Open baffle speaker pitfalls

[eddantes]

... I could be romanticizing the experience at this point.

This is a good point. There's a nostalgia at play here... Like the first time you expereince a thing - it's just so notable being novel, but in reality maybe hardly worthy of admiration. Still - I think we both agree that the key word here (that we both used) is "fun", and how can fun not also be enjoyable, and thus admirable.

 

[suur13]

Old thread, but still... do I understand correctly, recommended/needed crossover point depends on rather sidewall depth, not on the U-shape or H-shape ?

If depth was 40 cm. crossover point should be around 100Hz, but in case of 20 cm. one can go high as 250Hz ?

Reason being - depth makes kind of "tube" which acts like horn or TL and shorter "tube" lessens the effect ?

 

[ctrl]

sorry for late reply, hardly have time for the hobby...

do I understand correctly, recommended/needed crossover point depends on rather sidewall depth, not on the U-shape or H-shape ?

Yes, the cabinet depth is critical for U-, V-, and H-shaped open baffles (OB). Approximately, the cabinet can be considered a half-open tube (a lambda/4 resonator).

The resonance frequency can be estimated with the formula f = c / 4*Depth. For a cabinet with a depth of 0.4 m, this results in a resonance frequency of about 214 Hz, and for a depth of 0.2 m, about 428 Hz in a rough approximation (in reality, the resonance frequency is slightly lower than theoretically calculated).

For U- and V-shaped OBs, the crossover frequency should be well below the resonance frequency (as a guideline, I’d suggest 0.5 to one octave lower, depending on the slope of the crossover filter). So for U-shape cabinet with 0.4m depth, resonance frequency about 214Hz, one octave below would be 100Hz as XO frequency - as you suggested.

In H-shaped OBs, the resonance is symmetrically radiated, which doesn’t impair the dipole character as much, but for the same resonance frequency, it requires double the cabinet depth.

Reason being - depth makes kind of "tube" which acts like horn or TL and shorter "tube" lessens the effect ?

Yes, as mentioned above, for U-, V-, and H-shaped open baffles, the resonance frequency can be roughly estimated by considering them as a half-open tube.
An U-, V- or H-shaped OB design resembles a transmission line (TL) more than a horn when considering the effect of the air's radiation resistance.

In my next post I will provide a comparison of a 1m wide classical/flat OB and a 0.3m wide but 0.35m depth U-shape OB, the "folded version" of the flat OB, to show the differences in more detail.

 

[MAB]

sorry for late reply, hardly have time for the hobby...

I'm sorry to hear that!
Best wishes.

 

[wonderfulaudio]

@ctrl , thanks for putting in the hard work, you have shown very good revelations.
Do you know what is the directivity of a H frame, given that a flat baffle dipole gives 120 degrees, does the H baffle deviate from it?

 

[howardg]

I think that the premise of the OP is correct, that it is best to create a speaker without resonances as much as possible---whether dipoles or monopole boxes. I think that with my U-frames the pipe resonances are high enough and that the drive level is low enough, that it is not much of a problem.

The two multiway open baffles I have made each have U-frame bass sections; which are each about 7" deep. I feel that the fact that the U-frame shapes that enclose the back side of the bass drivers have as little extra lateral dimensions as practicable-they are almost pipe like- helps limit the variety of resonances that occur. The (passive) acoustic low-pass is not very steep, but it starts from a low frequency (about 40hz) in both cases (inspiration from Martin King). There must be a pipe resonance occurring in each case, but I do not notice it when listening, and I think that I am pretty sensitive to humps in the lowish midrange.

The first build is arguably the one with the least complexity and faults, though the woodwork is not simple. It is a full 4way with front and rear dome tweeters (slightly offset). Other drivers are 18" 32hzFs/ 12" 45hzFs/ 4.5" cone mid. Only the 18" is in the U-frame. The bass is different than a boxed woofer, but is pleasing to me (and others). The bass is strong to about 40hz, which is usually enough. I have a 3.3cuft NHT12" active sealed sub if I need to hear 30hz. The crossover is low order and not too many parts---the drivers are all well behaved and have a lot of overlap. The baffle is 1.75" thick and its slightly curved sides taper from 20" width at the bottom, to about 4.5" at the top. The overall sound-stage resembles that of the LX-521---which I only heard once in a largish sized hotel room.
The second build is a little bit of a problem child, mostly due to the fact that I stubbornly used some classic JBL drivers---perhaps unwisely. I plan to replace the 123a bass drivers in the U-frame with Beyma 12BR70s. The D216 8" low mids have been further tamed with an adjustable L-Pad---these do provide nice detail.

