Asymmetric Horn Mouths

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Has anyone thought about, or done some work on, horn mouths that are not symmetric? Where one side of the mouth extends further than the other? Is the effective mouth defined by a line drawn across from one side to the other, or is the effective mouth somewhere further back in the horn? For example, on a line drawn from the shortest edge to the nearest point on the other wall? A rather extreme example is given in the Cornu horn threads, such as the plan in this post:


Related to this, has anyone noticed whether the presence of a surface (floor or wall) against one side of the mouth of a horn significantly affects the tuning (apart from the SPL reinforcement)?
 
Founder of XSA-Labs
Joined 2012
Paid Member

Attachments

  • cornu-04.jpg
    cornu-04.jpg
    103.4 KB · Views: 351
  • 322366d1357653073t-ever-think-building-cornu-spiral-horn-now-you-can-true-cornu-24x3p5-plan25137.png
    322366d1357653073t-ever-think-building-cornu-spiral-horn-now-you-can-true-cornu-24x3p5-plan25137.png
    23.1 KB · Views: 366
Last edited:

GM

Member
Joined 2003
Yes, Bell Labs/W.E. did extensive testing [including anechoic] in the earliest days of sound system design on virtually every aspect of it and why there’s been so little truly new technology and/or improvement over the decades except in the materials and manufacturing sectors.

Of course local boundary conditions affect the horn’s design, performance!

Think of the horn mouth as being a large point source ‘driver’ with an effective ‘Sd’ [acoustical mouth area] of its physical ‘piston’ area plus its local boundary conditions, so adding boundaries increases both the horn’s effective ‘Sd’ and acoustical path-length and why a horn can be truncated quite a bit if folded up into a corner like the Klipschhorn.

Its acoustical axis, path-length is in general defined by the mean of the effective short and long side of the horn, same as cutting a pipe off at an angle or carving a Karlson slot out along its length. Note the dashed line painted on this early W.E. folded horn to aid the installers in ‘pointing’ it at the desired ‘sweet spot’.

GM
 

Attachments

  • wide-range.jpg
    wide-range.jpg
    138.1 KB · Views: 341
  • Like
Reactions: 1 user
Thanks for the insight, GM (can I call you G?) :)
I would classify those horns as having symmetrical mouths. If the dotted line is meant to indicate the "line of maximum SPL", it looks to me as if the effective mouth at HF is somewhat back in the horn.

Looking at xrk971's diagrams, the situation is complex. Each horn has a "short, straight" side and a "long, curved" side. At some point the "long, curved" side ceases to become a horn side and becomes a "local boundary condition". the question is, what's that point?

There's a further complication in that these horns are designed to be mounted on a wall, which adds another boundary condition, but I'm conveniently ignoring that for now... :)

I'll code up a model for the ripple tank simulator. I'll also see if this can be modeled in ABEC, I've been waiting for a reason to learn it.
 
Last edited:
Those are some fun looking horns!

The mouths appear symmetric, but they are not really. What the real issue GM is pointing out is not the symmetry of the mouth, it is the symmetry of length of the different sides of the horn. Which is the same as the symmetry of the horn mouth. In this fancy curled horn that has an "inner" short wall and an "outer" long wall, same as an "asymmetric" mouth. It is the measured path not the appearance of the mouth that matters.
 
Founder of XSA-Labs
Joined 2012
Paid Member
Jrenkin,
The inner wall (curved one) is actually the long wall. These horns are a lot of fun, and they sound great. Very easy to build if you are interested in trying it out. All the 1 dim models utilize a central average axial distance with corresponding cross sectional area. In the context of the 1 dim model, that is all that matters. Of course, that will not accurately model what is happening near the mouth when it expands fast like this, a 2 dim time varying model is needed. Now we are getting into computational fluid dynamics.
 
Jrenkin,
The inner wall (curved one) is actually the long wall. These horns are a lot of fun, and they sound great. Very easy to build if you are interested in trying it out. All the 1 dim models utilize a central average axial distance with corresponding cross sectional area. In the context of the 1 dim model, that is all that matters. Of course, that will not accurately model what is happening near the mouth when it expands fast like this, a 2 dim time varying model is needed. Now we are getting into computational fluid dynamics.

I fell into the visual trap, but at least I had the concept!

Where do I look to earn how to build one? I am sure the WAF is high..:no:
 
Thanks for the insight, GM (can I call you G?) :)
I would classify those horns as having symmetrical mouths. If the dotted line is meant to indicate the "line of maximum SPL", it looks to me as if the effective mouth at HF is somewhat back in the horn.

