Electrostatic Tweeter / Midrange - diyAudio
Go Back   Home > Forums > Loudspeakers > Planars & Exotics
Home Forums Rules Articles diyAudio Store Gallery Wiki Blogs Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

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
Thread Tools Search this Thread
Old 19th September 2015, 01:24 AM   #1
diyAudio Member
Join Date: Oct 2009
Location: Hamburg
Default Electrostatic Tweeter / Midrange

How large do we need to make a Electrostatic tweeter to match nicely with a conventional mid bass driver.

While I love my full range electrostatics I would also like a tiny speaker, and am no particular fan of moving coil tweeters, and ribbons typically require a 3 way.

If we target a 110 dB SPL to around what ever is needed to cross over with a good 4" to 6" mid-bass driver. How small can we make the electrostatic speaker, and still require no more than 30 W to drive it and about the same power to drive the bass. I assume we will go active.
  Reply With Quote
Old 19th September 2015, 03:11 AM   #2
diyAudio Member
Join Date: Feb 2008
Location: Jackson,michigan
About 3" x 9" to 4" x 12" is good.

Use a very good stator coating and run a bias voltage of about 7Kv to 8.5 Kv and you will reach your goals.
Or at least very close to them.

I got +105db under such conditions with only a 50-100 watt amp.
I have stated these figures several times in various threads.

I was using only a 5.25" driver so my low end peak was limited in order to be flat.
The ESL's could do much much more with a measured 110db than what the woofer could do to keep up.

That is the pretty much the performance I got with my desktop models.

If you electrically segment the stator you can still go wider like 6" and still have a wide and even horizontal dispersion.

This will allow a bit more Ooomph at the lower end, and have a lower reasonable crossover point.
Much more lower that about 800Hz to 1200Hz!!
I was using a D/S of about .072" and I have increased this somewhat on my newer version for better room for Diaphragm movement in the 200Hz to 300Hz range.

The photo is of my latest untested version (that I had made two years ago), designed to meet the above requirements and push the bias level even higher toward 10Kv-13.5Kv.

The next set of photos are of my test panel as described in the threads and the calibrated measurements of it the very day I pushed it to far and it finally gave up for good.

It really wasn't designed to be pushed as far as I did it, but it was a trooper to work well enough to do it several times and allow me to finally get some solid data.

The measurement were set at 95db and at a 10v peak signal into a 1:160 transformation ratio and i think my normal 6.8kv of bias.

Running the bias up to 8.5Kv-10 Kv range yeilded me another 3-4 db of efficiency with no change in the drive level.
I ran them at 10Kv for quite a while on this last test, But it was pushing the integrity of my stator coating until their final ultimate demise.

The efficiency of this setup at 6.8KV was measured at about 89-91db with a 5v peak signal feeding the step up transformers.

There was slight 3db peak slope in the otherwise flat response curve that accounts for the 91db and is probably some sort of room reflection of my test setup when I did the test.

Those conditions where refined for the calibrated curve as shown and the 89db for a 5Vpeak corresponds to the 95db level for a 10v peak input using a sine wave test tone.


Attached Images
File Type: jpg ESL2.jpg (60.4 KB, 389 views)
File Type: jpg Desktop ESL.jpg (74.7 KB, 389 views)
File Type: jpg ESL and Woofer.jpg (101.3 KB, 380 views)
File Type: jpg Octave smoothing.jpg (124.0 KB, 373 views)
File Type: jpg Third Smoothing.jpg (126.1 KB, 368 views)
File Type: jpg Raw.jpg (130.1 KB, 67 views)

Last edited by geraldfryjr; 19th September 2015 at 03:33 AM.
  Reply With Quote
Old 19th September 2015, 03:26 AM   #3
diyAudio Member
bolserst's Avatar
Join Date: Jun 2005
Location: Fort Worth, Texas
Originally Posted by owenhamburg View Post
If we target a 110 dB SPL to around what ever is needed to cross over with a good 4" to 6" mid-bass driver. How small can we make the electrostatic speaker, and still require no more than 30 W to drive it and about the same power to drive the bass. I assume we will go active.
If 110dB @ 1m is truly the requirement you would need a 23cm diameter ESL to cross over in the 2kHz - 3kHz range. If you can move the crossover a little higher (or lower your SPL requirement) you could manage with a 17cm diameter ESL. The output capability increases +6dB for every doubling of panel area.

Just to clarify, these area and SPL numbers are for a dipole ESL panel.
You can calculate them for yourself at this site:
Electrostatic Loudspeaker (ESL) Simulator

With typical 1.5mm D/S spacing, use 3500V polarizing voltage and 3500Vrms Drive voltage to get a good estimate of peak output capability for an ESL. Change dimensions and distance to see effect on SPL.

