[JohninCR]

Were talking 3x15" driver with high qts per channel doing bass job, you still don't think we'll have enough of low freq?

http://www.sonicflare.com/archives/vegas07-vision-of-the-future-emerald-audio.php

I guess it depends on what you mean by enough, but it looks like three of those drivers are only capable of producing ~90db at 40hz at full xmax on a reasonably sized baffle.

Don't forget the +6db of boundary reinforcement.  Anyway, you can't really look at OB bass in those terms unless you're talking about wanting to really "feel" the bass.  First, there's a 4db perception advantage of OB bass due to the increased directivity of OB bass (ie reduced reflections makes the direct radiation perceptively louder.  Linkwitz goes into some detail regarding this).  Also, because the bass roll-off is much shallower than the roll-off below F3 of boxes, our brains are better able to fill in what's missing.

[Duke]

A dipole typically has a roughly third-order rolloff below driver resonance:  6 dB per octave from dipole cancellation, and roughly 12 dB per octave because we're below the driver or panel's free-air resonance.

So this is steeper than some boxes, and shallower than others.

I used to subscribe to Linkwitz's theory that the low-frequency directivity of a dipole was responsible for its subjectively very good in-room bass smoothness.  Much as I respect Linkwitz's work, I now disagree with him on this point.

Note that the ear does not even register a low frequency tone until one or more cycles have reached it, so the concept of "direct sound" as opposed to "reverberant sound" doesn't really apply in the bass region.  By the time the ear registers the tone, we're past the direct-sound time threshold.  In fact, if the bass notes decay too quickly, the ear is unable to get a fix on what the pitch is.  One implication is that the oft-noted improved low-frequency pitch definition of a good dipole arises from another cause.  I theorize that this other cause is the in-room smoothing effect of multiple low frequency sources observed by Welti, Geddes, and others - for you see, a dipole is in effect two monopoles with the phase reversed on one of them.  James Kates documents the improved in-room smoothness of dipole bass, and while his work preceeded that of Welti and Geddes it did focus on how a dipole interacts with the room differently from a monopole to produce measurably (and audibly) smoother bass.

Duke

[Kevin Haskins]

There is just no way to get "high efficiency" and "dipole bass" in the same sentence (hey... I just did it!).

For simplicity, just think of it this way.   At the dipole roll-off you start "throwing away" output.    It cancels at the sides (hey... its a dipole).    The only way you make up for it is with power & driver output.   Real dipole bass requires lots of high stroke drivers, EQ, and wide baffles.   

 

[JohninCR]

Duke,

You obviously don't use woofers with an Fs above where you want the bulk of your output, so low 40's at a minimum.  It's also pretty obvious that reduced noise (in this case reflections) makes output subjectively louder, which SL quantifies at 4.2db.  In fact, I believe this is the primary factor contributing to increased clarity in the bass region for OB's.  The only "knock" I have regarding OB bass is that it doesn't result in the same kind of physical sensation of bass as boxes, though deep U-baffle subs do give back much of that feel with relatively minor increased room interaction.  I'd caution against being at odds with SL, because I've disagreed with him on several occasions and always been proven wrong, except about U-baffles being +6db at the bottom over similar sized dipoles, but he has since changed his position on that.

John

[JohninCR]

Duke,

After re-reading your post I think we agree, but the effect is explained in different ways.  Take boxed bass and OB bass out in a field and the benefit goes away.


Kevin,

I disagree, good sensitivity is possible, but high output in the bass region becomes problematic.  That is unless you're a subscriber to a requirement for ruler flat response to 20hz, for which boxes must take into consideration the room, and dipoles typically start banging against the limit related to the room's on axis dimensions.  Judicious placement can also be of significant assistance in the bass region....it's not just a baffle size thing.  In addition, U-baffles can be employed in compact forms that are comparable to box sizes that can enhance bass performance on par with boxes.

[Duke]

JohninCR,

Just for the record, my intention is NOT to knock dipole bass!  Three of the five speaker lines I'm a dealer for are dipoles in the bass region.  I just have a somewhat unorthodox concept of why dipoles do such a good job with pitch definition in the bass region. 

I don't know much about U-shaped baffle bass systems.  How deep of a U are u talking about?   Is the pathway stuffed or unstuffed?  Maybe there's an article or a thread that I should go read. 

Regarding the original poster's concept, I'm skeptical about being able to do a SET-friendly reasonable-sized dipole speaker that has decent bass extension without equalization (which would make it no longer SET-friendly due to increased power demands).  I am also skeptical about being able to get a SET-friendly impedance curve from the format he proposes.  But if the U-shaped box opens a window of opportunity in the bass region, that would be valuable information.

For a SET-friendly, mostly-dipole speaker in my opinion Dick Olsher has thought it through quite well in his BassZilla.

Duke

[ttan98]

Duke,

refer to this site for U-frame theory,

http://www.musicanddesign.com/u_frame.html

cheers.

