Jean Michel on LeCleac'h horns

[revintage]

Hey M10,

Check your sim one more time!

You say the throat piece is 68,5mm but your faulty sim is done with 65mm length. Because of this you get the wrong flare and angles.

I found length was 69mm so no big difference.

375(2441, 2482 etc.) was to be a a permantent magnet knockoff of the field-coil WE594. This one has a longer throatpiece with a flare of 160Hz. For some reason the piece was shortened resulting in the higher flare. For some reason the JBL-techs seems unaware of this as they mention the 2441 being 160Hz in their tech-papers.

[more10]

Originally Posted by revintage

Hey M10,

Check your sim one more time!

You say the throat piece is 68,5mm but your faulty sim is done with 65mm length. Because of this you get the wrong flare and angles.

I found length was 69mm so no big difference.

375(2441, 2482 etc.) was to be a a permantent magnet knockoff of the field-coil WE594. This one has a longer throatpiece with a flare of 160Hz. For some reason the piece was shortened resulting in the higher flare. For some reason the JBL-techs seems unaware of this as they mention the 2441 being 160Hz in their tech-papers.

You are absolutely right Revintage. I am a bit dissapointed that you didn't find the other mistake :-). The sim calculates the throat angle, not the mouth angle.



Cutoff is 182 Hz. From the schematic diagram I export the horn data, and it gives the mouth angle. It is 4.67 degrees.

Finally we agree Revintage!

Regarding the difference between JBL and WE, I suspect that JBL simplified (made cheaper) the construction by removing the snout, but kept the mouth diameter. This drawing shows original construction plans:

[whgeiger]

Originally Posted by vinylnvalves

I am proposing to make some adaptors to reduce the throat of my 250Hz Le Cleach horns down from 160mm (for Lowther) down to 52mm for a 2" compression driver. When i used the Le Cleach calculator i have found that the profile does not simply extend backwards. Looking at the spreadsheet it looks as if the natural expansion routine it is using will give a different profile for every throat size. Attached is a graph showing how the profiles differ - what should i do, curve fit a profile "that smooth" from my current throat backwards or ?? suggestions please, as i dont know how sensitive the throat portion of the horn is.
i will be making flanged fibreglass extensions which will bolt on to the back of the existing horns.

VnV,

Your adapter problem has been used as a test case for the horn adapter spreadsheet I developed recently. The attached MS/Word file documents dimensions of the adapter that you need. Unfortunately Excel files cannot be attached to messages here; so the formula used have been documented here as well. Comments as well as additions/corrections to this work are welcome.

Regards,

WHG

N.B., due to the product of parameters m * T being fixed by driver exit diameter and flare slope, the required long adapter length and low fc should enhance LF loading of the horn.

[vinylnvalves]

I have put the data provided into a plot - for comparison. To my uneducated eye it looks a little long, and out of family with Azura and other commercial JMLC horns. My experimentation with conical horns gave long results also.

I have been using curve fitting routines to back predict the existing horn, working out compression instead of expansion. To date all have a point of inflection around 100mm dia where the solution diverges.

[whgeiger]

What is the radius r_1 and tangent angle a_1 at the throat exit of the horn you are using?
I estimated A_1 = 6.86 deg @ R_1 = 80 mm based on JMLC-200 horn
Horns
There should be no abrupt change in profile at the point where horn and adapter are joined.
Regards,
WHG

[Jmmlc]

Hello,

Often peole speak of a "Le Cléac'h horn profile". The under-meaning of that is that for a given T, a given Fc and various throat diameters the profiles of the obtained horns should share a common profile.

This is not the case, for the obvious reason that for a given distance to mouth, the wavefront shape will not be the same if the throat diameter is a, 2a or na...

So the problem to build an adapter for a Le Cléac'h horn of a given throat diameter in order to use a driver having a very smaller exit diameter is not a trivial one.

In some cases the adding of the adapter will introduce the profile slope twill loose 1st order or second order continuity and this is IMHO a bad thing.

The problem is a bit simpler if the axisymetric horn is calculated for the smaller throat diameter and then cut in order to provide a removable adapter (the wavefront on the membrane of a loudspeaker having a large diameter is lesss well defined also... )

Best regards from Paris, France

Jean-Michel Le Cléac'h

[lbstyling]

Originally Posted by Jmmlc

Hello,

Often peole speak of a "Le Cléac'h horn profile". The under-meaning of that is that for a given T, a given Fc and various throat diameters the profiles of the obtained horns should share a common profile.

This is not the case, for the obvious reason that for a given distance to mouth, the wavefront shape will not be the same if the throat diameter is a, 2a or na...

So the problem to build an adapter for a Le Cléac'h horn of a given throat diameter in order to use a driver having a very smaller exit diameter is not a trivial one.

In some cases the adding of the adapter will introduce the profile slope twill loose 1st order or second order continuity and this is IMHO a bad thing.

