Speaker Build: Squawker 2.0


Introduction

After using the Squawker for some time as a practice guitar speaker, I became interested in improving the design. I wanted to retain its idiosyncrasies, but shift the bandwidth downwards to gain some low mids and lose the ‘fizzy’ treble content that isn’t of much use for electric guitar. Among other things, this would require a larger baffle. Here’s why.


Acoustics of Open Baffles

Loudspeaker drivers produce roughly equivalent output from the front and rear as they move back and forth, but these acoustic emissions are 180 degrees out of phase. This isn’t an issue for the higher frequencies, which have very short wavelengths, but below a certain frequency, the wavelengths become long enough that the sound waves will wrap around the baffle as they disperse into the listening environment. These out-of-phase signals then begin to destructively interfere, causing output below this critical frequency to disappear at a slope of 6dB/octave. You can easily push this frequency lower by making the baffle larger, but at some point, the baffle becomes impractically large.

It then becomes more practical to put the driver in a box, which effectively extends the baffle backwards and around the driver rather than straight outwards, creating a more manageable physical footprint at the same time that it mitigates destructive interference between front and rear output. This is one of the fundamental reasons why people put speaker drivers in boxes - to trap the out-of-phase rear emissions and prevent them from blending with the desired output from the front of the driver. You don’t have to use a completely sealed box; many open-baffle designs include wings along the sides but remain open at the back, extending the baffle without converting the design into a proper ‘sealed’ topology. This comes with its own advantages and issues, which may be considered in a future post; Squawker is and will only ever be a single, open baffle.


Build Process

I went out for coffee one morning and noticed a discarded 10″ woofer in my neighbor’s garbage, still mounted on its weathered plywood baffle. The garbage truck was making its rounds and approaching rapidly, so I swooped right in. The paper cone had a small tear in it, maybe half an inch, but the driver looked like it was in otherwise good condition. With cautious optimism, I brought it home.

Googling around, I ascertained that was an American-made Capehart driver. Capehart manufactured a popular line of console record players (combined turntable, radio tuner, and loudspeaker arrangements in a single enclosure) through the 1930s and 1940s.

The kicker here is that these console players essentially used an open-baffle topology for the loudspeakers. While the drivers were mounted in the cabinet, the cabinet itself was at least partially open on most models, and didn’t conform to the design considerations that we see in modern sealed and vented topologies. (The mathematical approaches that inform these modern designs would be developed by A. Neville Thiele and Richard H. Small, two Australian broadcast engineers to whom we are deeply indebted, in the early 1960s.) As such, the woofer would have been manufactured to perform well in this nominally open-baffle role, and would have been expected to reproduce a passable bandwidth for full-spectrum musical reproduction in conjunction with a tweeter. Score!

I sliced up some whitewood 2x4s to make little feet that I screwed into the bottom of the baffle so it could stand upright. I installed a jack plate in the top corner of the baffle, with a 1/4″ TS jack to receive input from a power amplifier. To maintain a sort of genetic continuity, this jack was cannibalized from Squawker 1.0, thereby formally retiring the first-generation unit in favor of bigger and better things. I lacked a hole saw and didn’t want to break out a router so casually, so cutting the required 3″ hole became a laborious process involving a large drill bit, a coping saw, and a few hours of carefully working away at the rough cutout with a rasp while I crushed a couple beers to keep my spirits up. I glued the jack plate in place with silicone adhesive and wired the jack directly to the speaker terminals with zip cord.


Outcome

I used a cheap mic preamplifier with a DI input to bring my guitar signal up to line level, and fed this to an ancient 40W mono power amp from Radioshack. I’d estimate that the driver’s RMS power handling tops out at 30W @ 4Ω, its nominal impedance, so this arrangement works fine.

At 12 o’clock on the amp, it’s plenty loud for practice, recording, and even small gigs, and the driver enters a surprisingly smooth breakup when pushed harder. Compared to sealed and vented enclosures, open-baffle speakers can have a sort of blurry, ambient characteristic to the way that they fill a room. In this case, that’s exactly what I wanted.

With the baffle being roughly 2′ x 2′, the unit is able to reproduce frequencies down to about 200 Hz before phase cancellation occurs between the front and rear emissions, as described earlier. This is a substantial improvement over Squawker 1.0, which dropped off a cliff more than an octave above that, at 450 Hz. While the guitar’s lowest fundamental is around 82 Hz in E standard, the bottom octave is of limited use in many contexts, and is often removed from the mix to leave space for bass instruments like kick drum and electric bass guitar. The majority of the guitar’s useful low-end comes instead from the second harmonic (starting at around 165 Hz on the low E).

I haven’t measured the high-end response, but I’d estimate that it has substantial output up to about 3 kHz; a tad dark compared to a modern guitar driver, but I’m definitely not losing sleep over it. This is the Squawker I’d always imagined.


Conclusion

Being a hi-fi woofer intended for full-range music reproduction in the living room, it lacks the inherent coloration and compression that is often expected from a guitar speaker; it has a sort of neutral quality to it that makes it a good ‘clean slate’ for tone shaping with pedals and the like. When driven hard with distortion, it ‘barks’ nicely, and has a surprisingly articulate character compared to the 10″ Eminence guitar driver in my practice amp. It’s not that it’s better or worse - it just sounds different in a way that I get excited about. If the tear in the cone is introducing any nonlinearities (and I’m inclined to think that it is), then goddamn it, I love them, and I definitely don’t plan on mending the tear.

Pretty cool what you can toss together with a lucky pull from the garbage!