Giant Subwoofer?

[TNT]

But a driver could be made behave like these specs by: "negative output resistance. However, this method also controls the moving mass and complianc"

http://www.aes.org/e-lib/browse.cfm?elib=3890


/

I love horns. I'm not knocking horns. But answer the question. For ultra-low bass, what is the most efficient?

You know what the answer is, so don't dodge the question.

And besides, what are you going to PUT in this uber-horn? A driver with an FS of 25hz? That's not going to cut it. To get that low, you'll need a driver with an FS of 8hz and a qts of 0.25. Drivers like that DON'T EXIST.

:: PB ::

 

[AKN]

Hi,

Go ahead then. Fire up hornresp or Mathcad, and let's see how you make a horn that's efficient below the FS of the driver. How do you do that?

I am not used to Hornresp yet, give me some time and I will shurely confess if I am wrong.
However inititial sims seems to confirm my and others pratical results that shows it's possible to extend a great deal below fs.

Just a thought:
If we take one driver and decrease its moving mass so fs goes up, why should that rise a given horn cutoff?
Edit: Make the response drop earlier?

Has any other good theoretical insight into this?

 

[AndrewT]

Hi,
just some ramblings about what might be happening around Fs.

A normal speaker is inductive on one side of Fs and capacitive on the other side (the +ve and -ve slopes on the resonant peak).

With a driver that is fully horn loaded the mechanical (air) resistance provided by the horn suppresses the inductive/capacitive parts and provides just resistance.
The horn efficiency hits 50% when the electrical resistance of the Voice Coil matches the mechanical resistance of the air in the horn. When this occurs the amplifier sees a load equal to double the Re of the speaker. I guess an 8ohm driver becomes a 14r (about 2times Re) load on the amplifier.

If this holds true near or above Fs, then it should also hold true below Fs. But this will only hold true if the horn still provides the correct air loading for the cone at the lower frequencies below Fs. That means a bigger horn.

Is any part of that close to the truth?

 

[Retsel]

Any large horn subwoofer will result in an extremely delayed bass response that will never match the main speakers unless if the main speakers are also delayed. I would never place any sort of delaying circuit in my precious midrange which I want as transparent as possible. I read up on horn subwoofers to the point that I was ready to build one, then that one and only issue stopped me.

The second issue with horn subwoofers is while they tend to have less distortion than many other subwoofer types because of the efficiencies involved, they have their own distortion problem. They essentially are a small sealed sub with a long horn at the end of them. I firmly believe that sealed subs cause the cone to deflex due to the pressure exerted upon the cone by the backpress of the box, thus causing distortion.

Infinite baffle subs, however, use very, very large boxes for venting the back wave which does not cause deflections of the cone by the forces of backpressure. This suggests to me that they will have lower distortion compared to horns, however, they will need more surface area/xmax to accomplish this.

The only additional technology that could be added to further improve the sound quality is to add a servo. However, servos can be added to any of these technology types.

Retsel

 

[Patrick Bateman]

Hi,




I am not used to Hornresp yet, give me some time and I will shurely confess if I am wrong.
However inititial sims seems to confirm my and others pratical results that shows it's possible to extend a great deal below fs.

You state that you have simulated results that contradict my statement. Post them. And if you're not using hornresp, then what are you using?

Originally posted by 4fun Just a thought:
If we take one driver and decrease its moving mass so fs goes up, why should that rise a given horn cutoff?
Edit: Make the response drop earlier?

Has any other good theoretical insight into this?

 

[SpeakerBob]

Oh really? Riddle me this. Which of the following has the highest efficiency at 10hz:

1: A pair of twelves in a big ol' horn, a la the Lab Sub
2: Twelve twelves in a infinite baffle, a la Tom Nousaine's Twelve Shivas Dancing

Hey Patrick,

What happened to the series tuned BP being the king down low ? IB's are a sealed alignment.

All comparisons to the labhorns should really be made for a group of 6 labs, as that is how they were designed to be used. Admittedly some are getting good results with pairs in corner loading, but that is room / placement dependant. I actually have another 4 labhorns cut but not assembled in case I get a room big enough to squeeze 6 in. Not likely but I can dream

Cheers,

Rob.

 

[SpeakerBob]

Any large horn subwoofer will result in an extremely delayed bass response that will never match the main speakers unless if the main speakers are also delayed. I would never place any sort of delaying circuit in my precious midrange which I want as transparent as possible. I read up on horn subwoofers to the point that I was ready to build one, then that one and only issue stopped me.

The second issue with horn subwoofers is while they tend to have less distortion than many other subwoofer types because of the efficiencies involved, they have their own distortion problem. They essentially are a small sealed sub with a long horn at the end of them. I firmly believe that sealed subs cause the cone to deflex due to the pressure exerted upon the cone by the backpress of the box, thus causing distortion.