More details on using the ESL calculator here:
Help with esl simulator

Attachment #2 of this post shows how to get a flat response using a series resistor:
Help with esl simulator
Attached Images
File Type: png ESL_SPL_Calc.png (26.9 KB, 78 views)

Last edited by bolserst; 19th September 2015 at 03:46 AM.
  Reply With Quote
Old 19th September 2015, 07:20 AM   #4
diyAudio Member
Join Date: Feb 2008
Location: Jackson,michigan
In reality it took about 100watts to get 110DB with a 6.8Kv bias.

My as measured efficiency was 89db for a 5V peak signal (close enough to 4Vpeak for a 1 watt reference).

So for 100 watts=20db 20+89=109db !!

And then for everytime you double a voltage it means +6db of more SPL.

Doubling of the bias is the easiest.

So if I can run a 14Kv bias than it would only take 50 watts at 20v peak to get to 110db or so.
I did this a couple of times but my stator coating was starting to fail at anything above 12Kv.

I was running a 8.5Kv bias and about 50v peak input when it happened (burned for the last time).

Anyhow 10.2Kv of bias gave me approximately +3db more so now my efficiency is now 91db as measured.

So from 5v to 40v (for a 100watt amp) this is a doubling of the voltage of three times for a +18db.

Therefore 18+91=109db as measured with my SPL meter in my test at 1 meter from the panel.

Being that I normally sit no more than a .5 meter from them you can add 3db to that final figure.

And IF I could run them at a full 14kv of bias that adds +6db from my nominal 6.8kv.

And add another say 2-3db from 40v peak of drive to 60v peak ( the equiv. of a 200-225 watt amp).
I was also using my Crown DC300a for the extreme tests,

Our total now comes to approximately 109+2.5+6=117.5db from a 3.25"x 9.75" ESL!!!

Double the surface area and add another 6db to that!!

So by the same token a 3.25" x 19.25" ESL could produce 120db with just a 150 to 200watt amp or 117.5 with a 100 watt amp (40v peak) !!!

And we would still have our 111.5db from only a 50watt amp.
I got this easily with my cheapy AWIA 80watt amp under normal conditions.

Only the little 5.25" Sony woofer I was using couldn't keep up.

That is why the graphs are set to 95DB.
That little woofer was doing all it could do below 100hz without hitting its stops and flying apart!!

From the very same single panel as shown in the photo I posted.
I didn't even have the added benefit of running two panels at the time.

The 115db range is what I know I had when I didn't even have an SPL meter at the time back in 2010 when I had First posted my info on this project.

I was running about 6.5Kv-7.5Kv of bias then and 20-25KV p--p across the stator's (as confirmed by my scope) for 20 minutes straight the First time it arced over and caught fire.

Like I have said before it was loud enough to scare even me out of the room at that level and do play a very loud guitar.

Now, I know that most probably don't like my ideas of pushing the voltages and I wouldn't expect the average DIY'er to do so either!!!

But it was the only way for me to find out what I wanted to know since nobody had any answers for what I was seeking to find.

I did all of this in the name of DIY ESL Research.

And yes the ESL Simulator is pretty much right on from the results of my tests.
Very nice work on that Calculator indeed!!!

But the question did arise of how small can we make an ESL and still get 110db at 1 meter with a 30 watt amp.

As that has been my criteria every since I started building ESL's in the First place from the beginning in 2003.

Cheers !!


Last edited by geraldfryjr; 19th September 2015 at 07:29 AM.
  Reply With Quote
Old 19th September 2015, 10:24 AM   #5
diyAudio Member
Join Date: Oct 2009
Location: Hamburg
Thanks guys. Thinking carefully I donít think we really need 110 dB but Iím going to keep looking at it as a reasonable safety margin.

Looking at your suggestions, your are both suggesting much larger D/S spacing than I thought.

bolserst: 1.5mm
jer: 0.072" == 1.8288mm

Having only made and ESL with 0.3 mm D/S spacing your proposals seem to be for large D/S spacing.

Looking at Piston Excursion calculator

With a 4" driver, playing 1000Hz at 110 dB SPL @ 1m we only need 0.22 mm Xmax. For for 500Hz, at 110 dB SPL we will need 0.8 mm.

Why not go for a 0.5 to 1 mm D/S spacing?
  Reply With Quote
Old 19th September 2015, 01:00 PM   #6
diyAudio Member
bolserst's Avatar
Join Date: Jun 2005
Location: Fort Worth, Texas
Originally Posted by owenhamburg View Post
...Why not go for a 0.5 to 1 mm D/S spacing?
I mentioned 1.5mm spacing because that is the most common spacing used.
I have built ESLs with 0.75mm spacing for use as RTR tweeter replacement.
Halving the D/S spacing halves the required voltages, but the max SPL is the same.
So smaller gap will not get you higher max SPL but it will allow you to get there with lower voltages.

You are correct that the motion of the membrane due to audio signal will be very small.
However, remember that the diaphragm will be deflected from its center position when bias voltage is applied. But, as long as you are able to build flat, tight tolerance stator and spacer assemblies you can work with the smaller gaps you mention.
  Reply With Quote
Old 19th September 2015, 03:19 PM   #7
diyAudio Member
Join Date: Feb 2008
Location: Jackson,michigan
Yes, in my tests the D/S of .072"-.075" of my original Panel was quite sufficient for most cases.
I used just standard thickness of plexiglass.

Even at the extreme levels it was fine for anything above 400hz to 300Hz or so.

But below 300hz and at 200Hz I got clipping of the Diaphragm in to the stator's as verified by my scope showing flat topping of the waveform of the signal coming off of my test microphone.

I verified this by moving the mic. father away from the panel to make sure that it wasn't the mic clipping.

So switched out the 1/16" frames for the 3/32" ones that I had already made and I had no issues of such clipping after that except when I hit the resonate frequency of the diaphragm in the 70Hz to 90hz range.

Also as mentioned I just simply raised the Bias voltage to compensate for the larger D/S and got the same SPL with the same drive voltage but no nasty clipping or slapping of the diaphragm into the stators.

It was loud and clean!!

Now, Yes 110db is quite loud!!!
But, It is not a unreasonable goal considering on how far away your listening position may be!!!

For instance my Desktop unit is only .5 meter away normally, But my room is 18 feet long and if I were to put them on the far end I will lose approximately 8-10db at my listening position at 14-16 feet away.

And even though I do like it loud when I am playing along with my guitar and other rare occasions, I have found that 105db is my comfort zone to set my mains at.
This means that they have to produce at least 114db, and when they were working I could get 109db clean at my listening position.

So, when I was listening to my Desktop ESL (just one) I had found the even at 90db of SPL it was more than enough to be comfortable and they could do 95db flat as my charts have shown at 1 meter.

Even the roll off at 100hz didn't seem to be much of a problem to cause any degradation of the music I was playing on the low end ( Rock Stuff with lots of Bass Guitar and clean Kick drums like Montrose , Spin Doctors and such) at a full 95db!!

And the clarity of ELP's "From the Beginning" was just astounding!!!

Especially, When I was only listening at an average of 85db to 88db I was able to use a bit more low end and to be flat down to 40Hz or so and it still played with much authority only it wasn't earth shaking!!!
But very very moving as the way ESL's are anyhow!!



Last edited by geraldfryjr; 19th September 2015 at 03:26 PM.
  Reply With Quote
Old 19th September 2015, 04:14 PM   #8
diyAudio Member
Join Date: Feb 2008
Location: Jackson,michigan
P.S. My new design is still set at 1.85mm or .072" but I can easily add an exact amount of more spacing should I need to when I do finally finish it and start some testing.

The coating on the stator's have already passed the HV tests I have given them for breakdown at 14Kv and above using a mock up test with a transformer connected and amplifier running as well as my variable HV supply.

I am hoping to get good results of a crossover down in the 300-600hz range but I may need to use a wider panel for this to happen smoothly below 600Hz or so.

I have a 6" by 18" electrically segmented stator in mind for my next design using some of my .062" TIG rod and one or two of my 8" subwoofers.

I will get some thinner .035" or .045" stock sometime in the future for more builds.

More on this later.
  Reply With Quote
Old 22nd September 2015, 02:58 PM   #9
WrineX is offline WrineX  Netherlands
diyAudio Member
WrineX's Avatar
Join Date: Mar 2004
Location: Den Haag
Maybe you should take a look at my sandwich esl ? yummie. its damn tiny has verry decent to high spl. and could even fit into an a large cone driver.

If tranformers where not that expensive i would persui this further. but im low on cash. and i want to someday produce exoctic loudspeakers for a decent price.

oh i missed the part about not wanting a 3 way. well the tiny esl i used could be crossed at i think 1000 i believe. thats still almost 2k lower then most tweeters. i can imagine a simple paper woofer can reach 1khz pretty easy (and beyond)

sandwich esl ?Yummie

Last edited by WrineX; 22nd September 2015 at 03:25 PM.
  Reply With Quote
Old 22nd September 2015, 04:12 PM   #10
WrineX is offline WrineX  Netherlands
diyAudio Member
WrineX's Avatar
Join Date: Mar 2004
Location: Den Haag
sorry i re read my own post and the small one should be crossed at 1.4khz. the bigger but still smal panek can take 900hz. maybe still not in your range... damned i get a feel i must do a revisit of these panels you got me stoked again
  Reply With Quote


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
electrostatic tweeter repair Sony SA-EX100 ric-paul Planars & Exotics 74 26th July 2016 11:48 PM
TWEESL 2020: an electrostatic tweeter with electrically delayed concentric rings Andersonix Planars & Exotics 6 27th November 2012 09:13 AM
Tweeter below Midrange okay ? Scott L Multi-Way 2 5th March 2011 08:44 PM
Interesting old Philco electrostatic tweeter peace brainerd Multi-Way 0 8th March 2010 10:56 PM
Long, narrow electrostatic tweeter for line array Bill Fitzpatrick Planars & Exotics 13 29th October 2004 08:33 PM

New To Site? Need Help?

All times are GMT. The time now is 09:24 PM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright ©1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2