Summer 2010

Pass The Bofu
Building a transmission line cabinet for the Pioneer 8-inch full range driver.
Article By Jeff Poth

Difficulty Level

  Now, dear reader, you may wonder how on earth one passes a Bofu, or indeed, what a Bofu is. The more faithful amongst you may remember from Modifying Speaker Drivers that the Bofu is a popular Pioneer 8-inch full range driver. This is a transmission line cabinet, one of a few styles suitable for this moderate efficiency, higher Qts driver. More to the point this line was designed by one Mr. Nelson Pass who you all may know from Threshold, Pass Labs, and First Watt. Indeed Nelson has participated in these very pages and is a huge proponent of DIYers the world over. Dave Dlugos of Planet10 hi-fi provides this PDF drawing for all cabinet parts needed. Both gentlemen gave their blessings on the use of their IP for this article, thanks guys! This design is a nice balance of cabinet and cost and size hence my interest in the project. I was already engaged in a more unique build with this driver, a small low-tuned box designed to provide a little boost for baffle step compensation. For most of us, the pass TL is the more worthwhile design. For interest, the small box is below- slot tuned to 25Hz and a Vb of about 25L.

Isn't This A DIY Mag?
This Is Someone Else's Design!
Well, there's not a lot new under the sun in speaker design. But, as you'll see, we have some (optional) flavor added. Amongst other things, the construction received a number of enhancements as well as being "beefed up". The box received a modified low diffraction baffle that also provides a modicum of horn loading, bracing not in the original design, and a removable base. Panel thickness was also increased. The removable base allows for stuffing variation, an important factor for optimizing the enclosure for various rooms. It does require a redesign of some panel sizes, however. All of these are optional, but definitely keep the bottom panel removable for adjusting stuffing, and the bracing is very easy to incorporate and buys you quite a bit of benefit in enclosure stiffness.

Transmission line enclosures and their predecessor the Acoustic Labyrinth have been around since the dawn of loudspeakers, but in the past 20 some odd years they've become much better understood thanks to the efforts of Martin King, George Augsburger, Dave Dlugos and others. A transmission line operates as a resonating enclosure, typically operating at quarter-wavelength to reinforce the output from the driver. This is similar in effect to a bass reflex but a transmission line has a little different behavior due to being a wavelength resonator rather than a mass resonant system (though it's often hybridized). Effectively we get some suppression of cone motion, deeper bass, and changes to the cone loading. TLs are known for having clean, smooth bass extension. For more details go to Mr. King's site at quarter-wave.com, t-linespeakers.org or Google.

Box Construction
The basic construction is made of generic decent quality 0.75-inch ABX ply. The front panel is of Lowes "Arauco" premium ply and the sheathing is 0.25-inch oak-veneered premium ply. The base and "Suprabaffle" is 13-ply Baltic Birch ply. I like using multiple materials not only because it lets me use up leftovers and scraps from other projects but because any resonant characteristics of the materials will be averaged out to a lower overall level of coloration. Construction was exceedingly simple. I used butt joints for everything, made sure the bracing rods were tightly affixed, and used right-angle clamps to help with the alignment during glue-up.

Several parts of this used laminated construction. The removable base, the "Suprabaffle", and the oak veneer panels all required lamination. I used regular Titebond II woodglue and a lot of clamps and scrap wood to ensure a proper lamination. The idea is to have a complete coating of woodglue, and force applied evenly on the whole surface. Give glue extra time to dry when using it for lamination purposes. I like to go 24 hours. This is because the glue moisture in the center of the laminate must travel much further to evaporate than on a normal 0.75 to 1.5 inch wide glue line. Nothing's worse than a delaminating panel that buzzes, so be patient. Once all was assembled a router and flush trim bit was used to clean up the edges.