Infinite baffle subs, however, use very, very large boxes for venting the back wave which does not cause deflections of the cone by the forces of backpressure. This suggests to me that they will have lower distortion compared to horns, however, they will need more surface area/xmax to accomplish this.

The only additional technology that could be added to further improve the sound quality is to add a servo. However, servos can be added to any of these technology types.

Retsel

Retsel,

I'm sure many people are looking to horns for home theater, and any dolby digital / dts processor will compensate for the delay within its setup options. Only 2 channel people will have the delay problem. As you know, I use digital xo's for my mains anyway , so it is not a problem and doesn't cause any extra degradation of sound for me.

I'm pretty sure I read somewhere that the air suspension of a small sealed box is more linear than the drivers suspension. The IB has to rely on the driver's suspension which 'may' not be too good. Of course this is my fuzzy memory so don't hold it as fact.

I'd love to see someone make a servo feedback horn - Sealed servo subs are great sounding but a little shy in the spl stakes.

Cheers,

Rob.

 

[SpeakerBob]

Just out of interest, when comparing IB's to other types of enclosure, shouldn't the volume of attic space / crawl space be allowed for the other enclosure as well ?

Seems fair to me. Andrew seems to be doing that with his horn.

Rob

 

[AndrewT]

Hi,
my first iteration will be without a back box.
If that works then it's saved a lot of work.
If Ret was right then it saves the cone re-shaping distortion as well.
But, I do not believe cone re-shaping to be a problem. I'm trying to imagine the load the cone feels just driving the horn.
The load/stress/strain in the cone at higher frequencies trying to accelerate the inertia of the outer reaches of the cone are a mighty big load as well. I think the manufacturers have this one nailed pretty well.

 

[AKN]

Hi Patrick Bateman,

I am posting two sims with a known box. William Cowan's Horn sub Jr, dual 10" driver 40Hz horn from here. I am simulating with two different drivers with extreme differense in fs. One Beyma 10" pro mid driver 10MI100, fs 89Hz. And one with Peerless XLS 10, fs 18Hz. The Beyma driver fs is well one octave obove cutoff.
I cannot see why this should not apply to a horn in a bigger scale.

Here is Peerless results:

 

[AKN]

And here Beyma results:

 

[Retsel]

Well people have different tolerances for the quality of their midrange. I recall that Bert Doppenberg built a large concrete stereo horn for a subwoofer under his Oris horns. After he built it he could not obtain a satisfactory integration with his mains without delaying his mains, which he found undesirable, so his concrete horns lay idle. He instead uses boxed subs which he sets in front of his now extinct concrete horn subs.

High x-max drivers have very flexible surrounds which will allow the driver to travel linerally over its x-max. These are the best drivers for use with infinite baffle subwoofer systems. Usually you use multiple drivers which helps to insure that the drivers are only being used over their linear range.

Infinite baffle "box" spaces are usually spare areas which can easily be used "as is." In my case I am planning to use a crawlspace when I finish renovating that part of the house. Horn subwoofers, on the other hand, require very carefully built boxes with the appropriate expansion. Even just a little flex in the horn or compression chamber will kill the dynamics of the bass. I experienced this with my Hedlund horns.

Retsel

 

[SpeakerBob]

Yep its funny as some people have different tolerances for their bass. I used to use 3 tempests in 200L stuffed / heavily braced boxes. The pair of labhorns sound so much better that I'd really not want to go back to sealed boxes. Unfortunately I'm probably going to have to use them for the bottom octaves as the labs don't get down there. GM's BLH sounds very interesting but I'm guessing absolutely huge!

Oh well, horses for courses.

Rob.

 

[Retsel]

Hey Rob, we agree.... I remember the time when my sealed box subwoofer was having problems so I set out to figure out what was wrong. With the driver still receiving a signal, I removed the driver from the box. Just after the driver was removed, the upper bass suddenly improved dramatically in sound quality (of course the lower bass disappeared because of sound cancelling). The box was clearly the problem in letting the driver do its thing. Your small 300 liter boxes would have such a problem.

I am convinced that the box is the problem. Horn subs attempt to avoid this problem by reducing the travel of the driver by the better coupling of the driver with the air. It still does have a box to deal with (the compression chamber), but less so because the driver does not have to move as much.

Infinite baffle speakers esentially does away with the box altogether by using a huge (infinite sized) box, hence the very large improvement in sound quality over smaller boxed subwoofers. I suspect that infinite baffle subwoofers will better horn subwoofers in sound quality, and allow for extension to very low frequencies if that is desired.

Retsel

 

[SpeakerBob]

Hmm I wouldn't call a 5' high per driver cab small. I doubt my tempests would sound any better in an IB tbh. How can you explain that my scans sounded great in a 120L per side enclosure, with a higher Q to boot ? The tempests were ~Q 0.6, with the scans around the Q 0.72 mark. Surely if it was 'small boxes' causing the sound problems then the scans would get it too ?

Also you were listening to your driver in a dipole config, not IB.

