Geddes on Waveguides

[Jmmlc]

Hello,

Sorry, I was not in Paris during several days.

You'll find in attached file a screenshot from Hornresp to simulate a TD2001 on a Le Cléac'h horn having a Fc = 320Hz.

Also you cab give a look to the message:

http://www.audioasylum.com/forums/hug/messages/12/128658.html


Best regards from Paris, France

Jean-Michel Le Cléac'h

Jean-Michel,

A little off-topic here, but could you post the input parameters you use to model the TAD driver in Hornresp? Thank you.

Rgs, JLH

 

[Jmmlc]

Hello,

The start for the modelisation of the TAD TD2001 driver is obviously Kinoshita and Locanthi's paper in JAES.

JAES convention 1978
preprint N0 1422 (M2)
"The influence of parasitic resonances on comrpession driver loudspeaker preformance".

Then I used a good measurement I did myself of the impedance curve of the TD2001 compression driver on a high quality horn that can be easily simulated (in Hornresp by example) .

In this context high quality horn means first : with very few refelcted waves from mouth to throat as this will be mirrored on the impedance curve.

Then, as you said, an iterative process will lead to the determination of the missing parameters.

You'll find in attached file, a comparison between the measured impedance curve (and response curve) with the simulated one under Hornresp.

Best regards from Paris, France

Jean-Michel Le Cléac'h

Getting an accurate set of parameters for a compression driver can be tricky business because virtually no manufacturer supplies enough information to get a good model. You have to do some tests for yourself. The way I do it has worked well for me. ...
...
Then you can take the back off and run the impedance again. This will give you a better idea of what the parameters are by fitting this - this usually gives you a good idea of wht the rear volume compliance and resistance are.

 

[Jmmlc]

Hello Paul,

During the last years, at the demand of owners of waveguides, I calculated several Le Cléac'h flared mouthes.

Even, one of those audiophiles, after having obtained a very good result with a medium sized waveguide with susch a Le Cléac'h mouth, asked me to calculate another Le Cléac'h mouth for a larger waveguide...

Just let you know that...

Best regards from Paris, France

Jean-Michel Le Cléac'h

Actually I thought SPL would decrease with increased radial distance but didn't factor in the increasing circumferential line length.

In the coming months I plan to build some really large waveguides but want to get the best overall performance from a given OD and am trying to understand if there is a better mouth flare than a simple circular radius. Does "burning" OD with a larger radius early in the curve have a benefit over a circular radius? The simple radius is easiest to construct, but a more complex curve (large radius early, decreasing radius at the outer edge) doesn't add much difficulty.

 

[Jmmlc]

Hello Earl,

Can you confirm that the quoted sentence is only valid in the Oblate Spheroidal system of coordinates (mu, theta, phi, t) but when coming back to our ( x, y, z, t ) system (= our Euclidean world) it is not an exact solution but an approximate solution.

For other readers: the OS waveguide can be consider as one parameter in the OS system of coordinates. The single coordinate mu is sufficient in the OS system to define the wavefront, but this is only after a scale factor was previously chosen.

This scale factor (it acts a lot on the shape of the orthogonal net of curves), can be considered as an additional parameter. So coming back from the OS system of coordinates to our Cartesian system of coordinates a wavefront is defined by a couple of parameter (mu and the "scale factor" which is often called a), thus the one parameter "hypothesis" is no more valid (but a good aproximation)... this explains also why the theorical wavefronts are not parallel-equidistant...

Best regards from Paris, France

Jean-Michel Le Cléac'h

The OS solutions are analytically exact in full three dimensions and as such they are the only true solution of the wave equation for a flared contour, all others being only approximations, some better than others.

 

[dlr]

Re: Re: Re: Re: Which position to you actually take on this?

If one does not listen to the detailed revealed during live performance as in small lounges, then they will not have appreciation of what detail they are missing, so it really does not make sense to convince them they are missing something becasue they really don't miss it. But because some people don't miss it or care about it, it does not mean the detail cannot be revealed?

So the point that I wish to make is, if you are satisfied with what you are hearing, it is meaningless to explore issues regarding revealing more detail.

First, one would have to assume that it's possible to completely re-create the acoustics present in your scenario, but we all know that's not possible. Every re-production system will suffer in many ways, so it's a straw man argument to even begin any debate from that standpoint. Your statement can only be considered as a comparison between reproduction systems, that's what's under consideration.

Your response still has nothing to do with your point stating that one (as in anyone) should fully expect there to be loss in transient detail (when comparing different reproduction systems) as you described for the situation. There are objective methods to analyze it. You still have not addressed this, you keep providing irrelevant anecdotes. Are you in essence retracting that statement as being inaccurate?

Dave

 

[soongsc]

Re: Re: Re: Re: Re: Which position to you actually take on this?

First, one would have to assume that it's possible to completely re-create the acoustics present in your scenario, but we all know that's not possible. Every re-production system will suffer in many ways, so it's a straw man argument to even begin any debate from that standpoint. Your statement can only be considered as a comparison between reproduction systems, that's what's under consideration.

Your response still has nothing to do with your point stating that one (as in anyone) should fully expect there to be loss in transient detail (when comparing different reproduction systems) as you described for the situation. There are objective methods to analyze it. You still have not addressed this, you keep providing irrelevant anecdotes. Are you in essence retracting that statement as being inaccurate?

Dave

You cannot prove otherwise either, we will just have to leave it at that unless you have substantial data. I know that I have enought to work on.

 

[dlr]

Re: Re: Re: Re: Re: Re: Which position to you actually take on this?

You cannot prove otherwise either, we will just have to leave it at that unless you have substantial data. I know that I have enought to work on.

I had and have nothing to prove, I simply asked on what basis you made your initial unqualified claim. Evidently it was all based on nothing but anecdote.

Dave

 

[Patrick Bateman]

Makes me wonder what would happen if we added foam/stuffing to other "ordinary" waveguides

I've tried the foam on severely compromised* tractrix horns, on semi-compromised OS waveguides, and my Summas have it.

IMHO, the improvements were most audible with the tractrix horns.

I'd love to see someone like Avant Garde license it. IMHO, it would be a huge improvement.

* By "severely compromised", I mean the aspect ratio was dramatically skewed, and the length of the horn was dramatically shortened. The waveguides I made were much closer to a "standard" ratio. My experiments on the tractrix were the most scientific, as they were basically flat before the foam, then I redid the crossover so they'd measure flat WITH it.

 

[Patrick Bateman]

Compare impulse response shape and you can probably understand what I mean. Putting something in front of the diaphram does help suppress some high frequency resonances, but you lose some transient detail. This is the same as putting foam in a horn. I would prefer trying to solve the problem without anything in front of the diaphragm that effect the impulse onset. But it's only my preference.

The foam is simply an acoustic low pass filter. There's nothing magical about it. It attenuates the sound from the diaphragm; more at high frequencies and less at low frequencies. The magic is the implementation; soundwaves reflected back down the throat are attenuated THREE times, due to a pathlength from diaphragm to mouth to throat and back out the mouth.

Keep in mind that the only way to filter HOMs is mechanically; EQ won't cut it.

 

[soongsc]

Re: Re: Re: Re: Re: Re: Re: Which position to you actually take on this?

I had and have nothing to prove, I simply asked on what basis you made your initial unqualified claim. Evidently it was all based on nothing but anecdote.

Dave

All my research was on direct radiating drivers, I see not point in , nor do I care to flood the thread with non-waveguide data. I think Geddes does not post all data that readers want either. The thing is you just got to some work on your own. Provoking talk simply will not do.

 

[soongsc]

The foam is simply an acoustic low pass filter. There's nothing magical about it. It attenuates the sound from the diaphragm; more at high frequencies and less at low frequencies. The magic is the implementation; soundwaves reflected back down the throat are attenuated THREE times, due to a pathlength from diaphragm to mouth to throat and back out the mouth.

