You can also find his whole PHD thesis here (in English):
http://publications.rwth-aachen.de/record/61507/files/Makarski_Michael.pdf
http://publications.rwth-aachen.de/record/61507/files/Makarski_Michael.pdf
No, I have not.
Yes, you are technically correct. I was considering that a larger aperture would necessitate a large moving mass. I don't see how it could be otherwise (unless someone deliberately did a bad design.)To be exact, isn't it the (possibly) higher Mms compared to motor force, not the size of the exit?
There are 2" exit drivers with different sizes of radiating surfaces, some have 3" VC, some have 4", some are ring radiators etc. That's why above statement is misleading, but of course I understand the meaning behind it (bigged domes usually don't go as high).
This may have some truth to it, but the fact remains that the response is wider than it would be if the waveguide were not present - in both cases. The De250 and the JBL are not so different that the effect would be so different, and I already said that I tested a TAD of almost exactly the same design as the JBL - it did not beam. As I have said, there is some narrowing at HFs, just never as much as the throat angle would dictate. And this all depends on how the phase plug is designed as it is a critical factor at these frequencies.
At Lower Frequencies Only
If the phase plug is designed to be placed at the apex of the bounding hyperbola of an OS horn throat, then ideally the wave front should be 'flat" at frequencies where wavelength is larger than horn throat dimensions. Sometimes a concave wave front is used to tweak performance.
At higher frequencies where horn air column and diaphragm modes occur, a coherent flat or spherical (OS) wave front geometry will not be maintained.
WHG
If the phase plug is designed to be placed at the apex of the bounding hyperbola of an OS horn throat, then ideally the wave front should be 'flat" at frequencies where wavelength is larger than horn throat dimensions. Sometimes a concave wave front is used to tweak performance.
At higher frequencies where horn air column and diaphragm modes occur, a coherent flat or spherical (OS) wave front geometry will not be maintained.
WHG
Last edited:
Thanks for the URL.
It is evident that this paper represents a considerable effort on the part of the author.
WHG
It is evident that this paper represents a considerable effort on the part of the author.
WHG
but we have seen instructions to match the driver exit angle
and I don't see how the velocity profiles measured by Makarski aren't showing a more nearly spherical front than flat - and a plane wave couldn't propagate in a conical section without changing shape over distance anyway
I agree that if you had a plane wave then you want to start your OS with 0 degree walls at the origin/symmetry plane - and maybe some compression drives do give more nearly plane wave at their exit - JBL D2 velocity surface plot?
and I don't see how the velocity profiles measured by Makarski aren't showing a more nearly spherical front than flat - and a plane wave couldn't propagate in a conical section without changing shape over distance anyway
I agree that if you had a plane wave then you want to start your OS with 0 degree walls at the origin/symmetry plane - and maybe some compression drives do give more nearly plane wave at their exit - JBL D2 velocity surface plot?
Last edited:
I don't recall what Makarski measured (other than it was not actually flat,) but the fact is that the Bob Smith phase plug (ASA circa the 30's), which JBL has claimed was the source for their design (See the AES paper by Fancer Murray circa the 80's.) Hence the design intent (at that time) was a flat wave front at the exit of the phase plug. A long conical section will tend to bend it into a more spherical shape but the details of this at HFs are very specific to the specific details of the design.
If the wave front is slightly curved, then it is best to start the OS contour some ways down its progression such that both the wall angles and the wave front better match each other. This was discussed in my very first paper on waveguides way back in 1991.
Other companies may in fact be deliberately trying to achieve a spherical wave front, that is certainly possible, but to my knowledge nothing has ever been published in the public domain stating that intent (my own patents excluded.) I therefor have to assume that designers are still doing what they always did, or they have their own trade secrets that none of the rest of us know about.
This is a public forum and even I do not disclose all that I know.
I was a reviewer for most (if not all) of Makarski's papers. He was doing quite good work and upon graduation went to work for some audio company in Germany. (I met him several times.) However, then all of his public work seemed to stop and I haven't seen anything in probably a decade. I would love to know what he is doing these days.
Earl,
Michael is active here:
Institute for Acoustics and Audio Technique
| Office Würselen
Here is a 2010 paper:
Waveguide for tall ribbon driver
Dear mr. Geddes,
I am designing a dipole tall ribbon based loudspeaker (Actually is a 2-way dipole system composed of 2 12'' bass drivers and tall ribbon tweeter in the middle. Just to have an idea of the whole project: SoulSonic Impact Limited Edition).
