An idea for a wide dispersion ESL

jerko · 26th November 2007, 05:38 AM

I'm new to ESLs, so I'm not sure if someone has done this before, but I have an idea for an ESL that would have good dispersion characteristics.

The wavelength of a 20 kHz sound wave at sea level is about 1.7 cm, so the idea is to have the stator broken up into multiple rings which are each about that half that wide (or whatever gives the desired dispersion pattern). The signal sent to each ring would be delayed by the time it takes sound from the center of the panel to get to that ring (digitally delayed, before it is ever amplified) and then sent to separate amplifiers (one for each ring).

I know the Quad ESL had something like this (except without separate amps), but my understanding is that the delays and the rings were spaced too far apart to give good dispersion at high frequencies.

I know it could work, but I am not sure about the various factors that would possibly need to be worked out (and I would like to hear if anyone has tried to). For instance, each ring needs an amp, so for a reasonably sized panel this could get expensive quickly, and each ring needing its own transformer just compounds the problem. Is it possible to make a class-D amp that would put out the voltage necessary for an ESL without having to use a transformer?

Another problem is what to make the stators out of. Printed circuit boards seem to be almost ideal, but getting a large one made is prohibitively expensive. I'm really at a loss on this question.

I'm sure there are plenty of other things I am not thinking of, so please let me know what those are too.

wrl · 26th November 2007, 06:02 AM

Just a random question. I've seen a lot of these posts now about creating ESLs with wider dispersion patterns. I'm wondering if achieving greater dispersion can be accomplished without losing some of the imaging or intangible qualities of ESLs in the sweet spot.

From my experience, martin logans and such don't sound quite as dynamic in the sweet spot as planer ESLs. A wide dispersion ESL to me sounds good in theory, but I feel like there has to be a compromise in something in order to achieve this.

-Wes

jerko · 26th November 2007, 06:39 AM

The reason why I want dispersion (and the only reason I can think of to want it) is to have a larger sweet spot.

The original idea was to make an array of small electrodes on the stator with lots of pre-processing before the amps (one amp for each little electrode) so that the speaker could reproduce each source of sound between the far left and far right with a spherical pattern. Such a speaker would be wall sized and expensive beyond belief, but would, when coupled with an appropriate sub, be essentially ideal and have a huge sweet spot.

Since almost all recordings are stereo, I decided it would be good enough to try to reproduce two point sources rather than hundreds.

Calvin · 26th November 2007, 07:58 AM

Hi,

as wrl pointed out, optimizing one parameter will effect another parameter in the quite opposite direction. Unsegmented panels sound more dynamic in the mid-highs than segmented ones. Curved panels and planar panels (only unsegmented ones!) sound basically the same, since the claimed wider dispersion doesn´t really happen in praxis.
The problem with small rings or strips is, that they represent a very high impedance, which has to be transformed with much larger transformation factors. Everybody skilled in building transformers knows that high transformation factors always(!) cost on bandwidth and dynamics. Unsegmented designs can work with factors down to 1:50 (often 1:75), segmented panels maybe down to 80 (often 1:100)
The older Quads worked with up to 1:270. Besides the problems of getting Your hands on such a transformer, it´ll be no doubt worse than a simpler tranny with a small transformation factor.
Besides the highimpedance prob such small rings or strips will suffer in efficiency. Think of reduced effective membrane area or increased and wasted capacitance (beneath the spacers).

I can only recommend -especially those newbies in ESL- to closely watch what the great names did....to understand why they did it the way they did and no other way. This alone will take You considerable time and maybe lead to the lesson, that those designs not only have their merits but can hardly be topped.
I still have to listen to a Quad I could like and I think that the concept has major flaws, but the engineering behind is straightforward and marvellous. Same is with MLs. Those are much more than some pieces of sheet metal and a film diaphragm.

jauu
Calvin

ps: here´s a little task for You if You like: ;-)
The point source!
Is it the one and only distribution character?
Is it the best distribution character in my living room?
Can a different dispersion character be superior?
If so under which circumstances?
Which distribution character could that be?

