gedlee said:
oh no
rather try and get the expert interested
those pictures are taken from that 1984 Japanese "joke" (in the words of Mr Hansen)
I am not sure but I am probably violating copyrights by reproducing them here
below is the image of the measured device
this was their first try, later "Scepter 2002" ( photo reproduced somewhere above) had different waveguide
best,
graaf
Attachments
graaf said:
oh no
rather try and get the expert interested
those pictures are taken from that 1984 Japanese "joke" (in the words of Mr Hansen)
I am not sure but I am probably violating copyrights by reproducing them here
below is the image of the measured device
this was their first try, later "Scepter 2002" ( photo reproduced somewhere above) had different waveguide
best,
graaf
I suspected that there was some sort of game going on here. I don't like to play games.
Hello David,
You are right for most part of your reply. The 0° "off axis response" is quite well predicted by Hornresp.
But on your simulation at 0° we can see a difference of level reaching 15dB between 2kHz and 20kHz when in my measurements it is generally no more than 6dB.
Give a look to the responses and waterfall obtained by Jzagaja on the pulse responses I obtained using a current source and a voltage source :
http://www.diyaudio.com/forums/attachment.php?s=&postid=1531292&stamp=1212610725
The original pulse responses can be found in text format attached to messages:
http://www.diyaudio.com/forums/showthread.php?postid=1527765#post1527765
this is with a current source (well in fact a 100ohms series resistor
and
http://www.diyaudio.com/forums/showthread.php?postid=1528568#post1528568
with a voltage source.
You'll find also, attached to that message, a graph resulting from simulations under Hornresp for the constant directivity response and the 0° "off axis" response for both the current source and the voltage source cases.
As we see linearity is increased at low frequency using a current source as we can see also on the measurement.
At high frequency linearity is also increased but the difference between the level at 2kHz and 20kHz is still 2 times larger than on the measurement.
May be that's a question on how the pressure is distributed at high frequency at the output of the phase plug. May be the high frequency favor the inner slots and thus the equivalent diameter of the source at that frequency is smaller.
Another explanation is that, unlike waveguides and largely open horns at the throat, HOM are benefitial to the Le Cléac'h horn. It helps in linearizing the on axis response at high frequency.
This will be somewhat of an irony ... the Le Cleac'h horn HOM friendly...
(But I know someone who'll disagree...)
Best regards from Paris, France
Jean-Michel Le Cléac'h
You are right for most part of your reply. The 0° "off axis response" is quite well predicted by Hornresp.
But on your simulation at 0° we can see a difference of level reaching 15dB between 2kHz and 20kHz when in my measurements it is generally no more than 6dB.
Give a look to the responses and waterfall obtained by Jzagaja on the pulse responses I obtained using a current source and a voltage source :
http://www.diyaudio.com/forums/attachment.php?s=&postid=1531292&stamp=1212610725
The original pulse responses can be found in text format attached to messages:
http://www.diyaudio.com/forums/showthread.php?postid=1527765#post1527765
this is with a current source (well in fact a 100ohms series resistor
and
http://www.diyaudio.com/forums/showthread.php?postid=1528568#post1528568
with a voltage source.
You'll find also, attached to that message, a graph resulting from simulations under Hornresp for the constant directivity response and the 0° "off axis" response for both the current source and the voltage source cases.
As we see linearity is increased at low frequency using a current source as we can see also on the measurement.
At high frequency linearity is also increased but the difference between the level at 2kHz and 20kHz is still 2 times larger than on the measurement.
May be that's a question on how the pressure is distributed at high frequency at the output of the phase plug. May be the high frequency favor the inner slots and thus the equivalent diameter of the source at that frequency is smaller.
Another explanation is that, unlike waveguides and largely open horns at the throat, HOM are benefitial to the Le Cléac'h horn. It helps in linearizing the on axis response at high frequency.
This will be somewhat of an irony ... the Le Cleac'h horn HOM friendly...
(But I know someone who'll disagree...)
Best regards from Paris, France
Jean-Michel Le Cléac'h
David McBean said:
Attachments
I have been follow this tread for quite some time but many of the post are above my head.
