The HOSS project was initially conceived as an attempt to match the sound quality and performance of my first omni project, yet with much improved aesthetics. After all, my wife and I have to live with these speakers and I felt they should look like a piece of fine furniture-a work of art both sonically and aesthetically.

Technically, this design is best described as a Hybrid Omnidirectional Speaker System (HOSS), not a true omni. I refer to it as a hybrid because it is omnidirectional for only a portion of its frequency range: through the crossover to the tweeter, at which point it transitions to a direct firing monopole. I believe the hybrid concept combines the best features of both omni and monopole speakers. One of the knocks often made on omni speakers is that they don't image well. The HOSS's transition from omnidirectional in the bass and lower midrange to a monopole upper midrange and treble contributes to a large, deep, and lifelike soundstage, while maintaining the excellent imaging of a small monopole speaker.

I decided on a modular, multi-box approach, with a sealed MTM sitting on bass bins. The sealed MTMs might work as stand-alone speakers, when combined with a single sub or perhaps even as surrounds. A modular system is also easier move and/or transport! Another goal was to position the drivers so their voice coils are "time aligned" to the listening position, with a seated ear height of 37" high, 11' away. I hoped this quasi "time alignment" effort would facilitate phase integration between the drivers.

I wanted to use high quality, yet affordable and readily available drivers and crossover parts. My search ended with the Dayton Audio Reference Series drivers, which are hard to beat from a cost/performance standpoint. Once I started refining the goals of the HOSS project, I realized that the RS150-4 was my midwoofer of choice. It has a very flat, distortion-free profile from about 200 Hz up to 2 kHz, would work well in small sealed boxes, and could be crossed above 1,500 Hz in an omni presentation. The baffles could be narrower and rectangular, and the drivers offset toward the front of the baffle. This allowed the tweeter and baffle to be placed further forward as well, which helps reduce diffraction effects from the midwoofer baffle edges. Positioning the tweeter further forward also helps increase its vertical dispersion.

I was extremely happy with the performance of the RS28 tweeter used in my previous omni project and saw little reason to depart from it with this one. It has one of the lowest distortion profiles available at any cost and can handle playing low enough to integrate well with the RS150 midwoofers.

The subwoofer driver had to go low, handle considerable power, have minimal distortion, work in a relatively small box, and play cleanly up to several hundred hertz. I was shooting for an F3 of around 30 Hz that with room gain would play well into the twenties. The RSS265HF 10" driver possesses all these characteristics, plus it is visually distinctive and very well built.

Early on, I decided that these speakers were going to have to be biamped. I did not want to make the bass bins bigger in order to fit plate amps in the boxes. In the long term, I plan to change over to a single stereo amp combined with an active crossover to create the subwoofer low pass. For the short run, I've elected to use plate amps and sit them on the floor behind the speakers. The ideal plate amp was the Dayton Audio 240 watt SA240 because it has a variable active crossover that can be set as high as 180 Hz, which in turn modeled well with the RS150 at its low end.

The box design was governed by aesthetics, the hybrid omnidirectional speaker concept, and the modular approach. As stated above, I wanted the drivers to be time aligned, which also significantly affected the resulting design. Enclosure assembly is a bit more complex than your typical box, due the tapered bass bins. You should probably have a table saw to accomplish it-a drill press and router are also very helpful. The chamfered baffles reduce the effect of edge diffraction. I posted detailed PDF plans in my HOSS album on the PE forum, for anyone interested.

Bass Bins:

I decided to try a transmission line for the bass bin, hoping it would deliver the low-end extension I was looking for, remain relatively small, and provide some over-excursion protection for the driver. It also promised to allow an attenuated back wave to exit the enclosure, thereby reducing distortion caused by the back wave radiated through the cone. A tapered bass bin allows the woofer to be placed near the floor, transitioning to a narrower width at the top to match the MTM section, yet still provide an acceptable internal volume for a TL. The tapered bass bin would also not visually overpower the MTM section, and the slanted front positions the woofer further forward so it is time aligned with the mids and tweeter. Fellow DIYer Paul Kittinger helped me with the tapered TL modeling.

Our final TL model has a 71" line length at a 20:1 taper, in a net volume of 31 liters. F3 is around 29 Hz, F6 around 25 Hz, and F10 about 20 Hz. The TL has six sections, with a terminus out the top back, and the horizontal TL partitions contribute to a very well braced and inert enclosure. The bottoms of the bass bin and the lower front section where the RSS265HF mounts are 1.5" thick MDF, and the rest is 3/4" MDF. The woofer is placed 14" along the TL line, about 20% from the end-a typical goal for TLs.

MTM:

I wanted to keep the MTM narrow in width to reduce its visual footprint in the room and give the speaker a more elegant overall look. The bottom enclosure in the MTM is where the MTM crossover is placed, so it is 1" taller to keep approximately the same volume as the top box. The boxes are sealed and the smaller one's volume is about 3.4 liters. The boxes are built from 3/4" MDF, though the baffles are 1" thick (3/4" MDF plus 1/4" HDF). Additionally, a cross brace is placed off-center but right behind the RS150 drivers. There is speaker gasket material between the brace and the back of the driver's magnet, which actually presses down against the cross brace. The top and bottom sections of both boxes are removable, using threaded insert nuts and cap head screws for easy access.

