Subs in small boxes - why & how?

[joyspring]

Dave H.

"To me a large efficient driver pushed by a very clean Class A or OTL amp probably has to be a cleaner sound than a driver trying to contain 2700 watts."

Not if the large, efficient driver is exceeding its linear excursion capability because of its underhung voice coil; that was one of the major points of my last post.

When the voice coil travels outside of the voice coil gap and magnetic flux field, the result is distortion. No Ands, Ifs or Buts. (but, but...).

From the Altec Thiele-Small parameters available at Todd's Unofficial Altec Lansing site, the 416s have an Xmax of 0.15" (3.8 mm) and a Vd of 19.2 in^3.

The NHT 1259 (a popular yet fairly inexpensive 12" subwoofer driver with an overhung voice coil) has an Xmax of 13 mm and a Vd exceeding 30 in^3.

The diminutive and cheap 12" driver can move nearly twice the air than the mighty 15" 416 can, all WITHIN its linear range.

Now WHICH driver can produce greater undistorted output at very low frequency?

BTW, the 12" driver in the Sunfire sub boasts a Vd greater than 159 in^3. That is greater than TWO Altec 3182s. That's right; Bob Carver's little 13 in cube can theoretically (and practically) produce greater acoustic output at extremely low frequency than an Altec 3182 in a 24 ft^3 enclosure!

Do the math -- it does work out.

Lastly, a Class A amp for a subwoofer? Even with 100+ dB 1 W /1 M LF drivers, you will easily reach the power supply rails of the amp resulting in ... distortion.

Remember the logarithmic relationship between input electrical power and output acoustical power?

Ten watts does not go a very far for any subwoofer, especially with modern high dynamic range digital media that may easily exceed 20 db peak-to-average and considering the extended LF content present in modern recordings.

2700 watts from Carver's Tracking Downconverter amplifier is an excellent and efficient means of preventing distortion; coupled with a driver that can actually produce high sound pressure levels at extremely low frequency, this easily translates into high-quality and low distortion sound.

BobR

[joyspring]

Ooops,

Computational error in the previous post; the 3182 has a Vd of 88 in^3, which is over half that of my estimate for the Sunfire 12" driver.

Still, the fact that a little 13" cube can match or exceed the undistorted acoustic output of a 3182 in a 24 ft^3 enclosure is rather staggering. I know which I'd rather have in my home / control room.

BobR

[dhook17322]

The Carver sub is an amazing product in a number of ways, to get 2 1/2" of travel from a ~12" driver is impressive. Their cost is also surprising (~$1130), not nearly as expensive as I would have expected. Your claim that they should have an order of magnitude less distortion sounds unrealistic to me especially considering they have a 2700 watt amp. To me a large efficient driver pushed by a very clean Class A or OTL amp probably has to be a cleaner sound than a driver trying to contain 2700 watts.
Not having heard them, I can't speak from experience. It seems like people are returning to older technology to achieve the quality they seek. I hope to do some experimenting this winter in comparing a Maelsstrom in their largest recommended enclosure to a 3182 in the proper vented cab. Have a pair of A-7's in my living room and the presence they achieve from a 10w/channel Marantz is startling to many people and very satisfying to me. Hope to experience 210's later this summer.
Best regards, Dave

[taudo]

Wow, a great deal of interesting info! Thanks, guys!

I seriously considered building a Bag End Infra 18" into my 828's - but decided to try using new 2235's first. If the result isn't satisfactory, the 2235's will go into some BR cabs I have - or perhaps into that experimental small-footprint VOTT I'm thinking about.

Speaking of which, has anyone built such a system - using a modern little cabinet with an Altec 811B/1"driver combo? (It looks like the same idea behind the AvantGarde systems.)

On eBay, there is a German firm selling new 1" Altec/400 Hz tractrix horns for a song. Might be an interesting experiment. I'm tempted... (as if I have time for _another_ project!)

taudo

[joyspring]

Taudo,

You ought to seriously consider a computer-aided loudspeaker enclosure modelling program (i.e. - LEAP) or consider consulting with a local enthusiast who has access to enclosure design software.