Keep in mind that the only way to filter HOMs is mechanically; EQ won't cut it.

It is probably the best compomise for most wave guides. It would be interesting to see impulses without and with the foam. Since HOM probably will not be significant in the range above 10KHz, EQ with foam probably brings back some filtered transient detail. But is it possible to reduce HOMs just by treating the area near the mouth?

 

[dwk123]

It is probably the best compomise for most wave guides. It would be interesting to see impulses without and with the foam. Since HOM probably will not be significant in the range above 10KHz, EQ with foam probably brings back some filtered transient detail. But is it possible to reduce HOMs just by treating the area near the mouth?

I have 6" of foam in my U15s, and while it's very tough to directly A/B due to the need to re-eq the system, IMHO it makes a distinct and worthwhile improvement. Not staggeringly jaw-dropping, but worth it. I originally had the entire horn filled, but it became much tougher to eq due to the HF attenuation (the U15 is a 60-degree horn, and so is somewhat deeper than the summa).
So, I can't 'prove' that I'm hearing a reduction in HOM rather than an attenuation in reflected waves for example, but it appears that partial treatment is still effective.

 

[gedlee]

Hello Earl,

Can you confirm that the quoted sentence is only valid in the Oblate Spheroidal system of coordinates (mu, theta, phi, t) but when coming back to our ( x, y, z, t ) system (= our Euclidean world) it is not an exact solution but an approximate solution.

For other readers: the OS waveguide can be consider as one parameter in the OS system of coordinates. The single coordinate mu is sufficient in the OS system to define the wavefront, but this is only after a scale factor was previously chosen.

This scale factor (it acts a lot on the shape of the orthogonal net of curves), can be considered as an additional parameter. So coming back from the OS system of coordinates to our Cartesian system of coordinates a wavefront is defined by a couple of parameter (mu and the "scale factor" which is often called a), thus the one parameter "hypothesis" is no more valid (but a good aproximation)... this explains also why the theorical wavefronts are not parallel-equidistant...

Best regards from Paris, France

Jean-Michel Le Cléac'h

Mr. Le Cléac'h

Your comments are incorrect and they exhibit a lack of understanding of the OS waveguide and the entire concept of "waveguide theory". There is nothing "approximate" about the OS solution (no matter what coordinate system one uses) and it IS NOT one parameter because this concept is without merit (read my response to Audio X-Press on this point). It is a fundamental requirement in physics that a solution in one coordinate system is valid in all coordinate systems, although it may be very difficult to describe in the other coordinate systems. The OS is EXACT in the OS system and hence it MUST BE EXACT in cartesian coordinates, although it would be very very difficult to describe in that particular system. You really do need to read my book; Chapter 6 - Wavguides.

 

[gedlee]

It is probably the best compomise for most wave guides. It would be interesting to see impulses without and with the foam. Since HOM probably will not be significant in the range above 10KHz, EQ with foam probably brings back some filtered transient detail. But is it possible to reduce HOMs just by treating the area near the mouth?

I have posted this information on my website in the paper on waveguide to ALMA.

HOM are most significant above 10 kHz.

I really do have to agree with the critiques of your posts here as you do make statements that are rather all-encompasing without giving any more support than thats what you heard. This is not in the vein of any posts that I make nor my whole approach to audio. Thus, this kind of posting is really not appropriate in my threads. They are other threads where the posters all take listening as the final word on everything, but here you will be challenged, as you have been, to support your beliefs with more that just anecdotal data.

 

[Patrick Bateman]

"Sounds" a bit like XT120/XT1086/XT1464 from 18sound ... anyone have heard any of these?

I schlepped out to the garage and took some pics for y'all.

In the pic is my Summa, the 18 Sound waveguide, a USD Audio waveguide, and the mold that I'm creating for my new Unities. Note the 18Sound waveguide is MUCH smaller than the Summa waveguide. Even worse, there's some kind of diffraction device in the throat
Excuse the crummy quality of the pix; I haven't bothered to remove the wax used for preserving the finish of the Summas during shipping. And there's a lot of sawdust on everything else.