Both of the above want a larger room, but are listenable in my 11'x14' space.

Also, I must say that I have heard H frames that had pronounced low-mid peaks that were obvious to me----though others that were present did not seem to mind---or did not want to say (I don't think I "said" either). Maybe these were crossed just enough higher to make the resonances more obvious? Also these had square H-frames that were 2" larger on the inside than the 18" driver frames, and I think that this likely set up various reflections that added to the sound produced by the driver.

I also made H-frames whose front enclosed shapes were geometrically identical to their rears, and identical to my "first build" 18" woofers mentioned above. I think that the contribution they make to the overall presentation of the low-mids and upper bass was less clear sounding than that of the U-frames, though they likely had more lower bass.

 

[Richard G]

Very helpful observations

So I need to re-read when I’m properly awake.,

I just realised that I can enlarge my room relatively easily, into a large enough space that dipoles should work very well. So I hope/ anticipate being able to employ some of your experience and learnings

Much appreciated

 

[howardg]

I think that the sound-stage effects of dipole, even some with very significant response or radiation/dispersion "irregularities", can be enjoyable. This does not mean that the problems don't need fixing, just that the spacial aspects can be special enough to be enjoyed despite the sonic "warts".

While it may not be a purist appeal, I think that the added soundstage effects of a dipole pair (even with "warts") could be compared to the enjoyable aspects of even a very modest stereo setup, when compared to mono. It may be harder to explain why some inexpensive stereos sound better than they should, once we remove our criticism "hats"---than it is to point out specific sonic flaws in nicer equipment (with those discerning hats on). Maybe the explanation lies in how much emphasis our brains/perceptual systems place on locating objects and creatures in our environment.

With the above said, I have monopole speakers that I enjoy alot also. Sometimes it is even better not to be quite as fully envolved in the sonic presentation. I recently completed an 3way 8" acoustic suspension pair (vintage AR woofers) that work wonderfully---quite near the front wall and corners. They do have controlled directivity facing forward. This is very handy in a small room.

I have heard a hybrid reflex/dipole, crossed actively at about 120hz, that must have been interacting with the room in the bass differently than a full dipole, but was quite entertaining---in a good way. I think that it was mostly the nice dipole planar tweeter that really made these sing. Having the solid bass there was good. The bass module was a bipole---with the woofers on the sides-so very different than dipole bass.

I think a very worthwhile experiment would consist of an active sealed bass module near the floor, monopole or front/back bipole, with a dipole element spaced about 30" on center above it, handling mids and highs. The bass module could be blended in over a very wide range, handling all of the baffle step compensation. I mocked up something like this, but with open baffle woofers below, one of which had lovely midrange sound (Faital pro 10FE200), the other being crossed lower, like a ".5" type of setup. The upper driver was a wide range vintage Fostex, which when spaced 28" on center from the Faital woofer, retained its single driver point-source appeal. When spaced closer, the combo sounded more like a convetional multi-way speaker. I think this may be due to the brain's implementation of source locating, known as the "precedence effect". The stereo woofers were directly below the single driver, not off "hidden" somewhere.

In the above scenario, I think that the mid/hf unit(s) could be dipole, or monopole, if room placement was problematic. If output demands were modest, a small single driver in a narrow baffle might be passively blended with the woofer module, maintaining the 28/30" spacing. This might share some of the spacial quailties of Linkwitz' LXmini, but have deeper bass if a larger woofer were used. Active might be necessary to deal with the dipole peak, though.

 

[Juhazi]

(10" sealed sub, 12" cone, planars, dsp)

 

[howardg]

I just realised that I can enlarge my room relatively easily, into a large enough space that dipoles should work very well. So I hope/ anticipate being able to employ some of your experience and learnings

Me too. I have a larger unfinished basement room that I have been using for projects and storage, but I am moving towards having a listening space there. Just a matter of shifting things around and doing finishing work.

