Need a Klipsch history lesson.

[m00n]

Why do some Klipsch have horns for midrange while some have regular woffer style midrage.

Example, Chorus, chorus II, Cornwall, etc etc. they have a horn midrange, yet the RF7s have a woofer midrange... Why is that?

Also, why horns in the first place? What are the attribues of horns that Klipsch see's as being so valuable to them? Why does Klipsch favor horns over regular woofers.

[yaffstone]

Heratige series used horns exclusively. Advantage is very high efficiency due to small size of driver, very low distortion due to small size of driver. Note that the bass bins in the k-horns, belles, and lascalas are folded horns by virtue of the cabinets while the bass drivers are large cones. High efficiency was very important in an age of limited power in amps. The more modern designs of the reference series start to use cones. I haven't looked at the specs but one can easily imagine that they no longer deliver 100 db at 1 W since power is cheap these days.

These lame reasons are my relatively uninformed take on the subject and are not to be construed to be pure fact, true as they may be, or not.

[m00n]

Ok... So essentually (sp?) horn help amplify db levels in small drivers?

How important is the shape of a horn? Do different shaped horns produce a different sound or do they just have an affect on db levels?

[michael hurd]

A horn allows a small diapragm ( simplified piston ) to move a very large amount of air with relatively small movement. Basically the horn allows a smooth transformation from the throat area ( small end ) to the mouth. ( high pressure at throat, low pressure at the mouth or big end )

The diaphragm is much larger than the throat opening, for example a 2" diaphragm compression driver with a 1" throat. This forces a relatively small amount of air through an even smaller opening, creating high pressure. This area of high pressure is transformed to a low pressure at the big end or mouth of the horn.

Horns can be designed with different angles of dispersion in the horizontal and vertical plane.

[m00n]

Interesting.. Very interesting. I would be curious to see what happens if I was to take some little computer speaker I have at home, build a box around it, choke the driver and build my own wood horn.

Is there a formula for the size of the compression hole to the size of the driver and horn?

[m00n]

----------------

On 3/14/2005 6:50:21 PM michael hurd wrote:

The diaphragm is much larger than the throat opening, for example a 2" diaphragm compression driver with a 1" throat. This forces a relatively small amount of air through an even smaller opening, creating high pressure. This area of high pressure is transformed to a low pressure at the big end or mouth of the horn.

----------------

The more I think about this, the more I want to ask... Why? Why have a smaller throat than diaphragm? Just to create a high pressure? Will having a throat the same size as the cone not create enough pressure? Seems to me that having restriction like this would alter the overall sound of the diaphragm. But.... maybe that's a good thing.

[WMcD]

Yikes, I post all sorts of articles and some still questions are asked which are answered in those articles.

Please see the Bruce Edgar interview of PWK. Dr. Paul describes the experiment of mating a small direct driver to a horn.

There is a lot more to it of course. It was actually Edgar who turned me on to it in the first place.

For low distortion you need low diaphragm (radiator) excursion, or movement. This is the goal. You have to keep your eye on the prize as they say.

One way of doing this and having good output levels is to have a very large diaphragm in free space. So even if the pressure generated per square inch is low, you have many square inches.

There is something else complicated going on. That is radiation resistance of the air; which is typically better with bigger size. Bigger radiators are better, without getting into something real complicated.

But the diaphragm would also have to have very low mass so that you can get it moving in time to high frequencies. This creates a contradiction. Big diaphragm in free space, low mass. (Admittedly, one way of doing this is with a big piece of cellophane, which is how electrostatics and Magneplaners work; driving the entire area is tricky though.)

The other way is to use a horn. The mouth (big end) acts like a big radiator with no mass at all. This is somewhat a thought experiment. One side of the mouth is looking at free space with air to influence (just like a big diaphragm would) and the other is looking into the horn.

So we've got the big radiating area issue nailed down. But like with the cellophane, we have to get pressure to it to drive it. The side of the thought experiment diaphragm of the mouth has to be driven.

Now, how do we create some pressure at the big end? The answer is the horn, which we here look at backwards. We taper it down to a small area (throat) where we can put a small radiator. Small radiator, small mass.

