Matthew Anker's

Electrostatic Loudspeaker Page

    Electrostatic loudspeakers boast many advantages over conventional magnetic loudspeakers. Electrostatic loudspeakers have more accurate response characteristics than conventional magnetic drivers, but yet they are not widely used. In order to take advantage of the electrostatic force, the speakers must use extremely high voltages with low currents. Electrostatic loudspeakers have outstanding clarity that the most serious audiophiles demand.

    Electrostatic speakers operate by interacting with charges of static electricity to create the movement necessary to produce sound waves. Electric charges can be either positive or negative, an object with no electrical charge is said to be neutral. A negative electric charge is produced when electrons are transferred from a negatively charged object into a neutral object. Positive charges are produced when electrons are lost. In an electrostatic speaker, the bias power supply induces a high voltage negative charge into the diaphragm that has almost no current. The bias power supply induces a positive charge on both of the stators that does not let the diaphragm change position. Since opposite charges attract, the diaphragm is held in-between the two stators because both stators are pulling equally on it.  In addition, "Static Stability" from the diaphragm under tension, also returns the diaphragm to a neutral position.
     When music is playing, neither the diaphragm nor the stators change polarity, but the stators do change their phase. The electrical waves produced by music and released by the amplifier are in the form of alternating current, a type of current that is constantly changing polarity. The rear stator is fed a waveform that is the inverse of what the forward stator is receiving. The signal inversion keeps the diaphragm from being pulled toward the forward and rear stator at the same time which would prevent the movement of the diaphragm. When the music plays the forward stator is fed a different phase than the rear stator, causing the diaphragm to be both pulled and pushed forward, this is called a push-pull configuration. As the music demands, the phase and voltage of those charges change at high speeds producing a movement that produces a sound wave that is heard by the ear.

    Electrostatic loudspeakers are made of three main parts, the diaphragm, the stators, and the spacers. Diaphragms are the only moving part of the electrostatic loudspeaker, or ESL for short. diaphragms are made from Mylar®, an extremely thin and light weight plastic film placed under tension. The diaphragm of an ESL must move the air inside a room in order to make sound. Since sound moves in waves, the diaphragm moves forward and backward in between the two stators. Most Mylar® diaphragms have a mass that is equivalent to a layer of air only 1/4 of an inch thick. ESL diaphragms are generally referred to as a massless driver compared to magnetic drivers which have a large mass. The diaphragm must have an electrostatic charge of several kilovolts to generate an electrostatic potential, and this is provided by the polarizing (bias voltage) from the step-up power supply. Because plastic is an excellent insulator, a layer of conductive material must be applied in order to conduct the electrostatic charge. Graphite, one of the softest forms of carbon, is usually applied by most hobbyists because of its high resistance properties and easy to apply powdered form. The graphite coating can conduct the electrostatic charge, but it may also produce electrical arcs between the diaphragm and stators. Graphite is only applied to one side of the diaphragm. Because static is a stationary charge it could power the speakers for hours after the high-voltage power supply is removed, but it cannot because of tiny electron leakage paths.

    The electrostatically charged diaphragm is driven by one or more stators. A stator must be both electrically conductive and acoustically transparent in order to operate. Stators are stationary, and are placed on both sides of the diaphragm. Because the stator must be acoustically transparent it has to have holes or slots in it in order for air to pass through, the most commonly used stator is a sheet of perforated metal. The perforations cannot be too large or else it would not be able to produce an even electrostatic charge. If the perforations were too small air would not be able to pass through and the ESL would be worthless as a speaker.  I have found that a perforation size of around 1/8" is optimal.  Electrostatic force is applied to the stators as a high voltage that varies at an audio rate, applied to the stators by the power amplifier. The only true static charge is on the diaphragm.

 If the diaphragm has a positive charge and the amplifier applies a negative charge to the stator, the diaphragm will be drawn to the forward stator. Since the rear stator has an opposite polarity of the forward stator, the diaphragm is also pushed away as it is being pulled by the forward stator. As the waveform of the music changes, so does the polarity and voltage (electrostatic force) applied to the stators. This provides for the different frequencies and volumes found in music. Most power amplifiers lack the voltage necessary to drive the ESL, therefore, a step-up transformer contained in the step-up power supply boosts the power amplifier’s voltage to several thousand volts.

    Spacers are used to separate the diaphragm from the stators. The spacers must be a very good insulator, and plastic is the unofficial standard.  The spacers cannot be too thick, or the electrostatic force in the stators will not be able to move the diaphragm, and sound will not be produced. If the spacer is too thin the diaphragm could be pulled into the stator causing the diaphragm to arc out until it rebuilds its voltage and it is drawn into the stator again.  However, this can be countered simply by insulating the stators by painting with high voltage insulating paints.  The typical range is 0.03" to 0.07".  One of the larger gap sizes used in a commercial product was 0.125" by Dayton Wright.  This required that the panels be housed in a bag containing sulphur hexaflouride gas that acts as a high voltage insulator preventing arcing. "Dayton Wright did this because in the 1970's he did not trust any of the insulation out there, he only found out later that it had other benefits such as giving the speaker dispersion. He could still have the spacing AND not use the SF6 but he wanted a higher bias voltage, so he had a choice gas, or insulation? He chose gas."(Ben Openshaw)  Spacers are usually made of Lexan®, a form of plexiglass, which can be glued together to form the shape of the speaker.

