David Flinn
ESL Information eXch.
Govers and Hortensius
Graz Graphix
Mitsubishi Polyester
The Audio Circuit


On the top of this page you can find an overview of all brands that supply membrane materials.




micron millimeters mil inches
1 0.001 0.0393701 0.0000394
4 0.004 0.1574803 0.0001575
6 0.006 0.2362205 0.0002362
12 0.012 0.4724409 0.0004724

Membrane material

Roger Sanders Please be aware that there is Polyester and there is Mylar. Cheap imported polyester film will not heat-shrink well, nor will it maintain tension over long periods like genuine Dupont Mylar. Also, polyester is normally annealed so that it will not heat-shrink very much, which is not what we want for ESLs. We here at InnerSound supply the special Mylar we have custom made for us by Dupont to Barry Waldron from the ESL Information eXchange (see link below). Barry sells the Mylar to amateurs for a very reasonable price. This Mylar heat-shrinks beautifully and maintains its high tension forever.

Michael Ehrlinspiel Mylar is Dupont's trademark. Technical this material is called PETP. In Germany this material is produced by Hoechst and is called Hostaphan

I use Hostaphan with 6 Ám thickness for my fullrange ESL's. I produced 14 pairs for friends and others. after a certain time of playing them (more then half a year depending of listening soundpressure) the membrane becomes smooth and therefor the rumbling bass and the rough mids vanishes and booth speakers have identical frequency response in the range of membrane resonance (ca.35 Hz). In 8 years I never had problems with destroyed membranes.

Actual I'm testing Hostaphan with 2 Ám thickness (but only for hybrids, because of mechanical limits). after 3 years of continuos play I have no problems. I tried to make a thermal treatment before connecting the membranes to the frame to avoid the needed playing time, but I never got reproducible results. Do you know about a successful way?

Membrane material for ribbon loudspeakers

Christian Groneberg (May 2005) Ribbons either use metal film or a sandwich of metal film and mylar for membrane. If you use a plain metal film, you have to use a voltage transformer, because the impendance would be to low. If you use a sandwich of mylar and aluminium, you have the possibility to apply the aluminium in several stripes, which increases imoendance to an amplifier friendy load. The Membranes always have to be crincled horizontally to prevent unintended membrane resonances. This is because the magnetic force and so also the magneto-electric Force on the membrane would vary from the edges to the midpoint.

I havent seen projects with coated ribbon-membranes. In some forums the use of coated mylar, like you have in your safety pack in your car has been discussed but I as far as I know the strength of the metal is necessary to stabilize the membrane horizontally.

Membrane glueing

Nick van Beek Concerning the sticking down of the diaphragm to the stator. I used after trying the glue was the double sided tape supplied with the 3M window insulation kits. I tried a no name brand but found it didn't stick well enough. If you clean the stator surface real well with Iso Prop and then you rub the tape down to smooth it out you wind up with a nice firm bond between the tape and the stator. Then you peel off the tape backing, lay the stator on the diaphragm, and rub that down. I found this to work very well and the nice thing about it is if something goes wrong with the diaphragm afterwards it's easy to recover. With a little effort the tape can be removed and you can start over.

Will this tape stand the test of time? I don't know but what I can tell you is that I had a 10 year old roll of the stuff and it was just as fresh as a new roll I bought at the hardware store. Additionally, I rebuilt my tweeter panels last winter using this tape and to date they are still working very well.

Membrane glueing

Moray James (May 2005) Hello: I recently contacted 3M Canada technical support services regarding suitable transfer adhesives for ESL diaphragm bonding. I had been given a product number by a fellow builder and wanted to check it out. The product recommended had been 467MP/468MP.

The 3M tech rep said that for shrink film purposes that the 467MP would probably be a good choice. Further I was told that for a stretched diaphragm that the 467MP would probably not have enough initial strength to hold the diaphragm. The recommended transfer was # 9485 which is an acrylic transfer that is 5 thousandth of an inch thick. 3M also make a two thou thick version #9482 but this is only about half the strength of the thicker adhesive.

I was reminded that with such adhesives that one must remember the rules of TIME, TEMPERATURE and PRESSURE. At room temperature this adhesive will take (under pressure) about 3 days to achieve full strength. Increasing the temperature to 70 degrees C (bonding and then allowing to cool) will result in a full strength bond in an hour. So if you plan to use a stretched diaphragm you will need to keep the film on the stretching jig for three days at room temperature.

