There is one question that never seems to get an answer: How can we predict and measure the strength of HOMs relative to the main signal? Diaphragm modes can be identified and when suppressed, are clearly measurable. How about HOMs?
Whilst I have no desire to get enmeshed in a debate about horn acoustics can I mention something about the Dr Rick Morgans research. By an accident of geography and AES membership I have encountered him on a couple of ocasions. He told me he had acquired a microflown (velocity measuring microphone) and that this was going to be traversed in front of the test objects.
I don't know if some measurements in his paper were done with this. Just thought I would mention it lest there is a reader assumption that acoustic measurements automatically equate with pressure microphones. I believe microflowns have rather limited bandwidth.
Keith
I don't know if some measurements in his paper were done with this. Just thought I would mention it lest there is a reader assumption that acoustic measurements automatically equate with pressure microphones. I believe microflowns have rather limited bandwidth.
Keith
Lynn Olson said:
It looks like that is what I'll end up doing. While I can most certainly match the directivity, I have no idea what the low frequency response is going to be on an Oblate Spheroid. I really need to get an idea of what the expected bandwidth is going to be. What good is it to match directivity if you have a frequency gap between the mid-bass and mid-range?
In no way am I questioning Dr. Geddes claims about the Oblate Spheroid. There is more than enough data to support what he has written in this thread. My problem is the difficulty involved in designing a good Oblate Spheroid wave guide that meets the needs of various applications. There is no design road map to follow. If I wanted to cover 700Hz to 7KHz with an Oblate Spheroid wave guide there is nothing to help me in deciding what angle, depth, or diameter to use.
Looking at the tractrix vs. Oblate Spheroid, the design processes are almost opposites. Designing for the tractrix is cake. I knew if you wanted to radiate into free space then the mouth circumference needed to be 1 wave length of the lowest frequency of interest when the horn length was between ¼ and ½ wave length. If the horn was a little shorter than this, then I knew how much bigger to go to achieve the desired bandwidth. Of course, you are stuck living with the resulting beam width restrictions and narrowing directivity of the tractrix. (This is what I'm trying to get away from.) Now on the other hand, you design the Oblate Spheroid with beam width and directivity in mind, but then are left to live with the resulting band width. Some one has a very sick sense of humor.
When the beam is widened for higer frequencies, then there will be a drop in SPL. Equalization is necessary. I do believe this has been mentioned before. The smaller the throat, the higher frequency directivity control you have. The angle of the guide gives the beam width. The equation that Dr. Geddes provided up front probably is the design that gives the least HOMs.
I still wonder how the mouth effects HOMs and low frequency response.
I still wonder how the mouth effects HOMs and low frequency response.
JLH said:
There are decades of experience with other horn types to draw on, but with the OS there appears to be only mine. Is there any doubt as to why there is so little "rules of thumb" to help in their design? And the whole theory IS completely new and what you knew in the past IS of little use - that is a fact. It's like starting all over again from scratch. Is it any wonder that people don't use my approach? Anyone but me that is.
Hence you really have no option but to try and learn from me if you want to go down this path. But everybody seems to want to argue and refute my claims, not learn from them.
The LF cut-off is basically NOT determined by the waveguide, but by the resonance of the driver. The OS does not have a strong LF boost in efficiency as do some other curves like the exponential. But those other curves are not CD. Hence you have to choose which you want - extended LF bandwidth or CD - can't have both.
The SPEAK models for the OS waveguide turn out to be pretty reliable. I don't have experince with any other programs - I don't need to.
soongsc said:
This is correct. One would at least need the phase of the pressure to do anything even if the vlocity direction were simply assumed. So there is actually nothing that you can derive from the pressure plots other than the fact that they are most definately not axi-symmetric and that they do contain HOMs.
soongsc said:
Mouth reflections will generate HOMs that reflect back down the device. If there is no foam to absorb them then they will be reflected again off of the diaphragm and radiated as sound, but with a terrible group delay.
gedlee said:
Dr. Geddes,
Thank you for your response. It had some very useful information. I do have a tentative design on paper. I think I will go ahead with it and keep an objective mindset. There are always problems with being the first in any area. Thanks for having the courage to continue the good fight.
