Greetings all--OB newbie here. I've been building box speakers for close to 20 years, and would like to tackle my first OB/dipole project.
I understand baffle diffraction step as it applies to a conventional box, but I figure it works a little differently with an OB. Instead of the longer wavelengths merely dispersing off-axis, causing a shelving down of energy on the forward axis at lower frequencies, with a dipole they wrap around from front AND back resulting in phase cancellation. This cancellation becomes more pronounced as frequency drops, resulting in a 6dB/octave highpass function (as opposed to a 6dB shelf, which levels out as wavelengths exceed 8x baffle width). This highpass function must be combined with any electrical filtration applied to the midrange driver, to yield a smooth 2nd order (or steeper) slope, which is symmetrical with the lowpass slope applied to the woofer.
OK so far?
What I need to know, before I start cutting wood, is how to estimate my baffle width vs. frequency for a targeted rolloff characteristic. Say I want to cross in the dipole mid at 200Hz (that being the F6 of my 2nd order active lowpass on the boxed woofer). I would then want an F3 right around there, which, combined with the F3 of the 1st order electrical highpass on the mid, would yield a 2nd order highpass slope.
The textbook formula for BDS in a conventional box is F3 = 4560/Wb in inches. Of course its more complicated than that, as other variables come into play, but it gets you in the ballpark. With an open baffle, will this formula yield the -3dB down point, or, due to phase cancellation, the -6dB down point? If the latter, how do I calculate baffle width for a given -3dB down point.
Of course I will measure at every step of the way, and model my crossover based on actual FR data, but I need a working baffle width to start with.
All advice and insights are greatly appreciated.