 

[howardg]

(10" sealed sub, 12" cone, planars, dsp)

What is the blond/lighter colored sculptural shape that is just in front of/to the left of the cone/planar?

Is the sealed woofer unit a stepped tower that is shaped like a (black) wedding cake?

 

[Richard G]

What is the blond/lighter colored sculptural shape that is just in front of/to the left of the cone/planar?

Is the sealed woofer unit a stepped tower that is shaped like a (black) wedding cake?

Your design derivation of a Gradient, Juhazi?

 

[howardg]

OK, now I am getting the idea:

Aino gradient - a collaborative speaker project

Ok folks, inspired by the discussions of room acoustics, sound perception and boundary reflections here at diyaudio.com, I hereby start a new thread for a collaboratory speaker project named Aino gradient The inspiration to this project is the legendary and revolutionary home stereo speaker...

www.diyaudio.com

 

[howardg]

I see the AINOgradient uses discreet DSP channels. I only have one miniDSP 2X4, and I even try to make all passive designs if I can. I am still learning about passive. Part of it is the challenge, part the portability for sharing at clubs etc. I do have a pair of DIY 4way monopole that really need the EQ in the bass, so they get the miniDSP.
A friend, who is much more technically inclined and capable than I am, is putting the finishing touches on a pair that have 6channels each side---which he can switch between monopole/dipole/bipole---most folks like the dipole the best, with one exception so far, who prefers the monopole---he uses big Harbeth at home, which are a conventional design.

 

[Juhazi]

A commercial speaker in USA is basically similar to my AINOgradient

Consonus Audio Products

Transcend horizons with Consonus sound system!

www.consonusaudio.com

 

[howardg]

A commercial speaker in USA is basically similar to my AINOgradient

Yes, I can appreciate the similarities.
And now we can get the GRS planars for a low price from Parts Express in the USA. The ones that I have heard sound good.

My first open baffle(2016) is shaped like that; but the bass is 18" U-frame, then 12" Bozak, then 4" Faital Pro, then (Vifa?)neo-dome tweeter that has a tiny wave-guide. Been thinking about swapping the Faital mid to a Scan Speak. I found that I like wrapping absorbent material around the back of the dipole mid---maybe it makes it more like cardioid? The old 12" Bozaks are lovely sounding extended range drivers. Reasonable price on Ebay here.

The sealed bass would have a different presentation than my 18"---and I think that I would like it. My 18"--- Tymphany manufacturer's samples--- have low QTS, and only 92dB sensitivity, but I still get good bass from it. The 12" Bozak has an Fs around 45hz, and maybe .45 Qts, and is not hi-passed. So, though not in the U-frame, it supports the bass quite a bit. I think system sensitivity may be about 87dB(all passive filters). I should run an impedance sweep on it--- I think that it must get down around 4ohms. Even the 40watt push-pull tube amp does not do good bass with it---kinda mushy.

 

[Juhazi]

Dipole woofers need huge EQ and thus high power from amps. The low mid should have some highpass to prevent excessive excursion (it increases IMD)
The major reason why I use sealed sub is that it is smaller than a dipole that could go below 30Hz

My room response drops fast below 20Hz

 

[howardg]

Yes! Size is a big factor for me too.
I think that using my 12" Bozak up to about 500-600hz works out OK because I do not boost it in the bass, or start rolling it off before it meets the Faital 4" midrange driver, which I think is marketed as a "full-range". The filters are low order, and I don't think that this speaker should be played super-loud; but I have played it reasonably loud in large rooms (church basement), and it worked well. The 12" does supplement the 18" to about 45hz or so, but the 18" does most of the heavy lifting in the bass.
These speakers are really pleasant to listen to. I have plans to make a version with a boosted 15" (so that it will be smaller) instead of the 18". I would just use a miniDSP 2X4HD, having the upper 3way be a passive section. I could use a sealed/boosted bass section for this one, but I suppose that the enclosure would need to be at least 2cuft to work with a 15". Maybe better use a 12" to save space.