Can someone check over my HF horn design to make sure I'm not crazy?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
"K402 does not appear to be conic..."

While rectangular, the initial expansion is conical, so it has very, very good dispersion in the top octave. It has a large roll-over at the mouth that they call tractrix.

"even so, I am pretty sure that if one ran the polar response of that K402 one would not find any lack of beaming,"

I've seen the polars, and you would be wrong.

"unless there was a diffraction at the throat entry or something in the driver causing HF dispersion."

It's called a phase plug that is flush with the horn entry.

"Can't seem to quickly find the AH! horns on a search, something came up but nothing definitive. I seem to recall reading about them before, maybe here?"

More likely at the Audio Asylum.

http://web.archive.org/web/20090223220729/http://www.acoustichorn.com/products/index.html

"Are these horns starting with a conic (straight side) expansion then converting to something else to limit issues at the mouth?"

Bill's horns are straight conical (some with a roll at the mouth).

retro vintage modern hi-fi: AH! Horns by Bill Woods

"Thirty years of designing loudspeakers, loudspeaker systems, and horns have resulted in the Acoustic Horn Company family of horns.

Ultra clear voice structure, delicate textures and explosive dynamic range set AH! horns apart from the rest.

As senior loudspeaker engineer at Yorkville sound, I have designed hundreds of transducers, horns and various high end systems for the past 18 years. Using the most advanced test equipment, the availability to test any driver or horn made today, has resulted in the AH! horn family.

If you desire the sound of a piano right in the room, a snare drum right in the room, the human voice so clear you can practically talk back to the system, then AH! horns are for you.

Honest test curves, full customer support in designing a system with your drivers, or our own systems, you will hear sound reproduction second only to the real thing.

No horn sound, perfect acoustic phase, constant directivity, and built to last more than a lifetime, AH! horns are a benchmark in critical playback systems. Do you have a Tad 4000 series, JBL 2440, or Altec 288 drivers? If so, then these horns are for you. AH! horns have coverage patterns of 40, 50, 60, and 80 degrees to fit your room requirements. AH! horns work especially well with the new B&C loudspeakers DCX50 cone diaphragm drivers.

Please give AH! horns a listen. Thank you. "
 
Last edited:
About Conical Horns
The conical horn is the simplest horn shape. The cross-section increases linearly, like a cheerleader's megaphone. Compared with all other horn flares, conical horns have a precisely defined radiation characteristic. There is no disruption to the wavefront as it moves out towards the horn mouth. The "horn sound" is nil with a conical horn. Because the wavefront has a smooth passage, it has perfect phase.

Another advantage is that the cross-sectional area in the vicinity of the horn throat increases more rapidly than for exponential horns. This thus reduces the sound pressure in the horn as well as distortions due to the compressibility of air.

What conical horns sound like....

Since conical horns have no curvature, the ever emerging wave is not deformed as it moves from the throat to the mouth. This means that there is a linear pressure change throughout the horn.

When this condition is met, the horn will have very low 'air column' distortion, or distortion cause by squeezing the air at the throat. Conical horn also have good phase response, due to the reason stated above.

In short, the conical horn amplifies the sound, with the least disruption to the sound wave.

The sound of a conical horn system has very little trace of the "honky" sound usually associated with horn. Because the horn has good phase response it can be crossed over with another horn or direct radiator very smoothly.

Directivity is another positive feature of a conical horn. The sound goes where you want, cutting down unwanted room reverberation.

The end result of all of this is that when you hear a horn system, it has a very lifelike sound. Bells and horns sound like the real thing. Nothing comes close the sound of a cello (especially the Mercury recordings with Starker.)

If you were to clap your hands together in an auditorium, you will hear the echo more clearly- you literally hear more information. This is due to a linear wavefront and linear phase.
 
Conical Horn Geometry
I’m often asked how horns differ from one another, and also how different types of horns differ in sound and design, such as exponential, tractrix, and new versions of these which carry the names of their designers. I’ve put together some observations, backed up by actual frequency response curves and measurements, so that people can decide for themselves what types of horns would work best for them.

