A few days ago I posted a message stating that I had built a subwoofer with deep bass response for about $125. Apparently there was quite a bit of interest for this sort of thing, because I have since been bombarded with messages asking for details. Well, since I've become tired of replying individually to all these, I'll set forth what I did here so everyone who cares can take a look, while reproducing bits that also appear in the rec.audio FAQ (which is worth your time to track down, you can get a copy from rtfm.mit.edu or by looking for the monthly posting in one of the rec.audio groups).
If you are serious about wanting to find out more about building your own speakers, I'd suggest first contacting Old Colony Sound Lab at (603) 924-6371. Their catalog lists tons of books on that and many other audio related subjects. Good books for beginners include David Weems Designing, Building and Testing Your Own Speaker Systems and Vance Dickasons's Loudspeaker Design Cookbook. The company I prefer to buy speaker drivers and parts from is Madisound, (608) 831-3433. Their catalog is very useful browsing for the speaker novice, since they give a couple of project plans and include full recommendations for most of what they sell, at least as far as enclosure design goes. If you really don't know what you are doing yet, make sure you stick with a sealed box design instead of a ported one--you've got a much better chance of getting it right.
(FAQ mode off) I wanted to build a subwoofer with response extending well below 30hz. I also had a specific size constraint for where I was putting it (it had to fit into the trunk of my car so I could haul it around or maybe even use it there). And I wanted to build a sealed box. Getting a sealed box below 30hz is quite difficult and requires a very specific set of driver characteristics to pull off. The driver I ended up using is the Madisound 1252DVC. This is a dual-voice coil 12" woofer, each coil is 8 ohms and handles 50 watts or so (yes, I have bottomed it out, so I know that is an accurate specification). This means you can drive the woofer with a stereo amplifier and it will present an easy 8 ohm load, or you can connect the coils in parallel and use it as a mono 4 ohm woofer (perfect for most car amplifiers). The driver is very well constructed; I guess that a commercial design using it would run at least $500. Most sealed box designs end up having a bottom 3dB down response point that is approximately twice the resonant frequency of the woofer (fs) used; since the 1252 has an extremely low fs of 15hz, most of the correct sealed designs with it have a 3dB down (f3) point of about 30hz.
I used Madisound's design charts, because after measuring the woofer I received, it was right on compared with the specifications. The charts given suggest a box size of 85-130 liters (divide by 28 to get measurements in cubic feet). This corresponds to a system Qtc of .96-.8, and a corresponding f3 of 32.2-30hz. Normally, a Qtc this high would produce bass that was "boomy", but since the resonant frequency of these designs are in the 40hz range, it's not very objectionable or noticeable (a 1-2dB peak in response at something like 100hz is tough to take and annoying; your ear is so much less sensative to low level variations at the deep bass spectrum of things at 40hz that it's not really a problem, IMHO--that variation is swamped by, say, room gain at that frequency).
The box I built, in order to fit in my trunk, ended up
being about 80 liters, on the small side of things. To try and make this
work as well as possible, I also bought two Scan-Speak flow resistive
vents (i.e.
fiberglass holes, similar to the Dynaudio variovent).
Dropping these in the cabinet lowered the resonant frequency a few hertz,
making its effective volume 90 liters (accordingly, I'd have to agree with
Madisound's description of "lets the woofer behave as if its in a
larger enclosure", but I don't think it was necessarily a better
solution then stuffing the box a bunch). I'd definately recommend making
a box for
home audio much larger, at least 100 liters, and 150 is not
out of the question (the ported version of the design uses a 142 liter
cabinet to get an f3 of 26hz). My box ate most of a 4'x8' sheet of
3/4" plywood, along with a bunch of 2"x2" wood using for
internal bracing. The woofer was $42, the plywood around $30, input
terminals ran $11 for a nice two-terminal set, I used about $5 of liquid
nails and 1 1/4" drywall screws, and I got some carpeting from
Crutchfield for $8 to cover the whole mess. After gingerly moving it
around a few times trying not to poke out the cone, I picked up a 12"
woofer grille at Radio Shack for $15
or so. Add it all up, throw in
some shipping and handling, and you've got the whole thing for about
$125.
So, just what is it that you have, then? A subwoofer that compares pretty favorably with most on the market, capable of a wonderful Terminator 2 performance (which I show off about once a week as people visit). The remainder of the fun is trying to figure out how to power the beast.
Looking at its frequency response curve, the 1252DVC tries to reproduce up to 2Khz, a very good effort for a woofer of its size, but doomed to failure anyway. I recommend crossing it over at under 150hz; mine works quite happily at 80hz. The crossover is probably the trickiest part of this whole thing. Normally you'll be told you can just put an inductor in series with a woofer to cross it over and roll off the treble. What you're usually not told is that the inductor value required for subwoofers is really, really big (for our 80hz example, it's about 17mH). When you start talking about inductors this large, their inherent DC resistance becomes a problem (Madisound's 18mH inductor, for example, has a DCR of 1.5 ohms, which is very much not something you can ignore in your design). Plus, a subwoofer like this one has a bunch of strange things happening in its impedance around the frequencies it is being used at (there's that big hump around resonance). Accordingly, I cannot recomment using a passive crossover for this subwoofer (or any subwoofer, IMHO, but that's a different argument for other designs). You really need an active subwoofer crossover; check, your receiver may already have a line level subwoofer output on it that's crossoved over, if so you are set (this is becoming especially common on home-theater components, that's how I derive my signal). Next you feed that line level output to a dedicated channel or two of amplification. I use a dedicated high-headroom 40w/ch stereo amplifier; you might also want to consider using one channel of a stereo amp for the sub and the other for a center channel speaker. As previously noted, one advantage of the dual voice coil woofer is the amplifier flexibility it offers. I definately recommend at least 100w total of power for driving this design if you want to take full advantage of it, but more then that is really wasted.
So, if you have a subwoofer output on some piece of your equipment already, and have an extra 100w or so of amplification (possibly from your last equipment upgrade), building this subwoofer will really only cost you about $125. You can scale that up accordingly if you have to buy another amplifier to drive it or an active crossover (check out Marchand for electronic crossovers). What you definately can't do is just hook this woofer up to you existing system, say by using the secondary speaker outputs on your amplifier and turning on speakers B to enable the subwoofer (trust me on this one--I've tried it, and it sounds really bad, to use this woofer without crossing it over). Madisound's woofers offer a great quality for what they charge, and you can design bass systems with them that put all but the best commercial designs to shame if you put some work into thinking the design through and into quality construction techniques.
Feel free to pass this message around, repost it, put it on your WWW page, or whatever--there's nothing here that's very innovative, I'm just passing on some of my experience. As long as you keep my signature below and this paragraph intact, you can do whatever you want to the rest (I'm trying to grab a piece of my 15 minutes of fame if I can). Let me know if you do anything with this, I'd like to see if it shows up anywhere. What you don't want to do is try to send it to Speaker Builder, because I'm including parts of this discussion in an article I'll be submitting soon as a followup to my previous cheap home theater article.