el`Ol · 26th November 2007, 10:16 AM

Thyratrons are cheap and can drive an ESL transformerless, but are too slow for Class D and too non-linear for other amps. KOON3876 works on a concept of array speakers where the single drivers are Class D driven, but not with full resolution, so only the array as a whole gives the full signal. You could ask him about his progress.

oshifis · 26th November 2007, 02:06 PM

Does a curved panel give wider dispersion? If it does, the solution is simple. Quad, ML and others use this approach, I think.

jerko · 26th November 2007, 05:17 PM

First, no, curved panels don't have a wider dispersion. Or at least what I have heard says that they do not have a significantly higher dispersion.

Calvin,

I think I might not have been as clear as I would have liked. I did not mean that the diaphragm would be segmented. The stator electrodes are the only thing that would be rings. I don't think that your points hold when the diaphragm is large but the stator electrodes are in segments (the stator itself (the thing the electrodes are attached to) would not be segmented). I agree that all of your points hold true for the design you thought I was talking about though.

el`Ol · 26th November 2007, 05:33 PM

What do you think about that?
http://www.freepatentsonline.com/20070242844.html

MJ Dijkstra · 26th November 2007, 06:25 PM

Quote:

Originally posted by el`Ol
What do you think about that?
http://www.freepatentsonline.com/20070242844.html

Hi,

I remember that audiostatic had made a hybrid esl with the esl being dome-shaped. They discontinued its production. I have been told that this was because of production problems, no official source, so maybe because of another reason.

jerko · 26th November 2007, 11:09 PM

Yeah, that looks absolutely impossible to produce.

26th November 2007, 05:38 AM	#1
jerko diyAudio Member Join Date: Jan 2005 Location: earth	An idea for a wide dispersion ESL I'm new to ESLs, so I'm not sure if someone has done this before, but I have an idea for an ESL that would have good dispersion characteristics. The wavelength of a 20 kHz sound wave at sea level is about 1.7 cm, so the idea is to have the stator broken up into multiple rings which are each about that half that wide (or whatever gives the desired dispersion pattern). The signal sent to each ring would be delayed by the time it takes sound from the center of the panel to get to that ring (digitally delayed, before it is ever amplified) and then sent to separate amplifiers (one for each ring). I know the Quad ESL had something like this (except without separate amps), but my understanding is that the delays and the rings were spaced too far apart to give good dispersion at high frequencies. I know it could work, but I am not sure about the various factors that would possibly need to be worked out (and I would like to hear if anyone has tried to). For instance, each ring needs an amp, so for a reasonably sized panel this could get expensive quickly, and each ring needing its own transformer just compounds the problem. Is it possible to make a class-D amp that would put out the voltage necessary for an ESL without having to use a transformer? Another problem is what to make the stators out of. Printed circuit boards seem to be almost ideal, but getting a large one made is prohibitively expensive. I'm really at a loss on this question. I'm sure there are plenty of other things I am not thinking of, so please let me know what those are too.

26th November 2007, 06:02 AM	#2
wrl diyAudio Member Join Date: Jan 2003 Location: USA	Just a random question. I've seen a lot of these posts now about creating ESLs with wider dispersion patterns. I'm wondering if achieving greater dispersion can be accomplished without losing some of the imaging or intangible qualities of ESLs in the sweet spot. From my experience, martin logans and such don't sound quite as dynamic in the sweet spot as planer ESLs. A wide dispersion ESL to me sounds good in theory, but I feel like there has to be a compromise in something in order to achieve this. -Wes __________________ Anything worth doing is worth doing right... and redoing to make it better... and again to fix it back the way it was.