I’m fine with serving as a stepping stone for others. However, I just want to make sure I’m heading in the correct direction and not misleading anyone.
If the angle between the velocity vector and the wall increases gradually before eaching the mouth, does it improve the situation or worsen it?gedlee said:
Dr. Geddes,
I have a question about matching the throat angle of the driver. I understand that we want the initial angle of our Oblate Spheroid wave guide to match the throat angle of the driver. What do you do if the compression driver does not have a throat tube and the phase plug is right at the mounting surface? See below image.
Would you just start the Oblate Spheroid with a 90 degrees throat? I think this could also apply to cone driven Oblate Spheroid wave guides too. Thanks.
Rgs, JLH
I have a question about matching the throat angle of the driver. I understand that we want the initial angle of our Oblate Spheroid wave guide to match the throat angle of the driver. What do you do if the compression driver does not have a throat tube and the phase plug is right at the mounting surface? See below image.
Would you just start the Oblate Spheroid with a 90 degrees throat? I think this could also apply to cone driven Oblate Spheroid wave guides too. Thanks.
Rgs, JLH
In the case above there is no coupling tube as there is in most compression drivers so you are correct that the starting angle should be 0° (you said 90°, but I know what you mean).
Who makes the driver that you have shown? The coupling tube is undesirable and I'd like to have a driver like you show.
There is no "correct" way to drive a waveguide with a cone, there are only tradeoffs. Since I don't do this, I haven't really studied it in depth, but I do know that I would not do this without a phase plug.
Who makes the driver that you have shown? The coupling tube is undesirable and I'd like to have a driver like you show.
There is no "correct" way to drive a waveguide with a cone, there are only tradeoffs. Since I don't do this, I haven't really studied it in depth, but I do know that I would not do this without a phase plug.
soongsc said:
You are not being precise enough. Do you mean "If the angle between the velocity vector and the axis increases gradually as the location gets closer to the wall"? What you said doesn't make sense. The velocity is defined at a point, there isn't just one and only in a very rare situation are they all the same.
gedlee said:
Dr. Geddes,
I found that image using Google search. However, I looked into it more and it appears to be another spin on JBL's Coherent phasing plug design. See patent below link:
http://www.pat2pdf.org/patents/pat6952874.pdf
In addition, the JBL Tech note link discussing the original design:
http://www.jblpro.com/pub/technote/tn_v1n21.pdf
I do not have any personal experience with the Coherent phasing plug drivers. Some people have referred to them as the Non-Coherent plugs after listening to them. I'm not sure what these people were hearing, or not hearing. I just know there are mixed opinions about them when mounted to conventional horns.
Yes, if I was going to try cone drivers with the Oblate Spheroid I would make a phase plug for it. Without one it would be quite good at creating HOMs and reflections.
Rgs, JLH
Hi Earl, JLH
You had mentioned that the HOM’s are hard to identify in measurements and it reminded me several times in the past of a visualization that might be useful in “seeing” them even if it doesn’t derive a number value. This is a 3d view of a “stack of polar’s”.
Below is a mathematical interference pattern, a large stack of “polar plots” for two equal amplitude sources who’s physical spacing begins at the bottom at 1/10 wl and ends at several wavelengths at the top. Granted in this “perfect noise free world” one has the largest, most clear case possible (and makes a pleasing graph haha).
If one wanted to find the lobes and nulls AND to see where there are frequency dependent things, things that track, these standout as chevron shapes on the sides in the head on view as below..
I would imagine that the HOM’s would have a similar appearance in that there number increases with frequency etc.
Earl, here is a driver with a phase plug ending at the exit, I think there just down the street from B&C. I have not tested this one yet but I use a couple of their other drivers.
http://www.faitalpro.com/products/schede/cd.php?id=502010200
JLH
I hesitate to bring up “the speed of sound” after the watery diversion but there is an issue with that kind of driver so far as its use for some things.
For the driver to produce a “plane wave”, all of acoustic the paths to the dome have an equal length (to the degree the dome is a piston with minimal phase shift between the center and outer region etc).
