Idiot question #4 - Horns

[filmboydoug]

Perusing web sites I see horn types of:

Exponential

Radial

Bi-Radial

Constant Directivity

1. Is one of these types a non-trademarked term for Tracktix?

2. Are any of these better for hifi applications? I assume the biggest difference would be their dispertion characteristics, but are any types better as far as distortion goes?

[DrWho]

Tractrix and Conical are two other types that you didn't mention that are different from the rest.

As far as differences...each has their own bandwidth over which they

are best suited. For example you don't want tractrix on the low end

because it ends up being much much bigger than a conical horn of the

same cutoff.

Constant Directivity is probably the only one not used in HIFI because

it requires extra filtering to achieve a flat frequency response (ahhhh

filters = bad...or so says the audiophool). It is more often used in

sound reinforcement applications where an even and wide dispersion

pattern is required - and the extra filter isn't audible (and is

usually built into the crossover unit).

As far as distortion - you would be hardpressed to claim any type of

flare is better than another because they are used in different

applications (so it's not apples to apples).

Just curious...why do you ask?

[filmboydoug]

Mostly trying to learn by osmosis. []

I have a habit of plucking a bit of information here and a tidbit there and forming conclusions. Sometimes incorrect, so I ask a ton of dumb questions to find out for sure. So Drwho, here is a question for you. After reading your above answer, I am wondering if Al's Klapperhorn (that was for Craig) is preferred by so many over the K400 because the tracktix curve is superior, or simply because Al's execution of the horn is better?

BTW, all my other idiot questions have been for a project I am still hashing out. I'm 99% sure I'll use an Eminence H290 horn for that.

[DrWho]

Seems we're in the same boat...I'm trying to learn by osmosis too []

There is some good argument to claim that a tractrix expansion is better in the midrange...you will notice that all of klipsch's new designs are using it too.

The reason the heritage series is still using exponential is 100% in honor of PWK - who very adamently preferred exponential over tractrix. Exponential has the best efficiency and PWK wouldn't implement anything that reduced it. But all theory aside in the end it comes down to which sounds better - and there was a lot of disagreement in that regard too []

So all that to say...it wouldn't be too far out there to claim that any of the differences could reside within the window of placebo...

[Xover]

Since I have been popping in and trying to answer some of your questions here I am again.

So all I really know about horns are those horns lubricated by testosterone. So I wouldn't be taking any of my advice in this area.

I like you many years ago studied horn designs and after it all my conclusions were just summed up pretty well by Dr. Who.

Keep picking other folks brains and sort out the chaff then let your own ears decide.

Last horns I used were EV 8HD with 1824M and T350 thought they were good. I did like the sensitivity of the T350. I did think that there were some very efficient direct radiating drivers that were better than the 8 HD/1824M but they were very expensive.

[WMcD]

You should go to www.dbkeele.com and download his paper Whats so Sacred about Exponential Horns. There is a lot compare and contrast.

By way of background:

The classic horn is exponential. The exponential equation dictates that cross section area doubles every X distance from the small end to the big end. The equations behind it assume the wave down the horn is flat, which really can't be true. The exponential equation really does not tell us anything about the shape of the horn itself.

A radial horn is usually exponential in expansion. But the side walls are flat. The top and bottom look like the bottom of a salad bowl. You could make something like it by cutting a salad bowl into quarters. This actually honors the thought that the wave is curved.

In a biradial horn, both the top and sides are curved. In recent times this refers to the bell (bigger end) of a constant directivity horn.

One joker in the pack is just how the curved wavefront intersects the sides of the horn structure. Because of drag, the intersecton should be right angles. The tractrix horn does that. See my post earlier today in Tractrix and Voigt.

To some extent most horns beam to some extent and it is not constant with frequency. The pattern narrows. This means there is some "gain" on axis at the expense of off axis level. If we need that constant directivity the conical and tractrix does that.

Conical horns have poor bass loading and Don Keele solved that. However, it looks like the old tractrix by Voigt did the same thing in a similar way, about 50 years earlier.

Drivers typically have poor performance (output) at higher frequencies. So the beaming in some designs compensate. This is controlled directivity. So actually when people say that constant directivity horns need equalizing, it is to so the beaming (which is not there) has to be made up for.

Eg. is you look at Keele's paper he tested the final product. The driver output had to be equalized to get the output of the horn flat. But that is to say the horn WAS flat and the driver was at fault..

Gil

.

[djk]

http://aa.peavey.com/downloads/pdf/qwp1.pdf#search='hughes%20horn'

[billyjoe72]

No such thing as a stupid question. Ask, listen and learn I say.