Dipole 3 Way / 4 Way - Why? Why not?

[Teh]

I’m really quite cautious about posting this since opinions vary to such a varied degree here on DIYaudio. Why? Well, we’ve got single driver proponents as well as many, many passive crossover proponents. As a result you end up with such varied opinion that a reasonable conclusion is difficult to reach, given the passion of such individual opinion. So, I guess I’ve got to establish my own parameters, and therefore, my own entrenched positions.

First, the objective: to build an open baffle system. What does that mean? Well to me it means a system that can accomplish all from stereo to 5 speaker surround system home entertainment. Unlike many, I actually enjoy 5 speaker surround music reproduction. Apparently, this is heresy. Still, that is exactly what live music is. With all of it’s flaws and all of it’s resonant inflammations.

That said I am interested in a much wider soundstage. As a result, I’ve decided to endeavor an open baffle system. I’m interested in one thing in particular; the elimination (or expansion) of the “sweet spot”. Why? Because who “always” listens to music, movies and the like in the most ideal position?

I am repulsed by all I’ve read so far; that in order to “hear” a system’s best you must, MUST, be in the ideal position within the listening room. So, if you have 5 others (plus yourself) in your listening environment at least five cannot possibly be in the “sweet spot” so to speak. As a result I see all discussions as a dismal solution to the true problem; that music, movies, and any other entertainment can be equally enjoyed by more than one. One, is such a disgusting solution. As a result, I view the OB alternative the only valid solution to adequate to widening the “sweet spot”.

With that I’ll propose the question with adequate parameters; I have a room that is 20 by 26 feet (I know, big!), second I already have the drivers purchased that I plan to use (they are already in use in an enclosed system). They are:

Tweeters: Audiom TD5 for all five units (db – 95)
Upper mids: Audion 6WM (db – 94.5)
Lower mids: Cabasse 2 - 21NDC (fs – 32, db – 91)
Low frequency units: TAD TL-1602 (two beneath front right and left front panels)

So, what am I trying to accomplish? Let’s just work with this: http://www.perfect8.com/ Pay attention to “The Point” that looks to be the closest to my objectives.

The question becomes; Why look to a sealed box solution (for the lowest octaves) verses the open baffle solution so popular of late? Well, regardless of who the system designer is they all recommend a seal solution for the lowest two, 2 ½, or three octaves. Then, as a result, I question the whole dipole process, right? That is for the lowest octaves, anyway. So, I ask; at what point does the dipole issue, of low frequency extension, become a non issue?

The major benefit of open baffle design is apparently in the region of 100/120 (anything below is monopole and dipole’s are incapable of room coupling required at such frequencies) to 2000hz. Primarily in the 100 – 1000hz range. Different techniques are now being utilized to capitalize on the “dipole sound” up into non-dipole frequencies by incorporating rear firing tweeters or other means.

I see a dichotomy, if you will. I find most fighting the need for monopole woofers to satisfy the region of 20hz to 60hz/80hz/100hz. As a result, I see many fighting to stay as low as possible with inordinate sized woofers (or arrays, should I say) while few are satisfied with output results. I guess this comes down to one question;

Why have not the many just relented and, instead of trying to implement an array of 12 12” low frequency woofers (in any room of decent size) to do the job, that is flat performance to 20hz, just implemented a serious, efficient, fast, accurate, low distortion, low Fs, low frequency woofer combination (stereo subs) to accomplish the obvious?

Am I missing something here?

Panel deflection and baffle step issues appear to be of no problem at all at these low frequency levels, right? So, why do I not see a better acceptance of what is possible and what is clearly not?

Much of the panel stiffness needed can be obtained by the use of Carbon fibers (in combination of Epoxy resins). Do we even see a little of its usage? No we do not!

I plan to build active speakers, made of Carbon Fiber reinforced MDF (inside with shredded carbon fiber, outside with woven carbon), sculpted as “The Point” (referenced above), to accomplish the same type of result. Will it be as acoustically dead? I’m not sure, but I know it will be close at worst.

The crossovers will be based upon the Linkwitz active crossover and equalization/correction networks – one per speaker. All speakers will be powered by onboard Tri-Path amplifiers (60 watts per driver) and resolved with Jenson Transformers for a balanced signal delivery and integrity. All delivered from a Lexicon MC-1 carried on Mogami 2534 cable to its destination. The low pass will deliver the Linkwitz Transform circuit to bring low frequency (Front Left & Right) performance to its most accurate.

