Horn-loading Ribbon Tweeter?

[mefistofelez]

I have been unable to find much information on this topic. Anyone has some literature/experience?

Thank you,

M

[rcavictim]

I was contemplating making some rectangular horns out of hardwood for a pair of Dayton PT-2 planar tweeters. The reason is that I need to mount them back of the baffle to time align the diaphragm with the mid range drivers in one of my experimental designs. I was just going to 'do it' and see what happens. That is project 8,547.

[VvvvvV]

I don't know what the point is sonically, but to make one probably best get hold of a local enthusiast wood Turner,

Here's a strange link the though http://vincent.brient.free.fr/round_horns.htm

[LineSource]

Horns are pressure to volume translators that work best with a transducer that has a stiff radiator and a low Qts for tight control of the air mass.

The principal virtue of a horn resides in the possibility of presenting practically any value of acoustical impedance to the sound generator. As a side benefit, it provides directivity contol to the dispersion of acoustical energy. A wafeguide is a horn with the directional characteristics being the prime criteria over optimum driver loading. Many planar transducers are fair candidates for horn loading.

A true ribbon, built with 2-10 micron thick aluminum supported at two ends, is better suited for a waveguide...a very shallow horn with wide radius to control dispersion and diffraction. Some designers use a waveguide to both control ribbon dispersion, and reduce the effects of a deep magnet cavity.

Geddes has a few recent papers on waveguides.

[rcavictim]

Originally posted by LineSource
Horns are pressure to volume translators that work best with a transducer that has a stiff radiator and a low Qts for tight control of the air mass.

The principal virtue of a horn resides in the possibility of presenting practically any value of acoustical impedance to the sound generator. As a side benefit, it provides directivity contol to the dispersion of acoustical energy. A wafeguide is a horn with the directional characteristics being the prime criteria over optimum driver loading. Many planar transducers are fair candidates for horn loading.

A true ribbon, built with 2-10 micron thick aluminum supported at two ends, is better suited for a waveguide...a very shallow horn with wide radius to control dispersion and diffraction. Some designers use a waveguide to both control ribbon dispersion, and reduce the effects of a deep magnet cavity.

Geddes has a few recent papers on waveguides.

Thanx for this encouraging news. Now all I need is time to experiment.

I come into this speaker design with a background in microwave antennas. To me a waveguide is usually many wavelengths long and is used to move radio frequency energy with very low loss from point a to point b. If the end of the waveguide contains, steps or tapers, parallel plates, etc. for the purpose of launching a wave into open air while shaping it's field distribution or pattern it becomes either a horn or a lens.

IMO, I do not think that these short tapered shapes that are placed on the front of a tweeter should be called "wave guides". They are probably short horns. IMO the treatment on the front of the Dayton PT-s planar tweeter is a short horn. JBL was correct in calling some of their "wave guiding" contraptions "acoustical lenses".

[mefistofelez]

LinSource wrote:

As a side benefit, it provides directivity contol to the dispersion of acoustical energy. A wafeguide is a horn with the directional characteristics being the prime criteria over optimum driver loading.

Yes, this is exactly, what I am after. Could you please give me some examples of the:

Many planar transducers are fair candidates for horn loading.

Thank you very much,

M

[bear]

my take on the loading of a ribbon or other non compression driver with a horn of any sort is this:

If you start with a driver that has a flat response and add a horn, you will get a response that rises - highest at the most effective frequency of the horn, which is usually a function of the expansion. The ability of any horn to be "flat" without ripples is related to the mouth size's relationship to the expansion rate
(which in turn is related to some "frequency"). So that's a bump in the response at the bottom end of the useful response range.

Generally speaking the best candidates for horn loading are drivers that are non flat in free air - having a rising response. The ability to be "stiff" as in the case of a compression driver (the air load *not* being so great as to overcome the force on the diaphragm) is what permits the wide range operation of the compression driver loaded into a horn. In free air it acts like a rather bandwidth limited tweeter... (thus the stiff diaphragm and BIG high Tesla magnet structure, limited excursion)

In the case of the compression driver, the horn acts like a lever whose motion increases with frequency in exact ratio to the falling response of the diaphragm... thus the output of the horn system is increases as the frequency drops, while the excursion of the diaphragm is not - whereas in a direct radiator the excursion increases as the frequency drops. In the compression driver the *force* of the diaphragm's motion is transformed into air motion by the horn's loading - a strong short force is transformed into a large not so strong force... a lever.

The other useful application for horn loading, and the one that has been most often employed (generally speaking) for ribbon drivers is to use the horn to "extend" the useful LF response of the ribbon below where it would otherwise start to roll off free air (or baffle mounted). So, if ur ribbon runs out of headroom or response at say 2.5kHz, adding a horn might get you down an octave or less in additional bandwidth, with some headroom benefits as well at the 2.5kHz area... as mentioned it also serves to control the dispersion/polar response to some extent.

Keep in mind, that as far as I can tell, at sufficiently high frequencies the horn isn't doing very much and the driver tends to
act like it is sitting out in free air, and the polar response tends to narrow as the frequency goes up...

Quite frankly, I've seen polar plots of some new waveguides and I am unsure how they manage to keep the polar response "flat" - in the case of the one I saw it was "flat" in a 30 deg dispersion out to ~20kHz... which is fairly remarkable actually. Of course 30 degrees is still narrow compared to what most drivers are doing except at the HF limits of their response...

Many ribbon drivers have been made with horns, the most famous are the Decca Ribbons, Kelly Ribbons and Romagna Ribbons - all of the same heritage and vintage.

If ur thinking ur going to get the same sort of efficiency that you get with compression drivers and horns, it's not likely.

_-_-bear

[kelticwizard]

Josef Merhaut wrote an article about building an electrostic horn in 1968. It was published in the Journal of the Audio E ngineering Society. I have it. I have not scanned it, but I will and send it to you via Email if you want.

The horn is 20" long and 18" wide. It has flat response from 19 KHz down to 1500 Hz, and is 3 dB down at 1000 Hz.

I have not scanned the article yet, but I will and Email it if you wish. Anyone who wants it can Email me.

[bear]

That ESL horn loaded was *essentially* a compression driver set up... he used multiple diaphragms, with a phase compensated entry to a single throat entrance... but it is/was a nice project...

_-_-bear

[MarkMcK]

This thread started off with the idea of horn loading a PT2 tweeter. My only problem with that idea is that I do not like the low end sound of the PT2. The sub 9 kHz response is where this tweeter is weakest. The problem is not a lack of response, but rather poor quality. Most of the bottom end output is caused by resonance or diaphragm vibration mode and is smeared well into the decay portion of the transient response.

Don't take this the wrong way. The PT2 is not a bad tweeter, just difficult to work with. Horn loading is unlikely to cure the problem and is likely to make the problem more audible.

Good designing and good building,

Mark