For the past 30 years I have designed loud speakers for professional applications and more recently for home HiFi. While designing at Yorkville Sound and in the course of my career, have built and tested numerous exponential and tractrix horns. This includes hybrid profiles using a tractrix and a conical which were created to fulfill specific tasks. In professional audio, horns have to work as intended. They cannot be of unlimited size - in fact, there exists an obvious commercial advantage to building the smallest horn to do a given job well. Tractrix or exponential horns are rarely used in professional applications where clarity and coverage are an issue.

In the early 80's I built a four sided conical prototype to be converted to a bi-radial design. I was astounded by how good it sounded so I measured and tested it. This opened the door to the study of conical horns. Some of the characteristics of conical horns are….

•Good directivity or power response - the hallmark of the conical horn.
•Conicals keep the sound wave at 90 degrees to the wall of the horn at all times as the wave exits the horn.


...After Hughes (Peavey, Also Morse)

(this defines a conical.) The sound wave is not disrupted by the walls of the horn. For these reasons, conical horns have the best sound of any horns for home HiFi Use.

The accidental conical horn

The tractrix profile has been used in commercial products and in DIY. All exponential, and tractrix horns tend to beam as they go higher in frequency due to the squeezing of the side wall of the horn. This results in a "honky" sound and some of the high frequency information is lost, when listening off axis. This is the case no mater what type of mouth correction you might use. The tractrix profile itself is a form of mouth correction.


My first tractrix in production 1989


I have experimented extensively with all types of horns, using absorbing materials, mouth geometry, adding a foam ring at the horn’s mouth, cutting slots, and also with acoustical lenses. (see photo) I have discovered that these devices have, at best, an ameliorating effect, often redistributing the anomalies they try to correct to other places in the passband instead of eliminating them. The acoustic lens holds the most promise.



It is interesting to note that RCA acoustic laboratories analyzed the conical horn flare and mouth termination issues in the 1950's. I am lucky to be in possession of the very lab sample test horns, of which my work is based. See picture.




In listening tests, the rollback using pink noise is very slight. Few individuals have the time, money, space, and technical ability to build and compare using state of the art hardware and software. (I use a TEF analyzer, ATB B&K ,AKABAK, AJ horn software, etc).

I can, however, suggest a simple way to make some experiments for yourself, to see how horns differ from each other in a very rudimentary way. Choose a compression driver, and make a number of very short horns of only 4-5 inches in length ---really horn adapters-- for your driver.

The most critical part of a horn is the first few inches. This is often over looked. Start with a section of straight pipe. Listen to that. It will be very loud, and very colored. Next, try an exponential horn of the same length, then a tractrix - they will be progressively less colored. Finally, try the conical section. It will be the most natural sounding and uncolored.

Musical instruments such as trumpets and trombones are not conical. They are designed to produce harmonics, the very thing we try to avoid in a sound reproducing device. The more the horn is squeezed, the more the harmonics. On the other hand, voice producing devices such as megaphones are always conical.

Horn design is always a game of balancing the factors of length, angle, directivity, low frequency cutoff. There are many ways to tweak a horn’s response. One is to add an exponential throat adapter to a conical horn. This colors the sound, but improves the measurements. Another is to add a secondary flare to a conical to deal with the internal reflection issues.

One other area which deserves discussion is the crucial issue of the beginning of the horn- the throat- and the drivers being used. Compression drivers all have a horn beginning within the driver which you cannot see. The internal geometry of the driver and its phase plug has a direct relationship with the horn.

Sometimes a small exponential throat will be of benefit, sometime it will not. Once again it is a question of degree, and audibility.