The problem is a bit simpler if the axisymetric horn is calculated for the smaller throat diameter and then cut in order to provide a removable adapter (the wavefront on the membrane of a loudspeaker having a large diameter is lesss well defined also... )

Best regards from Paris, France

Jean-Michel Le Cléac'h

Hi,
considering your comments here, what are your feelings on the following situation-

A choice between the following-

A JMLC profile 350hz horn with a 1.5" throat (7 degree angle at start of throat) meeting to a 1.5" exit driver with a 0 degree exit angle.

Or a JMLC profile 350hz horn with 2" exit and a 1 inch long adapter to bring the angle down to 0 degrees in order to meet the driver exit angle correctly.

This is a real choice for me as I am looking to buy a 350hz JMLC horn from autotech and meet it with a JBL 2452 driver (0 deg exit angle), but Im concerned about having a 7 or 8 degree 'kink' in the throat.- Would this not render a nice smooth profile pointless to some level?

Regards,
LBS.

[Jmmlc]

Hello lbstyling,

For drivers having an exit angle very different than the throat angle of a calculated hypex adapter, I recommand the use of a quadratic throat adpater.

You may find drawing ( page 8) and formula (page 10 )at:

http://www.excelsior-audio.com/Publi...WhitePaper.pdf

Best regards from Paris, France

Jean-Michel Le Cléac'h

Originally Posted by lbstyling

Hi,
considering your comments here, what are your feelings on the following situation-

A choice between the following-

A JMLC profile 350hz horn with a 1.5" throat (7 degree angle at start of throat) meeting to a 1.5" exit driver with a 0 degree exit angle.

Or a JMLC profile 350hz horn with 2" exit and a 1 inch long adapter to bring the angle down to 0 degrees in order to meet the driver exit angle correctly.

This is a real choice for me as I am looking to buy a 350hz JMLC horn from autotech and meet it with a JBL 2452 driver (0 deg exit angle), but Im concerned about having a 7 or 8 degree 'kink' in the throat.- Would this not render a nice smooth profile pointless to some level?

Regards,
LBS.

[Jmmlc]

Hello,

I wrote previously :

Originally Posted by Jmmlc

For drivers having an exit angle very different than the throat angle of a calculated hypex adapter, I recommand the use of a quadratic throat adpater.

You may find drawing ( page 8) and formula (page 10 )at:

http://www.excelsior-audio.com/Publi...WhitePaper.pdf

More information can be found in Hughes 's paper :

http://www.excelsior-audio.com/Publi...201999-09).pdf


Best regards from Paris, France

Jean-Michel Le Cléac'h

[whgeiger]

Originally Posted by Jmmlc

Hello,

Often peole speak of a "Le Cléac'h horn profile". The under-meaning of that is that for a given T, a given Fc and various throat diameters the profiles of the obtained horns should share a common profile.


This is not the case, for the obvious reason that for a given distance to mouth, the wavefront shape will not be the same if the throat diameter is a, 2a or na...

This is high school algebra stuff no matter what symbology is used, yours, mine or someone else’s. The wave front area at the throat is represented as the coefficient [St] in Salmon’s formula [Sl = St*(Cosh(m*l)+T*sinh(m*l))^2]; thus, different horn throat size, different horn profile.

So the problem to build an adapter for a Le Cléac'h horn of a given throat diameter in order to use a driver having a very smaller exit diameter is not a trivial one.

Not trivial, but simple still. Two angle-radius pairs, one for the driver, the other for the horn, must be matched while following Salmon’s horn profile. In the horn neck the differences in methods, use of ordinal vs. curvilinear dimensions, is trivial and well within the ‘noise’ of manufacturing tolerances; only in the horn bell, are the differences in dimensioning methodology significant. As one angle-radius pair fix the product of m*T the second sets the ratio between them; thus there is only one adapter solution available when Salmon’s formula is used. So the problem is only, just slightly more complex than trivial. Of course a clothoid segment could be used to make more solutions available.

In some cases the adding of the adapter will introduce the profile slope twill loose 1st order or second order continuity and this is IMHO a bad thing.

Here in the horn neck, there is little difference between wave front profile and its span dimensions. What is important is mitigating the diffraction edge created by a mismatch between horn slope and that of a compression driver. For that we may (or may not) sustain a minor change of slope derivatives, which for most cases, will represent an un-measurable difference.

The problem is a bit simpler if the axis metric horn is calculated for the smaller throat diameter and then cut in order to provide a removable adapter (the wavefront on the membrane of a loudspeaker having a large diameter is lesss well defined also... )
Best regards from Paris, France

Jean-Michel Le Cléac'h

The only way to get a matching horn is by specifying the desired throat radius and slope angle required. This intern fixes the product m*T. That is a mathematical fact of Salmon’s equation irrespective of the coordinate system used to implement it. Note that once [fc] is specified the value of [T] becomes fixed as well.

Regards,

WHG