Building a large curved ESL - diyAudio
Go Back   Home > Forums > Loudspeakers > Planars & Exotics

Planars & Exotics ESL's, planars, and alternative technologies

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
Reply
 
Thread Tools Search this Thread
Old 3rd October 2013, 01:41 AM   #1
diyAudio Member
 
wreckingball's Avatar
 
Join Date: Sep 2013
Location: Michigan, US
Default Building a large curved ESL

Hello everyone, long time lurker here to this great site.

As the title suggests I want a pair of large, curved ESLs, but of the only two people that I know that will build ESLs to order, I am not finding exactly what I am looking for...

Russ Knotts of justrealmusic builds them, but he only curves them to 22 degrees, and I am looking for more like 30, and the largest panels he makes are 16"x 60", and I want one a full 18" wide. Plus I don't think there is a return policy as this would be considered a custom build.

Buen (username) of Crescendo makes panels that wide, but there are a few things he does witch go against my desired outcome. Firstly, he constructs his ESLs using vertical spacers, and I want to use horizontal ones, preferably using double sided tape. Secondly, he uses aluminum coated diaphragms instead of clear, and to me it makes them appear as a wall instead of a window, if you know what I mean.


A little background: I own flat ESLs that are 14.5 inches by 42 inches that sit atop my (10 inch Peerless) 8 foot long tapered DIY transmission line. Now I love my t-lines, but my ESL stators are getting rusty, and most importantly, I am looking to get away from having my head in a vice just to enjoy the listening position. Plus I have just finished a home theater with a projector, and when I have people over, I'd like them to be able to enjoy the sound too, not just me.
Attached picture for reference:E-stats and screen.jpg


I'm sort of using Charlie's (Jazzman's) blog as a general guide, as it is very thorough and includes a supply list and links, but I have a few questions, (actually quite a few, lol).


  1. The ESLs I want to build will be 18 inches by 60 inches, with a frequency response that goes relatively flat to around 200Hz (x-over around 150-180), that's quite a bit larger than Charlie's 12”x48” (I believe), and a lower crossover point. Will the same step-up transformers and bias supply work for a design that much different, or do I need to completely rethink this?
  2. He uses spacers that are 3/8 of an inch wide, I found some double sided tape of the same thickness that is ž inch wide, does anyone see potential problems using this width?
  3. Doing a curved design, do the spacers between the diaphragm and the rear stator need to be a different thickness than the front as to avoid it hitting the rear stator?
  4. I noticed that quite a few ESL builders don't have their stators perforated all the way to the edge, they have a perimeter of around one inch from the edge that is not perforated, is this necessary, or do I need to just make sure the edges are wrapped with insulator tape of some sort?
  5. I plan on building a jig curved to 30 degrees for an 18 inch panel, but the jig needs to be around 22 inches wide (so I have room to tape horizontally), does anyone know an easy way to calculate this?


Well that's more than enough questions for now, but I'm really determined to fabricate these, not only to save a few dollars, but more importantly to have that satisfaction that comes from the accomplishment.


So to all you stat guys, please leave your answers/comments/further questions for me if you'd be so kind, if all goes well, and I get this ball rolling, I'd like to document any and all progress.




cheers
-wreck

Last edited by wreckingball; 6th October 2013 at 01:59 AM. Reason: misquoted measurement
  Reply With Quote
Old 3rd October 2013, 03:11 AM   #2
diyAudio Member
 
Join Date: Feb 2008
Location: Jackson,michigan
The wider you make the panel the worse the beaming issues become.

Segmenting the diaphragm into narrower vertical sections will help with this.

Sorry, I don't have any experience with curved panels.

This thread may help you to get some ideas for your project,
Both the builder and the owner of these very large curved ESL's are members of this forum as well,


Building an Electrostatic Speaker

I get good performance down to about 200HZ with a D/S of about 1.5 mm to 1.85mm with a 3.25" wide diaphargm on my little panels.

