The Attic Subwoofer was created to use up four 8" drivers that I had laying around from the defunct 4-Active Tower speaker project (Seas L21 RNX/P click here for data sheet) The design goals were:
   (A) to cover the lowest audible octave (e.g. down to 20Hz) and
   (B) to have a relatively flat frequency response because the listening area is large with vaulted ceilings and room gain is minimal.

Luckily I had at my disposal a large adjoining attic space so there was no restriction on the internal volume of the cabinet. Can't we all be so lucky! After doing some Unibox modeling I discovered that I could indeed meet my design goals with these drivers using a 600 liter volume (21.2 cubic feet). Wow, that's big!

Below are some plots of the initial UniBox modeling of the system:
This was looking good so far, except I didn't like the rapid increase in cone excursion below 20 Hz. I recalled reading and article about filter-assisted vented alignments so I decided to see what would happen if I combined this vented response with an active filter.

I tried second and fourth order Butterworth filters and found that, with a second order high-pass filter at 20 Hz, I could get the following performance:
This was looking good! By adding the 2nd order high pass filter I could reduce the infrasonic cone excursion. This would, in essence, protect the drivers from getting damaged from ultra-low frequencies. I also liked the look of the nice and symmetric cone excursion plot around the tuning frequency. With this alignment I could take advantage of the high efficiency and passband extension of a vented system and get pretty good extension for an 8-inch driver. Cool!

But... How How to build a 20+ cubic foot box??? How should I put the box in the attic but have the drivers and ports flush with the wall, playing into the room??? I solved this dilemma by designing a narrow baffle that stuck out from one corner of the box. I planned to mount the drivers and ports in the narrow baffle and fit only this part between two studs of the attic space, through a hole that I would cut into the drywall, and into the living area at about 8 feet off of the ground.

So I had a rough plan for the location. What about bracing the box? I'm someone who likes to try new things, so I decided that I would create the internal bracing from a network of PVC tubes that would be joined together with glue and attached to the walls of the enclosure. With a network of lots of tubes, I should be able to provide lots of stiffness but would not add too much weight. Oh, did I mention that I had to construct the sub in pieces and then assemble them together in the attic space? What fun! Here are some photos of the construction journey:
MDF and particleboard pieces. Testing the fit using tape to hold it together in the garage. One of the Seas L21 RNX/P drivers is sitting on top.
Looking down into the yawning cavern...
Building some pipe connectors from blocks cut from 2x4s and starting to connect the bracing system.
Pipe sections connecting the front baffle required some special bends. At this point my fingers are blue from all that PVC primer! At right is the detail of the supports for the three port tubes.
More partial assembly photos...
Up in the attic, starting to put it all together...
By cutting the drywall just wide enough to let the baffle slip through, the rest of the enclosure was hidden from the living area view.
I used this sub for several years and enjoyed the fact that it didn't take up any floor room. The response was pretty good and you could definitely appreciate the low octave. The system wouldn't blow you away becausae it could produce only about 105-108dB before the drivers were excursion limited. I still found this level of output to be very satisfying.

There were only a couple of annoying problems that I encountered. The PVC bracing system didn't really turn out as good as I had hoped. Cutting and gluing all of those pieces together took a loooooong time. Also, I wasn't able to find an adhesive that would stick to both the PVC and the wood blocks and give me a really strong bond. None of these really turned out to be too much of a problem in the end.

After extended listening I discovered that internal box resonances could be heard through the ports. This wasn't too much of a surprise because I didn't bother to add any stuffing on the walls. Since the box was so huge, a tone around 100Hz-150Hz could be heard, droning on. I knew that this couldn't be a port resonance because the diameter of the ports was large enough to keep the vent velocity quite low.  I never bothered to open up the box and add stuffing and I just lived with the resonance issues since they really were not all that noticeable.

All in all, this project was a great success!
If you want to learn more about "filter assisted vented alignments" like the one I used in this project, I suggest reading a great article at the Elliot Sound Products web site.

If you have access to the Journal of the Audio Engineering Society, you might find the following to be interesting as well:

Keele, Jr., D. B, "A New Set of Sixth-Order Vented-Box Loudspeaker System Alignments", JAES volume 49 (August 1974).

Tim J. Mellow, "A New Set of Fifth and Sixth-Order Vented-Box Loudspeaker System Alignments using a Loudspeaker-Enclosure Matching Filter: Part II", Audio Engineering Society Convention Paper, Presented at the 112th Convention, May 10-13 2002, Munich, Germany.