For a given amount of "extra width", what is the most optimal/efficient profile for the termination of a horn or waveguide mouth? Obviously looking to minimize on- & off-axis aberrations and other diffraction effects.
I think it was mentioned somewhere that a plain radius would not be best, but rather something more like an Euler spiral with a smoother derivative curve.
And would it make any difference, profile-wise, if the mouth were terminated into a flat baffle vs free-standing and needing a wraparound?
I think it was mentioned somewhere that a plain radius would not be best, but rather something more like an Euler spiral with a smoother derivative curve.
And would it make any difference, profile-wise, if the mouth were terminated into a flat baffle vs free-standing and needing a wraparound?
The flaring of conical horns is discussed here (empirical):
http://www.xlrtechs.com/dbkeele.com/PDF/Keele (1975-05 AES Preprint) - Whats So Sacred Exp Horns.pdf
This method however introduces a discontinuity which may not be desirable.
Probably there is no general solution and there is a dependency on the profile of the rest of the waveguide. Tractrix - Wikipedia seem interesting for termination on a flat baffle. Perhaps check the Ath4 thread?
http://www.xlrtechs.com/dbkeele.com/PDF/Keele (1975-05 AES Preprint) - Whats So Sacred Exp Horns.pdf
This method however introduces a discontinuity which may not be desirable.
Probably there is no general solution and there is a dependency on the profile of the rest of the waveguide. Tractrix - Wikipedia seem interesting for termination on a flat baffle. Perhaps check the Ath4 thread?
There are two ways you can approach it. The tractrix/LeCleach/smooth curve method reduces the axial reflection but it tends to draw the low end out wide with it.
They did some investigation in the ATH thread. I remember one MABAT did that curled 270 degrees from the driver diaphragm and it looked fantastic out to huge angle, like behind it even. It was a continuation of the flare, maybe OS?, but that last 180 degrees wasn't all that much different than a simple roundover.
Pattern control also went much lower in frequency than you would expect given the mouth diameter.
Pattern control also went much lower in frequency than you would expect given the mouth diameter.
Last edited:
What you want is a smooth curvature change, which also means that a flat baffle termination and a free-standing termination require different solutions. A flat baffle has zero curvature, so you need the curvature of the profile to smoothly decrease to zero at the point where it connects to the baffle. That requires some "extra width" compared to a free-standing termination, which can be of a smoothly increasing curvature all along (to a point where most of the sound energy already "left the walls").
mabat are you saying its better to begin increasing the curvature from the start of the horn rather than the final 1/3? won't this lead to something like an exponential horn where the directivity narrows in the HF?
Sorry, I was talking specifically about a termination alone, whatever the initial profile is. But if you start mildly enough, you can do so quite near the throat.
What I think is the optimal waveguide profile, including a natural flat baffle termination, is probably already well known: http://www.at-horns.eu/release/OS-SE Waveguide.pdf - it's as smooth as it can be (note the curvature curve on Fig.5, page 7)
What I think is the optimal waveguide profile, including a natural flat baffle termination, is probably already well known: http://www.at-horns.eu/release/OS-SE Waveguide.pdf - it's as smooth as it can be (note the curvature curve on Fig.5, page 7)
Last edited:
my current project is making a horn for the M200 midrange compression driver for PA usage (400 - 4kHz) so I don't think I can use the OS-SE profile due to poor low frequency loading. My investigations so far have been around conic exponential horns as per the Keele paper, but trying to avoid discontinuities. I haven't run any simulations yet with sharable results yet though as I have been writing software to generate the files for simulation in Akabak.
figure 5. I find amusing how it ends up looking like a Qaudratic throat waveguide with mouth termination:
http://www.excelsior-audio.com/Publications/QTWaveguide/QTWaveguide_WhitePaper.pdf
My current horn becomes conic so I think I can probably join it to the OE-SE for mouth termination.
figure 5. I find amusing how it ends up looking like a Qaudratic throat waveguide with mouth termination:
http://www.excelsior-audio.com/Publications/QTWaveguide/QTWaveguide_WhitePaper.pdf
My current horn becomes conic so I think I can probably join it to the OE-SE for mouth termination.
Then why not just add the SE term from the OS-SE formula (starting with Ls/q). It's pretty universal: SE term onlykipman725 said:
I think that the Qaudratic throat was actually an approximation of the OS profile.
Last edited:
It's possible but at the cost of a beaming response. This is also an OS-SE profile: Acoustic Horn Design – The Easy Way (Ath4) (you don't have to go to such extremes though).
Awesome, thanks. I definitely remember seeing this years ago and just couldn't find my way back.
And of course there is always the option of soft termination. AKA "Mad Dog" termination (foam at the mouth). Many ways to do that and the results are certainly audible.
This is what the pioneers did when stuck with a too shallow a depth for the desired [bass] horn response, so truncated it as required to get the critical speech BW loaded, then made a baffle with the necessary area/shape to fill in the LF and in a few cases for the HF:
800 VOTT
http://greatplainsaudio.com/wp-content/uploads/2016/06/A7.early_.jpg
Damp the junction perimeter and/or the whole baffle as desired for HIFI/HT apps.
Can you point to some of the ways? GM has also mentioned this but I haven't found any how-to's specifically for conical horns.
Surprising how far into the horn it goes. It's too dark to see the interior edge of the foam but presumably the sheet thickness is tapered toward the throat for a smooth transition.(?)
Thanks.
Well, a basic square/rectangular conical horn [DSL Synergy type] 'flips' out where the horn's long enough to meet the 1 kHz? point with the rest loading to 250 Hz? [don't remember ATM, so instead of the flip out, make a big enough baffle to load it down to the desired Hz using one of the OB calculators.
Some folks like to cover the transition with open cell foam while others stick down the horn ala Dr. Geddes, while me n' some others just damp the perimeter of the horn [Pano's 'Mad Dog'] or in this case the perimeter of the baffle or cover the whole baffle in some cases.