I plan to use a waveguide in front of the tall ribbon loudspeaker. Current prototype is based on curved panels with radius of 40 cm. The throath is rectangular with the width of 2,7 cm approximatelly (1,05''). This is more or less defined with the width of the ribbon (2,5 cm) and a little clearence between ribbon and magnets on each side of the iron support. So, the magnet gap is of app. 2,7 cm and so the waveguide throath should be. At least I suppose so, in order to not risk any diffractions, am I right?)
The opening angle is approximatelly 115 degrees, the mouth width is app 82 cm (app 32'') and the waveguide depth is app 23,5 cm (0,92''). The waveguiedes are made of plexiglass (8 mm or 0,31'' thickness)
The bonus of current curvature/contour is that it represents also a loading for the ribbon driver in the lower midrange zone. The ribbon works down to app. 600 Hz without issues (despite having crossover set at app 2 Khz/6 db/Oct because of the such evident gain in midrange zone). This is quite important since the bass drivers are 12'' and mounted very near the floor (in order to not have problems with the room dip in the higher bass region). However, sometimes I have a feeling that some higher frequencies can be heard as they are bouncing the waveguide or as they are ''sticked'' to waveguide panels. The other problem is a very narrow sweet spot and the last one is the need of an additional parallel notch filter set at 2,5 kHz to compensate a peak of app 3 dB).
I am not sure whether the problem is in the curvature of the waveguide or some other details. So, I have done some ''research'' on Horn/Waveguide supports and came across with you paper (http://www.gedlee.com/downloads/What is a Waveguide.pdf) and some similar (Audio-X-Press: Horn Theory).
I have read that the best waveguide contour is actually Oblate Spheroidal or OS waveguide: HOM problems should be minimized.
However, I have some questions related to the construction of such OS waveguide:
- the angle/radius of the curvature in the throath
- the opening angle of the sidewalls/panels...
- any other issue to consider?
Thank you very much for your help.
Dear mr. Geddes,
I am designing a dipole tall ribbon based loudspeaker (Actually is a 2-way dipole system composed of 2 12'' bass drivers and tall ribbon tweeter in the middle. Just to have an idea of the whole project: SoulSonic Impact Limited Edition).
I plan to use a waveguide in front of the tall ribbon loudspeaker. Current prototype is based on curved panels with radius of 40 cm. The throath is rectangular with the width of 2,7 cm approximatelly (1,05''). This is more or less defined with the width of the ribbon (2,5 cm) and a little clearence between ribbon and magnets on each side of the iron support. So, the magnet gap is of app. 2,7 cm and so the waveguide throath should be. At least I suppose so, in order to not risk any diffractions, am I right?)
The opening angle is approximatelly 115 degrees, the mouth width is app 82 cm (app 32'') and the waveguide depth is app 23,5 cm (0,92''). The waveguiedes are made of plexiglass (8 mm or 0,31'' thickness)
The bonus of current curvature/contour is that it represents also a loading for the ribbon driver in the lower midrange zone. The ribbon works down to app. 600 Hz without issues (despite having crossover set at app 2 Khz/6 db/Oct because of the such evident gain in midrange zone). This is quite important since the bass drivers are 12'' and mounted very near the floor (in order to not have problems with the room dip in the higher bass region). However, sometimes I have a feeling that some higher frequencies can be heard as they are bouncing the waveguide or as they are ''sticked'' to waveguide panels. The other problem is a very narrow sweet spot and the last one is the need of an additional parallel notch filter set at 2,5 kHz to compensate a peak of app 3 dB).
I am not sure whether the problem is in the curvature of the waveguide or some other details. So, I have done some ''research'' on Horn/Waveguide supports and came across with you paper (http://www.gedlee.com/downloads/What is a Waveguide.pdf) and some similar (Audio-X-Press: Horn Theory).
I have read that the best waveguide contour is actually Oblate Spheroidal or OS waveguide: HOM problems should be minimized.
However, I have some questions related to the construction of such OS waveguide:
- the angle/radius of the curvature in the throath
- the opening angle of the sidewalls/panels...
- any other issue to consider?
Thank you very much for your help.
kodrib
1) there is no radius at the throat, it is horizontal (if looking at the contour in cross section laid horizontally.)
2) the opening angle should match those of the driver exit angle - in your case there is none so the angles should be horizontal.
3) lots of issue to consider especially with a square ribbon. The OS waveguide theory does not cover this situation so everything is a guess that can only be resolved with measurements. The best coverage of this topic that I know of is my book which you can download from my site - chapter 6. (Not easy reading however.) Consider a Prolate Spheroidal surface for this shape. It is not exact, but much better than using OS in this case.
