Beyond the Ariel

Lynn Olson said:
Two rather simple reasons I favor the Le Cleac'h profile over others.

1) I can buy them right now, from Azurahorn in Australia, Music Concrete in France, and Experience Music in Memphis, Tennessee. These aren't the only people building them, but they're the ones I've been in communication with. All of the products look well-built, measure exceptionally well, and have a good track record from buyers and listeners.

Looking at pics of a le'cleach horn, it seems to have a lot in common with an OS waveguide.

Perhaps I am oversimplifying things, but as I see it, there are three big features in an OS waveguide:
1: The majority of the waveguide is a plain ol' conical horn
2: There is a LOT of attention paid to matching the mouth with the exit angle of the compression driver. As someone noted previously, Geddes critiqued someone's spreadsheet which ignored the exit angle of the compression driver. This is a very big factor in the OS waveguide. For instance, if you have a sixty degree conical horn and you mate it to a compression driver with a six degree exit, there is a massive discontinuity where the two meet. The OS waveguide is designed to minimize this discontinuity.
3: Geddes uses a massive roundover at the waveguide exit to mate the waveguide to the baffle. The roundover in the Summa is 2" IIRC.

To make a long story short, when I look at the pics of the Le'Cleach waveguide, I see a lot of the same features. The body of the horn is conical, but there's a lot of attention paid to minimizing diffraction at the throat and mouth.

Basically most horn guys are going about it all wrong. They get wrapped up in arguments about the profile of the horn, when they should be worrying about improving the mouth and throat termination of their horns. Changing the profile of a horn may improve the quantity of sound, but I'm worried about improving the quality of sound. The surest way to improve the quality of sound in a horn is to improve the mouth and throat termination.
 
Re: foam details?

moray james said:
Patric iwas waiting for mile B to try some foam in a set of Transylvania tubes but it's been a while. I think I would like to give it a try in a set of home brew k-tubes. Could you provide details on the appropriate foam characteristics for wave guide applications? Thanks.

I posted some measurements I did with and without the foam about 2.5 years ago here
Keep in mind that the measurements were for dramatically narrow tractrix horns. The results are even better on a real waveguide.

Dr Geddes uses a big ol' bun of the stuff, which is tres expensive. This is because you get better results that way.

I am a cheapskate, so I use layers of it, and build it up layer by layer. Not optimum, but far less expensive and wasteful.

There's a reason the Summa costs as much as it does. No corners are cut.

Here's the part number I use, it's about a hundred bucks for a lifetime supply.
 
Patrick Bateman said:


Considering all the positive press Dr Geddes has garnered with his OS waveguides, I am dumbfounded that the foam plug hasn't garnered more attention. Honest to God, it's the single biggest improvement I've ever made to my horns and waveguides. The difference is NOT subtle, it's a revelation.


I am quite sure I read it somewhere that the used of foam plug/lens has been copyright by Dr Geddes. If this is the case, it will be impossible for Lynn to build something with it and publish the build on the net.
 
Re: Re: foam details?

Patrick Bateman said:
Dr Geddes uses a big ol' bun of the stuff, which is tres expensive.

I am a cheapskate, so I use layers of it, and build it up layer by layer. Not optimum, but far less expensive and wasteful.

[/URL]

I bought a "big ol' bun" from Dr. Geddes. I originally filled the horn (Unity from Lambda). I also tried just the throat third, plus a plug that goes all the way to the phase plug on the TAD 2001. That treatment gives a great deal of the benefit, with 20% of the foam. This may not be true if the horn termination is sharp.

Sheldon
 
Soundstage

Retsel.
I spread the speakers apart to 8 ft. & toed them out to about 5 degrees toe-in for some improvement. the left speaker is about 4 ft. to the glass window behind at 45 degrees. The right speaker is over 5 ft. to the French doors.
Both amps are SS. No wings are used, however I had added what I was told sounded like an improvement.
To simulate a rolled edge to the baffles I had added 4" diameter cardboard carpet tubes that were opened up to slide over the sides of the baffles. Tonight, after removing & replacing them several times, I came to the conclusion that what my wife & son-in-law said was an improvement was the collapse of the soundstage to a more "coherant" grouping between the speakers. Be that as it may be, removing the tubes almost doubled the width of the soundstage.
As for capacitors & inductances, I have splurged on the best I could find at Audiosound. The crossovers are simple 6 dB, first order. The cone tweeters are XO at 5,000 Hz, the #1 woofer is xo at 300 Hz & the #2 is XO at 150 Hz. The 12" full range has no XO.
Thanks for your help. Your questioning of possible side wings led to what I'm sure was most of my problem, the round tubes around the sides of the OBs.
dobias
 
As an experiment, if you could toe the speakers in so the axes cross just in front of the center listening position, you should achieve a wider "sweet spot" and might improve imaging in other ways by reducing reflections from the side walls.
If you have the capability within your set up, you can check your progress by selecting mono mode and going for the narrowest phantom image.
 
The parallel discussion on the Geddes thread is worthy of close reading - I recommend everyone go over and give it a visit. Having worked in the Tektronix Spectrum Analyzer business unit on analyzers that went out to 26 GHz, I can affirm that termination is very important in minimizing time-domain reflections and FR ripples. Another business unit made the Time Domain Reflectometers, otherwise known as Cable Radar, extremely useful in seeing just how good your connections really are.