In the driver shown, one see’s a geometry that appears likely produce a converging wavefront at the exit and this could be particularly icky on a conical horn and possibly on an OS as it wants a plane wave. I see your later comment which seems possibly relevant.
Check out the BMS drivers too, I guess good enough for JBL to use and re-badge as there own.
http://www.audioheritage.org/vbulletin/showthread.php?t=6368&page=7
I’ve had very good luck with the 4550 and 4552 in several products.
Best,
Tom Danley
You had mentioned that the HOM’s are hard to identify in measurements and it reminded me several times in the past of a visualization that might be useful in “seeing” them even if it doesn’t derive a number value. This is a 3d view of a “stack of polar’s”.
Below is a mathematical interference pattern, a large stack of “polar plots” for two equal amplitude sources who’s physical spacing begins at the bottom at 1/10 wl and ends at several wavelengths at the top. Granted in this “perfect noise free world” one has the largest, most clear case possible (and makes a pleasing graph haha).
If one wanted to find the lobes and nulls AND to see where there are frequency dependent things, things that track, these standout as chevron shapes on the sides in the head on view as below..
I would imagine that the HOM’s would have a similar appearance in that there number increases with frequency etc.
Earl, here is a driver with a phase plug ending at the exit, I think there just down the street from B&C. I have not tested this one yet but I use a couple of their other drivers.
http://www.faitalpro.com/products/schede/cd.php?id=502010200
JLH
I hesitate to bring up “the speed of sound” after the watery diversion but there is an issue with that kind of driver so far as its use for some things.
For the driver to produce a “plane wave”, all of acoustic the paths to the dome have an equal length (to the degree the dome is a piston with minimal phase shift between the center and outer region etc).
In the driver shown, one see’s a geometry that appears likely produce a converging wavefront at the exit and this could be particularly icky on a conical horn and possibly on an OS as it wants a plane wave. I see your later comment which seems possibly relevant.
Check out the BMS drivers too, I guess good enough for JBL to use and re-badge as there own.
http://www.audioheritage.org/vbulletin/showthread.php?t=6368&page=7
I’ve had very good luck with the 4550 and 4552 in several products.
Best,
Tom Danley
Attachments
Tom
I'm not sure how to read the figure that you show. To me a plt of SPL versus angle and theta is extremly useful.
Gotfried Bueler did a study of the HOM and found that "they had a negligable effect on the polar response" of which I have no doubt. But thats missing the point. They could easily have a pronounce audible effect and still have a small effect on the total field. I discussed this with him, but "audibility" was not his thing.
I am not a fan of tangential phase plugs as they can't really do whats required.
JHL
The JBL patent concerns manufacturing, but the actual design is classic - based on Bob Smith's work. I showed in my patent on phase plugs that there was an error in Bob Smith's approach.
I'm not sure how to read the figure that you show. To me a plt of SPL versus angle and theta is extremly useful.
Gotfried Bueler did a study of the HOM and found that "they had a negligable effect on the polar response" of which I have no doubt. But thats missing the point. They could easily have a pronounce audible effect and still have a small effect on the total field. I discussed this with him, but "audibility" was not his thing.
I am not a fan of tangential phase plugs as they can't really do whats required.
JHL
The JBL patent concerns manufacturing, but the actual design is classic - based on Bob Smith's work. I showed in my patent on phase plugs that there was an error in Bob Smith's approach.
That Faital HF10TX compression driver looks interesting. I wonder why they chose 24 slits for the phasing plug. When designing a radial phasing plug it is general practice to use a number of slits that is a prime number to prevent resonance. They could have used 11, 13, 17, 19 or 23 slits.
Rgs, JLH
Rgs, JLH
salas said:So HOMs are the ''Quarks'' of audio? They surely exist and are making their thing but are elusive to display? Interesting. We have to call up some CERtaiN guys in Europe, maybe they have some testing ideas...
There must come a point at which you just say, I know that these are bad and I'm going to do everything that I can to get rid of them. Thats my approach. Someday I ma get arround to developing a test for them, but right now I have found that the more that they are reduced the better the sound. Thats all I need to know for now.