The dipole panel shall be decoupled from the bass unit by urethane gaskets and an adjustable angle to suit the listener.

So, my question is; Take fault! Tell me where this can go wrong? What is wrong with the drivers, amps, possibilities? Why cannot a full spectrum, outstanding, result not be obtained?

One difference of note; The sub section shall be shaped like a flat ended football (above view) with all amps mounted in the back of the subwoofer section.

[CeramicMan]

Well, as frequency decreases the required driver excursion increases in order to achieve the same SPL. In an open baffle configuration, low frequency sound partially cancels itself out – not entirely, but partially – which means that equalization is required and cone excursion goes up significantly to compensate for the losses.

At midrange frequencies (say: 150Hz ~ 2kHz) it's no big deal: who cares if the cone moves 1mm p-p instead of 0.5mm? But at the 20Hz to 100Hz range, distortion kicks in because the cone often moves a lot more even with sealed or ported boxes.

Furthermore, because frequencies like 20Hz or even 40Hz are near the limit of human hearing, even moderately low distortion can be a lot louder than the fundamental tone. Check out Linkwitz Lab - he's a big fan of open baffle designs and even he uses ordinary boxed subs.

[Teh]

"Furthermore, because frequencies like 20Hz or even 40Hz are near the limit of human hearing, even moderately low distortion can be a lot louder than the fundamental tone. Check out Linkwitz Lab - he's a big fan of open baffle designs and even he uses ordinary boxed subs."

I am well a aware of Siegfreid Likwitz' site. I've spent much time there. And, you're right, he does recomend his "Thor" for frequencies below 40 hz, now he has revised that to 60 hz. His site is incredibly complete and very, very, well documented.

I'm just having a problem being one fish in the school, that's all. If you need a "Sub" then why do the drivers from 60/80/100 hz and below have to be specified? In fact since he himself has made it clear that dipole sound disappears at roughly 80-100hz why not bypass woofers/design that can't do the job adequately at those frequencies? Furthermore, it is stated in his site that frequency propogation below roughly 80 hz ALL become monopole, isn't it? Isn't that precisely what Projest 8 is doing? They are bypassing the weekest link of the "example" dipole system, aren't they?

As we design our own systems, why should we not do the same?

[CeramicMan]

Sure, why not try it then? Part of the thing about DIY speakers is that you can do all sorts of things that aren't possible from a commercial/mass-market point of view.

You'll just need to make sure that the baffle is big enough, and if it sits directly on the floor then that will be a tremendous help as well. The floor produces a mirror image of the baffle and the speaker, effectively doubling its size. If you find that the baffle needs to be 1m (~3') wide, there may be engineering considerations too, e.g.: bracing, so that the open baffe doesn't flex.

Have you checked out the Edge ?

[Ian J]

'The major benefit of open baffle design is apparently in the region of 100/120 (anything below is monopole and dipole’s are incapable of room coupling required at such frequencies) to 2000hz. Primarily in the 100 – 1000hz range.'

That's very interesting - though some people definitely try to go lower with dipoles (~40Hz) as they say that's where dipoles sound really good...

I can't remember where I saw it but there is an early Gradient design with large OB mid which went down to ~200Hz which sat atop a floor-firing boxed woofer. Apparently this eliminated floor-bounce and had the best directivity pattern for a monopole bass helping blending with the dipole mid. Something to bear in mind.

Ian

[pikkujöpö]

[i]

I can't remember where I saw it but there is an early Gradient design with large OB mid which went down to ~200Hz which sat atop a floor-firing boxed woofer.

Ian [/B]

This is the model you are referring to ?


More information can be found from their site,look for products then click vintage.

http://www.gradient.fi/

At gradient they seem to believe in the quality of dipole bass and closed speaker boxes.Gradient Revolution is one proof of that quality.

[Timn8ter]

This is a great thread for me since I've been experimenting and contemplating where to go next with my Dakini OB (a.k.a. Adire DDR). Your idea makes a lot of sense to me.
I built the Linkwitz dipole woofer to fill in the bottom and while it's nice it still doesn't handle the super deep stuff well. Also, I found a first order roll off starting at 125Hz for the Extremis woofers to clean things up considerably. Now, I'm thinking of building sealed dual Shiva woofer sections for each channel (since I have 4 of them anyway) which, surprise surprise, is getting close to the Perfect 8.
Reading about the Perfect 8 I see it says the subwoofer can be used dipole. Does this mean it's not wired dipole by default?