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Wide dispersion speakers using full (wide) range drivers	ozziozzi	Full Range	1	28th December 2008 06:13 AM
Sub placement to counter wide dispersion?	kyrie48	Subwoofers	3	20th September 2007 04:33 AM
2 tweeters for vertical dispersion?	Ang	Planars & Exotics	2	19th April 2007 04:10 AM
speaker dispersion	malcom_xavior	Car Audio	2	14th December 2004 09:23 PM
Baffle width vs dispersion	Vg	Multi-Way	5	13th October 2003 06:34 PM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read


	Please consider donating to help us continue to serve you. Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving

26th November 2007, 06:39 AM	#3
jerko diyAudio Member Join Date: Jan 2005 Location: earth	The reason why I want dispersion (and the only reason I can think of to want it) is to have a larger sweet spot. The original idea was to make an array of small electrodes on the stator with lots of pre-processing before the amps (one amp for each little electrode) so that the speaker could reproduce each source of sound between the far left and far right with a spherical pattern. Such a speaker would be wall sized and expensive beyond belief, but would, when coupled with an appropriate sub, be essentially ideal and have a huge sweet spot. Since almost all recordings are stereo, I decided it would be good enough to try to reproduce two point sources rather than hundreds.

26th November 2007, 07:58 AM	#4
Calvin diyAudio Member Join Date: Nov 2004 Location: close to Basel	Hi, as wrl pointed out, optimizing one parameter will effect another parameter in the quite opposite direction. Unsegmented panels sound more dynamic in the mid-highs than segmented ones. Curved panels and planar panels (only unsegmented ones!) sound basically the same, since the claimed wider dispersion doesn´t really happen in praxis. The problem with small rings or strips is, that they represent a very high impedance, which has to be transformed with much larger transformation factors. Everybody skilled in building transformers knows that high transformation factors always(!) cost on bandwidth and dynamics. Unsegmented designs can work with factors down to 1:50 (often 1:75), segmented panels maybe down to 80 (often 1:100) The older Quads worked with up to 1:270. Besides the problems of getting Your hands on such a transformer, it´ll be no doubt worse than a simpler tranny with a small transformation factor. Besides the highimpedance prob such small rings or strips will suffer in efficiency. Think of reduced effective membrane area or increased and wasted capacitance (beneath the spacers). I can only recommend -especially those newbies in ESL- to closely watch what the great names did....to understand why they did it the way they did and no other way. This alone will take You considerable time and maybe lead to the lesson, that those designs not only have their merits but can hardly be topped. I still have to listen to a Quad I could like and I think that the concept has major flaws, but the engineering behind is straightforward and marvellous. Same is with MLs. Those are much more than some pieces of sheet metal and a film diaphragm. jauu Calvin ps: here´s a little task for You if You like: ;-) The point source! Is it the one and only distribution character? Is it the best distribution character in my living room? Can a different dispersion character be superior? If so under which circumstances? Which distribution character could that be?

26th November 2007, 10:16 AM	#5
el`Ol diyAudio Member Join Date: May 2005 Location: Bavarian Forest	Thyratrons are cheap and can drive an ESL transformerless, but are too slow for Class D and too non-linear for other amps. KOON3876 works on a concept of array speakers where the single drivers are Class D driven, but not with full resolution, so only the array as a whole gives the full signal. You could ask him about his progress.

26th November 2007, 02:06 PM	#6
oshifis diyAudio Member Join Date: Mar 2004 Location: Budapest, Hungary	Does a curved panel give wider dispersion? If it does, the solution is simple. Quad, ML and others use this approach, I think.

26th November 2007, 05:17 PM	#7
jerko diyAudio Member Join Date: Jan 2005 Location: earth	First, no, curved panels don't have a wider dispersion. Or at least what I have heard says that they do not have a significantly higher dispersion. Calvin, I think I might not have been as clear as I would have liked. I did not mean that the diaphragm would be segmented. The stator electrodes are the only thing that would be rings. I don't think that your points hold when the diaphragm is large but the stator electrodes are in segments (the stator itself (the thing the electrodes are attached to) would not be segmented). I agree that all of your points hold true for the design you thought I was talking about though.

26th November 2007, 05:33 PM	#8
el`Ol diyAudio Member Join Date: May 2005 Location: Bavarian Forest	What do you think about that? http://www.freepatentsonline.com/20070242844.html

26th November 2007, 11:09 PM	#10
jerko diyAudio Member Join Date: Jan 2005 Location: earth	Yeah, that looks absolutely impossible to produce.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)