Looking at xrk971's diagrams, the situation is complex. Each horn has a "short, straight" side and a "long, curved" side. At some point the "long, curved" side ceases to become a horn side and becomes a "local boundary condition". the question is, what's that point?

There's a further complication in that these horns are designed to be mounted on a wall, which adds another boundary condition, but I'm conveniently ignoring that for now... :)

I'll code up a model for the ripple tank simulator. I'll also see if this can be modeled in ABEC, I've been waiting for a reason to learn it.

You’re welcome!

I guess, for sure I’ve been called far worse!

Well, it’s the acoustical axial path-length exit angle, so as the horn’s output begins collapsing [beaming] in the HF, this is its center-line, but due to horn damping and local boundary conditions its acoustical output is falling, so not necessarily its ‘line of maximum SPL’.

Correct, the frequencies increase as one moves back in time to the beginning of the wave-form [moving towards the throat], so all the HF response is back at the throat stub.

I assume the Cornu horn is a foreshortened [truncated] one since it’s designed to be wall mounted, so where the WLs are long compared to its perimeter, the individual horns sum as one and slowly become individual point sources, same as in a stacked speaker array. As such, each mouth abruptly ends at the four corners plus however much the pipe end correction of each acoustically lengthens/expands them.

I’ve never built one, but assume that it’s necessary to damp down any HF coming out of them to keep from having four phantom tweeters around its perimeter. Then again, it may add some interesting ambiance.

When mounted to a wall, it’s in 2pi space, so the horn loads down around an extra half octave due to doubling the effective mouth areas via the mirror effect.

GM
 
Has anyone thought about, or done some work on, horn mouths that are not symmetric? Where one side of the mouth extends further than the other? Is the effective mouth defined by a line drawn across from one side to the other, or is the effective mouth somewhere further back in the horn? For example, on a line drawn from the shortest edge to the nearest point on the other wall? A rather extreme example is given in the Cornu horn threads, such as the plan in this post:



Related to this, has anyone noticed whether the presence of a surface (floor or wall) against one side of the mouth of a horn significantly affects the tuning (apart from the SPL reinforcement)?

Be Yamamura - back in the 90s - designed several such horns and they were very, very good. He made them from cork! The ones I heard were front loading Lowther drivers, and I had some at home for a while. I had a pair which front loaded a smaller driver and lived happily with these and a prototype stereo amp from the same designer. [The drivers were also rear loading a conventional horn]

The pic shows the first prototype....much bigger ones followed!!:)
 

Attachments

  • horn be 1st. cork:pm lowther.jpg
    horn be 1st. cork:pm lowther.jpg
    670 KB · Views: 205
...
Well, it’s the acoustical axial path-length exit angle, so as the horn’s output begins collapsing [beaming] in the HF, this is its center-line, but due to horn damping and local boundary conditions its acoustical output is falling, so not necessarily its ‘line of maximum SPL’.

Ah. That's what I meant, I just said it poorly. Thanks.

I assume the Cornu horn is a foreshortened [truncated] one ...
I’ve never built one, but assume that it’s necessary to damp down any HF coming out of them to keep from having four phantom tweeters around its perimeter. Then again, it may add some interesting ambiance. ...

Yes indeed, they are truncated. The builders have found it necessary to add significant damping in the driver chamber and sometimes in the throats of the horns. The damping has proven effective in damping the resonances and suppressing much of the HF.

I can see how the differing path length of a curved horn may matter for a HF horn, but I don't see that it will greatly affect the performance of a truncated horn intended for bass reinforcement, where you can bend, fold, spindle and mutilate the path without greatly affecting the performance in the lowest octave or so. The length matters as much as the profile, which is where we came in - how to determine the effective length of a horn with an asymmetric mouth.
 
  • Like
Reactions: 1 user
Be Yamamura - back in the 90s - designed several such horns and they were very, very good. ...

I see what you mean about the asymmetric mouth on the front horn. The effective length can probably be estimated quite accurately, according to GM:
Its acoustical axis, path-length is in general defined by the mean of the effective short and long side of the horn, same as cutting a pipe off at an angle or carving a Karlson slot out along its length.

The problem with determining the effective length of the Cornu horn is: what's the effective length of the longer, curved side of the mouth? At what point does it cease to be considered as part of the horn mouth and begin to be a "boundary condition"?
 
frugal-phile™
Joined 2001
Paid Member
The problem with determining the effective length of the Cornu horn is: what's the effective length of the longer, curved side of the mouth? At what point does it cease to be considered as part of the horn mouth and begin to be a "boundary condition"?

My 1st guess would be where the normal to the long curve intersects the corner of the square + an end correction.

dave
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.