The perforations along the edge is just a matter of the material that you are using.
Having the holes will help to reduce any extra wasted capacitance.
However, They also give you less area for the tape to properly bond to as well, So you may want to consider using a wider tape than 1/4" around the perimeters or add some extra clamping forces using your final frame structure.

Here is a Radii Calculator that may help you,

http://www.handymath.com/cgi-bin/rad2.cgi?submit=Entry

FWIW

jer

Last edited by geraldfryjr; 3rd October 2013 at 03:35 AM.
  Reply With Quote
Old 3rd October 2013, 06:48 AM   #3
beun is offline beun  United States
diyAudio Member
 
Join Date: Feb 2010
Wreck,

If you really want to have a transparent membrane it's possible for me to make that. I actually have this material in stock for a width up to 30". Using horizontal spacers on a curved panel is definitely possible but makes stretching a whole lot more difficult as really only a vertical stretch is possible. This also means that a hand stretch is really the only option left unless you want to built a dedicated stretching rig for only two speakers. The hand stretch also makes the use of double sided tape a virtual impossibility, apart from the fact that double sided tape actually has a shorter lifespan that the mylar tape I use. Martin Logan can get away with it because they built a lot of the same, for custom production this becomes price prohibitive.

The size of 18" x 60" is actually relatively large for a panel that only has to go down to 200Hz, something this size should easily make 100Hz. If 200Hz is the only requirement than 12" wide should be more than enough.

The stators are custom punched and deburred according to your specified size, for manufacturing reasons an unpunched area around the outer edge is required otherwise the panel will start to warp, this area is also useful to mount the spacers on.

The front an rear spacer thickness is identical, you don't want to introduce more asymmetry than there already exists. The spacers need to be made from a fairly hard material, you don't want them to deform because of membrane tension, I usually use polycarbonate (Lexan) for the spacers around to edge where the membrane is mounted on and 1/8" acrylic for the non structural ones in the center.

The reason for the vertical spacers is better horizontal dispersion and ease of manufacturing although I also do believe that even tensioning in both planes ultimately make a better speaker.
  Reply With Quote
Old 3rd October 2013, 08:11 AM   #4
Calvin is offline Calvin  Germany
diyAudio Member
 
Calvin's Avatar
 
Join Date: Nov 2004
Location: close to Basel
Hi,

it seems to me that Your design process is prone to non-success.
Iīd suggest You first specify the requirements of the final speaker, as useage, possible size range, possible positioning, etc.
If I look at the attached pic, I know my ideas about a HomeTheatre capable ESL would certainly not start at: "I want an 18" wide panel", but rather at: "What dynamic range do I want, how many persons will sit side by side, how much space do I have towards backside, side and distance to audience, etc. etc."
A panel is easily chosen too large for a certain listening distance and staging is lost completely. For example will a 18" panel require approximately 12-15Ft minimum listening distance.
My rough guess for a solution would be a hybrid ESL with an only 8"-12" wide panel, crossed around 300-500Hz, assisted by a slim dipole tower and a dedicated powerful Subwoofer (or multiples, or SBA/DBA). Otherwise I hardly see any chance to achieve the required dynamic range for HT. Although this requires more effort regading the required amount of material, but itīll make building easier and raises the chance of success considerably.
Still it will be difficult if not impossible to achive such wide dipersion that 3 or 4 persons may be served equally well. For HT with ESLs a centre channel is rather a must than an option.

jauu
Calvin
  Reply With Quote
Old 4th October 2013, 02:10 AM   #5
diyAudio Member
 
wreckingball's Avatar
 
Join Date: Sep 2013
Location: Michigan, US
Hey guys, sorry I didn't respond earlier, I had an unexpected occurrence.
Quote:
Originally Posted by geraldfryjr View Post
The wider you make the panel the worse the beaming issues become.
Segmenting the diaphragm into narrower vertical sections will help with this.
I understand. But doesn't having the panel curved negate this to a certain extent? Seems to me I've seen Capaciti mention this on several occasions..
I'd like to avoid segmenting if possible, as I have seen many mentions of this causing the dreaded 'venetian blind effect'..unless of course you are talking about something else... hmmm...

