Multiway off-axis irregularity - causing midrange distortion (Long post)

[akunec]

I was thinking of building 3 way speakers with a pair of 4 inch tang band drivers covering 250-8000hz, a scan speak tweeter for 8-20khz, and a pair of 10 inch drivers covering 30-250hz. I want to use such a high 8000hz xo frequency as 8000hz is about 10dB lower then 2k on the equal loudnesss contours.

I am really concerned with lobing effects. What would be the best driver arrangement? I was thinking MTMWW or TMMWW.


Thanks
Alex

[Ron E]

Originally posted by jesper
I'd like to have no ceiling reflection, but since it can't be avoided (unless you go for line source or dipole), I prefer undistorted midrange ceiling reflection. This means a fullrange driver instead of a tweeter, crossed over at 500hz.

/Jesper

1, You can damp out the ceiling reflection, and BTW floor and side wall reflections are just as damaging to the image. A 1/3 octave measurement at the listening position will reveal to some extent how much the colored off-axis sound is a problem

2. You can use a coaxial speaker and you will get no lobing problems at all.

3. If you are really concerned about directivity, look into high quality constant directivity horns. In a reasonably large room, they can be very hi-fi.

[rick57]

Ron

I'm trying to find high quality DIY constant directivity horns

Can you suggest any theory links, or any designs?

Thanks

[454Casull]

Not DIY, but www.gedlee.com

[rick57]

Yes Earl knows quite a bit, but as you say, not DIY.

I don't think there are any public domain CD formulae.

[jeff mai]

Originally posted by rick57
I don't think there are any public domain CD formulae.

There may not be any public domain formulae, but there are known formulae. All the math for his waveguide is in Earl's AES paper. I used Excel to model it against Charles Hughes' Quadratic Throat Waveguide (there isn't much difference - and there is a typo in the latter that made it tricky!) which also has published math. I believe all the math is there in Keele's and Ureda's papers too. None of it is easy reading, but if you have the desire you can figure out how to plot the curves.

[Joules]

I've been woried about lobing errors also. All this is very interesting.
I found an acoustic simulator that is sum what helpful in getting a feel for these sort of problems. It's a 2D wave tank JAVA applet http://www.falstad.com/mathphysics.html. there is alot of cool stuff there.

One thing I began to realize is that at these freqs. and size of drivers (5.5 ,6.5, 7 ect) and the listening distances, that the mid drivers are not point sources. With a sourse hight of, say 6.5 inches the the lobing error problems of two mids about 6 inches apart tends to diffuse the lobing errors some what.

So sticking with one T and two M's whats the best solution?

[rick57]

Are you referring to "Acoustic Waveguide Theory"?

That allows you to design CD bass/ mid/ HF horns?

Cheers

[jeff mai]

Originally posted by rick57
Are you referring to "Acoustic Waveguide Theory"?

That allows you to design CD bass/ mid/ HF horns?

Yes, that or the "In Practice" follow-up. I don't know because I loaned out my copies.

I'm not too sure about CD bass horns. I've never seen any concern about CD below 500Hz in a horn.

If you want to discuss this further, I suggest starting a new thread. We're thread-jacking.

[dwk123]

Originally posted by rick57
Are you referring to "Acoustic Waveguide Theory"?

That allows you to design CD bass/ mid/ HF horns?

Cheers

Well, at the risk of one more slightly thread-jacking post I'll chime in. I think that discussion of controlled directivity is actually rather on-topic, as it is one of the best ways to solve the problem the original poster is asking about.

Briefly, *any* CD horn reduces to conical as you move away from the throat, and in fact the easiest and most common way to do a CD horn is simply to make it straight-sided. Lots of folks do this and it works OK, particularly in cone-driven mids.

Earls work is aimed at achieving the best possible pattern control, and a large part of that is avoiding discontinuities in the flare, particularly right at the throat. Thus, his OS waveguide is a forumlation that matches a conical flare at the mouth to the exit geometry of the driver at the throat.

There is really no such thing as a CD bass horn. Pattern control starts to break down where the dimensions of the device become comparable to the wavelengths involved, which for real-world devices that fit in the home is in the 250-500Hz range. Below this things go omni and there isn't much you can do other than ensure that there is a smooth transition from controlled to omni radiation.

BTW - I believe that Earl said in another forum that he never patented the OS waveguide forumlation. As a private user/DIYer it doesn't matter though, as patents only apply to commercial use. You are free to copy/clone anything you want as long as it's for your own personal use.