The small radiator can create high pressure (and volume of air flow) because it is looking into the pipe of the horn. Here the small driver has low excursion because it can push on the air in a small area in front of it. The small driver is an effective pump.

Let's crank that the other way, the ususal way, travelling from the small end of the horn to the big.

At the small end we have high air pressure applied to a small area. Small excursion, low distortion, and relatively light weight diaphragm.

Then we allow the wave to grow as it travels toward the big end of the horn. It is an exponential growth meaning that the area doubles every X distance. Then the wave is at the mouth where big area of the wave has been traded off for low pressure. But that, again is okay, we've got big area at the mouth.

I've explained that a bit backwards (at first) and focused on mouth size and then gone into the horn as it narrows. A very important issue is the relative size of the mouth which is not usually explained. We can dwell on it:

Let's look at the LS bass unit. Its mouth is about 4 square feet. It would take, roughly, six 15-inch drivers in a direct radiator bass bin to have the same radiating area. But we excite the small end of the LS with only one such driver. The theoretical bass bin would be much larger overall.

Similarly, let's look at the K-400 midrange. It has a radiating area at the mouth of about 100 square inches. The driver on the small end has a radiating area (diaphragm) of about 3 square inches.

There is a lot which can be done with the overall shape of the horn to make the wave coming out the big end more directional. That is a much longer story.

Gil

[m00n]

Sorry Gil. I don't venture up in 2 channel land all that often. I must have missed your posts.

[JohnA]

Moon,

Gil did a great job, as usual, without being too technical.

I want to add one thought I think you will follow. There is a compression ratio in horn drivers just like in an engine. Driving a 3/4" diameter opening with a 2" diaphragm (K-55-V) gives you that compression ratio. Within limits, the higher the compression ratio the higher the efficiency, just like in an engine. Putting a ratty, 3" computer speaker at the mouth of a horn will work, just like a cheerleader's megaphone. Putting a chamber between it and the horn that causes compression works better.

[formica]

----------------

On 3/14/2005 9:43:21 PM m00n wrote:

Sorry Gil. I don't venture up in 2 channel land all that often.

----------------

Gil's more of an engineering type... so you are better off at checking "Technical Questions" for fellows like him...

[J M O N]

Moon:

Horns have (at least) the following advantages over direct radiators (cones):

1. Higher efficiency

2. Lower distortion

3. Greater dynamics

The physics behind why this is true have already been touched upon by previous posts, so I won't get into further.

[WMcD]

I must say my mea culpas for some mildly cranky comments which were in response to our buddy mOOn. Sorry.

This coming weekend I'll post Don Keele's article on optimum mouth size. It will make most sense to technical types.

Horns are a type of finite length transmission lines. For optimum transmission we have to look at what is going on at both ends.

The paper by Keele about use of T-S parameters told us about the sending end at the throat. Now we have to look at the receiving end, at the mouth, which radiates SPL into space.

Also, if things are not correct at the big end, the wave is reflected back to the small end.

Gil

[colterphoto1]

well spoken Gil- I understood that!

Michael

[Deang]

Since we are on the subject, I have an issue that I'm really struggling to understand.

I've been studying the JAES article on the Jubilee, and reading just about everything under the sun on compression drivers and the various horn types. One thing I can't get a handle on is how anything good can come about by using any of the drivers I've looked at in a two-way design. All of the drivers I've looked at so far (designed for low crossover points) drop like a rock after 10kHz, and in fact, most of them have response curves that look like crap after 8kHz. Sometimes they say the plot reflects the response with such and such a horn, sometimes they don't. Apparently the "right" horn can stretch out the response of the driver, because if it doesn't, I sure don't see how you make a decent two-way with any of them. I don't see any short cut formula for prediciting the behavior of any given driver/horn combination, and it's extremely frustrating.

I would love to try to figure out a nice top end (horn/driver) for the LF section of the pro-Jubilee (KHJ-LF). I can afford some nice drivers and horns -- what I can't afford is an Engineering degree!

[MrMcGoo]

Dean,

The material used in the compression driver affects the frequency response. Hence, the lightness of the titanium in the Reference series allows a wider frequency range to be covered. Lower mass allows a higher frequencies to be reproduced.

