Panel Structure

 This section deals with a closer look at the panel structures proposed for the Quad Electrostatic Loudspeaker. In practice, only one of these was put into final production, and, as it happens it was the mechanically simplest of all the possibilities canvassed in the original patent.

Let's start with an overall view of the generic panel in cross-section, not to scale, and with various dimensions (like the diaphragm thickness) exaggerated, so that we can "see" the structure clearly in concept. The view shown below is from the U.S. Patent 3,008,014 of November 7, 1961.

I refer to this as the 'generic' panel design, since it embodies in one schematic, most of the essential ideas that Williamson and Walker wanted to incorporate in a possible production speaker. There were alterations to this, of course, some of which improved actual performance, and some of which were not only improvements in performance but improvements in so far as production techniques and material costs were concerned.

While we're on the subject, the principle behind the conductive strips 25 is to feed the charging voltage to the periphery of the "wells" so that the edges of the diaphragm are supplied with a charge source at every point, rather than at a single point. A lot of modern constructors of electrostatic speakers use a single contact, and this works, provided that they also use a relatively low resistance diaphragm and a larger diaphragm-to-stator spacing than in the Quad ESL. The designers of the Quad ESL knew that they would have about 15 Watts of amplification available to them (not much by modern standards), and so their stator-to-diaphragm gaps had to be as small as practically possible to gain higher efficiency. These smaller gaps also meant more uniform electric field. They also used a very high resistance diaphragm to ensure uniformity of charge across all elemental areas of the diaphragm (following F.V. Hunt's maths). To achieve uniform charging they, therefore, had to introduce the "conductive strips" 25 to ensure uniform and rapid charging of such a high resistance. In my opinion, this is one of the "secrets" that sets the Quad ESL apart from modern designs in its purity of sound - virtually distortionless.

Examining the diagram a little more closely we see that the conductive strips are located inwards from the extreme edge of the stator, beyond the insulating strips 26. This was initially proposed to prevent any leakage between the diaphragm and the plates which would cause a voltage drop across a free part of the diaphragm. Such a drop would prevent the full polarising voltage from appearing on the diaphragm and would render the performance unpredictable (differential charging = differential driving forces = Bad sound). This is not how the panel was built, in the final production models, but we have a couple of steps to go, before we get to that.

As a slight detour, shown below is a curiosity which never found its way into production. This is a section of a panel that theoretically would have been produced with thickened areas, lines or beads which would only vibrate at low frequencies, and the less "massive" areas at high frequencies. Perhaps it was considered as a possible bass panel design?

 For Figure 4, above, please note that all structures are denoted by the same codes as shown for Figure 3 , except that Code 21' now denotes the thickened portions of the proposed diaphragm. This particular design was considered so that:

"Where the diaphragm area is small compared with the wavelength radiated, the frequency response may be equalized by loading parts of the diaphragm either by thickening or by adding button-shaped portions of other material...
At the higher frequencies these loaded parts remain stationary and vibrational displacement occurs mainly in the intermediate areas of low mass, from which areas effective treble radiation takes place. At the lower frequencies the diaphragm vibrates as a whole, the increased mass lowering the bass resonance frequency."

Below, in Figure 5, we see a section of what essentially went into production in both the treble and bass panels with a few minor wrinkles added.

Doesn't look very different, but there are some subtle changes that improve both the performance and the production economies of the device.In the construction shown as per Figure 3 there is waste (stray) capacitance in areas not contributing to the sound output. The plates 22 and 23 plus the conductive strips 25 and the insulation 26 acting as the dielectric. The modified construction of Figure 5 shows selectively metallized areas 34 painted onto the inner surface of rigid sheets of insulation (the stators). These metallized areas stop short of the diaphragm edges so that no portion of the plates is engaged by the conductive strips. Hence, this capacitance is eliminated, and the speaker is more efficient. The distance 37 between the edge of each metallized area 34 and the inner edge of the strip 25 on the same side of the diaphragm needs to be large enough to prevent ionizing discharge, but also small enough to allow a fringing field to be established to the limit of the "free" diaphragm.

This design is practically the exact conformation of the Treble and Bass Panels as we know them, with the modification in the case of the Bass Panels that the selectively coated (metallized) areas of stator are on the outside of the panel.

In what might be considered one last "wrinkle" in the design, consider Figure 6, shown below:

Again, some subtle differences that seem to make a difference in the production of the device, and its costs over a large volume of units produced. Notice the references to the use of “strips of low grade insulation”, which in the Treble Panel takes the form of a combination of masking paper and gray paint. The “masking tape” is also there to damp the high frequencies in the diaphragm. The diaphragm and a small gap to the stator edge, as per Figure 5, is also used to lower the incidence of use of high grade insulation. Also considered was the possibility of coating the area inside the conductive strips with a conductive material and leaving the area outside as bare diaphragm. In the end, from observation of the actual panel materials, it is clear the Quad opted for coating the whole diaphragm. This would undoubtedly be quicker in production, since there would be no need to mark out, or mask the area to be coated. Any small leakages would be drained away through the HT supply. I suppose this is one minor improvement that one could make to the diaphragms if one is re-building the panel, since you probably don't have the time and production constraints that Quad had. The contacts to those conductive strips that are referred to in the patent as “metallized” strips, in practice became a couple of rivets making contact with a carbon "frame" in both types of panel. In the Treble Panel, as noted previously, these "low grade insulations" (like gray paint) actually disguise the fact that the carbon tracks are there at all.