Quote:
Originally Posted by geraldfryjr View Post
This thread may help you to get some ideas for your project,
Both the builder and the owner of these very large curved ESL's are members of this forum as well,
Building an Electrostatic Speaker
Thanks jer, I have gone over this thread several times, he does some fantastic work! I am not however looking to build a full-range ESL, I'm looking to build a hybrid with a crossover point below the critical midrange (below 200 but above 150) for reasons that Russel Knotts states about transfos only being able to cover a certain amount of FR without strain. Source: Transformers

Quote:
Originally Posted by geraldfryjr View Post
I get good performance down to about 200HZ with a D/S of about 1.5 mm to 1.85mm with a 3.25" wide diaphargm on my little panels.
Wow. Are you using shelving EQ to accomplish this?

Quote:
Originally Posted by geraldfryjr View Post
The perforations along the edge is just a matter of the material that you are using.
Having the holes will help to reduce any extra wasted capacitance.
However, They also give you less area for the tape to properly bond to as well, So you may want to consider using a wider tape than 1/4" around the perimeters or add some extra clamping forces using your final frame structure.
Good to know. Thanks for this info. However, I must not have made it clear that I intend to use the 1/4 inch spacers for non-perimeter purposes, I'll be using 3/4 inch for that.

Quote:
Originally Posted by geraldfryjr View Post
Here is a Radii Calculator that may help you,

Calculator for Radius of an Arc

FWIW

jer
This will help a bunch, thanks bro..
  Reply With Quote
Old 4th October 2013, 03:05 AM   #6
diyAudio Member
 
wreckingball's Avatar
 
Join Date: Sep 2013
Location: Michigan, US
Quote:
Originally Posted by beun View Post
Wreck,

If you really want to have a transparent membrane it's possible for me to make that. I actually have this material in stock for a width up to 30". Using horizontal spacers on a curved panel is definitely possible but makes stretching a whole lot more difficult as really only a vertical stretch is possible. This also means that a hand stretch is really the only option left unless you want to built a dedicated stretching rig for only two speakers. The hand stretch also makes the use of double sided tape a virtual impossibility, apart from the fact that double sided tape actually has a shorter lifespan that the mylar tape I use. Martin Logan can get away with it because they built a lot of the same, for custom production this becomes price prohibitive.
.
Thanks for responding beun, and thanks for your offer to build these for me, I appreciate it. But for learning's sake, for satisfaction's sake, and for the feeling similar to when you saw that CLS way back when you said to yourself "I can do that".. I think I am going to roll my own here.
I understand that hand stretching would be a real PITA, but unless costs become problematic, I think building a stretching rig is exactly what I'm going to do, plus I think I'll be building more than just two, as I have friends that need to replace their aging panels too (we all bought the same flat panels I have at the same time [from Barry Waldren, I believe]).

Quote:
The size of 18" x 60" is actually relatively large for a panel that only has to go down to 200Hz, something this size should easily make 100Hz. If 200Hz is the only requirement than 12" wide should be more than enough.
I used "quarter wavelength in size relative to lowest frequency desired" rule of thumb here. I did it in my head so maybe I calculated incorrectly, I'll have to revisit.

Quote:
The front an rear spacer thickness is identical, you don't want to introduce more asymmetry than there already exists. The spacers need to be made from a fairly hard material, you don't want them to deform because of membrane tension, I usually use polycarbonate (Lexan) for the spacers around to edge where the membrane is mounted on and 1/8" acrylic for the non structural ones in the center.

The reason for the vertical spacers is better horizontal dispersion and ease of manufacturing although I also do believe that even tensioning in both planes ultimately make a better speaker
That's interesting, I could have sworn that I read Capaciti (username) state that using a differing spacer thickness between the diaphragm and the rear stator would be acceptable, if not desirable (and I think the ESL cookbook too).. I understand your thinking on this though, I'll have to revisit some of this forums archives, and borrow a copy of the cookbook back from my buddy.

So you seem to think that foam would deform (or compress) from diaphragm tension? I wonder if anyone's tested this...you?