Berylium is better than tintanium, because it is lighter. The down side to beryliun is that if the metal fractures, the dust from the fracture is very dangerous.

Bill

[WMcD]

Yes, it is a difficult subject. Part of it is that some of the information is misleading. The data sheets fail to put it simply enough and fail beat you over the head with the facts:

1) Driver high freq output is poor. 2) Horns help equalize that by beaming (but constant directivity horns don't beam). 3) The situation is still poor with just about any horn and we have to equalize by active or passive crossovers.

Look at the EV and the Atlas sites or most others with a lot of data, and specifically the specs for their midrange-treble drivers. There is one curve for the response on a plane wave tube and one for the output of a horn plus the driver.

The plane wave tube has a diameter the same as the driver throat connection. A microphone is place just at the side of the tube at a small T junction. The tube is long (up to 25 feet) and there is some absorptive matter at the far end.

The mike gives us just the output of the driver in a matched condition. The plane wave tube acts like a perfect horn, with no reflections and no Fc cut-off (for the most part).

The measured output of the driver on a plane wave tube always shows some bad high freq roll off. It usually starts at 3 or 4 kHz. This is the true output of the driver. Yet on horn, the overall response in free space goes higher.

This just goes to show what I've often spoken about. Horns can be used to narrow down the beam and give some on axis gain.

There is still trouble though. They practically never get up to 15 kHz with any flatness.

But in some cases (usually the expensive drivers with 2 inch exits and sophisticated phase plugs) mid-band response (with horn) is up at 110 or 112 dB and it rolls off to 104 someplace near 15 kHz. That looks promising.

What can be done is use a passive network which attenuates the midband but not the very top end. That way it is possible to get overall flat response at about 104 dB to match the bass horn. We have to beat down the midrange to get that.

It took me a while to understand these factors. Once you do, you see them implied in many places.

PWK speaks of controlled dirctivity in his midrange to squirt out certain (high) freqs.

The T-35 plots show very much narrowing.

The point of the Jubilee in the bass horn is to get the mouths facing forward to beam.

The ST-350 constant directivity horn tweeter uses a tailored passive crossover.

Don Keele actively equalizes the driver to test his prototype constant directivity horns.

Even direct radiator woofers beam at high freqs. This is why on axis response can go up to 1 kHz. Off axis response is poor, though.

Gil

[Deang]

Sounds to me like the best solution in a case like this is biamping with active networks and EQ. It almost begs the question though - why bother?

[maxg]

Ya know Dean I have been thinking. Maybe we could make a mock 2 way using something like a concentric driver (ala Tannoy) or even something like the drivers used on an Avantgarde Solo.

Rigging up a tweeter within a midrange driver would get around all these problems in one fell swoop - although it would enforce the use of a circular horn opening which is not really in keeping with the design you are looking for.

Course if we could then put this within the bass bin......

[boom3]

----------------

On 3/15/2005 8:15:36 AM maxg wrote:

Ya know Dean I have been thinking. Maybe we could make a mock 2 way using something like a concentric driver (ala Tannoy) or even something like the drivers used on an Avantgarde Solo.

Rigging up a tweeter within a midrange driver would get around all these problems in one fell swoop - although it would enforce the use of a circular horn opening which is not really in keeping with the design you are looking for.

Course if we could then put this within the bass bin......

----------------

There have also been attempts to drive one horn with a mid and tweeter unit on the same manifold. I think the modulation distortion would be excessive. If you look at Paul's paper on the K-400, he shows a curve from a tweeter mounted in the mouth of the mid, and the curve shows comb filtering from two drivers trying to occupy the same space at the same time, i.e. in the crossover region. That would have been true of the old K-5-J set-up where T-35s or University 4401's were often mounted on flanges at the mouth of the mid to make a three-way system.

A possible approach to a 'unified' top end could be an adaption of the Beveridge acoustic lens, which is used to load and control the output of an electrostatic panel.

http://beveridge-audio.com/Technology.htm

It seems to me that the narrow constriction acts as a 'phase plug' at least in one dimension and creates a virtual driver at that plane. I hasten to add that the Beveridge lens is of course patented.

I have heard the Beveridges and they and the Klipschorn are the two best speakers I have ever heard.