If a driver has an exit a half inch, you will likely get to 12kHz, but a driver with a 2 inch opening will be lucky to reach 8kHz. Directivity decreases as the throat diameter increases. Also, if a 2 inch exit diameter driver is being used on a horn expected to go up high, the issue of mouth reflections coming back into the horn pales compared to the very real problem of high frequencies getting “stuck” across the 2 inch throat opening of the horn

As you can see, designing horns is actually a very complex equation involving parameters that all need to be addressed so that an optimal design results. There is no such thing as an “ideal” horn. But certain basic flare profiles work vastly better than others if the best sound quality is the most desired result.

Some general horn examples under test.
I have hundreds of pictures, hundreds of horns, but this is enough for now.








AH 700 with and without flare and throat...it's not as though I missed it somehow...


A measurable difference, but not audible.


Barely audible, but does one sound better than the other?


In this case the off axis is better without the rollback.
This driver has 3" throat in the driver.it is a very early Altec 288


a 40 degree horn with a 2" driver.


Keele angle rollback




An example of a small exponential throat section added...


a 90 degree horn vs a 50 degree horn, both with rollback




Itawa horn, and odd drivers I use to cover a complete range of drivers when testing a horn
Also A multiple flare horn after Olson

We stand on the shoulders of giants...

Olson
Holland and Newell


I built the Holland horn soon after his paper came out. the gray one was cut down from a 24" mouth version
Morse
Geddes


Here is 90 deg conical with and without rollback. Also a tractrix, and a tractrix mouth coupled to a conical throat
 
Megaphones have always been accused of the "Rudy Valle" sound - accentuating a narrow band.

Of course if you open the conic section, it starts to look like a speaker cone... open it more and it is a flat disc.

A discontinuity in the throat, going between whatever the expansion of the driver's throat and some conic expansion will create response anomalies and diffraction. I'm certain you know this.

Then too there is the issue of what happens at the mouth, once the horn is no longer there... JMLC did a lot to get this part figured out. Sorry that he is no longer with us. :(

Geddes oblate spheroid, and what was done at Peavey are attempts to minimize that effect while getting the benefits in terms of polar response that one might realize with a very wide horn with straight sides. Of course, the wider the horn, the less the loading. That's intuitively obvious by considering the difference between a long tube and a flat baffle. The horn is somewhere in-between? :D

Back a few decades ago now, when Bruce Edgar had rediscovered the Tractrix expansion, he used to do a demo. The demo consisted of his speaking through a standard conic horn (aka megaphone) and a tractrix horn - both of similar acoustic/dimensional qualities. He did this to illustrate just how different they sounded. Indeed, the tractrix appeared to show horn gain without audible anomalies, whereas the conic horn was, well, a megaphone.

Looking forward to viewing the images that did not appear - suggest attaching them wia the attachment manager, rather than putting them inline...
 
Last edited:
"Indeed, the tractrix appeared to show horn gain without audible anomalies, whereas the conic horn was, well, a megaphone."

I call BS.

If this would be true, why is the best sounding horn from Klipsch (K402) a conical expansion with a large roll-over?

The reason why a megaphone sounds 'horny' is because of its finite size, the lower part of the voice range abruptly losing its loading.

Go buy the 18 Sound conical elliptical horn and then go buy the Faital Pro tractrix elliptical horn and have somone talk through both of them.
 
About Conical Horns
The conical horn is the simplest horn shape. The cross-section increases linearly, like a cheerleader's megaphone. Compared with all other horn flares, conical horns have a precisely defined radiation characteristic. There is no disruption to the wavefront as it moves out towards the horn mouth. The "horn sound" is nil with a conical horn.
Cross sectional area increases quadratically with length. Horn honk has more to do with an impedance mismatch at both the throat and mouth. A proper mouth termination, like a big rollover, fixes that.
 
"Horn honk has more to do with an impedance mismatch at both the throat and mouth. A proper mouth termination, like a big rollover, fixes that. "

So does crossing it over at a higher frequency than it unloads at.

"The reason why a megaphone sounds 'horny' is because of its finite size, the lower part of the voice range abruptly losing its loading."
 
"Indeed, the tractrix appeared to show horn gain without audible anomalies, whereas the conic horn was, well, a megaphone."