All can offer very accurate small-signal response calculation but more importantly, estimate large-signal behaviour where at very low frequency, where those calculations become really significant.

With accurate driver and enclosure data, you can estimate maximum displacement-limited and power-limited output inside and outside the passband. For example, so what if you can tune a 416 down to 24 Hz in a vented, 9ft^3 enclosure if it will only produce 80 dB SPL before hitting its linear displacement limits... hopefully you're catching my drift ;-)

While final fine-tuning is invariably necessary, computer-aided modelling will quickly point out what is practical and what is a pipe dream WITHOUT actually spending the time and money to build and measure a system.

You may want to check out the book Bullock on Boxes by Miami University mathematics professor Robert Bullock III (who recently passed away, RIP). Inside is an Apple II program BoxResponse which can calculate the displacement and power limited acoustic output of second-, fourth- and sixth-order alignments. FWIW, I ported it to Pascal and later to C in the early 1990s -- I'll try to dig up the source code, provided I can find it.

Best of luck with your subwoofer project; the JBL 2235s are very capable drivers though they're a bit large (and require a larger enclosure) for the performance they offer compared to some of the newer, sub-specific 12" drivers, particularly the NHT 1259.

BobR

[taudo]

BobR,

Thanks for the info - I find your input invaluable!

I have used a couple of modeling programs available on the web. Indeed, that is why I decided to try the 2235 instead of the 2245 most prefer. The modeling program indicated to me that the smaller speaker would lower, better response in the 7 cu ft available to us in the bottom of the 828.

WRT surprising 12" subs, right now I have a couple of Sony HT subs working in tandem with my A7's. I haven't pulled one apart to determine what driver is in it, but the things really surprised me when I hooked them up! It actually requires very low settings of gain and crossover frequency to blend with the A7's. The 12's are perfectly capable of supplying _way_ too much bass if set too "high". Set right, however, they sound far better than I ever expected! Their low end augmentation really complemented and improved my VOTTs. That demonstration convinced me to build in the 2235s.

I also took your advice and got some good convection cooled pro-audio amps to power these things. So far, I'm quite pleased!

Thanks again!

taudo

[GM]

>I have used a couple of modeling programs available on the web. Indeed, that is why I decided to try the 2235 instead of the 2245 most prefer. The modeling program indicated to me that the smaller speaker would lower, better response in the 7 cu ft available to us in the bottom of the 828.
====
Hmm, both the various modeling programs I use and real world experience tells me the 2245H 'blows the doors off' a 2235H. Not that the 2235H isn't a fine driver, but materials technology has come along ways in 20yrs and you can hear/measure the improvements.

A pity the Aura 1808 is no longer available as it will 'pound' a 2245H 'flat' in every way but price, and with superior SQ to boot. I built two subs using these for a large HT and we measured 108dB/16Hz and >126dB/100Hz (pegged the meter) 45ft from the screen.

GM

[joyspring]

GM:

The Aurasound drivers are excellent, and the 1808's replacement - the NRT18-8 - has TWICE the Xmax of the JBL 2245H.

However, the Aurasound NS18-992-4A is particularly interesting:

Aurasound NS18-992-4A specs (pdf)

The Xmax is 50mm! It is of course a higher-Qts driver and has lower sensitivity but is capable of far greater undistorted output at sub-30hz range (the distortion curve on the .pdf spec sheet bears this out).

The somewhat smaller Vas will offset total enclosure volume though it does appear upon cursory examination of the Thiele-Small parameters that this driver would do best in a slightly overdamped system and yield excellent very low frequency performance.

It would require a very large amplifier though judging from the specs and guesstimation would even surpass the JBL 2242H in maximum undistorted very low frequency SPL.

BobR

[GM]

BR,

Yeah, both of these are better suited to sealed than the 1808, espcially the -4A, but both require either dual or quad large displacement PRs if vented, while the 1808 can eek by on a long vent, and with its low Qts, far better damped. That, or a very low/steep XO point to keep pipe resonances out of the passband. Factor in pricing/efficiency and neither of these make a whole lot of marketing sense to me. Since I haven't seen them listed in any system, I wonder how many of either they've sold Vs the 1808.

GM