I understand your reasoning for vertical spacers, and you are most certainly correct that it makes for superior horizontal tensioning, I just don't know if it matters that much is all, I know I'm not sure, that's for sure..haha

Again, thanks for your response, seems to me I have a lot to learn, but you know what? I'm looking forward to it!

cheers
-wreck
  Reply With Quote
Old 4th October 2013, 04:02 AM   #7
diyAudio Member
 
wreckingball's Avatar
 
Join Date: Sep 2013
Location: Michigan, US
Quote:
Originally Posted by Calvin View Post
Hi,

it seems to me that Your design process is prone to non-success.
Iīd suggest You first specify the requirements of the final speaker, as useage, possible size range, possible positioning, etc.
If I look at the attached pic, I know my ideas about a HomeTheatre capable ESL would certainly not start at: "I want an 18" wide panel", but rather at: "What dynamic range do I want, how many persons will sit side by side, how much space do I have towards backside, side and distance to audience, etc. etc."
Hi Calvin,
Thanks for you no BS response, your opinion carries much weight (to/with me).
What I originally planned was to have an ESL that would cover all but the lowest FR (below 200-180 Hz) and be able to play at 'concert' levels (very loud) without the distortion associated with conventional dynamic, point source speakers, and without straining.
I must tell you, that all of your responses have me seriously reconsidering this project, but as they say: "how do you know if you don't ask"

As for usage, I don't separate my home theater from my 2 channel listening categorically, I am indeed seeking to make a compromise. ie: I am willing to sacrifice some dynamic range if I can still enjoy these ESLs for 2 channel music. And yes, I plan on making an ESL center channel in the future, but I wanted to get some opinions about the mains first.


Quote:
A panel is easily chosen too large for a certain listening distance and staging is lost completely. For example will a 18" panel require approximately 12-15Ft minimum listening distance.
My rough guess for a solution would be a hybrid ESL with an only 8"-12" wide panel, crossed around 300-500Hz, assisted by a slim dipole tower and a dedicated powerful Subwoofer (or multiples, or SBA/DBA). Otherwise I hardly see any chance to achieve the required dynamic range for HT. Although this requires more effort regading the required amount of material, but itīll make building easier and raises the chance of success considerably.
Still it will be difficult if not impossible to achive such wide dipersion that 3 or 4 persons may be served equally well. For HT with ESLs a centre channel is rather a must than an option.
Calvin, my trasnsmission-line FR covers anything that a dedicated subwoofer would as far as FR goes in the lower registers. It goes into the very low 20's, and covers the mid-range up to and including the 500 HZ range (current x-over point), so I see no need for a mid-range tower.
I have seen picrtures of your setup with your mid-tower, and it looks great, but like I said, I don't think my TL is going to have a problem covering the FR that the proposed ESL can't handle.

That being said, are you suggesting that an ESL will be too weak to faithfully reproduce the dynamic range that presents itself in a home theater setting? If so, I would very much like to hear you opinions on this, (more to learn maybe?)

As to my desired dispersion for ESL design, I would like very much to be able to let 4 people sitting side by side on a couch to enjoy the same (or close to) the same sounds I am hearing. Also,I never did mention that the listening distance in my home theater room is around 12 to 13 feet.

Again, any and all answers/comments/further questions are welcome.

Help me design/build this thing!

cheers
-wreck


cheers
-wreck
  Reply With Quote
Old 4th October 2013, 06:59 AM   #8
diyAudio Member
 
Join Date: Feb 2008
Location: Jackson,michigan
Hello,
I have a paper to explain this in a more physical detail and I will try to find the link on it or just post the paper itself as it explains it more in a better way, In a physical way as Roger Sanders explains it here,

AudiogoN @ CES: Roger Sanders electrostatic loudspeakers designed contrary to most popular theories. - YouTube

#2 Beun can respond to those details as I just work with smaller models for now.

I will be applying all of this to a larger model once I get the room to build one.

#3 It is dealing with the very high voltages that becomes very difficult.
If you don't apply those idea's to a panel and it fail's, It becomes very costly to (re)build and get it right, right off of the bat!!

