Eminence Beta 15 ripole loaded

[bigwill]

Recalling TS numbers from memory, this driver has a Qts of 0.59, Fs of 35Hz, Xmax of 4mm and an efficiency of around 95dB.

In a ripole, theoretically the Fs will be lowered and Qts raised somewhat, both parameters going in the right direction for dipole use...


Also, this driver has a huge, quite broad peak around 2Khz, but in a ripole alignment this would possibly be tamed due to the geometry of the thing

Basically I want to make a relatively efficient passive speaker system of around 95dB /m to drive with medium power tube amps and I wonder if two ripole beta 15s per side would create satisfying bass?

 

[bigwill]

As a side note, the quality of bass to me is more important that sheer quantity, I just want to hear bass lines and get SOME kind of tactile sensation on drum hits.

 

[Scottmoose]

It'd probably be decent enough to ~40 - 50Hz, so I'd then XO to a driver actually designed for something other than LF duties at about 200Hz (or 500Hz -my two favourite XO points). The Fostex FE206E run sealed or in a front-horn springs to mind & if the HF isn't good enough for your taste, then bring in a supertweeter at ~10 - 12KHz. Forget trying to run the Betas up past 500Hz. Vocals & midband are not what 15in drivers are good at.

 

[bigwill]

It'd probably be decent enough to ~40 - 50Hz, so I'd then XO to a driver actually designed for something other than LF duties at about 200Hz (or 500Hz -my two favourite XO points). The Fostex FE206E run sealed or in a front-horn springs to mind & if the HF isn't good enough for your taste, then bring in a supertweeter at ~10 - 12KHz. Forget trying to run the Betas up past 500Hz. Vocals & midband are not what 15in drivers are good at.

That's roughly what I hope to do - cross them to a wideband, possibly a fostex in the 200Hz range and add a tweeter on top as needed.

Subjectively speaking, is it possible to guess from your (or anyone's) experience that the bass from the Betas would provide a solid musical foundation and not a 'thin' sound?

 

[Rudolf]

Recalling TS numbers from memory, this driver has a Qts of 0.59, Fs of 35Hz, Xmax of 4mm and an efficiency of around 95dB.

For a Qts of 0.5 efficiency will be down 6dB at Fs. In open baffles of acceptable size the 6 dB dipole rolloff will start at least below 100 Hz. So at 35 Hz your efficiency will have dropped to 85 dB (if you are lucky). In a fully passive system you will not be able to get better than that.

In a ripole, theoretically the Fs will be lowered and Qts raised somewhat, both parameters going in the right direction for dipole use...

That´s right, but the price you pay is efficiency. My own tests have shown that a driver in a ripole configuration will be roughly 6 dB less efficient than the same driver in a H-frame for most of the passband. Only at the lowest frequencies will the ripole be on par.

Basically I want to make a relatively efficient passive speaker system of around 95dB /m to drive with medium power tube amps and I wonder if two ripole beta 15s per side would create satisfying bass?

You will have a hard time to achieve that. May be the Eminence Alpha 15A is a better choice. Did you read MJKs paper on a two way OB speaker design?

 

[bigwill]

How would the Alpha or Beta 15 do in a U-Frame?

 

[bzfcocon]

I use the Alpha15 (one per channel) in Y baffles(sort of U baffles narrowed at the back) and I am quite happy with the sound. You get a bit more efficiency (the drivers sees a larger baffle), but the polar pattern is probably not a dipole anymoore (it becomes cardioid)

The one big issue is the TL resonance of the enclosure which needs to be dealt with(around 220 Hz in my case). There are several ways to do that:
- equalize it out, like I do. This works best if the resonance is not in the passband (I cut the Alphas low, at 120Hz).
- use the "right" stuffing as JohnK describes for NaO on his great website: http://www.musicanddesign.com/u_frame.html

Personally, I found the stuffing approach tedious - also because my measurement system is not good enough.

 

[Rudolf]

How would the Alpha or Beta 15 do in a U-Frame?

Just reckoning since I have neither driver: They should work.
The U frame will be more efficient than the ripole (even more than a H-frame of same size), but you will be limited to a lower frequency threshold at Fs of the driver.
Regardless of the frame you use - its always efficiency vs. lower frequency limit. You can´t push one of it without sacrifycing the other.

 

[bzfcocon]

The U frame will be more efficient than the ripole (even more than a H-frame of same size), but you will be limited to a lower frequency threshold at Fs of the driver.

Just to make sure I understand you right: is this because of the additional mass loading of the driver and therefore the lowering of Fs in a ripole ? Isn't there a similar mass loading in a (deep enough) U-Frame ?

Are we talking about the efficiency at the lowest frequency in the passband (which forces everything above to be equalized down) or about the "average" efficiency - which increases towards the dipole/TL resonance ?

Or do you mean something else ?

 

[Rudolf]

Just to make sure I understand you right: is this because of the additional mass loading of the driver and therefore the lowering of Fs in a ripole ? Isn't there a similar mass loading in a (deep enough) U-Frame ?

I should have said: The U frame will be more efficient than the ripole (even more than a H-frame of same size), but you will be limited to a HIGHER frequency threshold at Fs of the driver. And yes, this is due to the additional mass loading. But you would need a VERY deep U frame to get the same amount of mass loading. Consider that the opening area of a really mass loaded ripole is typically 1/4xSd. For a U-frame it would typically be 1.5x Sd. And the U-frame would only be loaded from one side.

