[Duke]

A number of people have asked me to make a sealed version of the Swarm, and I've been looking for ways to lower the price of the Swarm.  At present, the vented Swarm 2.0 is $2500 for the set, and even then I'm not making much on 'em. 

I think I've found a solution.  The sealed Swarm, let's dub it "version 3.0", will come in around $1700, and will include the same 1000-watt amp as version 2.0. 

Version 3.0 will trade off some bottom end extension and about 6 dB of headroom relative to 2.0, and won't have a nice veneered finish.  Footprint will also be larger.  But it will be $800 less expensive, and these days that counts for something.   Driver will be a 10" woofer with enough linear excursion to handle more power per woofer down to 20 Hz than the kilowatt amp can put out.  The enclosure will be a Qtc = .5 sealed box.  The anechoic -3 dB point is about 50 Hz, and the in-room -3 dB point is in the upper 20's.  Excursion is sufficient that it's probably safe to use the amp's single band of parametric EQ to boost the bottom octave a bit, if desired.  Sound quality should be very competitive even before the advantages of scattered multisubs come into play. 

Duke

[launche]

This is good news Duke.

Please post pictures of the new system when available.


Thanks

[Duke]

Thanks, launche.

Pics are probably a couple of weeks away.  They'll be squattier, plainer boxes than the Swarm 2.0, but that mini-tower format is a large part of what was driving the price. 

[Duke]

Here's what a quartet of Sealed Swarm subs looks like:



The amplifier is a shelf-mount 1,000 watt unit that can be switched between 110 and 220 volt operation and features a 4th order low-pass filter that can be varied between 30 and 200 Hz; a single band of parametric EQ; bass boost switch; and continuously-variable phase control.  It does not have speaker-level inputs, but I can provide an adaptor for $66.00.

From the Sealed Swarm page that will be uploaded to my website soon:

The less expensive sealed version of the Swarm follows the same basic principles as the vented version, though inevitably it can't quite match the gentle 3 dB per octave rolloff of the low-tuned vented boxes. Still, the low-Q sealed box has excellent impact and pitch definition, and its extension is sufficient for most music applications. The parametric EQ in the amplifier can safely be used to extend the bottom end, as the drivers can handle the full output of the amp down to 20 Hz without bottoming out.

Specifications:

Type: Four-piece powered multisub system
Woofer: 10" long-throw, rated at 600 watts each
Tuning:  Sealed box, Qtc = .5
Subwoofer Amp: 1000 watts, 4th order filter (30-200 Hz variable), 1-band parametric EQ
Typical System Bandwidth: 28 Hz to 200 Hz in-room (before EQ)
Dimensions:  14" tall by 14" wide by 14" deep
Weight: 35 pounds each + 30 pound amplifier
Pricing: $1700 in black wrinkletex finish, direct sales only

[TRADERXFAN]

What is "wrinkletex" finish?  And what is the sensitivity?

-Tony

[Duke]

The finish is wrinkly; not smooth.  It's a spray-on finish that keeps the cost down.  It's similar to this, but I don't know if it's the same product:

http://www.speakerhardware.com/prodimages/dt_closeup.gif

Parts Express uses the same finish on their kit subwoofers boxes.

Each 4-ohm sub module has a calculated efficiency of 82.3 dB/1 watt/1 meter, but the rolloff has already started at 100 Hz so a lower number, like 81 dB, might be more realistic.  The theoretical combined efficiency of the four subs would be 6 dB higher.  Gain from boundary reinforcement is probably about 3 dB, but then we lose 2 or 3 dB because with the recommended placement their outputs are adding in semi-random phase rather than in-phase.  Below the frequency of the lowest room mode their outputs will add in-phase.  My guess at a reasonable real-world estimate for the quartet would be 86 dB/1 watt/1 meter.

86 dB efficiency driven by a thousand watts implies 116 dB capability, right?  Well, not really.  While the subs have adequate mechanical excursion to not bottom out, they will have some thermal compression despite their combined power handling capacity of 2400 watts.  So 114 dB is more realistic.  Also, some bass boost would be required to be "flat" to 25 Hz in-room because the 6 dB per octave rolloff of a low-Q sealed box is steeper than "room gain" can offset.   I would expect the practical upper limit at 25 Hz to be around 110 dB.

[JoshK]

Is this a 1 channel amp?  If so, are you wiring the subs together in series/parrallel?  I am curious about the semi-random phase thing.  What exactly does that mean?

