The electrostatic, because its diaphragm is so thin and light, offers exceptionally good transient response and reproduction of subtle, low-level musical detail. And, because it is a true push-pull device (i.e., its diaphragm is, by design, driven from both the front and the rear), the ESL operates in a linear fashion. Typically, gross distortion results only when the driving amplifier clips into the speaker, or when, in an attempt to play the speaker louder than its design allows, its step-up transformer reaches a point of saturation.
On the negative side of the ledger, the ESL does require passing the amplified musical signal through a transformer, which can introduce its own colorations and non-linearities. Also, some ESLs are prone to a condition known as arcing: Under the conditions of stress induced by playing an ESL loudly, it is not uncommon for an electrical spark to jump between one stator and the diaphragm (a phenomenon exactly analogous to lightning), burning a minute hole in the diaphragm and, over time, ruining it.
As for the planar magnetic, its strengths are similar to those of the ESL--although the addition of several feet of wire and an adhesive coat make for a somewhat more massive diaphragm, limiting this designs transient capabilities by comparison. But the planar magnetic requires no step-up transformer or bias voltage supply, and it has the added benefit of being an extremely manageable load for most amplifiers. However, the most specific drawback of the traditional planar magnetic is that it is a single-ended (as opposed to push-pull) device: As the diaphragmÕs physical excursion increases, the voice grid moves further away from its optimal location within the permanent magnetic field (at least in one direction). Thus, at the very instant when this speaker is called upon to reproduce large-amplitude waveforms, it is least able to do so without distortion.
In many ways, a ribbon driver can be an excellent performer: the moving element (the ribbon itself) is extremely light, allowing good speed and transient performance as well as freedom from coloration. And there is no significant physical structure on either side of the ribbon's radiating pattern. The ribbon's main problem is not one of performance but of application: it cannot be used to reproduce low frequencies. To create a moving element large enough to generate frequencies lower than a few hundred Hz would mean moving opposing magnetic poles so far apart that they would no longer exert a sufficient magnetic field over the entire area of the ribbon.
Also, when a ribbon is operated at frequencies approaching the element's own resonant frequency (which is naturally quite low, due to its high compliance), the ribbon element stretches and bows to a point where it is no longer within the magnetic gap. To get around either of these problems means to move the permanent magnet structure from the edges of the element to one entire side of the element, and/or to bond the element to a host; diaphragm, such as a sheet of mylar, and to clamp that diaphragm around its perimeter. In either case the driver is no longer a ribbon; it is, in fact, a planar magnetic. To date, no one has succeeded in creating a full range ribbon loudspeaker.
The Linear Field Transducer
Eminent Technology's Linear Field Transducer, introduced as the LFT, represents a new approach to the design and construction of a high-quality loudspeaker*. It builds on the strengths of the above designs while eliminating many of their drawbacks.
* The design and construction of the LFT is patented.
The construction of the LFT -8 begins by laminating a very thin sheet of aluminum foil to a half-mil-thick sheet of Mylar. A voice grid pattern, created by means of CAD (Computer-Aided Design) technology, is silkscreened onto the foil side; the remainder of the aluminum--the part not covered by ink from the silkscreening is chemically etched away, in a manner similar to the etching of traces on a printed-circuit board. The ink is then washed away, leaving a voice grid of near-perfect uniformity. This technique results in a diaphragm/voice coil grid that is still less than one mil in total thickness, and also permits relatively narrow spaces between the individual traces, so the diaphragm can be evenly driven over its entire surface.
The Magnet/Frame Structure
The magnet/frame structure developed for the LFT-VIII is also unique. Eminent Technology builds its strip magnets into individual steel channels, the size and shape of which have been carefully designed to help focus the magnetic flux lines and concentrate the strength of the magnetic field on the appropriate area of the diaphragm/voice grid. These channels are then welded to steel crossbars , which in turn are bolted to the frame that holds the diaphragm in place.
Interestingly, one of the biggest challenges faced in creating a true push-pull dynamic speaker was not a design consideration but rather a matter of construction difficulty: to assemble a perfectly rigid structure with very powerful permanent magnets at the front and the rear, both sides opposing each other with tremendous force. It was not until Eminent Technology developed a special method for this assembly procedure that the Linear Field Transducer became a reality.
By applying such new techniques to planar loudspeaker construction, Eminent Technology has been able to eliminate many of the flaws inherent in earlier designs. The use of a welded channel-and-crossbar frame dispenses with the need for perforated sheet metal (an off-the-shelf material presumably used for reasons of economy and ease of manufacture.) thus greatly improving dispersion, especially at high frequencies.
Since it is now possible to have a powerful, precisely aligned magnet structure on both sides of the diaphragm, true push-pull operation has been achieved: Regardless of the degree of excursion the diaphragm undergoes, the voice element is always optimally positioned within the magnetic field. The result is extremely linear performance throughout the audible range, with a profound increase in dynamic range and an absolute minimum of distortion.