Apr 21, 2015 at 10:25 AM
Why can the diaphragm collapse semi-permanently onto one of the stators, which forces enables it to do so?

A diaphragm with no tension is inherently unstable. Precisely inbetween the two stators, equal force pulls it in both directions since the distance to the two stators is the same making it quasi-stable in that position. Due to manufacturing tolerances, it will not be precisely in the middle, and even if it were even the slightest deflection would move it closer to one stator. When it is closer to one, it is attracted to it. Even if the stator is neutrally charged, electrostatic induction will cause the electric field between the neutral stator and negatively charged diaphragm to induce a positive charge in the surface of the stator thus creating an attraction to the diaphragm.

Tension provides an opposite force to the instability so that when the diaphragm is slightly deflected in one direction, the forces of the tension are greater than the electrostatic attraction to the stator and the diaphragm is pulled back to an equilibrium point. So with enough tension, the quasi-stable system turns into a system with a wide region of true stability.

Apr 21, 2015 at 10:41 AM
A little update. We did finish our esl panel (physics project). Tried making our own transformer etc. It did play, though not very loud. Just tried connecting the panel to my srm-T1. It does play music very quietly. I'm at 12-2 o'clock and even then it is not loud at all. The sound is distorted as well, especially playing bass I think. It seems it is because parts or even most of the diaphragm has collapsed on one stator. Either the food film is bad or the tension wasn't high enough. Or both. That is my hypothesis at least.

What is the size of your ESL?  What spacer thickness do you use?  What is the bias voltage?  How much is the step up ratio on your transformer?

Apr 21, 2015 at 3:48 PM

A little update. We did finish our esl panel (physics project). Tried making our own transformer etc. It did play, though not very loud. Just tried connecting the panel to my srm-T1. It does play music very quietly. I'm at 12-2 o'clock and even then it is not loud at all. The sound is distorted as well, especially playing bass I think. It seems it is because parts or even most of the diaphragm has collapsed on one stator. Either the food film is bad or the tension wasn't high enough. Or both. That is my hypothesis at least.

What is the size of your ESL?  What spacer thickness do you use?  What is the bias voltage?  How much is the step up ratio on your transformer?

You can have the details, but consider the project a partial failure.

Size: 23x60 cm (approx.)
Spacer thickness: 1mm (approx.)
The material used is double adhessive foam tape. The configurations is 2 cm tape around the edges, a strip of 1cm tape in the middle.
The diaphragm is coated with rinse aid, which seems to work.
The diaphragm is made of food film ~11um
Bias voltage: Between 500-1000 volts
Step-up: 2x1600/1x200 in 3 phase iron core configuration, with small and very poor inductors used for general use in the physics class.

Also tried my SRM-T1 for driving the panel (580 volt pro bias).

Why can the diaphragm collapse semi-permanently onto one of the stators, which forces enables it to do so?

A diaphragm with no tension is inherently unstable. Precisely inbetween the two stators, equal force pulls it in both directions since the distance to the two stators is the same making it quasi-stable in that position. Due to manufacturing tolerances, it will not be precisely in the middle, and even if it were even the slightest deflection would move it closer to one stator. When it is closer to one, it is attracted to it. Even if the stator is neutrally charged, electrostatic induction will cause the electric field between the neutral stator and negatively charged diaphragm to induce a positive charge in the surface of the stator thus creating an attraction to the diaphragm.

Tension provides an opposite force to the instability so that when the diaphragm is slightly deflected in one direction, the forces of the tension are greater than the electrostatic attraction to the stator and the diaphragm is pulled back to an equilibrium point. So with enough tension, the quasi-stable system turns into a system with a wide region of true stability.

I've got to look into electrostatic induction, somehow I thought the diaphragm wouldn't be subjected to any force when there'd be no potential between the stators as that'd mean no electric field between the stators eg. no forces acting on the charge of the diaphragm.
This seems to be wrong.
I find it difficult to understand this concept. Is it much to ask that you break it down (when there's voltage potential between the stators and when there is no potential) or refer me to some litterature?

Apr 21, 2015 at 4:33 PM
You can have the details, but consider the project a partial failure.

Size: 23x60 cm (approx.)
Spacer thickness: 1mm (approx.)
The material used is double adhessive foam tape. The configurations is 2 cm tape around the edges, a strip of 1cm tape in the middle.
The diaphragm is coated with rinse aid, which seems to work.
The diaphragm is made of food film ~11um
Bias voltage: Between 500-1000 volts
Step-up: 2x1600/1x200 in 3 phase iron core configuration, with small and very poor inductors used for general use in the physics class.

Also tried my SRM-T1 for driving the panel (580 volt pro bias).

I've got to look into electrostatic induction, somehow I thought the diaphragm wouldn't be subjected to any force when there'd be no potential between the stators as that'd mean no electric field between the stators eg. no forces acting on the charge of the diaphragm.
This seems to be wrong.
I find it difficult to understand this concept. Is it much to ask that you break it down (when there's voltage potential between the stators and when there is no potential) or refer me to some litterature?

OK - the truth and nothing but the truth.

ES drivers are in principle perhaps the simplest to make - of all the generating priciples.

They are equally the most difficult to make them work right. Because they have to be geometrically exact and have to keep that geopmetrical perfection to work - no matter what.

I will tell you the true anecdote regarding the ES headphones by Neil Pollock, published in Wireless World in late 70s ( I shared the link ). Three students , me included, saw the article the same day. One had his drivers made from phenolic prototype printed circuit board and driven by some transformers in two days - and it fared no better than yours, despite being much smaller in size. The second did a bit better, in about two weeks - also transformers - but to stay on the polite side, let's say he demonstrated ES could potentially work...

I had to get a pro forma invoice, order the Aquadag, wait till it arrived, cleared the customs - before I could do "anything", it was almost two months. Remember, that was in 1980-81, snail mail, Yugoslavia. I did prepare in between the acrylic stators, spacers, etc - but could only proceed after getting Aquadag.

First, I used transformer. It worked - but was not an improvement over Audio Technica ATH-7 electret headphones I owned at the time as I have hoped for.

Enter amplifier. It took me almost 6 or so years, building various amps, before I found something that finally did satisfy me. And better drivers, etc, etc.

If you can, I suggest going to see Quad latest ESL speakers - they differ from their immediate predecessors primarily in how the stators are strenghtened against curving, resonances, etc - and thus achieve even better loudness and better bass, while at the same time having less distortion.

In short, the very first thing an ES driver must NOT be is - a propeller. It should be one precise plane, not twisted and curved. And making this reality is hard - VERY hard. Sennheiser went to untold extremes in their Orpheus - Quartz Glass electrodes (gold plated for conductivity ) that are about the most plan/flat thing on earth, encased in some hard wood from Africa usually used for - propellers in aircraft. Because it resists the deformation better than almost anything else, regardless being hard to machine well. No wonder it was lots of \$\$\$\$\$.

Regarding literature - http://www.amazon.com/Electrostatic-Loudspeaker-Design-Cookbook-Sanders/dp/1882580001  There was a series of articles on ES drivers in The Audio Amateur, later (reprinted) in The Speaker Builder ( google is your friend - perhaps they are still available as download )  - both by Sanders ( practical guy) and Ronald Wagner (scientific guy) http://www.hificollective.co.uk/books/bk5004.html - and it was more than a mild acrimony between the two. Yet both are very helpful at understanding why and how ES works and what can be made to make practical designs that can be duplicated by a dedicated DIYer.

You can wrap it any way you like - the EXACT method of HOW  it is made in top commercial designs that were and are successful in the market will most likely be never  revealed - it can be one's lifetime work to figure out how to make that driver that was just a bit "off" finally "perfect" - would you spend your entire work life just to be able to give everything away in a forum - for free ?

This is the primary reason why ES is not more widespread.

And we have not even touched the amps driving these drivers...

Apr 21, 2015 at 4:47 PM
Thanks, fine reading, what you say makes sense.
I do not have the time to respond properly but I have read the cookbook by sanders. It was a fine starting point. Don't know the other book, will look into it.

I've now read the wiki page on electrostatic induction. Makes much more sense why the diaphragm can collapse onto a stator.

Apr 21, 2015 at 8:55 PM
For ESL, we normally keep the spacer thickness to diaphragm width ratio to not more than 1:100. That means if you're using 1 mm spacer, you would want to keep the width of the diaphragm to not more than 100 mm or 10 cm. Your width of 23 cm is way too much and it's rather difficult to make the diaphragm stable. You can try to inject some silicone between stator and diaphragm every 10 cm horizontally. This helps as a spacer on every 10 cm.

For bias voltage, you might want to try at least 1,200V. You also want a high step up ratio transformer to amplify your music signal. You can try using a 110-0-110V primary and 3V secondary transformer. If the transfomer is good enough, it'll make good sound.

Apr 22, 2015 at 12:01 AM

I've got to look into electrostatic induction, somehow I thought the diaphragm wouldn't be subjected to any force when there'd be no potential between the stators as that'd mean no electric field between the stators eg. no forces acting on the charge of the diaphragm.
This seems to be wrong.
I find it difficult to understand this concept. Is it much to ask that you break it down (when there's voltage potential between the stators and when there is no potential) or refer me to some litterature?

When the two stators are at the same potential, then there would be no field anywhere in the system if the diaphragm were removed. However, remember that the diaphragm is charged to a high positive potential (accidentally said negative in my previous post). If you have stators at 0V sandwiching a 580V diaphragm then there is electric field between the diaphragm and each stator.

Electrostatic induction allows for a force to exist between a positively charged object and a neutrally charged object by moving charges around inside the net neutrally charged object while keeping its total potential at zero. The stable point is directly in the center between the two stators with no applied potential on the stators, because the electric field on each side of the diaphragm is identical and opposite in sign.

Now put a potential across the stators, let's say +-20V. Now one side of the diaphragm has a potential drop of 560V, and the other side has a drop of 600V. If the diaphragm is initially in the center between the stators, a force will exist pulling it towards the 600V side. To keep the post free of math, I will leave out the analysis of where the diaphragm finds a new quasi-stable point, but suffice it to say it is not dead center between the stators anymore, thus putting music between the stators makes the diaphragm mechanically reproduce the audio.

In both cases, zero potential or high potential between the two stators, the same electrostatic induction instability will exist and must be cancelled out by having sufficient tension.

Apr 22, 2015 at 1:11 AM
For ESL, we normally keep the spacer thickness to diaphragm width ratio to not more than 1:100. That means if you're using 1 mm spacer, you would want to keep the width of the diaphragm to not more than 100 mm or 10 cm. Your width of 23 cm is way too much and it's rather difficult to make the diaphragm stable. You can try to inject some silicone between stator and diaphragm every 10 cm horizontally. This helps as a spacer on every 10 cm.

For bias voltage, you might want to try at least 1,200V. You also want a high step up ratio transformer to amplify your music signal. You can try using a 110-0-110V primary and 3V secondary transformer. If the transfomer is good enough, it'll make good sound.

That's why I put a strip of tape in the centre (to support the diaphragm)
In the right light you can see how the diaphragm doesn't hold tension and is deformed. Barely visible on the images below.

Apr 22, 2015 at 1:16 AM
Yes, I can see that you don't have enough tension. You can try using a hair dryer to heat shrink the film more. It might help.

Do you use tape to permanently bond the two stators together?

Apr 22, 2015 at 1:19 AM
Still you can inject some silicone to further help the stability.

Apr 22, 2015 at 1:32 AM
Thanks for the tip

May 10, 2015 at 12:43 PM
Hi all

I'm planning to build electrostatic headphones.The thing is I would like to make a 2 stage project.First will be the headphones and then later an amplifier.So is there an easy and cheap solution for the amplifier with decent sound.I'm thinking of some sort of cheap output transformers to use with the amps I have for regular speakers,or maybe some cheap amp kit from Ebay etc. to get Me started ?

May 10, 2015 at 1:01 PM
Hi all

I'm planning to build electrostatic headphones.The thing is I would like to make a 2 stage project.First will be the headphones and then later an amplifier.So is there an easy and cheap solution for the amplifier with decent sound.I'm thinking of some sort of cheap output transformers to use with the amps I have for regular speakers,or maybe some cheap amp kit from Ebay etc. to get Me started ?

Don't mean any disrespect, but this is the same as asking: I would like to build a Formula 1 car. First it will be the chassis and then later the motor. So is there....cheap  off Ebay to get me started ?

Eeerrr - no. I DID - surprisingly - find an esl amp kit - by none other than Helmut Becker (Audio Valve ). It ain't cheap, though...

You still need a transformer, box, connectors - and KNOWLEDGE to work with high voltages. These things can be lethal if handled improperly.

May 10, 2015 at 1:49 PM
Thank you for an honest answer.I have some experience with diy amplifiers,tube circuits etc. What I'm after is maybe a "cheap" amp kit (Ebay 15 Euro/Dollar class D amps for example) or a pair of reversed mains transformers, that could handle the load to test the headphones and listen to them for a while,until I have time to make decent more expencive amplifier.If none exist I will have to hold on for a while,save some money and make the amplifier first.

May 10, 2015 at 1:53 PM
Thank you for an honest answer.I have some experience with diy amplifiers,tube circuits etc. What I'm after is maybe a "cheap" amp kit (Ebay 15 Euro/Dollar class D amps for example) or a pair of reversed mains transformers, that could handle the load to test the headphones and listen to them for a while,until I have time to make decent more expencive amplifier.If none exist I will have to hold on for a while,save some money and make the amplifier first.

Sorry, no 15\$ Class D amps capable of driving ESLs...