I just want to ask some stupid question if the waveguide is benefit on dynamic speaker. The reason I ask because I have a pair of Jensen A12 fieldcoil that I really like the tone of it. There is not much low frequency from this 12" speaker so I plan to mate it with 15"woofer that play from 500-600Hz below and use Jensen A12 from 600Hz up. So I wonder if a short waveguide will benefit the Jensen A12.
I just want to ask some stupid question if the waveguide is benefit on dynamic speaker. The reason I ask because I have a pair of Jensen A12 fieldcoil that I really like the tone of it. There is not much low frequency from this 12" speaker so I plan to mate it with 15"woofer that play from 500-600Hz below and use Jensen A12 from 600Hz up. So I wonder if a short waveguide will benefit the Jensen A12.
Jmmlc said:
Hi Jean-Michel,
Many thanks for the additional data. The measured on-axis SPL response of the 320Hz Le Cléac'h horn certainly does seem to hold up very well at high frequencies. Unfortunately there is not much that I can do about it, as far as the Hornresp directivity model is concerned
.In the attachment to your Post #523 it is interesting that the simulated directivity polar patterns are generally considerably narrower than the measured patterns, which I would have thought would have boosted the predicted on-axis response above that of the measured response at 20000Hz, but apparently this is not the case.
http://www.diyaudio.com/forums/showthread.php?postid=1528921#post1528921
Thanks again for the great info.
Kind regards,
David
Hello David,
Concerning the attached file:
http://www.diyaudio.com/forums/attachment.php?s=&postid=1528921&stamp=1212595234
I did a mistake when regrouping the different individual graphs on a single one (unfortunately, Diyaudio allows only 30 minutes of edition for the messages after it appears on the forum).
As correctly written, the measured directivty curves were obtained by me on a Le CLéac'h horn having a cut-off of 320Hz.
The mistake concerns the 2 sets of curves obtained using Hornresp which are, in fact, related to a Le Cléac'h horn having a cut-off of 160Hz (here the AH160 is simulated). The directivity curves for the waveguide are obtained on a simulation of a OS waveguide having the same length, throat diameter and mouth diameter as the 160Hz Le Cléac'h horn.
As the 160Hz is considerably longer than the 320Hz when using the same 1 inch driver it is not surprising that its directivity is more pronounced in the high frequency.
I should modify the file...
Best regards from Paris, France
Jean-Michel Le Cléac'h
Concerning the attached file:
http://www.diyaudio.com/forums/attachment.php?s=&postid=1528921&stamp=1212595234
I did a mistake when regrouping the different individual graphs on a single one (unfortunately, Diyaudio allows only 30 minutes of edition for the messages after it appears on the forum).
As correctly written, the measured directivty curves were obtained by me on a Le CLéac'h horn having a cut-off of 320Hz.
The mistake concerns the 2 sets of curves obtained using Hornresp which are, in fact, related to a Le Cléac'h horn having a cut-off of 160Hz (here the AH160 is simulated). The directivity curves for the waveguide are obtained on a simulation of a OS waveguide having the same length, throat diameter and mouth diameter as the 160Hz Le Cléac'h horn.
As the 160Hz is considerably longer than the 320Hz when using the same 1 inch driver it is not surprising that its directivity is more pronounced in the high frequency.
I should modify the file...
Best regards from Paris, France
Jean-Michel Le Cléac'h
David McBean said:
Hi Jean-Michel,
Many thanks for the additional data. The measured on-axis SPL response of the 320Hz Le Cléac'h horn certainly does seem to hold up very well at high frequencies. Unfortunately there is not much that I can do about it, as far as the Hornresp directivity model is concerned
In the attachment to your Post #523 it is interesting that the simulated directivity polar patterns are generally considerably narrower than the measured patterns, which I would have thought would have boosted the predicted on-axis response above that of the measured response at 20000Hz, but apparently this is not the case.
http://www.diyaudio.com/forums/showthread.php?postid=1528921#post1528921
Thanks again for the great info.
Kind regards,
David
gedlee said:
What about this one:
V.A. Hoersch: "Non-radial Vibrations Within a Conical Horn", Physical Review Feb. 1925, pp 218-224 ?