I call BS.

If this would be true, why is the best sounding horn from Klipsch (K402) a conical expansion with a large roll-over?

The reason why a megaphone sounds 'horny' is because of its finite size, the lower part of the voice range abruptly losing its loading.

Go buy the 18 Sound conical elliptical horn and then go buy the Faital Pro tractrix elliptical horn and have somone talk through both of them.

Not to mention conical will likely require constant-directivity conjugation, so Dr. Edgar's voice was effectively low-passed.
 
I looked at the 18 sound horns, they are hardly conical in any traditional sense. They represent essentially a refinement of the oblate spheriod concept as far as I can see. The addition of a flat expansion after an elliptical or spheroidal does not in my book yield a "conic horn".

The early Danley Unity horn design is/was an actual conic section, although I always questioned the entry point of the compression driver.

State-of-the-art in "horn" design has certainly gone beyond Tractrix at this point. But the SOTA is not a traditional straight sided cone either. I don't think that the newer horn designs ought to be referred to a "conic" at all, as they are not, except (imho) in a limited way.

_-_-bear
 
Eh?
Something is amiss here.

They are not IN the post. Nor are there active links. Not on my machine/browser.
So, what are you trying to say?
What good would archive.org be?
your post is only days old, I never saw any graphics in the post.
am I the only one who does not see the images?? <---- is everyone else seeing them??
 
Ok, so the wayback site has most of Bill Woods' site there.

Are you Bill Woods?
If not, what is the purpose of cutting and pasting his words so extensively in your posts?
If you have a point to make, why not just make it yourself?

Also, I do not see any measurements thus far for the claims made?

In addition there remains the issue of how the transition from the expansion of the typical compression driver to the "conical horn" was handled. The expansion of the typical compression driver is intended to match the expansion of the typical exponential horn, any variance from that is going to result in a curve that is something other.

The abrupt end of a horn has been shown by JMLC and others to have negative results, even when one might think the wave has been "launched"...

So, I'm entirely unclear on all of this. Care to straighten this out in your own words?

_-_-
 
"If not, what is the purpose of cutting and pasting his words so extensively in your posts?"

Bill woods is brilliant, and known industry wide. His opinions should carry more weight, I also agree with him.

My HF horn building career is very limited compared to Bill. Every horn was either conical or parabolic (a little like the JBL 2397). After 1980 it got less expensive to buy good horns, so I only did two designs after that (hand-laid fiberlass is too time consuming and expensive, even if you have a chopper).

My reply in post #9 was clearly not understood by the OP, and things went downhill from there.

I wish I had known Bill's site was going down, I would have downloaded it all. So much was not archived by Wayback.
 
Well understood or not by the OP, I do not know.

One of the things to do when someone who does not know asks is to provide more than just the single example, like a page from Olsen or Beranek without comment. Fill in the spaces with explanation.

I recall now, Bill Woods' site after having looked at it. As with many things that you can only see a picture of, what he was offering did not appear to hold too much promise. And, as I never got a chance to hear any, nothing more happened here. It's interesting to note that Jonathan Weiss of Oswald's Mill is championing these same designs now...

But you made one comment, or quoted one, about a phase plug in the throat?
What is/was that exactly??

Yes, Wayback seems to skim the surface, shame.

_-_-
 
Not surpricing, from the Oswalds Mill Audio site:
Oswalds Mill Audio said:
Our Designers


Our acoustical designer Bill Woods has been designing horn loudspeakers professionally for over 35 years. He is one of a handful of the world's top experts on horns. Once Bill has set the acoustical parameters for a proposed speaker, such as the size and flare of the horn or horns, enclosure volume and geometry, selection of drivers, etc, OMA's industrial designer David D'Imperio is tasked with creating a visually arresting design that maintains it's acoustic integrity. The history of loudspeaker design shows that aesthetic beauty and acoustic excellence have rarely occurred in the same package. D'Imperio creates an individual identity for each speaker design, following from its acoustic goals
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.