1) there is no radius at the throat, it is horizontal (if looking at the contour in cross section laid horizontally.)
2) the opening angle should match those of the driver exit angle - in your case there is none so the angles should be horizontal.
3) lots of issue to consider especially with a square ribbon. The OS waveguide theory does not cover this situation so everything is a guess that can only be resolved with measurements. The best coverage of this topic that I know of is my book which you can download from my site - chapter 6. (Not easy reading however.) Consider a Prolate Spheroidal surface for this shape. It is not exact, but much better than using OS in this case.
N.B. #1 & #2
1) The c/o slope is to shallow for a displacement limited ribbon.
2) The gossamer ribbon does not behave the same as the robust diaphragm of a compression driver when placed in a horn that generates relatively intense LF back-waves..
Regards,
WHG
1) The c/o slope is to shallow for a displacement limited ribbon.
2) The gossamer ribbon does not behave the same as the robust diaphragm of a compression driver when placed in a horn that generates relatively intense LF back-waves..
Regards,
WHG
Dear Geddes,
Thank you for your comment and directions. I will have a look at your book. However, I would kindly ask you for some additional references or hints regarding the proposed Prolate Spheroidal surface. I guess you have in mind the waveguide with a quadratic section like the throat is quadratic. I have found the following reference, http://www.excelsior-audio.com/Publications/QTWaveguide/QTWaveguide_WhitePaper.pdf, however, I am not sure if this can be applied to square ribbon drivers.
Thank you very much.
Best regards, Borut.
Thank you for your comment and directions. I will have a look at your book. However, I would kindly ask you for some additional references or hints regarding the proposed Prolate Spheroidal surface. I guess you have in mind the waveguide with a quadratic section like the throat is quadratic. I have found the following reference, http://www.excelsior-audio.com/Publications/QTWaveguide/QTWaveguide_WhitePaper.pdf, however, I am not sure if this can be applied to square ribbon drivers.
Thank you very much.
Best regards, Borut.
Borut
The paper you referenced is just a simple approximation to an OS contour.
The only reference that I know of to the Prolate Spheroidal waveguide would be my own. No one else that I know of thinks in these terms.
Basically a ribbon is not an ideal source for a waveguide, so anyway you go is going to be an approximation and guess as to how it will come out. Then, only a good set of comprehensive measurements will tell you how good (or bad) your guess has been.
People are forever looking to "fit" a waveguide to whatever situation that they find the design in. That is not how I approached the problem. I asked: "What shapes allow for ideal wave propagation and what types of sources are required." This is a very different question with very different answers. You are trying to "fit" the answers to your question (the first) with answers to my question (the second one.) The "fit" is not a very good one.
The paper you referenced is just a simple approximation to an OS contour.
The only reference that I know of to the Prolate Spheroidal waveguide would be my own. No one else that I know of thinks in these terms.
Basically a ribbon is not an ideal source for a waveguide, so anyway you go is going to be an approximation and guess as to how it will come out. Then, only a good set of comprehensive measurements will tell you how good (or bad) your guess has been.
People are forever looking to "fit" a waveguide to whatever situation that they find the design in. That is not how I approached the problem. I asked: "What shapes allow for ideal wave propagation and what types of sources are required." This is a very different question with very different answers. You are trying to "fit" the answers to your question (the first) with answers to my question (the second one.) The "fit" is not a very good one.
Dear Mr.Geddes,
Thank you very much for fast response. As you already mentioned I was not able to find any useful reference related to Prolate Spheroidal. However, going through the chapter 6 of your book (available on your website) I could find only 4 basic characteristics of the Prolate Spheroidal (chapter 6.3 Waveguide Geometries). Since I do not have technical backround from this field it would be very helpful to get the idea if there would be any scheme showing the crosssection of the so mentioned PS. I learn by examples. I would very appreciate any hint or help.
Best regards, Borut.
Thank you very much for fast response. As you already mentioned I was not able to find any useful reference related to Prolate Spheroidal. However, going through the chapter 6 of your book (available on your website) I could find only 4 basic characteristics of the Prolate Spheroidal (chapter 6.3 Waveguide Geometries). Since I do not have technical backround from this field it would be very helpful to get the idea if there would be any scheme showing the crosssection of the so mentioned PS. I learn by examples. I would very appreciate any hint or help.
Best regards, Borut.
This might help you, Borut. You can download the pdf file.
Acoustic waveguide for controlled sound radiation - Geddes, Earl Russell