Suprabaffle
You can design your own Suprabaffle, however you like, but it does affect response. The idea is to increase the apparent baffle width, and/or control baffle edge diffraction. The more softly the transition occurs from forward radiation to omnidirectional (as the wavefront wraps around the cabinet edge), the less "ripple", or frequency response roughness there is. This is why you see roundovers and bevels on the edges of so many speakers. Bevels are slightly less effective than roundovers but either is much better than a 90 degree angle. Size matters; the larger the roundover/bevel you can apply the better. 0.25-inch does extremely little for example. Increases in the width of the baffle lower the frequency at which this takes place. Drivers tend to become more directional with increasing frequency, so the very common "skinny baffle" approach can be useful for minimizing diffraction too. It's a complex issue and this is as deep as we'll delve here. I can recommend the shape shown, but fabricating it is something for more advanced woodworkers.

A "normal" build of this cabinet would have simply a circular hole via a router and circle cutting jig, or a jigsaw. One could flush-mount the driver with the router. The baffle used was done using 1.5 inch laminated Baltic Birch ply, and the shaping done with a router with a chamfer bit. I pre-drilled holes on the face of the baffle with forstner bits before chamfering the driver cutout, so as to be able to mount short pieces of dowel rod with T-Nuts (see discussion of T-nuts here). This allows a secure, repeatable mounting of the driver from behind the baffle, with minimal additional edge diffraction. Better still from a diffraction perspective would be to use screws that don't breach the interior bevel, but they would have to be extremely short and thus lacking in strength. Suprabaffles were lined up and drilled to allow repeatable mounting to the rest of the cabinet, again, using T-nuts. With the driver mounting from behind the baffle using strong repeatable fasteners is crucial, otherwise a driver failure would mean you'd have to junk the cabinet. After sanding I sprayed with appliance epoxy in black. I'm a fan of this finish, as it's extremely durable and can look very good when done correctly. Unfortunately I haven't managed a perfect result yet, but I'll get there sometime.

Bracing
Braces were simple strips of ply, placed to achieve specific goals. There are strips in the corners of the folds (where the line bends) to act as diffraction devices. These scatter high frequencies within the line to help ensure they don't make it out of the end of the vent. Braces were also placed along all edges to help stiffen the front, top, and rear panels. Finally braces were incorporated into the front section to couple the side panels to each other, where the panels would be most likely to have resonant modes (large sections of panel with nothing connected to them, read: the middle of the panels). I also used these to provide additional surface area for the removable base to rest upon.

Removable Base
The base is three layers 0.75-inch ply, one large Baltic Birch layer to act as a plinth (stable base) for the speaker, and two layers of the cheap stuff that are carefully cut to size to fit inside the four walls of the enclosure. You can see gasketing applied to the bottom edges of the main enclosure assembly.

One could do the removable bottom differently, but this design provides a lot of ballast to guarantee the enclosure doesn't tip over if some kid pretending to be a rhinoceros rams into it like the little monsters did to my poor Zigmahornets.

Now What???
Just a quick impedance plot.

The Frequency response plot is contained in the PDF at the start of the article. This doesn't quite capture the amount of extended bass available from the enclosure due to measurement technique. One can expect solid response to 35Hz and even a bit below. These are amongst the few designs I've worked with that are fine without a subwoofer.  You can see that as built, the TL operates at about 25 Hz (the impedance minimum (the valley between the peaks)). This is nice and low, higher tunings tend to have more audible "group delay", which is the lag between the woofer output and that of the line. The lower the tuning, the lower in frequency group delay becomes high, and the less audible it is. I didn't notice any significant disconnect in the sonic signature, where badly tuned TLs and bass reflex boxes will often sound sloppy.

Yeah, that's right, the sound. It's what it should be. These are very nice loudspeakers. I started using them with the stock B20 but moved to a heavily modified version as seen at this link. Currently I'm running them with phase plugs and treated per the article, but also have tried a coaxial tweeter mounting. These could use assistance from a supertweeter for the very top end, but I rather enjoy them for what they are. These do not have a lot of top end but are extended enough to be satisfying. More importantly, the midrange is superbly rendered, with a rich, musical character and very impressive dynamic authority. They can reproduce realistic levels and even when exceeding their limits, they never sound harsh or strained. Bass is a big, ballsy affair, and you will find yourself itching to get up and boogie.

Nuff Said.