[JohninCR]

JohninCR,

Just for the record, my intention is NOT to knock dipole bass!  Three of the five speaker lines I'm a dealer for are dipoles in the bass region.  I just have a somewhat unorthodox concept of why dipoles do such a good job with pitch definition in the bass region. 

I don't know much about U-shaped baffle bass systems.  How deep of a U are u talking about?   Is the pathway stuffed or unstuffed?  Maybe there's an article or a thread that I should go read. 

Regarding the original poster's concept, I'm skeptical about being able to do a SET-friendly reasonable-sized dipole speaker that has decent bass extension without equalization (which would make it no longer SET-friendly due to increased power demands).  I am also skeptical about being able to get a SET-friendly impedance curve from the format he proposes.  But if the U-shaped box opens a window of opportunity in the bass region, that would be valuable information.

For a SET-friendly, mostly-dipole speaker in my opinion Dick Olsher has thought it through quite well in his BassZilla.

Duke

Duke,

U-baffles require some form of damping inside the cavity to function properly.  Without it, the air behaves as a lumped mass and the terminus becomes the actual source, so you give up 1/2 of the added travel distance for the rear wave.  When you break up that lumped mass behavior, the effective delay for the rear wave then becomes the distance from the driver to the rear edge + the distance from the rear edge to the front plane.  In effect, from a bass extension standpoint, a 1ft deep U (only wide enough to mount the driver) becomes the equivalent of a 4ft wide flat baffle.  That's why I absolutely disagree with Kevin's old school assessment that OB's have to be big.  As you get away from an identical structure for the front and rear output from the driver, you change the radiation pattern away from the dipole figure 8 toward a cardioid pattern.  In practice, I've found the difference to be minor other than to widen the front area of prime output, mostly because room placement doesn't permit a true dipole radiation pattern anyway.

I agree with you regarding power and EQ, though 1.5 years ago I did come up with a way around it and built several pairs of compact OB's with sufficient bass extension driven only by a single 2 watt SET without EQ.  It uses a single coax drive unit, and the approach I used was a U-baffle with a Helmholtz slot immediately behind the driver, which is the entrance to a much longer pathway.  The slot filters out only very low frequencies (like a Helmholtz bass trap), but instead of absorbing them, they travel the expanding pathway.  This results in a rear wave delay of almost 2m, sufficient for phase alignment of that portion of the rear wave to actually reinforce the front wave, negating dipole cancellation down to about 30hz.

John

[JoshK]

All the theory aside for a moment, the most practical way to implement something like this is with a modern plate amp that has built in xo.   Then look to John K's site on modding a plate amp to include EQ and modify accordingly for the type of OB you are trying for (U, flat, etc).  Then your SET amp can run the MT and let the plate do the hard work.

[JoshK]

For a SET-friendly, mostly-dipole speaker in my opinion Dick Olsher has thought it through quite well in his BassZilla.

Duke

Indeed this is a good solution to the problem.  I think our fellow ScottF has something of a variant of this solution without the tweeter. 

[mcgsxr]

I bought the Reckhorn Xover to run my mains and subs with different amps, and trim the frequencies accordingly.  Cheap and cheerful!

[wikin]

The baffles are now made wider to counter skeptics of poor bass 

http://www.audiocircle.com/index.php?action=gallerylastup&cat=0&pos=0

[Rudolf]

The baffles are now made wider to counter skeptics of poor bass 

wikin,
making the baffles wider surely will help for the lower woofers, but next to nothing for the top one IMHO. If you don´t need the side baffles that high for aesthetical reasons, you could easily stop them at the height of the waveguide.
In general the depth of the side baffles does more for you than the width. Your sketch doesn´t show exactly the value of the angle between the side wings, but you could make that 90° or even less without creating anything like a back cavity (which might create some resonance peak).

Rudolf

[wikin]

I figured the overall shape gives a balance between asthethics and functional. Going 90deg will make the effective path length longer but maybe less sexy 
Anyways since there's no model available I'll just have to reverse engineer that cross over using a DCX2496.

cheers.

[el`Ol]

A suggestion for an inexpensive passive multi-way 4 Ohm 95dB OB speaker:
Ciare CH250 without whizzer, crossed at 1kHz (no dispersion problems) to this open-back 2" soft dome:
http://www.intertechnik.de/index.html/JTI2bmF2aWQlM0QxNzY4JTI2bGFuZyUzRGRlJTI2c2lkJTNEbjQ3MWEwMGQ0YWE5MzMlMjZiJTNE.html?basis=1174&detail=4417&suchwort=
It has excellent measurements according to K+T.
BG Neo 3 PDR as supertweeter. Has a 95dB peak at 10kHz in the open back variant. Unfortunately rolls off a bit towards 20kHz.
Would cost a bargain, but would be big to achieve 60Hz.