This is why removing the bug-screen from a prosound compression driver is important - even the most open metal meshes are only 50% open, leaving the other 50% to reflect in both directions - back to the phase plug, and back to the horn mouth. Both reflections are obviously very undesirable in the most critical part of the acoustic path. The so-called "damping" claimed for these meshes are a crude attempt to limit LF energy to the diaphragm and provide an extra measure of diaphragm loading.

The entire path from diaphragm to horn mouth has to be carefully analyzed (and measured in the time domain) to discover sources of reflection and diffraction. This is pretty straightforward with an impulse measurement, where the time dimension directly correlates with physical dimensions (343 meters/second). This is the acoustical analog to using a TDR to examine a microwave cable for defects - in this case, we're looking for problems in transmission from diaphragm to free air, such as reflections, diffraction, and multipath, all of which appear as "glitches" in the time domain. For some reason, although this is standard practice in the microwave domain (and the wavelengths are comparable), this test protocol is almost unknown with compression drivers and horns.

As mentioned on the other thread, the most fruitful areas for improvement are close to the diaphragm (rear chamber reflections, phase-plug diffraction/nonuniform emission, and uneven transition regions into the horn) and the edges of the horn mouth where it make the transition into free air. Each area constitutes the "ends", or potential sources of reflection, diffraction, or multipath of the driver/horn system.

The whole preoccupation with "profile" over the last several decades has drawn attention away from the more critical problem of analyzing potential problems in the time domain. The standard commercial practice of 1/3 octave smoothing has also made it much more difficult to see narrowband comb-filtering resulting from time-domain problems.
 
If you want to be clever, repeat the impulse measurements over a range of emission angles - say, +45 through 0 through -45 degrees horizontally. If you want to be really clever, put the loudspeaker on a turntable, do a large series (20 to 50) of 0~3 mSec impulse measurements vs emission angle, then plot the results as a 3D waterfall graph, with the depth dimension representing emission angle.

You'll then be repeating the results of the JVC research team, who published a paper in the Audio Engineering Society Journal in the early Eighties using just this technique. This gives a powerful representation of the wavefront as it emerges from the loudspeaker, showing diffraction and reflection in a immediately obvious visual form, looking like ripples on a pond.
 
Soundstage

Russell Dawkins ,
It must be because of my furniture arranged to the corner TV,
your suggestion to use a mono signal wound up causing me to arrange the OBs in a straight line with zero toe-in. I never would have thought of going that far.
Anyhow, with all the help in this thread, I now have a large soundstage & the listener sweetspot is also larger.
Thanks everyone,
dobias
 
All

A few key points to be made.

The OS waveguide contour is public domain. The Bi-speriodal has an issued patent.

The use of foam in a waveguide is patent pending and has been through its first office action. Based on that I can say with certainty that 1) it will issue, and 2) it will cover most uses of foam or other materials placed inside of a horn or waveguide.

As to the Declerq comparison to the OS, I would say this. I have not seen anyone do an OS waveguide right (myself excluded of course) so very very few people have heard what they can do when done correctly. Lynn seems to be impressed with what hes heard with the Declerq, but he has, at best, heard an OS only for a brief period in a less than optimal setting. And Yes, OS's can be used without foam, but why? I give a license to use any foam plug that I sell. You just can't use one that you don't buy from me. Thats fair, isn't it?

While the OS contour is not patented there are an awful lot of trade secrets that I do to make these devices work to their full potential.

I have heard a lot of horns and systems and all I can say is that I would put my designs up against anything out there for sound quality at high output levels. There simply is nothing that I have heard that can compare - and I'm talking at any price - Avantgardes, JBLs, TADs.

I have been very busy with another 'invention" that the manufacturers are eating up. Haven't been around here much. Not likely to get less busy in the near future either.
 
Lynn Olson said:
Two rather simple reasons I favor the Le Cleac'h profile over others.

1) I can buy them right now, from Azurahorn in Australia, Music Concrete in France, and Experience Music in Memphis, Tennessee. These aren't the only people building them, but they're the ones I've been in communication with. All of the products look well-built, measure exceptionally well, and have a good track record from buyers and listeners.


Lynn, I was just perusing another forum and saw this, I thought you might be interested as it mentioned two of your sources.


Azura Horn is going to close down.

John
 
I've been in touch with Martin, and he's willing to sell me a pair of AH-550's with a mounting plate for the Altec 288/GPA 399 driver. He also sent photos of same, shown below. I'll be buying a pair of these in the next week or two, along with either Altec 288C's (refurbished) or GPA 399's. Then I'll start measuring with MLSSA.

The old Altecs use a three-bolt pattern, while I guess the newer ones use four(?). Anyway, I'll ask Martin that the mounting plate fits either new or old versions, so I'll be measuring and auditioning many different versions (aluminum vs Pascalite diaphragms, tangential vs plastic surrounds, circumferential vs Tangerine phase plugs, different sizes vs no rear chamber, Alnico vs ceramic magnets, etc.). Since the bandwidth of interest is only 800 Hz to 7 kHz, I won't be trying to "stretch" the HF to 15 kHz, as in many other horn systems.
 

Attachments

  • ah550_1.4"_flange.jpg
    ah550_1.4"_flange.jpg
    73.5 KB · Views: 1,365