Further it was recommended that the diaphragm film be wrapped over an edge (corner) to further enhance bond strength as the shear properties exceed those of the peel. I hope that this information is usefull to you and to readers. Best regards Moray james.


Hostaphan RE film, data sheet (PDF 117 KB)

Dupont Mylar film, data sheet (PDF 122 KB)

Dupont Mylar C film, data sheet (PDF 122 KB)

Dupont Teflon film, Electret-Manufaction about modern electret-manufaction with Dupont`s film as main material for the process and a thyratron for shooting it with electrons. (PDF, 156 KB)

Questions and answers

Jean Luc Trudel (November 2003): Hello to you. Here in Canada, the only type of Mylar I can get, is Type A. It's .00048 inch thick. Do you think it will be good for ESL membrane? Jean Luc Trudel

Rob Mackinlay (November 2003): Hi Jean Luc, The Dupont A type film in .00048 (12 micron) will work but will need to be annealed (heat treated) if it is to retain tension for any length of time. The mass of thicker films such as this will not give the best transient response but will still be better than most conventional dynamic drivers.

We have a 3.8 micron film which is proven to work well and has been sent to a number of ESL enthusiasts in your country. The response of these thin films has to be experienced to be believed. If you wish for us to send you a small sample please visit our website for the necessary contact address etc. best regards Rob

Please contact The Audio Circuit if you can provide an additional answer.

Arthur Vered: Has anyone tried 1/2 or even 1 thick metal audio tape as ribbons for the membrane of an ES? What would be the resistence between two points (say a few inches apart) on such tapes? I think the tape is thin enough. It's already coated. Can it be done?

Please contact The Audio Circuit if you can provide an answer.

Ole Thofte: I just want to ask if you are aware of a Finnish product called EMFi (electromechanical film) which is a sort of electret polymerfilm with a permanent charge built in. using this film might make it possible to build an ESL-speaker without the high-voltage.....? Does an ESL-speaker need to have a planar membrane? I'm experimenting at the moment with a dynamic driver using a large cone (250 cm tall!) as membrane. I'm going for the omnidirectioanal quality. A similar membrane might be possible as an ESL - or? Best wishes, Ole Thofte, Copenhagen

Barry Waldron: Electrostatic loudspeakers can take many forms. Modern day devices follow one of three plan forms. One is the flat (planar) cell; another is the curved cell; the last is the arced array. For the discriminating audiophile who desires the best image and transient response, the planar ESL is unbeatable. Room interaction is minimized. This and first arrival times that are not contaminated by reflections create an image that in nearly holographic in form. The curved cell, popularized by Martin Logan, widens the sweet spot to the detriment of achieving a pristine image. Moreover, the diaphragm travel is not linear. The membrane expands in the forward direction and collapses as it moves to the rear. This contributes to increased distortion. The arced array is a method that uses multiple flat panels placed about an arc. This creates multiple sweet spots. This architecture was popularized by SoundLab. The method works well but is only practical if listeners sit at the focal point of a stereo pair. Imaging also suffers due to increased room interaction.

Omnidirectional sound has its following as evidenced by the fact that Bose continues to sell speakers that bounce the sound all over the room. With the above information in mind, I do not understand how any audiophile who is serious about the ability to create an accurate illusion engine in his/her home, can want an omni directional system. Literature from manufacturers point to the fact that their products place you inside the orchestra. I guess this is alright, although I thought I purchased an admission ticket to sit in Row A center! I hope this is of help. Barry Waldron

Michael Ehrlinspiel : Concerning finish electret polymer: Two things may be critical.

  1. is the used material comparable to mylar in mechanical cases.
  2. electret microphons uses in the range of 1.5V to polarize the membrane. I think it should be very difficult to make a electret membrane that keep charges comparable to 5000 V systems in ESL's.

Concerning membrane shape:

Typically you are free in dimension. martin logan uses bended esl. in praxis it is limited by production affairs. If I understood your construction of a 250cm loudspeaker, you force your membrane only in the mid, so that you have only uncontrolled movement in the outer ranges. this is comparable to Manger- or Mission NXT- technic but with the difference of not calculating this uncontrolled movement.

Please contact The Audio Circuit if you can provide an additional answer.