMF 'equation'. I don't think DF has been mentioned in the threads fo far. Also of interest is the reference to driver motion following cessation of the electrical signal as the source of back EMF. I would think a more floppy suspension as that of AS speakers would be more prone to this even though it is naturally damped by the sealed box. He concludes that separate, active filter networks (an expensive option) are the best way to address this.

http://sound.westhost.com/biamp-vs-passive.htm

Attached are 2 1980's patents intended to addres back EMF. One of them

raises the spector of audible passive crossover 'ringing' effects his patent is designed to address.

It's all about the music

Carl

Carl's Custom Loudspeakers

[soundminded]

Wow, that was enlightening Pete. Thanks for sharing all that knowledge about reverse EMF with us. I'm sure now everyone understands it.

>Come on,

>Newtonian physics is taught at the junior high level where I

>come

>from.

Where, in the schoolyard at recess? In high school physics, most students don't get beyond F=ma. Only rarely when the best highschool math/science honors students take advanced courses do they get to see Newton's laws as differential equations as they are taught to college freshmen majoring in science and engineering.

>

>Now I'm really done here

>

I'm always grateful for the little things in life.

[Pete B]

>>Thanks. I thought your linked post was very articulate.

>

>>

>>Stored energy is fully acounted for in the complex input

>>impedance, including mechanically-stored energy. I don't

>know

>>why this is hard for people to grasp?

>>

>>http://www.aural.org/audio/articles/speakers_by_design/

>>

>>-k

>>

>>www.kenkantor.com

>>

>>http://kkantor.spaces.live.com

>

>Here's another reference to a less technical article wherein

>the author takes a slightly different view of the model by

>looking at the circuit from the driver's perspective.

>Amplifier damping factor is also raised as a variable in the

>back EMF 'equation'. I don't think DF has been mentioned in

>the threads fo far. Also of interest is the reference to

>driver motion following cessation of the electrical signal as

>the source of back EMF. I would think a more floppy suspension

>as that of AS speakers would be more prone to this even though

>it is naturally damped by the sealed box. He concludes that

>separate, active filter networks (an expensive option) are the

>best way to address this.

>

>http://sound.westhost.com/biamp-vs-passive.htm

>

>Attached are 2 1980's patents intended to addres back EMF. One

>of them

>raises the spector of audible passive crossover 'ringing'

>effects his patent is designed to address.

>

>It's all about the music

>

>Carl

>Carl's Custom Loudspeakers

Hi Carl,

This is a fairly tortured analysis, one thing he completely ignores

or misses is that the impedance seen looking back into the crossover

that he describes is purely reactive, he should have the imaginary j

operator in front of his impedance value:

http://sound.westhost.com/biamp-vs-passive.htm

Reactive impedances do not dissipate energy and this is a flaw in his

analysis. He is making much out of nothing there, speaker designers

take into account the impedance of the crossover network. The

crossover impedance is required to obtain the required frequency

response shaping, it is really very simple. And if you remove that

impedance you will have a different speaker than that intended by the

original designers.

The basic problem in understanding here, as I see it, is that people

involved in this discussion need to study and understand electro-

mechano-acoustical models and analogies as described by Beranek,

Thiele/Small, et al.

What they would get from this is an understanding that the

differential equations for mass/spring/damper systems in

mechanics/acoustics are the same as those for LCR systems in

electronics. And a speaker motor has a fascinating characteristic

in that it actually allows us to "see" the mechanical system on the

"secondary" as its electrical analogy from the primary side

including all of the energy storage and back EMF characteristics.

Now the term electro motive force suggests motion in a mechanical

system, and there is an analogy that stores energy in an electrical

system that being reactive components. They are analogous, and the

differential equations are the same. Back EMF is not something

"extra" that is left out in the model of a speaker. People who truly

understand loudspeaker modelling and electro-mechano-acoustical

analogies are completely aware of this fact.

I have left out a lot of detail, and as Ken has suggested people have

to study and understand the basics in order to get past the "back EMF"

confusion.

I did not read the patents, and just because the term back EMF is used

does not mean there is an error. People who use it and claim that

there is something more to it than the motional impedances seen reflected

into the primary are the ones in error. A negative output impedance

amplifier for example can be used to provide more damping and actually

apply a reverse voltage to stop a driver faster. This has been in

the literature for many years and all it does is provide a lower Qes

and therefore also Qtc.

[Pete B]

It is difficult to find AES articles online and this would not be my first choice, but this is just one of probably thousands that detail the complexities of loudspeaker modelling:

http://users.ece.gatech.edu/~mleach/papers/zobel/zobel.pdf

There are many others that are must reads, for anyone serious about loudspeaker modelling.

It is interesting to note that Leach was the first to introduce the correct circuit element, a gyrator, to represent the electro-mechano speaker motor reciprocity behavior. This was published in another AES article. Further, his AES paper on the need to cascade the electrical crossover response with the driver electro-acoustic response was another landmark paper.

Here's a good one, required reading for those who wish to understand enclosure filling. Many on the Bass list years ago should have read this, unfortunately it contradicted the thoughts of many of the elders there who thought they had the answers:

http://users.ece.gatech.edu/~mleach/papers/Filling.pdf

Ah, here we go, while this paper is about horns, it provides an accurate model that demonstrates the fundamentals of loudspeaker modelling, including the effects of back EMF. Refer to Figure 2:

http://users.ece.gatech.edu/~mleach/papers...r/HornPaper.pdf

I studied this paper by Leach in great detail around the time it came out in 1979 in order to better understand the methods. I used this paper here as an example because Leach has his articles online, there were the papers by Thiele and Small that cover direct radiator loudspeakers using similar modelling methods.