DIY Electrostatic Speaker

[turgid]

Hello,

I should have posted this on the 17th of May, back when I finished working on my ESL and played it for all of 2 minutes before it snapped and started leaking charge.

If I delay things any further, memory will start polluting those 2 minutes with superlative words, so posting the schematics and also some pictures. It took about 1 min. for the diaphragm to charge up, I expected more time for that.

As for the sound - I was pretty convinced that the very small dimensions of the active area - 12cm by 23 cm would cause a considerable frequency roll-off anywhere by 700-1KHz but this was not the case. What stood out to me were the clarity, distinction and sonic spacing of the instruments that made it sound very very open. It sounded huge considering the size of the speaker. I thought there would be a drastic roll-off as already mentioned but the ESL suffered serious lowered decibel levels at texturing the bass guitars but each note stood out clean and clear from the rest. I couldn't hear any bass drum except for a low pop sound(probably due to the diaphragm hitting the stators).
The reason the bass was more is probably due to the slightly thicker diaphragm material that I used(more on that later).
Vocals were the strong suit and the details were crisp. The treble sounded awfully bright and harsh at first but settled down for a clean sound. There was also a noise floor and the space between notes was covered with this noise.

Will post more impressions upon request albeit from memory. I wish I had made a video of it before it snapped. I will put up some more pics tomorrow.

Cheers,
Azaan

 

[thx911]

forgive my ignorance - what is ESL?

 

[turgid]

forgive my ignorance - what is ESL?

Electrostatic loudspeaker

They work on a slightly different principle from the electro-dynamic loudspeakers. Also the sound quality is much much better relatively with next to no audible distortion(if precisely made)

Electrostatic loudspeaker - Wikipedia, the free encyclopedia

Cheers,
Azaan

 

[linuxguru]

More details of circuit and construction, please? What are the specs of the main drive trafo? Bias supply? Did you coat the diaphragm and stators with an insulating coat (say, acrylic clearcoat)?

You probably need a 'helper' woofer to take care of (say) 40 Hz to 1 KHz, and use the ESL for everything above that, much like the Cadence Audio 2-ways.

 

[captrajesh]

More details of circuit and construction, please?

What is the diaphram made of? AFAIK, the latest designs use Iridium Oxide impregnated PTFE. What have you used here?

 

[turgid]

Ok here goes - The diaphragm is made of household plastic cling film which is elongated and brushed with liquid dish soap, the stators are perforated SS. For stator insulation I used Johnson's baby powder first in a liquid paste form. Once dry, put more and more powder on the stator. I wouldn't recommend this method as this might be one reason why the diaphragm snapped. I plan on using a clear coat of primer(w/acrylic) in my next build.

The circuit is not completely presented in the schematic. I used a 4.7ohm resistor to connect the trafo out (primary as reversed). As evident I used a R-ACT method for creating a VCT. The BIAS supply is a normal voltage ladder circuit that puts out around 2 KV DC theoretically which I soldered on the perf-board using 1N4007 diodes and 10nF, 2KV capacitors(the high capacitance probably caused the harsh treble at first). 6 diodes and equivalent capacitors as in the pic. The trafo is a 240V to 3V step down in reverse.

Thx for your replies linuxguru and captrajesh but I am just trying the principle and not re-inventing the wheel here. As for hybrid designs, I will be looking into them shortly and also prefer wire ESLs.

I will put up rest of the pics in the next post.

Many thanks,
Azaan

 

[captrajesh]

Your thread piqued my interest because I had Cadence Arista Two Way Hybrid Towers which use ESL panels as Tweeters.

I read somewhere that the membrane used in their panels was produced by some German firm based on a confirmed order for 10 kilometre long roll.

 

[Thad E Ginathom]

And now you find out that all you needed was some cling film and soap!

But seriously. turgid, it is just amazing that you have done such a thing. Really great.

 

[ranjeetrain]

The diaphragm is made of household plastic cling film

First off, great to see such a genuine DIY effort.

I have a suggestion, if I may. How about trying out thin aluminum film (Ordinary aluminum foil used to to wrap stuff in the Kitchen).

 

[linuxguru]

The difficult part is the diaphragm - it has to be thin and lightweight, but also stiff and capable of creating sufficient force to move air over its entire area. It has to be conductive enough to distribute charge evenly over its surface, but with high-enough impedance to ensure that charge does not migrate dynamically during operation (which will increase distortion).

The current materials of choice are PTFE and PETP, with conductive coatings of Graphite or Indium Oxide. These would be mostly out of reach of a DIYer, but alternatives are possible, including anti-static aluminium- and graphite-coated plastic bags used for shipping circuit boards and similar.

Your approach of using liquid dish soap has to be commended for novelty and simplicity, but it will require repeated application and renewal at periodic intervals.

Edit: It might be worth considering a sandwich of conductive-coated plastic (either the black or shiny kind), placed between two sheets of clear plastic laminating film with glue (used for laminating documents, ID cards, etc.) and laminating it flat with a laminating machine. This shouldn't cost more than about Rs.50 to 100, depending on the size of the panel. You could also stick in thin wires longitudinally and in contact with the conductive plastic bag for the bias. The only drawback is that the laminating material is fairly heavy, and the glue makes it heavier.

 

[omishra]

Great!
:clapping: yes this would be future for speaker technology. Now companies make it, diy would be nice to see.

 

[turgid]

The difficult part is the diaphragm - it has to be thin and lightweight, but also stiff and capable of creating sufficient force to move air over its entire area. It has to be conductive enough to distribute charge evenly over its surface, but with high-enough impedance to ensure that charge does not migrate dynamically during operation (which will increase distortion).

The current materials of choice are PTFE and PETP, with conductive coatings of Graphite or Indium Oxide. These would be mostly out of reach of a DIYer, but alternatives are possible, including anti-static aluminium- and graphite-coated plastic bags used for shipping circuit boards and similar.

Your approach of using liquid dish soap has to be commended for novelty and simplicity, but it will require repeated application and renewal at periodic intervals.

Edit: It might be worth considering a sandwich of conductive-coated plastic (either the black or shiny kind), placed between two sheets of clear plastic laminating film with glue (used for laminating documents, ID cards, etc.) and laminating it flat with a laminating machine. This shouldn't cost more than about Rs.50 to 100, depending on the size of the panel. You could also stick in thin wires longitudinally and in contact with the conductive plastic bag for the bias. The only drawback is that the laminating material is fairly heavy, and the glue makes it heavier.

Laminating film is a better choice than cling film. The cling film is even heavier and need a lot of elongation. Cling film.is also thermally unstable and dish soap or PVA glue will suck any moisture from the atmosphere. This is one reason I did not carry the ESL with me to Kochi. Anti-static spray is a really good choice for coating and is said to be an easy find in computer stores.

As for the constant charge over diaphragm - this can be attained by connecting a 10Mohm resistor in the bias to diaphragm connection. This is present in my circuit diagram.

Thanks for your interest...

Regards,
Azaan

 

[turgid]

More details of circuit and construction, please? What are the specs of the main drive trafo? Bias supply? Did you coat the diaphragm and stators with an insulating coat (say, acrylic clearcoat)?

You probably need a 'helper' woofer to take care of (say) 40 Hz to 1 KHz, and use the ESL for everything above that, much like the Cadence Audio 2-ways.

I do not like the idea of the cross-over frequency being anywhere near 1KHz, there is a certain sonic-transition between the dynamic woofer and the E-stat that needs to be smooth as possible. I feel the cross-over frequency should be below 500Hz. (provided the ESL is at least 4-6 times the size of mine)

I also would not like the ESL to just act as a tweeter as they mostly marvel in the mid-range.

Many thanks,
Azaan

 

[linuxguru]

Laminating film is a better choice than cling film.

Cool - please try it if you can. The maximum size I've seen here is about legal paper size (~11" along the longer axis). I think that's not a huge limitation, it still gives a pretty useful panel size. A thin copper wire can be routed in a zig-zag, non-inductive pattern to make contact with the conductive portion (either antistatic spray or black antistatic plastic film) between the laminates.

Regarding the crossover frequency, whatever sounds best is fine. The ESL will be bass-shy below a certain frequency, since it's essentially an open baffle. Maybe 500 Hz is good crossover, maybe even lower, maybe higher - it can always be found by trial and error. It's best to divert the high-power components of the signal into a conventional cone woofer (maybe 2 drivers) and let the ESL handle the portions with limited physical excursion (mids and higher).

If you're thinking of making a small-volume commercial offering, you may want to get in touch with Sreekanth (soundsgreat) at Sanjaynagar. He has vast experience with drivers, cabinets, crossover and all aspects of making very high-quality 2- and 3-way speakers. If you can collaboratively bring out a reliable and durable 2-way hybrid, that would be great.

 

[turgid]

Many thanks for your informative reply sir.

I wish I could stay back and contact sreekanth but I am flying to Belgium to pursue my masters. I am already in touch with a few people there like Geert Vijncke and 'Kevin' who are masters at making ESL hybrids. Amsterdam is just an hour from Antwerp, my place of residence, and I intend to meet the infamous Rob De Lugt.

Best regards,
Azaan

 

[turgid]

Here are 2 more pictures.

 

[turgid]

Hi Mod,

Request you to change the name of this thread to 'DIY Electrostatic Loudspeaker from Household stuff' for accounting post#2 in this thread. I can't do this myself.

Many thanks

 

[turgid]

Hi guys.. I am planning to jump into planar magnetic speaker soon...

Hope that pays off better

 

[arj]

Will post more impressions upon request albeit from memory. I wish I had made a video of it before it snapped. I will put up some more pics tomorrow.

Cheers,
Azaan

Sir, Hats of to you for your innovation , passion and perseverance and wishing you the very best in taking this through ! :clapping:. Really impressive

 

[turgid]

Sir, Hats of to you for your innovation , passion and perseverance and wishing you the very best in taking this through ! :clapping:. Really impressive

Oh why thank you! But i didn't take it all the way, had it played for more time I would have at least tried to post from FR charts and SPL levels...

To be honest with you, this took a lot of time and effort for making and patience behind ordering of components while the result was only good for a few minutes. :sad:

This is one reason I am shifting to other types of Planars. :licklips: