2235H as subwoofer, again

I know this topic's been hashed out quite a bit, because I've been reading all o the threads containing "2235*" and "subwoofer". I took in those threads and did some modeling now that there's a PC in the house again. (Hey, it's at least competent for running Unibox, even if the Powerbook is better for everything else!) One alignment that seems almost too good to be true I'll call the GW alignment after Gordon Waters, who suggested it here.

According to Unibox, a single 2235H in a 160L vented enclosure tuned to 25Hz with a 15.24cm diameter (i.e. 6" round) and 42.9cm (about 1'7") port gives an F3 of 29.1Hz and an F10 of 21Hz (re: 80Hz, which is the Fc of my active crossover). Moreover, assuming no power compression - not an entirely invalid assumption, given that these subs are for a domestic system that won't often be often run at full tilt for long periods of time - they will give me 104dB at 20Hz and simply will not hit their mechanical limits until under 20Hz when fed 200W. (Obviously, a highpass at 20-25Hz would be required to protect the woofers.) Two of them should do 110dB at 20Hz, which as a practical matter means that they'll be loafing when playing real music.

Given the source I'd expect it's one of, if not THE, best alignments for the 2235H in a modestly powered (by modern standards, at least) home audio system. But surely I must have something wrong, because all that looks too good to be true. I double checked the T/S parameters I entered into the spreadsheet, and they're all exactly as on the JBL document. So what's going to turn this alignment into a pumpkin at midnight? I really don't care about port ripples at 380Hz....Unibox doesn't show me group delay, for instance...

BB6P suggests two alignments. The one in black is basically B380 w/ BX63(A). Vb = 117.6 l., Fb = 24 Hz, F3 = ummm, 26 Hz. :p

The one in blue is Vb = 187.5 l, Fb = 22 Hz, F3 = 46.85 Hz.

Gordon's alignment in red, assuming I entered the design parameters correctly. BB6P says F3 is 39.43 Hz at 6 cuft. (170 l.), and 40.27 Hz. at your slightly smaller 160 l.

Interestingly, B380 has better group delay, and is not excursion limited at all at rated 150 W. It hits 8 mm. at 20 Hz when driven with 200 W.

See also: http://www.lansingheritage.org/html/.../1983-subs.htm

Quote:

---

Originally Posted by DS-21

Given the source I'd expect it's one of, if not THE, best alignments for the 2235H in a modestly powered (by modern standards, at least) home audio system.

---

D.B. Keele suggested a 4.55 cu ft volume tuned to ~ 29 Hz as the naturally flat alignment for a 2235H. This alignment should be considered the best compromise. End users can deviate from that specification as desired. For more background information on the B380 read the technical brief for the B460 on this website as well as "A New Set of Sixth-Order Vented-Box Loudspeaker System Alignments" by D. B. Keele. Note that the B380 and B460 are actually fifth-order systems. The group delay in those systems is affected by the BX63/BX63A.

The alignment suggested by Gordon Waters works fine for typical home use.

Thanks for the gut check! It's interesting that the BB6 and Unibox give slightly different results. I didn't mention it but I'm pretty sure my models include a light dose of polyfill inside the box, as show in the upper right corner of the Unibox output. Does BB6 Pro account for box losses and such? That said, on paper I like the GW alignment better than the BB6P recommended one, because Gordon's has a little bit more area under the usable curve, and the advantages of the larger box/lower tuning seem to be primarily in the region below the Fc of a reasonable high-pass...

I have read the B460/B380 article thanks to links in earlier threads, but at this point I'm not interested in adding another box to my signal chain. Speakers can be furniture, but black metal boxes are just clutter. (That's the main reason this iteration of my system has gone from almost two vertical feet of separates to a 4.25" high digital receiver with a pair of 1.75" high monoblocs for the sub.) Still, I probably will break down and buy a Behringer Feedback Destroyer at some point in the future. That box will give me the ability to match the EQ curve of the BX63A or any other fixed EQ, but not in the immediate future. Besides, honestly I'm not sure that a 26Hz F3 a good fit for this room. Its transfer function is such that an F3 in the mid-30's with a shallow rolloff below that point is optimal. I suppose that means eventually I'm going to have to buy a pair of JBL W15GTi "car" woofers, which model just about perfectly for this room in large sealed boxes. And model wretchedly for any car install of which I can conceive...

Quote:

---

Originally Posted by DS-21

Thanks for the gut check! It's interesting that the BB6 and Unibox give slightly different results. I didn't mention it but I'm pretty sure my models include a light dose of polyfill inside the box, as show in the upper right corner of the Unibox output. Does BB6 Pro account for box losses and such?

---

Yes, those are all "normal" damping, i.e., fiberglass on all but the front baffle.

Please understand, I'm not advocating any particular alignment in the presentation above. Like you, I thought it interesting to compare results from different box programs, and to see how the performance of the same driver can be "tuned" by the box design.

I think I'd go with Gordon's design in your case, as well, now that we know a bit more about it. :)

The alignment I did, should have marginally better group delay than the 187 liter box. I've been reading around, stuff by Dan Wiggins of Adire Audio and such... and I'm starting to become more reticent to recommend huge boxes for drivers, due to the protraction of group delay errors.

But, that said, in a typical small/medium room, the alignment I proposed, will give a good "match" with typical "room gain". In contrast, the B380, while great in a large or open space, has a much sharper "knee" to the response curve, and as such, will commonly get "overly bassy" in the 30-40 Hz region, while not having as much "percievable" VLF output (20-25 Hz) as my alignment, despite the lower f3.

Basically, making the rolloff as shallow and as "smooth" (ie, freedom from kinks in the rolloff curve) as possible, seems to be the best overall plan for most home use. If you wanted to still "smooth" the curve further, it's possible by just using a slightly lower port frequency. However, I still think that the tuning given, is probably the best compromise between group delay and in-room response integration...

Regards,
Gordon.

Very nice post Gordon.