Four posts made of 5/8" round acrylic rod separate the boxes. The distance between the baffles was the subject of considerable testing. Midwoofers facing each other integrate well, but determining the proper spacing was a challenge. It took a lot of testing to get it right. If they are too close, you get peaks in the frequency response in the passband. At the same time, the closer the drivers are, the better they integrate with each other and with the tweeter, which is suspended on a baffle between them. The boxes are spaced 9.5" apart in the MTM section, which allows these three drivers to be closer together than would be feasible in a traditional front-firing MTM.

Tweeter:

I did a lot of experimentation and testing with the tweeter baffle configuration when developing my first omni project. I built upon that experience further in the HOSS design. The baffle became narrower to increase off axis performance and improve power response. The baffle dimensions loosely follow the classical golden ratio of 1.618:1. The tweeter is offset both horizontally and vertically in the baffle. These were all steps taken to smooth the tweeter response and attenuate diffraction artifacts. The tweeter baffle itself is also offset vertically between the bottom and top midrange boxes so that the distances from the driver to the baffle edges of the top and bottom boxes were all different, again to spread out diffraction artifacts. The baffle's position, both vertically and front to back, has a significant effect on the tweeter and midwoofer responses.

I wanted the MTM to be an uncomplicated passive 2-way design, and decided on a relatively simple Linkwitz-Riley 4th order acoustic crossover. A zobel filter was used to control impedance and help shape the midwoofer output in the midrange. A 2nd order electrical high pass was added to the RS150 section, crossing to the RSS265HF at around 180 Hz. This ensures that the midwoofers won't be driven easily into over excursion, even at fairly high SPL levels. Not having to perform lower bass duties helped midrange clarity and openness considerably. Connecting the RS150-4 drivers in series presents a nominal impedance of around 6 ohms. The RS28's crossover takes place at about 1,560 Hz.

For the bass, the variable crossovers on the 240 watt plate amps are set to about 160 Hz for my room. This allows them to integrate well with the passive high pass on the MTM section. You have the flexibility to adjust the crossover point up to as high as 180 Hz and also adjust the gain for optimal performance in any room.

If there is one thing I'm asked the most about these speakers, it is how I finished them. It was a laborious process, primarily due to my use of Santos Rosewood veneer-an oily wood that likes to bleed though the finish. I know now that the ideal treatment for it is probably hand rubbed oil followed by wax. That said, the result was worth every minute of effort.

The veneer is two ply wood, not paper backed. It was applied using contact cement and then sanded smooth with up to 400-grit sandpaper. I applied three or four coats of dewaxed shellac to the veneer, lightly sanding after each coat. Next, I brushed on high gloss polyurethane. I used "Old Master" oil-based poly, applied liberally to one side at a time, with that side left facing up for a few hours. This allowed the brush marks to settle out and the wood pores to fill in. Even with sanding between coats, it probably took another three or four coats of poly to seal the veneer's natural oil. I masked the edges of the opposing sides before applying the poly, to protect them from drips, etc. When allowed to dry flat, this poly provides a very high gloss finish that looks sprayed on.

The gloss black painted sections were done with common rattle cans. First, I applied several coats of dewaxed shellac to seal the MDF, sanding between coats. You should have a glass-like finish after these coats. Then I applied a primer, then several coats of high gloss paint, followed by a clear coat. My recommendation is that whatever time the manufacturer says is needed for drying between coats, triple it. The biggest issue that I had was trying to sand or paint the next coat before the previous application was completely dry.

The HOSS was the highest scoring speaker at DaytonDIY2008. In the sound-only categories and in the originality and craftsmanship categories, no other speaker scored higher. The cost of the Dayton Reference Series drivers, crossovers parts, plus plate amps for this project, however, were lower than just the drivers alone in the other top scoring speakers. This is a testament to just how good these drivers and products are.

The HOSS project has been my most successful speaker design to date. All of my goals were met and expectations exceeded. I'm very pleased with how they sound, for both music and HT. I had a lot of great advice and patient help from fellow DIYers as I've endeavored to learn about designing and building speakers. The HOSS project would not have happened without that help and without a willingness to pioneer, to try something new and perhaps fail at it. I encourage all DIYers to expand their horizons and think "outside the box." Challenge yourself with something new, be it in your design, carpentry, or finishing skills. You will undoubtedly learn valuable lessons and create something memorable.

Dan Neubecker is a Landscape Architect who lives with his wife of thirty-one years in Bloomington, Indiana. He is a graduate of Michigan State University, but became a converted Hoosier fan after residing in Bloomington since 1979. He has no training or background in electronics, but has had a passion for audio since his college days. Besides loudspeaker design, Dan enjoys golf and has been a golf clubmaker for many years.