For quite a long time manufacturers of ESL's have been divided in two camps : either using flat, conductive sheets(e.g Quad, Martin Logan) or stretched wires(like Acoustat, Audiostatic). Each way has it's own advantages and field of application. For example, Martin Logan has succeeded in building optically appealing, "airy" designs which is obviously a very important selling point. On the other hand Quad focused in making full range designs without the use of crossovers, trading visual transparency for other design considerations, while Acoustats have been known for their reliability.
Building an electrostatic loudspeaker has many challenges. First, it should be protected from dielectric breakdowns, as voltages involved are in the order of thousands of volts. This is also an important safety counter-measure. One way of doing so is to provide insulating layer of plastic. Perforated metal sheets can be powder coated. However coating stators to resist high voltages is a difficult task. Designs employing stretched wires are often more robust to breakdowns. Main reasons are absence of sharp edges and a uniform layer of insulation.
Building an electrostatic loudspeaker has many challenges. First, it should be protected from dielectric breakdowns, as voltages involved are in the order of thousands of volts. This is also an important safety counter-measure. One way of doing so is to provide insulating layer of plastic. Perforated metal sheets can be powder coated. However coating stators to resist high voltages is a difficult task. Designs employing stretched wires are often more robust to breakdowns. Main reasons are absence of sharp edges and a uniform layer of insulation.
Another quite radical idea to prevent arcing was to "sandwich" panel within a thin film and fill interior with gas which has higher breakdown voltage than air. Such speakers were manufactured by Dayton Wright. However the gas was heavier than air, affecting high frequency response as well as escaping confinement over time. A more successful approach is to manufacture sheets from perforated printed circuit boards and put laminate side towards the membrane, i.e with copper facing outwards. Because dielectric constant of PCB material is quite high loss of sensitivity is insignificant. The spark must travel much higher distance to outer side of stator thus likelihood of arching is reduced. This is the solution used in well-known Quad ESL 63 and some other models.
Another problem of flat ESL's is beaming. When wavelength exceeds dimension of sound source it starts to become increasingly directional. For example, if radiator width is 20 centimeters it becomes more and more directional above approx. 1.7 kHz. There are several solutions to this problem. The first approach is to use separate panels for different frequencies as used in Quad ESL 57. Second is to create an imaginary point source by delaying sound propagating through membrane. Again, this was very successfully used by Quad in model 63 and above. The third approach is to curve the panel as employed by Martin Logan. The last approach I am going to discuss is specific to stators built from stretched wires. Wires can be grouped electrically and then fed with different frequencies. This is called electrical segmentation. Each segment may be optimized so that it's width is less than wavelength radiated, thus achieving good polar response along horizontal axis. A tall, narrow element improves dispersion in vertical direction. Reflections from floor and ceiling are reduced as well. This is often called a line source. In reality it's only an approximation to a line source because of finite width of the speaker.
Disclaimer. The author of this site is not responsible for, and expressly disclaims all liability for, damages of any kind arising out of use, reference to, or reliance on any information contained within the site. While the information contained within the site is periodically updated, no guarantee is given that the information provided in this website is correct, complete, and up-to-date.