    In theory, electrostatic loudspeakers should have perfect frequency response because their low mass should be able to drive the air directly. In theory, this would be true, but in practice it is not. An ESL’s frequency response suffers at extremely high frequencies and low frequencies. The high frequency problems are caused by the mass of the diaphragm, thinner diaphragm material would enable you to produce higher frequencies.  High frequency linearity is a function of the impedance matching transformer's leakage inductance that must be kept to a very low value. The transformer's inductance interacts with the speaker's capacitance to form an LC resonant circuit. The resulting resonant peak must be placed well above the audio spectrum in order for the speaker not to sound bright.
                                                                                                                                                                             "Brightness" is also caused by beaming. This phenomenon occurs with both narrow and wide dispersion electrostatic loudspeakers. Beaming is caused by the bundling of higher frequencies that are not allowed to disperse because their wavelengths are smaller than the shortest dimension (width) of the speaker. This perceived high end rise or brightness is the result of a concentration of energy from these higher frequencies.

Low frequency linearity is the second weak point of an ESL's frequency response.  Low frequencies are affected by phase cancellation and the size of the driver.  Phase cancellation is caused by the fact that an ESL is a dipole driver. A dipole driver is a speaker that is not in an enclosure and both sides are open to the air. When the diaphragm is pulled toward the forward diaphragm it forms a vacuum on the other side. This vacuum causes the air pushed by the forward motion to be drawn into the vacuum, effectively canceling out the sound produced. The surface of the ESL must be very large and have an extremely high bias voltage to produce a low frequency with an acceptable volume and to avoid phase cancellation.  It is possible to correct the frequency response with equalizers.

    Electrostatic loudspeakers require much higher voltages than conventional magnetic drivers, so specialized electronics are required to produce the high voltages necessary to drive ESLs from a conventional power amplifier. Most conventional power amplifiers can only produce around ninety volts at their peak voltage, which is several thousand volts short of the voltage required to drive an electrostatic loudspeaker. To increase the voltage enough to drive an ESL requires a step-up transformer and a bias power supply.  The step-up transformers serve two purposes in ESL electronics, to increase the voltage and match the impedance of the speakers to the power amplifier. As the signal from the amplifier moves into the transformer’s primary windings it produces magnetic lines which pass through magnetically conductive material known as the core. The magnetic lines pass through the core and move into the secondary winding. Because there is a greater number of turns in the secondary winding, a higher voltage is produced. The turns ratio of a standard transformer for ESLs is one to fifty. In other words, for every one turn in the primary winding there are fifty turns in the secondary winding. If a common amplifier produces a output voltage of one-hundred volts, the transformer will increase that to five-thousand volts.

    The bias power supply produces voltages of several kilowatts, and the power supply must be designed so that voltage is adjustable depending on the needs of the ESL.  The supply is connected directly to the diaphragm and produces a negative charge, and it is also connected to the secondary winding on the matching transformers.  It is connected to the matching transformer in order to provide a return path for the bias voltage. 
 

    There are many different designs of electrostatic loudspeakers, and each have their own distinct advantages and disadvantages. Curved ESLs are designed to counter directionality in electrostatic loudspeakers. The construction of a curved ESL is similar to a flat panel ESL, but it is in many ways harder to assemble. Because the diaphragm will always be flat under tension, it would contact the rear stator and short the speaker. The diaphragm is only stretched along the vertical axis of the ESL also.  Spacers must run horizontally across the ESL in several places. The spacers must also follow the contour of the curve. It is also harder to make the curved shape because you must bend the Plexiglass® or Lexan® spacers. To produce the curves, the spacers and perforated metal must be glued together on a curved table. The stator/spacer sections can then be glued to the diaphragm and then to each other to form the curved cell ESL.

    Rigid wire stator ESLs use stiff wires running vertically down the front and back of the ESL on a plastic framework similar to the spacers used to support the diaphragm. The rigid wire stators give the ESL more open area, but are harder to build. The other problem is that the perforated metal gives a normal ESL most of its strength. The frame is constructed out of Plexiglass® or Lexan® just like the diaphragm spacers, except there are internal spacers running horizontally across the center of the frame. The wire is usually one-sixteenth of an inch in diameter. There are usually about ten wires per inch. The other advantage of this type of ESL is that designs with very little stray capacitance can be built by only placing wires where the diaphragm can actually move. This makes sure that almost no energy from the power amplifier is wasted on sections of diaphragm that cannot move.

    Hybrid ESLs are designed to overcome the low frequency problems associated with full-range ESLs. They use a standard magnetic woofer to produce the low frequencies, which means that a smaller ESL panel can be built that uses a lower bias voltage and less spacing. The smaller ESL panel will work more efficiently on a smaller amplifier. The woofer enclosure can be built to the hobbyist’s needs or preferences, and the most common enclosures such as sealed-box and tuned-port are usually used. Some more exotic box designs such as a horn enclosure or a transmission line enclosure can be used. The horn design is very efficient, but requires a large amount of space sometimes only found in concert halls or churches. The transmission line design makes the sound waves released from the rear of the woofer travel through the equivalent of a extremely long tube which wipes out many of those waves.

    When the electrostatic loudspeakers are built, and the electronics is connected there are several setup procedures that are different than conventional magnetic loudspeakers. First, the matching transformers should be connected to the bias power supply. Once the power supply is on, the negative charge should be reaching the diaphragm, and the positive charge should be on both stators. If there is a hissing sound, the panel is burning off excess graphite. This should go away after several hours.   If the diaphragm is making a flapping noise it is because of insufficient tension. This can be corrected by re-shrinking the diaphragm with a heat gun. If the ESL is sitting quietly, connect the matching transformers to the four ohm outputs on a tube amplifier, or the 8 ohm outputs on a transistor amplifier.  After it is connected to the amplifier play some music, turn up the volume until the sound is audible. If the sound is very faint, shut down the power supply and increase the bias voltage by connecting the matching transformers to a higher voltage tap on the power supply.