Single driver back-loaded horns

I was smitten with the "Single Driver bug" many years ago, and have so far constructed and built a variety of cabinets for crossover-free, single driver operation. In my experience, the combination of a high-end driver designed to reproduce the entire audible frequency range and a well-designed cabinet is invincible in the cost/benefit department. Indeed, it could be argued that horn systems are the only type of DIY loudspeaker to potentially save money, compared to commercially available systems.

Apart from the sound economics of embarking on a project of this kind, there are some other distinct and desirable properties to said systems:

1. High sensitivity (perfect for low-power amps).
2. Very life-like dynamics and transients.
3. Low failure rate for DIY projects (few sources of error).
4. Extremely transparent and 'open' sound.

However, there are potential inherent downsides. It can hardly be argued that:

1. Horn systems are generally more physically imposing than direct-radiating multi-way systems.
2. Horn systems often require more careful placing in the listening environment.
3. There are certain features to the overall sound from single-driver horn systems that might not appeal to everyone. They are prone to peaks and dips through the frequency range, due to break-up modes of the cone, mechanical crossover properties, resonance, etc. Also, high frequency roll-off from about 14kHz-17KHz and upward should be expected.

Generally, though, I find that most people are overly concerned about flat response curves over the entire 20 to 20K area. Listening rooms will inevitably introduce resonance (eigentones) of their own, and it is easily demonstrable that a given loudspeaker configuration will sound differently when it is moved from room to room in a house. On top of this, reduction of hearing down to -20db in the area from 12KHz and upward is considered normal for adults. On my last listening test, I was down -30db at 15KHz on one ear, and was still rated 'normal'.

In this author´s opinion, the subjective realism of reproduced sound should be considered as (or even more) important, and most multi-way systems have a long way to go before they can match a properly housed crossover-free single-driver system. This is, I believe, largely due to the lack of a crossover network. Almost every multi-way system have a crossover point between 200 and 5000 Hz, and this is where our ears are most sensitive to distortion.

Background

My previous project, The "Big Fun" horn (published in EW&WW January 96), received a lot of praise from DIYers around the world. However, while I was very satisfied with its sound reproduction capabilities, I still felt that a lot could be done to optimize the design.

For those of you not familiar with the design, here is a brief recap:
The 'Big Fun' horn is a constant-width tractrix back-loaded horn with a cut-off frequency of 32 Hz, throat area of 100 cm2, mouth area of 4225cm2, and horn length of approximately 3,2 meters. Horn mouth exits at back and sides. It measures roughly 41cm*119cm*90cm (whd) when built from 22mm materials. You'll realize that 'Big' is a fitting description.

Goals

Based on experience from this and other projects, and some newly gained information on the art of compromising bass horns, I aimed to realize a design that compromised more on size and less with regards to cabinetry and overall design criteria, hopefully while still retaining overall sound reproduction capabilities.

• Cabinet size: footprint size was initially restricted to ca 60cm*30cm, which should be acceptable for most living-room environments.
• Horn expansion/dimensions should be calculated for corner placement.
• Horn efficiency should be optimized through careful selection of curvature, throat area, and folding geometry.
• Horn length should be less than 3 meters, to avoid audible time delay of bass going through the horn.
• Fc (cut-off frequency) should be set such that the compromises in horn length and mouth area are well within accepted rules of thumb.
• The horn should not be of constant width, as this impairs efficiency. A horn's cross section should ideally be round, alternatively square, for optimum efficiency.
• Mouth geometry should allow for a mouth as large as possible, and preferably provide exit at the back of the cabinet, to obtain good coupling to room corners.
• The horn should be as easy to build as possible.