That is the one of only reasons I have not built a larger panel so far.
For me (personally) I want the most I can get out of it!

I do keep in mind what is the cheapest way that I can go about this.

Not because I am cheap, But, Because it is not cheap, And in the long run if you do it wrong, it gets very costly !!!
And, especially when you get bad results, Then you have to do it all over again.

I don't Cator to Russ, Not that he is any type of bad person, But he doesn't, (as far as I know) is not of the DIYaudio community.

I have talked to him a few times, But lets leave it as that.

I have talked to Roger Sanders many times and that was great!!!!
It would be Great if Russ would chime in once in a while!!!!

Anybody can produce perforated panels and sell them.

Anyhow, My panels have been made to be taken apart and put back together again and again with the kind of consistency that I rarely need to re-tension them.

In the last infamous test's, I spent more time repairing the panel than I could have just building a new one!!!

I have tried many D/S spacing incrementing the D/S thickness by about .005" to .010" each time while keeping note on what the Bias voltage and driving voltage was.
And how loud it is of course !!

I used a microphone to monitor the panel while I was driving the panels past its limits.

This is not old news to these threads.

I detected clipping of the diaphragm into the stator with the microphone at a distance of about 2 feet from the panel.
Even at 200Hz it was more than to my liking and I like it loud !!!!

At 300Hz at the same setting (shifting the signal generator) it was even louder and there was no clipping of the diaphragm!!!

According to a displacement calculator it was somewhere in the +115 db SPL range before it finally burst into flames at 1meter and still kept playing for another 10 minutes because I let it and watched it burn from over excitement!

True story here at DIYaudio!!!!!

That is how I know.

All of that from a 3.25" X 9.75" diaphragm at about 6Kv of bias and 20to25Kv p-p across the stators.... you do the math.

If it had not been for that one weak spot in the design when it was built in 2003, as it was designed for only half of that voltage to begin with, it would still be playing today.
Therefore I built a new one and that is where I am Today.

The D/S of that original panel was about .075".

Now those were some extreme voltages but on the average I can get about 105db easily at 1 meter out of it with only a 20Vp-p input to the transformer (1:160 to 256 or so) with a 6.8Kv bias to boot.

If I could have doubled the bias at 13.8Kv or so this would have cut my amplifier demand in half (6db).

Imagine what it would be like for a larger panel?!!!!
Every time you double the surface area you gain 6db (Imagine that)!

So far I have only been able to get safely to 10Kv of bias so, it is not that easy to do, But the forces do still apply.
That was with my old panel with a D/S of .075"

We shall see what this new one does.

FWIW

jer
  Reply With Quote
Old 4th October 2013, 08:26 AM   #9
Calvin is offline Calvin  Germany
diyAudio Member
 
Calvin's Avatar
 
Join Date: Nov 2004
Location: close to Basel
Hi,

You donīt need to sacrifice on dynamic range just for home useage.
If You look at measurements of most dynamic home speakerīs capabilities, even larger models are hardly capable to achieve more than 100dB at lower frequencies and typically less than 110dB at higher frequencies, at 1m distance.
This SPL comes with typically 3-5dB of compression and loads of distortion.
If executed carefully and precisely a 10"x40-50" panel may achieve up to ~110dB at 3-4m distance from ~250Hz on up without compression and at THD levels of a dynamic speaker performing at less than 1W of power.
110dB@4m SPL translates to 122dB@1m for the classical measurement setup distance. The difference is lower in a room with boundaring walls, but still the dynamic range is closer to a compression horn than a direct radiator.
But its offering much more authentic sound over a much larger bandwidth.