There are at least calculations of equipotential surfaces for different modes, but the assumption of a loop (zero pressue) at the mouth of the horn is not correct.
In the article Hoersch refers to an article by G.W. Stewart from 1920, where he suggested "the precense of other types of vibration within the horn", i.e. that non-radial vibrations may exist. In my ears, this sounds like he is talking about higher order modes. Hoersch attempts to show that they do exist by calculating them.
Mason also shows how to calculate HOMs in uniform tubes in his book "Electromechanical Transducers and Wave Filters" (Van Nostrand, 1942).
In your papers you refer extensively to the books by Morse, which also cover HOMs.
I'm not trying to discredit your work here, but I find reason to believe that others have hypothesized (and calculated) the existence of HOMs in horns long time ago. It has obviously not been taken into account in mainstream horn theory, and noone except you have investigated their effects on sound quality, as far as I know.
Regards,
Bjørn
Attachments
Kolbrek said:
I was not familiar with the Hoersch paper, but yes from what little you show it is possible that he was talking about the HOMs for a conical horn. I would like to get that whole paper if you have it. Caculation of the Eigenvalues for the modfioed Legendre functions is not trivial and I'd like to he how he did it.
Uniform tubes, ala Mason, are not horns and the presence of HOM in uniform tubes has been known for a long time. It was the presence of HOM in uniform tubes that led me to believe that they had to exist in all horns, but Websters EQ could not define this.
In none of Morse's texts or papers are HOMs discussed. I don't understand why you think that they are.
Hence I think that you are partially correct in that Hoersch may have been the first to discuss HOM, but specific to conical horns. However, as I have said, ANY horn based on Websters Equation (virtually all of them in practice) CANNOT have a mathematical formulation for HOMs and hence one can only "hypothesize" their existance and guess at how they might function. I was clearly the first to formulate the mathematics for how one can calculate the HOM for a wide variety of waveguides, including the conical. The significant point here is that there is a marked difference in the conical horn and the OS in that the conical does not have coupled HOMs - it does not generate HOM within the contour itself. This aspect of waveguides was clearly identified by me and once again I do believe that I was the first to develop this highly important aspect of HOMs.
gedlee said:
The math in your book was over my head.
From what I've gleaned studying your work, it appears that the key to avoiding HOMs is to avoid any reflections.
For instance, a change in angle in a horn will cause reflections; this is why diffraction horns are loaded with HOMs.
A narrow angle conical horn would have fewer HOMs than a wide angle conical horn, simply because the transition from the throat to the conical horn will require a less abrupt change in angle. For instance, a compression driver with an exit angle of six degrees would mate better with a conical horn that has a coverage angle of fifty degrees, versus one which has a coverage angle of ninety degrees.
From looking at the OS curve, it's fundamental principle appears to be that any change in angle must be as smooth and as gradual as humanly possible.
Right now I'm making waveguides for a new project, and I'm *forced* to use a mouth that's not round. And I'm paying a LOT of attention to making the transition from circular to triangular as smooth as possible, based on what I know about the OS waveguides.
my new project
Magnetar said:
Huh? Magnetar, why talk in riddles? What specific crossover attributes are you referring to? And what constitutes "hugely increases"? Would that be by 10%, 2 fold, 100 fold? And over what baseline, and what distortion components? Based on your experience I'd expect that it's possible you have something substantive to add here. If so, let's hear it. Otherwise, please spare us the snide asides and junior high school testosterone displays.
Sheldon
I think all books are advertised in this manner. Since I have already expressed my interpretation of your data, and some possible sources causing HOMs, probably some reasonable resoponse would be more convincing that I will find more of what will lead me to believe such math in your book will actually allow estimation of homs of any horn/guide shape. Right now the information shown at you web site only shows wave guide with and without foam. There really is no information showing precisely which HOMs are from the guide/which are from the driver; neither is there information showing that HOMs of the guide are correctly by your equatiuons for at least two different shapes. It there was a series of tests that was ever conducted which people can follow to confirm findings, it would also be more convincing that the technology is realy more advanced. We are not afraid of work you know.gedlee said: