QUAD ESL-63 (988-2905 etc.) on steroids

I will cover som facts that you didn´t know about one of the most famous speaker ever.. hopefully someone else will contribute with some hard core information as well? Lets se... :)
Everyone know that this speaker is VERY different from all other speakers ever made!
There is no other speaker that comes even close in how it´s designed.
I have had those since the beginning of 1990´s and now I have 9 pairs.
I also have different types such as: Beveridge 2SW2, esl 57, DM70, Stax headphones..
Together with MTaudio i have refurbished many speakers and done some experimentation, which gives this thread the name "steroids".
-Is it possible to make this speaker better?
-Is there compromises made from the beginning?
-Is the newer ones (2912) different and better than 63?
-Does 989 and other 6 panel QUAD´s have a built in bug?
-Why did it take over 25 years for mr Peter Walker to develop this speaker?

I will cover some facts that you didn´t know.:eek::eek:
First a measurement of the square wave response of a sligtly modyfied ESL63
As you can see it´s not perfect.. but in the video I will show you what happends when the microphone is moved around 1" or 2,5 cm...
 

Attachments

  • QUAD.jpg
    QUAD.jpg
    849.5 KB · Views: 915
Instead of video here is some screenshots, unfortunately i don´t have my own unechoic chamber so when the microphone is moved around you will see how much the room actually is contributing. I tried to reduce reflexes from boundry surfaces a little. Behind the speaker there is another room that is well damped, and in the microphone surrounding there is some dacron panels.
So how to proceed to improve or modify this speaker with no unechoic chamer?

As you know this speaker is time coherent and has a fairly straight frequency response quite good polar responce, and ultra low distortion, in fact up to 10 times lower than most dynamic speakers.

If we should try to improve this gem we have to have an idea on what to improve and why and how?
  1. I want to improve the polar responce to get rid of the small sweet spot without destroing the other properties
  1. I want to improve on the frequency response.
  1. I want to extend the dynamics, and play at least 15dB higher
  1. I want to play down to 15Hz -3dB
And yes i have actually already done all this!
And sometimes I ask myself... was it worth 2000 hours of time studying and experimenting, a lot of money and effort and relations...(still is)?
Maybe..? Soundwice yes!
Hopefully you don´t need to put all that time and effort :)

So first, how to proceed to improve or modify this speaker with no unechoic chamer?
 

Attachments

  • kombo.jpg
    kombo.jpg
    85.8 KB · Views: 873
I got still 2 sets of 63 one SN3000 and one 33000 so pretty old and rather new :) both there panels are screwed.

-Is there compromises made from the beginning?
Yes Glue :) besides that sometimes the HV breaks but it is not as common as the panels that just fail all over.

-Is the newer ones (2912) different and better than 63?
The might changed isolation and glue and maybe the transformers. but they use the same panels as far as i know. they can all be changed out. they did make the fram more solid. something that is really was needed. i do hope they made the construction simpler as well, since the quad esl63 is a pain in the *** to put together, most retarded mechanical design ive seen so far.

-Does 989 and other 6 panel QUAD´s have a built in bug?
Yes.. glue :) all panels have the same problem and it might be because of the high voltage in general. the panel itself or the coating never fails (never seen) but the glue of the stator fails. then it will come close to the membrane and will burn holes an rattle all over. so it will destroy the panel in the end.

i got a few videos on my channel about the ESL63 if you interested. i did not finish them as of yet :( still looking for a coating that works like the quad one did. got a batch working like i want but have to scale up and see how it holds on my own set.
 
Wow this is interesting. I recalled audition the Esl 63 with Gradient subs & yes
what kept us away was realiablity issues as well as limited spl.
Would be interesting to hear & learn from your adventure

Cheers

with the use of the gradient sub or any other open baffle sub, you could filter the quad a little earlier, like 75 hz. that would save allot of power going into the quads otherwise. i bet you can crank up some dB's more out of them this way. or filter even a bit higher 100 should be possible to put out at least 6 dB more.
 
Instead of video here is some screenshots, unfortunately i don´t have my own unechoic chamber so when the microphone is moved around you will see how much the room actually is contributing. I tried to reduce reflexes from boundry surfaces a little. Behind the speaker there is another room that is well damped, and in the microphone surrounding there is some dacron panels.
So how to proceed to improve or modify this speaker with no unechoic chamer?

As you know this speaker is time coherent and has a fairly straight frequency response quite good polar responce, and ultra low distortion, in fact up to 10 times lower than most dynamic speakers.

If we should try to improve this gem we have to have an idea on what to improve and why and how?
  1. I want to improve the polar responce to get rid of the small sweet spot without destroing the other properties
  1. I want to improve on the frequency response.
  1. I want to extend the dynamics, and play at least 15dB higher
  1. I want to play down to 15Hz -3dB
And yes i have actually already done all this!
And sometimes I ask myself... was it worth 2000 hours of time studying and experimenting, a lot of money and effort and relations...(still is)?
Maybe..? Soundwice yes!
Hopefully you don´t need to put all that time and effort :)

So first, how to proceed to improve or modify this speaker with no unechoic chamer?

btw that square wave being higher in the middle and not drop instantly does it not mean it lost some of its high frequency component ?


btw you wishes are not gone happen 15 db louder and down to 15 hz...

so for the 15 db you already need more then 4 times the surface area, then you want to go instead of 40 Hz 15 hz, that doubles that once more.. well so with at least 8 times the surface area your good to go :) haha thats one hell of a speaker.


side note: the panels resonance is way above 15 hz as well, so might need to lower that to, and that might result in the membrane sticking to one of the stators :(
 
Last edited:
Since you are working on that Sisyphean task, you may want to consider some measurement software. It will make your life a whole lot better. You'll want to use a MLS or swept sine quasi-enchoic impulse response to get the room effects out of your measurements. You'll be able to see what you are doing and not constantly being confounded by your room interactions.

Good luck with your endeavors.
Sheldon
 
Off late I’ve been mucking around with hv bias on my Ml speakers & I’ve discovered that in a way they’re like tubes. There’s a sweet spot involved as well. When your in that sweet spot everything just falls into place beautifully. At lower voltage everthing is still there but they’re a little dull. At voltage over the sweet spot everything is bright & clear but in comes some hardness to the sound. I find this fascinating. Not sure if this would apply to the Quads though

Cheers
 
Member
Joined 2004
Paid Member
I want to play down to 15Hz -3dB

Find your woofer array that can accomplish this, and keep raising the xo point (as Wrinex says) on the active xo to the ESL until it can keep up. What happens at 15Hz?

Really looking forward to your insights on this speaker as there's a newly rebuilt (by Wayne Picquet) one from ~'94 with the Gradient subs in the family.
 
Off late I’ve been mucking around with hv bias on my Ml speakers & I’ve discovered that in a way they’re like tubes. There’s a sweet spot involved as well. When your in that sweet spot everything just falls into place beautifully. At lower voltage everthing is still there but they’re a little dull. At voltage over the sweet spot everything is bright & clear but in comes some hardness to the sound. I find this fascinating. Not sure if this would apply to the Quads though

Cheers

I too am fascinated by your report. But could you please tell us whatever your might mean by "everything falls into place" using readily defined descriptive terms which effectively communicate your observations to persons besides yourself.

B.
 
Hi Bentoronto
What I meant was when HV of panel are in the sweet spot instruments, vocals, percussion etc becomes clear & precise. Layering becomes better with background instruments being clearly heard. I also notice the leading edge attack on drums, crashes of cymbals, sticks sound all become very realistic. There’s also a by product if panel HV is low you’ll notice that bass will be fat & slow. Goes without saying as you bring up the HV, panel volume will increase so you’ve got to bring volume back down to compare the results of the tweaks. Whatever the changes it can clearly be heard.

Cheers