[Duke]

Josh, yes it's a one-channel amp with the subs wired in series-parallel.  A second amp could be added for stereo operation, in which case we'd wire each pair of subs in series because the amp is only rated down to 4 ohms.

The semi-random phase addition of their outputs is a result of spreading the subs around the room asymmetrically.   At frequencies where the distances between subs is greater than 1/4 wavelength, their outputs won't be entirely in-phase, but they won't be entirely out-of-phase either.  Maybe semi-random phase is the wrong wording.  In contrast if you cluster the four subs together their outputs will add in-phase throughout the bass region - maximizing output but trading off in-room bass smoothness.  Of course to maximize output a far more cost-effective solution would be to use a single large ubersub.

[JoshK]

Interesting.  I guess I hadn't completely thought through the distributed bass in terms of the phases.  Makes sense.  For some reason I thought you needed a way to control the phase of each sub to make them in phase but that wouldn't work at more than one freq or more than one spot.

[Duke]

Josh, I was mostly trying to describe the factors that come into play when trying to figure out the system's approximate real-world efficiency (Traderxfan used the word "sensitivity", which in my mind implies voltage sensitivity referenced to 2.83 volts input, and I took the liberty of converting back to "efficiency" which is a misnomer but the convention is to use that word when describing the 1 watt/1 meter SPL).  Anyway normally a doubling of drivers raises the efficiency by 3 dB, so two doublings would be 6 dB, but this assumes their outputs add in-phase; if they added in random phase (which I think more closely approximates the behavior of scattered multisubs in a normal-sized room) two doublings of drivers would only raise the efficiency by 3 dB.

Note that it takes the ear at least one cycle to even detect the presence of bass energy, and more than one cycle to hear pitch, which implies that in typical rooms our ears cannot differentiate the first-arrival energy from early room reflections in the bass region (unlike higher up, where the ear processes the first-arrival sound and subsequent reflections differently).  So in the bass region, any phase coherence in the original signal has been destroyed by room reflections by the time we begin to hear it. 

[JoshK]

Well, I am thankful, as you made it much clearer in my mind. 

[TRADERXFAN]

Me too.   
quite in-depth explanation.

-Tony

[poseidonsvoice]

Duke,

Me 3.

What 10 inch driver are you using? Any links to the amp/parametric EQ unit? $1700 may seem like a lot to some people, but realize this is (4) subwoofers completely finished, plus the amp/parameteric EQ, plus the expertise of Duke, plus the multisub/Floyd E. Toole/Earl Geddes philosophy all thrown in.

I like this better than your original 6.5 inch ported version.

You have a winner 

Anand.

[Duke]

Thank you all very much!

Anand, here's the amp that I use:

http://www.parts-express.com/pe/showdetl.cfm?Partnumber=300-811

I'd rather keep the woofer I use under wraps for now, but will say that it has a fairly advanced motor for a subwoofer driver.  Much more than that and you DIY guys would have all my secrets! 

One feature of my boxes that is of minor interest is the inputs are on the bottom.  This way, the sub can be turned so the grille faces the wall and you don't see input cables hanging out the back.  This gives a little bit more flexibility in achieving asymmetrical placement, as two of the subs can be placed woofer-to-the-wall and the other two can be placed woofer-out.  If the room lighting is a bit dim, it's not even obvious that the grill is sometimes facing you and sometimes not. 

I figure my competition would be four small subs that have a 4th order lowpass filter and a low-Q sealed box characteristic or room-gain-complementary vented box alignment, along with a bit of EQ capability.  At this point I don't know of any subs in the $425 ballpark meeting this description. 

Note that Geddes' subwoofer integration protocol is quite different, being optimized to get a smooth blend with his very-low-Q sealed box mains.  I'm going for steeper slopes in the crossover region and the ability to integrate with a fairly wide variety of mains, but his technique is optimum for his mains.

Comparing the 10" sealed Swarm with the original 6.5" vented Swarm, the original went deeper without equalization and its frequency response curve was a better complement to typical boundary-reinforcement-induced gain in the bass region.  The sealed version has better impact on kickdrum an such, and bass boost feature of the amp along with the parametric EQ (neither of which were available on the original) gives more real-world flexibility in tailoring the sub to its environment.   I still make the 8" vented Swarm 2.0, which plays deeper and louder than the sealed Swarm and has a smaller footprint, but it's also eight hundred bucks more.

Duke