Personally I think the heavy cone , big x-max type drivers just don't do 'it' in the 40 - 80 Hz area. I've had pro 15's in my room sounding way better than the tempests too. (Beyma LX60), again light cone, smaller x-max.

Makes me wonder why horn loading the lab12's makes them sound so good as they are of the heavy cone etc family.

I can only really comment on what I've heard though.

Cheers,

Rob.

 

[SpeakerBob]

Re : Bert Doppenberg,

Don't forget he's got products to sell - including boxed speakers to go beneath his horns...

I try not to use commercial info to make my mind up on things.

Rob.

 

[Patrick Bateman]

Hey Patrick,

What happened to the series tuned BP being the king down low ? IB's are a sealed alignment.

I haven't changed my tune; a bandpass or a acoustic lever is more efficient than a sealed box of the same volume.

Or to put it simply:

12 Shivas in bandpass boxes could get you more SPL than 12 shivas in sealed boxes. But who on earth is going to put 12 bandpass boxes in their house? Hence, IB is a compelling alternative.

All comparisons to the labhorns should really be made for a group of 6 labs, as that is how they were designed to be used. Admittedly some are getting good results with pairs in corner loading, but that is room / placement dependant. I actually have another 4 labhorns cut but not assembled in case I get a room big enough to squeeze 6 in. Not likely but I can dream

Cheers,

Rob.

You understand that 12 LAB12 woofers in sealed boxes will have nearly the same efficiency at 20hz as six Lab Subs right? The Lab Sub unloads at 32hz; below that point it's efficiency is little better than a sealed box. In fact Danley recommends a high pass at 35hz for this reason.

I'm not knocking the labsub; just saying that it's not designed for sub-35hz bass.

:: PB ::

 

[Patrick Bateman]

And here Beyma results:

Kudos for doing the modeling. I'm a little mystified by the results. I model horns using Martin King's FLH worksheet & mathcad. I generally see very large swings in response from different woofers.

IIRC, Martin's model doesn't take into account corner loading. I have a feeling that explains why two radically different woofers have similar response in hornresp.

 

[Patrick Bateman]

Hey Rob, we agree.... I remember the time when my sealed box subwoofer was having problems so I set out to figure out what was wrong. With the driver still receiving a signal, I removed the driver from the box. Just after the driver was removed, the upper bass suddenly improved dramatically in sound quality (of course the lower bass disappeared because of sound cancelling). The box was clearly the problem in letting the driver do its thing. Your small 300 liter boxes would have such a problem.

It's only natural to blame the box. Have you ever considered that the box you're sitting in is the problem, and not the box the speaker is in?

When you took the speaker out of the box, you changed the directivity of the speaker. It went from a monopole to a dipole. And guess what Rob's horn and your dipole both had in common? That's right, controlled directivity.

Once you learn to control the directivity, you can have high efficiency with the natural sound of a dipole.

I am convinced that the box is the problem. Horn subs attempt to avoid this problem by reducing the travel of the driver by the better coupling of the driver with the air. It still does have a box to deal with (the compression chamber), but less so because the driver does not have to move as much.

Infinite baffle speakers esentially does away with the box altogether by using a huge (infinite sized) box, hence the very large improvement in sound quality over smaller boxed subwoofers. I suspect that infinite baffle subwoofers will better horn subwoofers in sound quality, and allow for extension to very low frequencies if that is desired.

Retsel

To my ears, IB doesn't sound much better than sealed. But dipole and horn bass are definitely something special.

 

[GM]

Go ahead then. Fire up hornresp or Mathcad, and let's see how you make a horn that's efficient below the FS of the driver. How do you do that?

:: PB ::

GM, I just wiped a little drool off my chin

Seriously though, kudos for modeling the subs!

Greets!

You're welcome! The point I was trying to make though is that the horn is an acoustic lever, so if it expands at the ~correct rate and terminates into a large enough mouth for the desired BW the driver's size/Fs/efficiency is ~irrelevant WRT its LF gain BW. Both my designs will be efficient below the driver's Fs if correctly built.

Maybe a more extreme example will convince you assuming you trust MJK's programming......

RadioShack's discontinued 40-1271 8" ~fullrange driver:

Fs = 83 Hz
Re = 7.9 ohms
Le = 2.1 mH
Sd = 214.084 cm^2
BL = 3.032 N/A
Vas = 32.85 L
Qes = 3.3
Qms = 9.45
Qts = 2.446

With its high Fs and incredibly high Qes/Qts, hardly a candidate for a decent OB, much less a ~30 - 300 Hz bass-horn, or is it?

Vr = 6541.385"^3
Vf = 48.594"^3
L = 212.446"
St = 21.817"^2
Sm = 8014.947"^2
M = 1

For a variety of reasons it's not worth the time/$$ to build one, but my experience with smaller horns with even higher Fs drivers than this horn's Fc-Fs BW spread tells me it will work pretty much as predicted up to at least 300 Hz.

GM