Wvaeguides and edge diffraction frequency

[dsrviola]

Since I didn't get much response (just one) over on the PE board, I thought I'd try over here. Here's my query: Can anyone point me to a formula or a discussion on how to determine what the diffraction frequency of a given baffle width with a waveguide of a given depth? Or is it depth and width of waveguide combined? Would the surface radius of the waveguide be the same as baffle width? As an aside, is edge diffraction determined by the distance from the baffle edge to the surround of the tweeter or the center of the diaphragm?

[5th element]

As far as I understand it the diffraction ripple is based off of the width of the cabinet and has nothing really to do with the size of the driver doing the radiating. Baffle step generally starts just below the area where you get the traditional peak/dip combo from diffraction. Now you can reposition the driver anywhere on the front of the enclosure to perhaps minimise the peak/dip ripple, but it will not alter the baffle step frequency, or the frequency where the ripple occurs. I may be wrong on this though!

I would assume that with a wave-guide the width would then take into account the extra length as is created by the wave-guide going 'in' itself. Of course above a certain frequency the wave-guide redirects the far, off axis energy, into the central region of constant directivity. As a result if you only use the wave-guide mostly within its region of constant directivity, less energy will actually reach the cabinet edges so the effects of diffraction should be considerably less.

[Paul W]

Diffraction can occur at at the WG mouth as well as the baffle edge...really anywhere in the path from the acoustic source. Not certain, but think I read that WG mouth diffraction becomes a consideration when the roundover radius is less than 1/4 WL (go big to go low). Then there is the matter of degree...on an infinite baffle, with the same edge treatment, a 90 degree WG will have a different mouth diffraction profile than a 170 degree WG.

Another thing to think about is "waist-banding" on the low end where the pattern for ~straight sided WGs narrows just above the area where directional control is lost.

Hornresp is good for modeling "classic" horn & WG flares, to "roll you own" flare, try AxiDriver. AD has a steeper learning curve, but you can simulate any reasonable baffle and/or flare.