I just bumped to this thesis published in 2010 By Juha Holm
http://lib.tkk.fi/Dipl/2010/urn100200.pdf
http://lib.tkk.fi/Dipl/2010/urn100200.pdf
This thesis concentrates on applying the Finite Element Method (FEM) to virtually prototype waveguides. The theory of FEM and its usability in acoustics is reviewed. Also theory for horn directivity is discussed. Major emphasis is on reviewing and developing a method for visualizing modelled and measured directivity in a comparable manner.
Good thesis, except that one of the conclusions:
This is not true in general, although it would certainly be true in the example that he studied. In an OS waveguide, for example, the directivity is much wider than what an equivalent piston of the throat would dictate. The throat geometry accounts for this effect and the throat geometry in the reference thesis is quite different than an OS waveguide.
This is not true in general, although it would certainly be true in the example that he studied. In an OS waveguide, for example, the directivity is much wider than what an equivalent piston of the throat would dictate. The throat geometry accounts for this effect and the throat geometry in the reference thesis is quite different than an OS waveguide.
In or about 1991, I:
1) "formed a parametric model of a sound Waveguide surface" (i.e. an OS) with theta as a parameter
2) "simulated a sound field" (using the techniques shown in my book)
3) "determined the frequency dependent spatial distribution" (polar response using said techniques)
4) varied "at least on parameter" (theta) "to change the sound Waveguide surface to adjust" the polar response.
I guess that I infringed the patent. Hope they don't find out.
I found this image of an anamorphic lens designed to project a widescreen image from a distorted version recorded on a standard narrow format film. It would seem that things are different though maybe because it is many thousands of wavelengths across and where acoustic waveguides are comparable to a single wavelength.
Regarding the choice to be waveguiding lower midrange frequencies I find it curious that the LeCleach profile seems to coincide as a practical method in both horn theory and waveguide theory.
Regarding the choice to be waveguiding lower midrange frequencies I find it curious that the LeCleach profile seems to coincide as a practical method in both horn theory and waveguide theory.
Regarding the view that directivity should be held until 'formed' with a suitable mouth circumference before commencing a termination, is there a benefit, or need to have also held it for a quarter wavelength by that point?
Attachments
Last edited:
Out of the Blue
Of course, 'prior art' forms no basis for a case of patent infringement. For the case of an infringing patent, that's another matter.
This and similar shape optimization regimes being explored in Sweden as well, deliver fully formed finite horns, the shape of which materially departs from the locus of a hyperbola at all points except those in the immediate vicinity of the throat entry.
As curve generation starts from a conical horn the methodology is capable of generating a hyperbolic horn as well when given the appropriate parameter settings. However when optimization is pursued, the resulting profile is not that of a hyperbola.
In any event, I was quite surprised to see a patent covering shape optimization of an acoustic horn, because of the work being independently pursued in the same time frame by groups of researchers in several countries, the U.S. included.
That is why I took the time to post it here.
WHG
In or about 1991, I:
1) "formed a parametric model of a sound Waveguide surface" (i.e. an OS) with theta as a parameter
2) "simulated a sound field" (using the techniques shown in my book)
3) "determined the frequency dependent spatial distribution" (polar response using said techniques)
4) varied "at least on parameter" (theta) "to change the sound Waveguide surface to adjust" the polar response.
I guess that I infringed the patent. Hope they don't find out.
Of course, 'prior art' forms no basis for a case of patent infringement. For the case of an infringing patent, that's another matter.
This and similar shape optimization regimes being explored in Sweden as well, deliver fully formed finite horns, the shape of which materially departs from the locus of a hyperbola at all points except those in the immediate vicinity of the throat entry.
As curve generation starts from a conical horn the methodology is capable of generating a hyperbolic horn as well when given the appropriate parameter settings. However when optimization is pursued, the resulting profile is not that of a hyperbola.
In any event, I was quite surprised to see a patent covering shape optimization of an acoustic horn, because of the work being independently pursued in the same time frame by groups of researchers in several countries, the U.S. included.
That is why I took the time to post it here.
WHG
Last edited:
WHG,
It seems if constant directivity is anyone's major goal that is really the simplest horn shape to create. A conic section horn with a transition from the compression drivers throat angle to the conic sections angle and then a large round-over to take care of the inevitable diffraction at the end of the flare. Couldn't be easier than that.
It seems if constant directivity is anyone's major goal that is really the simplest horn shape to create. A conic section horn with a transition from the compression drivers throat angle to the conic sections angle and then a large round-over to take care of the inevitable diffraction at the end of the flare. Couldn't be easier than that.
The curve that may be used to acomplishe this mission is a spline and it is not clear that the methodology for its use is as simple as you suppose. WHG
The curve that may be used to acomplish this mission is a spline and it is not clear that the methodology for its use is as simple as you suppose. WHG
Andrew,
That is my thinking also. a simple tangent at the throat angle blending to a tangent at the conic angle chosen. If you are talking about a spline angle throughout the waveguide I see no reason to call it constant directivity. This is the problem, constant directivity is just a concept, it is not a reality, just more semantics as always, A conic section is the only thing that can give a true constant directivity angle and that only works up to 1/4 wavelength at that. In my eyes it is a BS argument really, there is no such animal if the entire horn is not a conic cone and up to a full wavelength at the lowest frequency required. Otherwise we are just playing with words.
That is my thinking also. a simple tangent at the throat angle blending to a tangent at the conic angle chosen. If you are talking about a spline angle throughout the waveguide I see no reason to call it constant directivity. This is the problem, constant directivity is just a concept, it is not a reality, just more semantics as always, A conic section is the only thing that can give a true constant directivity angle and that only works up to 1/4 wavelength at that. In my eyes it is a BS argument really, there is no such animal if the entire horn is not a conic cone and up to a full wavelength at the lowest frequency required. Otherwise we are just playing with words.
I've built a few elliptical os waveguides and conical horns by now and I'm not sure I'd say it's easy. Maybe I'm picky but they aren't as constant as I would like either. You can get the -6dB curve to be flat and look nice on a contour plot, but the 0-30deg curves aren't flat and I've noticed on many waveguides/conical horns that the 30deg response has a tendency it increase in level in the 4khz-8khz region to the point where it's the same or louder than the axial measurement. I asked why this was a few days ago in this thread but no one offered any ideas. I have an idea or two but until I start modeling/building again I'm not sure. Like Earl, I eq my waveguides to the listening axis, so maybe the 0-30deg irregularity is a moot point. It would be nice to have near perfect cd (to me that means perfect horizontal curves as you move off axis while maintaining the beamwidth) so I can experiment with different angles of toe in while keeping the same response on the listening axis. The SEOS waveguides come close to this in the horizontal. To me they look really short, and because of this I wonder if they have more wall curvature.
Nate - it would be a lot easier to analyze your waveguides if you provided a polar map. From that I might be able to comment.
PS. How do you make your waveguides? Fiberglass?
PSS. a constant radius from the throat to a conical section is the patented Quadratic Waveguide by Peavy. Not having as smooth a second derivative as the OS it will have more internal diffraction.
PS. How do you make your waveguides? Fiberglass?
PSS. a constant radius from the throat to a conical section is the patented Quadratic Waveguide by Peavy. Not having as smooth a second derivative as the OS it will have more internal diffraction.