The superior principle

Back in 1985 I could listen to an ESL for the first time in my life and I´ve continued to follow this principle. Even though the ESL on show wasn´t a first class device after my todays ideas, its performance was so much better than anything I had experienced before, that I immideately understood, why someone could be willing to spent a lot of money on a HiFi...... Over time it would be much less money spent as if he built a concert hall in his garden and had musicians flown in ;-)

Because out of a sudden Salomon Burke materializied in all his body size right in front of me and sang: "A change is gotta come...!" Well, how true he was

 

ESL-principle
ESL-principle

Electrostic Loudspeakers -short electrostats or ESLs- work with ultra-leightweight diapghragms, that follow the music signal with much greater ease and precision than any common dynamic loudspeaker ever could.

The sound quality of an electrostatic transducer is therefore extremely good and authentic. No other sound generating principle comes close to an ESLs qualities.

 

diaphragm
diaphragm

The diaphragm, a ultrathin plastic film, is located midways between two electrodes, called the stators. The electrodes may take up nearly any shape and size, be it round, rectangular or curved. Spacers between diaphragm and stators hold the membrane at a constant, centered distance to the stators. To function, a charge must be impressed upon the membrane. For Headphones, Microphones and very small ESL-panels so called Electret-films may be used, which come with a ´built-in´ charge.

HV-charging supply
HV-charging supply

For Loudspeaker-panels the charge is provided by a dedicated (highvoltage) HV-power supply (marked with the + and - in principle-pic).

The membrane needs to be coated weakly conductive so that the charge is distributed even over the entire membrane area.

 

Audio transformers + HV-supply
Audio transformers + HV-supply

It requires high signal voltages to create a sufficiently strong, homogenous electrical field between the stators. The stator electrodes are connected to the amplifier via a audio transformer, that transforms up the lowish amplifier signal voltage to the required HV-level. The developing electrical field directly acts upon the charge fixed on the membrane und vibrates the membrane depending on the music signal. Because the field acts evenly upon the whole membrane area, the membrane doesn´t need to be stiff but can be chosen from a soft highly dampening material.

 

effects of different film gauges
effects of different film gauges

Almost exclusively PET-films (polyester) of the brands Mylar and Hostaphan are used in thicknesses from 1/50mil for headphones up to 1mil for large bass-panels. The majority of manufacturers use 1/4mil to 1/2mil thick films.

I use 1/7mil and 1/6mil thick films. Human hair in comparison is 10-20times as thick in diameter!

The mass of films of less than 1/4mil is so low that it doesn´t play any role in the audio frequency range. Electrostats have been named massless speakers because of this. Thicker films exhibit a rolloff toward the upper bandwidth limit. Compared to this even a lightweight 1" dome tweeter suffers from mass-related effects beginning at much lower frequencies of a just a couple of kHz.

 

From the beginning I concentrated on electrostatic hybrids, which are a combination of an electrostat for the mid-highs and a dynamic bass speaker.

The reason simply is, that the ESLs superiority restrics to a wide-bandwidth range from approximately 150Hz up to >20kHz. Lower than 150Hz the ESL loses most of its advantages. As soon as the audio signal requires diaphragm excursions of more than +-1/50" all the glory disenchants. Dynamic bass speakers have a clear advantage in this frequency range.

They exhibit a larger Force-per-Area-unit product and can achieve higher soundpressure levels (SPL)  with considerably less membrane area.

Woofers using dynamic drivers can therefore build more compact.

Due to their construction details the freeair resonance (Fs) of the drivers is well controlled. Typically one finds quality factors (Qt) of less than 0.5, which asks for a casing to pimp bass reproduction.

amplitude- and phase response
amplitude- and phase response

The drum-like tightly tensioned ESL-diaphragm on the other side suffers from hardly any damping of its resonance, which leads to amplitude peaks of up to +10dB at the resonance frequency. Does anybody really expect a bass system exhibiting a Qt >>1 to sound clean and precise? One can control or at least lower the resonance effect by mechanical segmentation with varying segment sizes or by using damping material, but this costs on ( the anyway terribly low) efficiency, and/or sonics and/or optics.

The only factor saving a fullrange ESL from desaster is its typical working as an open-baffle dipolar speaker, which reacts differently upon room resonances and which affects booming much less.

The tendency to a boomy one-note-bass and to thin sounding lower mids is still audible though. A dynamic bass driver also working as dipole sets the mark for a ESL with regard to size, dynamic range, precision, efficiency and low distortion to unreachable levels.

In a frequency range higher than 150Hz ESLs rule. They use their large and extremely leightweight membranes, their highly linear motor and their wide-bandwidth-characteristic of up to 7 octaves to full advantage.

The required membrane excursions become very small so that stator-stator-distances of less than 1/10" are possible. Small distances of the stator-electrodes increase the transducers efficiency and maximum SPL.

The dynamic range against low signal levels (also called detail or resolution) is the traditional top-level domain of ESLs.

Its membrane is coupled ideally to the surrounding air, its tiny mass doesn´t play any significant role over the entire audio frequency range and the linear drive system works holohedral all but directly on the air.

A ESL extracts already finest details of the music, when the dynamic spaeaker has not even realized that a music signal knocks at its terminals. Neither paper, metal, ceramic or even diamond as speaker diaphragm material would help. Those finest dateils, naturally presented account to a great deal for the authentic and ´real´ sonic impression.

The large scale dynamic range usually was and mostly still is a downside of ESLs. One should note that this applies rather only to fullrange-ESLs and some inferior incarnations of hybrids.

Well designed a ESL can not only keep up with dynamic drivers but comes even close to large horns.

The small esl-panel for example (50"x10", from >200Hz) was measured by Prof. Anselm Goertz at Aachen Universitys anechoic chamber (RWTH Aachen). It reached a maximum SPL of 110dB at 4m distance. This figure translates to 122dB for a global distributing sound radiator, standardly measured at 1m distance. This number doesn´t account for the linear dynamic increase of an ESL, compared to the thermal compression by a heating voice coil of a dynamic speaker. It also doesn´t account for the fact that the dipolar cylindrical wave is a longthrow sound radiator, which keeps up the SPL over a greater distance than the shortthrow global or lobing wave. Compared over distance the distribution of sound is much more even.

The distortion level of the esl at this high maximum SPL remained below 0.3% over its entire working frequency range (not accounted for the B&K-Mic-capsules own THD of 0.15-0.20%).This is at least one order lower than even the best dynamic drivers.

At usual listening volume levels the esl-panel may exhibit lower distortion levels than the attached amplifier and electronics. A good and stable working 50W amplifier would already be sufficient, but a 100W amp could drive the panel beyond full SPL.

The big problem of hybrid-ESLs is the mating between the ESL-panel and dynamic bass system to form a homogenuos entity.

If done consequentially, the effort You need to put in especially for the bass system is high and costly. At least more complex and costly than for the usual closed -or even worse- reflexing boxes.

If neither the acoustical filter flanks nor the distribution character of the two branches merely match, You can´t expect a seamless transition from ESL to bass. This becomes clearly audible, as sluggish, dull sounding bass.

This sonic impression created the mar of a slow heavy paper diaphragm not fitting the fast massless film membrane.

The fairytale is repeated over and over again by magazines and audio-forums, by people whose knowhow about ESLs ends with the term "big and expensive".

The comparing picture of a race between a formula-1 race car against a battle tank is quite seducing, but factually it is still wrong.

Mass and acceleration are not the centre of the problem.

The problem is consequently tackled and the solution to the ESL-Bass integration problem is shown in both purist electrostats.