Are we talking about the efficiency at the lowest frequency in the passband (which forces everything above to be equalized down) or about the "average" efficiency - which increases towards the dipole/TL resonance ?

I have compared two Visaton W250 drivers in a ripole (middle) and a M-frame (right):
[IMGDEAD]http://rudolffinke.homepage.t-online.de/audio/Dipol/W250/Dipole_B.jpg[/IMGDEAD]

For this diagram the ripole was standing on the floor and the M-frame was lying. Mic position was 30 cm from the front opening at floor heigth:
[IMGDEAD]http://rudolffinke.homepage.t-online.de/audio/Dipol/W250/Vergleich%20M-frame%20Ripol_2.gif[/IMGDEAD]

As can be seen, efficiencies meet at the lowest frequencies in the passband. I hope the pictures can explain my point better than my words.

 

[bzfcocon]

I see.

In this particular case, one could think that it would me more advantageous to make use of the extra 5dB efficiency of the M-Frame (and maybe just lift the lowest end a bit) rather than using the ripole which is less efficient throughout.

I am very interested in this problem because it has been pointed out by some that "fixing" efficiency issues by heavy equalization is always done at a cost. In other words, the "genuine" efficiency of the speaker system alone IS important (as it allows for significantly more headroom and less stress for drivers). Other people seem to think that, as long as there's enough power and the drivers can take it, equalization won't hurt.

Did you make a comparison of how the two configuration sound - of course, equalized on the same response curve?

 

[theAnonymous1]

Speaking of equalization, is there some sort of analog active circuit that can be used for a dipole sub; like a specialized Linkwitz transform?

I have an active digital setup, but I do not like adding lots of gain in the digital domain to compensate.

 

[bzfcocon]

Equalization would highly depend on the actual raw response, but usually a shelving low pass circuit is used (that's also how I do it), combined with a notch filter, if necessary.

It's all on Linkwitz' site:

http://www.linkwitzlab.com/filters.htm

 

[theAnonymous1]

Thanks, I think I will try an analog shelving low pass filter. Any notch filtering I need can be done digitally.

 

[Rudolf]

I see.

In this particular case, one could think that it would be more advantageous to make use of the extra 5dB efficiency of the M-Frame (and maybe just lift the lowest end a bit) rather than using the ripole which is less efficient throughout.

That´s exactly what I would think too!
But we have not compared the excursions yet. It may well be that the ripole will move the same amount of air with a bit less excursion.

I am very interested in this problem because it has been pointed out by some that "fixing" efficiency issues by heavy equalization is always done at a cost. In other words, the "genuine" efficiency of the speaker system alone IS important (as it allows for significantly more headroom and less stress for drivers). Other people seem to think that, as long as there's enough power and the drivers can take it, equalization won't hurt.

I don´t see efficiency as the main limiting factor. As long as you are not restricted to tube flea power, maximal (linear) excursion is the parameter to keep in mind.

Did you make a comparison of how the two configuration sound - of course, equalized on the same response curve?

I have not done that comparison yet. In fact, I would rather measure distortion values than do a listening comparison. Limited to < 200 Hz almost all drivers sound the same to me (when equalized).

 

[bzfcocon]

That´s exactly what I would think too!
But we have not compared the excursions yet. It may well be that the ripole will move the same amount of air with a bit less excursion.

I might not see the whole picture here, but isn't increased efficiency usually associated with a reduction of the excursion for a given SPL ? How can the same 2 drivers work less efficient and simultaneously require less excursion for the same SPL ?

As you said, in a ripole, the radiating area is reduced compared to U or other baffles. Even if, in this way, the driver moves more air per surface unit, this is balanced by a proportionally reduced radiating area.

Again, i am no expert, please fell free to contradict me.

I have not done that comparison yet. In fact, I would rather measure distortion values than do a listening comparison. Limited to < 200 Hz almost all drivers sound the same to me (when equalized). [/B]

Oh, sorry, you are right of course. What I had in mind was a listening comparison in fullrange, that is, coupled to the same set of speakers taking care of the rest of the spectrum.

Listening to bass units alone is indeed not very revealing, except maybe for unpleasant obvious resonances.

 

[bigwill]

Thanks for your replies everyone, they are very informative.

If I were to buy a single Beta 15 and put it in a U-Frame would I be able to get a feel for what it would sound like without having two per side?

 

[Rudolf]

bigwill,
look at the left speaker in my dipole collection in post #10. That´s my 15" in a H-frame. One of those per side is plenty enough for my 5x4.2 m room.
So I´m quite sure that a single Beta 15 in a U-frame will give you a reliable first impression.

 

[bigwill]

Excellent I hope it works, clean tight bass is just so awesome to listen to

 

[theAnonymous1]

Rudolf, a few questions for you...

Is the driver baffle in your H-frame centered or offset? It looks centered, but I don't always trust my eyes.

Do you prefer the H-frame over your M-frame; and if so, why?

If I wanted to build a H-frame with two drivers, would it be advantageous to put a separator between the two drivers, kind of like two single H-frames stacked?

Thanks