Larger Panels wonīt add much more SPL-max or nil, since the SPL is finally limited by air voltage breakdown and typically a larger panel will be crossed at lower freqs also. A larger panel beams more than a slimmer panel and requires larger listening rooms and -distances and considerably more amp-power. Also it is more difficult to build (the width is more critical than the height here, talking of curved stators). Think of sourcing the membrane film (Iīve got for example an excellent film, but that is just 10" wide).
Regarding beaming. Starting at low freqs the panel begins lobing. The lobing getting stronger up to ~7-10kHz. Its just above that range that the curvature widens the lobe again, but it remains considerably less than the angle of curvature. Anybody talking of 30° curvature equalling 30° of dispersion is talking marketing babble. Regardless of (practical) dimensioning the panel will be beaming strongly. The best conditions for 3 to 4 persons listening would be sitting in a row on axis, one behind the other 1x1x1x1, followed by 1x1x2 and 2x2. I use an ESEL with a 10"-50", 30° panel at home for Video too and we donīt listen with more than 2 nerds side by side.

The dynamic range a panel can supply for, requires a quite potent lower-mid/Bass partner. Even more, in most films the bass is vastly overpowering everything else. So there should be quite a lot of headroom.
As seamless transition from the panel to the lower-mid/bass requires a similar distribution character, I chose a dipolar tower for the range from ~50-250Hz.
Besides achieving the right distribution character, the dipole-tower covers a minimal width and footprint area and also allows for enough membrane area to cope with the panel. Using drivers of the 5"-8" class one can easily generate huge dynamics well down to 50Hz. A dedicated woofer can take over the rest from ~60Hz to the infrasonic range.

A curved sheet panel may give You greatest dynamic range, but a flat segmented wire stator comes close and will be easier to build (way less worrying about insulation ... and man, insulation will grow You some gray hair ). Itīll cost less and will be more amplifier friendly. A flat membrane allows for heat treatment with probabely all the films offered for ESLs, while a curved panel may only be tensioned mechanically, unless You use a film with certain special qualities (afaik officially nothing left of that stuff on market any more). A mechanically tensioned film may take up to half year of time to break in and settle at the final values.

If You still decide for curved punched metal sheet stators, than be prepared to put a helluf preparation time and work into the things if You want SotA results. And do Yourself the favour to get the right and the best materials.
3/8" or 3/16" is not a question, but the road to hell.
And be prepared to maybe be in need of a different amplifier, as most solid state amps are incapable of pressing the ultimate out of the speaker

jauu
Calvin

Last edited by Calvin; 4th October 2013 at 08:30 AM.
  Reply With Quote
Old 4th October 2013, 12:55 PM   #10
diyAudio Member
 
bear's Avatar
 
Join Date: Apr 2002
Location: New York State USA
No matter what you build there are built-in compromises. The questions end up being which compromises are you willing to accept, and which not?

I am not impressed at all with the HF dispersion of horizontally segmented ESLs.
Of course, my experience may be at odds with others. Martin Logan being the prime example and probably first in the water. I'm talking here about their larger models, not just the thin small ones.

If I wanted horizontal dispersion, I'd build multiple vertical segments. The "thinness" of the segment divisions is less important than one might think, since in reality the acoustic center of the tall ESL membrane works out to be the center of the membrane. Unfortunately it doesn't really work quite as we would wish, not the single flat and planar diaphragm with uniform motion that is imagined.

So no matter what you can't get the acoustic centers super close in the vertical situation. Similarly, in the horizontal case, you have the HF centers separated by an even larger distance. Think of tweeters spaced by 12" or so?

The *idea* is that with the horizontal curvature and the horizontal segmentation that the HF output is spread uniformly across that horizontal surface. But, does it really work out that way when measured?

_-_-
__________________
_-_-bear
http://www.bearlabs.com -- Btw, I don't actually know anything, FYI -- [...2SJ74 Toshiba bogus asian parts - beware! ]
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Curved Vs. Flat ESL panels? markusA Planars & Exotics 31 9th November 2011 04:38 AM
Jig and stretch for curved ESL? markusA Planars & Exotics 56 26th January 2011 10:25 AM
curved esl question jerryo Planars & Exotics 3 6th June 2010 07:47 PM
2 Small Panels or 1 Large One - ESL audio.ESL Planars & Exotics 24 3rd November 2009 08:18 AM
Curved ESL MrPin Planars & Exotics 20 12th January 2006 12:06 PM


New To Site? Need Help?

All times are GMT. The time now is 09:46 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright Đ1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2