Here's the URL for the 18 Sound:
http://www.eighteensound.com/index.aspx?mainMenu=view_product&pid=178

Here's the URL for the USD waveguide (horns in the car!):
http://www.usdaudio.com/products/wg-bc300.php

Here's the URL for my new waveguides:
http://www.diyaudio.com/forums/showthread.php?threadid=117537&goto=newpost

An externally hosted image should be here but it was not working when we last tested it.

 

[soongsc]

I have posted this information on my website in the paper on waveguide to ALMA.

HOM are most significant above 10 kHz.

I really do have to agree with the critiques of your posts here as you do make statements that are rather all-encompasing without giving any more support than thats what you heard. This is not in the vein of any posts that I make nor my whole approach to audio. Thus, this kind of posting is really not appropriate in my threads. They are other threads where the posters all take listening as the final word on everything, but here you will be challenged, as you have been, to support your beliefs with more that just anecdotal data.

I have asked specific questions about your data, and was never answered. So I really don't know how technical the thread can get other than say what you want and present what you want. If you may have missed my questions, I'd be glad to address them again. I think we can focus on the technical issues rather than personal issues. People that bluntly criticise lack of data are the ones that do not present data themselves. If you asked for some specific data, I would probably run the test because I see you provide more explanation and data than many others. I have posted data in other threads, and welcomed criticism. received none pertaining to the data, just generlization like "there must be some other things causing this" etc. and these types of comments are generally made by the same people.

 

[soongsc]

HOM - Higher Order Mode, its a term that I coined to define waves that propagate in a waveguide that do not go down the axis, but travel by bouncing off of the walls. They are not predicted by the Horn Equation, so most people didn't even know that they existed (I was the first person to hypothesize there existance). The Waveguide Theory predicts them, and low and behold, it turns out that they are quite significant to audibility. Minimizing them yields a far better sound quality. But with "horns" its not possible to minimize them because you don't know what to do - the equations aren't rigorous enough to predict them so they are simply ignored.

While I understand this phenomena, I have never read anything that correlates prediction and measurement results. Is your paper to ALMA a presentation for or a paper form? I could not find any title on your site that seems related to this. Having looked throught the presentation titled "Sound Quality Improvements in compresson Driver Systems", it is addressed the HOM source could be from the driver itself, but does not address which is more significant, nor is there any way to distinguish between HOMs from guide or HOMs form driver. The frequency plots do indicate that it acts like a filter. This same filter also cuts out detail transients. Whether eq will linearize the effects caused by the foam, there is not data showing that.

Analysis on diaphragm vibration mode indicate that the diaphragm modes are very similare in nature to drum modes. When these modes interact with the compresson driver phase plug, we should be able to see the so called HOMs that coorelate with certain Axial modes of the diaphragm. I have not used compression drivers because this part of the design does not seem mature, but it's under investigation. Improve the diaphragm modes, and you reduce HOMs. I might just consider treating them like direct ratiating diaphragm to reduce the modes.

 

[soongsc]

It is also possible to design the phase plug of compression driver in such a way that the slots do not line up with the diaphragm modes, thus reducing HOMs.

Purely from a geometric point of view, I cannot see the guide generating HOMs, unless there is a structural design problem.

 

[tinitus]

Just curious, but how big is the foam plug ... does it start at the throath and fill the whole waveguide, or does it only fill half the waveguide. or ... ?

Soongsc, a foam plug or other material in a waveguide may not be the same thing as a cloth in front of a dome tweeter or midrange cone

 

[David McBean]

You'll find in attached file, a comparison between the measured impedance curve (and response curve) with the simulated one under Hornresp.

Hi Jean-Michel,

Just out of interest, how close is the 0 degrees off-axis directivity response prediction to the measured SPL response result shown in your attachment? (From the shape of the curve I am guessing that your attachment gives the predicted constant directivity or power response).

Thanks in anticipation.

Kind regards,

David