DAYTON WRIGHT ASSOCIATES LIMITED
The 3 r d Generation
VERY LOW DISTORTION
POSSIBLE ONLY IN A CONSTANT CHARGE
DENSITY ELECTROSTATIC SPEAKER
WIDE FREQUENCY RESPONSE
32 Hz - 24 kHz +/- 4db
HIGH SOUND LEVELS
AS HIGH AS 115 dB 1 m ON AXIS
HIGH POWER HANDLING
700 W TO 1 kW AMPLIFIERS NO PROBLEM
HIGH SLEWING RATE
OVER 100x GREATER THAN OTHER E.S.L.'s
50 W = 103 dB 1 m ON AXIS
LONG EXCURSION CELLS
1/4 in. ELECTRODE SPACING
on aluminum models - or,
on wood trimmed steel frame models
TIG welded frames
UNAFFECTED BY DUST
XG-8 Mk 3 is the only 3rd Generation Full Range ESL!
WHY ELECTROSTATIC SPEAKERS?
By drawing on the field experience gained with the previous units, together with four additional years of development work, Dayton Wright Associates Limited is the first company manufacturing a Full Range Electrostatic Loudspeaker to produce a Third Generation System. All the advantages of the Mark I and Mark 11 Units have been retained, while dramatically reducing the charge up time.
The Mark III cells are all of the constant charge density push-pull type. The cell halves are injection moulded of a high impact plastic, and are heavily ribbed for maximum strength. Bonded to the grid section of the cell halves is a conductive silver film, forming the stationary electrodes of the speaker cell. Once again, the spacing between the fixed electrodes is increased at the periphery in order to smooth the electric field gradients at the edge of the diaphragm. Centrally located between the two parallel bays formed in each cell, are high hystresis loss foam plastic inserts, located there to damp out standing waves in the cell diaphragms.
After testing, each of the ten full range cells used in the XG-8 Mark I I I are mounted into a plastic 'bin', located so as to be tangent to the surface of a sphere. The mounting mastic is located so that operation of the cell places the elastomeric joint so formed, in shear loading, avoiding any possibility of rattles.
The cabinet is a thick aluminum extrusion developed especially for the speaker. The four sides are Heliarc'ed together, then given a satin black Anodized Pebblegrain Finish, producing a hard scuff-resistant coating.
The 'bin', carrying the ten full-range cells is attached to the cabinet by caulking/bonding it into a groove in the cabinet. The complete cabinet is then lined with a dense foam mastic bonded to the aluminum in order to damp out vibrations. A special connector with gold plated pins/socket contacts is used to connect the speaker to the Matching Transformer Bias supply unit.
The result is a strong durable cabinet/speaker unit, which can be used in this form as a beautiful complement to modern furniture, or, with the addition of one of the wooden trim kits, to match a more traditional style.
One only need look at the increasing number of manufacturers who are resorting to electrostatic speakers for the mid/high frequency drivers in higher priced speaker systems to become aware that the electrostatic speaker permits a low distortion type of operation that only can be a goal of other speaker types. Dayton Wright Associates Limited is the only company which offers true full range electrostatic loudspeakers which can be operated at high volume levels, levels which in some instances are in excess of those possible with cone-type units.
Not only is the design of a low-distortion cone type unit difficult, it is even more difficult to maintain the design in production. On the other hand, while design of the electrostatic unit is at least as difficult, its very simplicity in the 'driver cell' area permits very high standards of uniformity in production.
In a cone-type speaker, a major amount of the driving force is used to move the physical mass of the cone back and forth. Only the residue moves the air to produce sound. Thus, anomalies in the mass loading of the cone structure cause corresponding anomalies in the sound; the dog is thus being wagged by what should be the tail. Then too, the driving force is developed in the voice coil while it is dispersed over the entire area of the cone; the cone must be strong (and therefore heavy) to transmit this force from the voice coil to the dispersion points.
In the electrostatic speaker the driving force is uniformly distributed over the diaphragm, as is the radiation load, thus we can say that in effect the driving force is applied directly to the load. The diaphragm is not required to transmit force from one point to another, it is simply there to locate the H.V. charge surface at a point averaging midway between the fixed electrodes. The diaphragm is made to be very thin, having negligable mass. As a consequence most of the force developed on the diaphragm moves the air, and only a very small amount is used to move the mass of the diaphragm.
The result is a system which has an exceptionally linear driving system,
coupled with a very low mass piston (diaphragm) coupled to the air. The
net result is very low distortion.
XG-8 STATE OF THE ART
By coating the fixed electrodes in an electrostatic driver with a medium resistance insulation; protection against
arcing can be achieved. What happens is that at signal voltages below the arcing point, there is negligible voltage
drop across the insulating coating, substantially all of the signal voltage appears across the electrode gap, producing a driving force on the diaghragm. But when the voltage breakdown of the gap is reached, there is sufficient current flow to cause an appreciable voltage drop in the insulation, reducing the voltage across the gap and preventing an arc - but also reducing the driving force on the diaphragm.
In other words the loudspeaker self-limits. Since the voltage at which
breakdown occurs will be dependant in part, on
spacing, manufacturing variations will give rise to a variation of this 'critical' signal voltage level-rounding off the limiting effect.
But the result will be a unit lacking dynamic range - major transients
will not be reproduced accurately.
If however, we operate the electrostatic cell in a gas having a very high voltage breakdown, we can get much higher outputs without having to resort to insulated electrodes. Let us cite a hypothetical example: If we were able to increase the breakdown voltage four times, we could use four times the bias voltage; which would be the equivalent of increasing the-driving power 4 x 4 or sixteen times; that is, it would increase the efficiency sixteen times (12dB); we could also increase the driving signal voltage four times; which would mean an increase in driving power of sixteen (12 dB). Thus we have increased the permissible output level 256 times (or 24 dB) before overload. By utilizing other properties of the gas it would be possible to increase this further.
HIGH SOUND LEVELS POSSIBLE ONLY
USING PATENTED DW DESIGN
Only the XG-8 units, of all the electrostatic units can therefore develop the high slewing rates normally found only in cone or horn speakers.
This is one reason that the XG-8's are capable of producing 115 dB levels 1 meter on axis, while most other 'pure' electrostatics are limited to about 100-103dB. Interestingly enough, the gas used behaves in a much more 'ideal' manner than air, causing lower distortion.
At low frequencies, the effective piston area is therefore 9 square feet, that is, the area of the front diaphragm; rather than the aggregate of the cell diaphragm areas. With a displacement of plus or minus 0.122 ins. a total displacement of over 370 cu. ins. would be possible.
LOW ACOUSTIC GRADIENTS & LOW GLARE
Through experimentation with different size electrostatic (and cone)
units we have found that "Acoustic Glare" is an all too real effect in
small source speakers. While some systems deliberately bounce a large portion
of the sound off a wall, achieving in the process, a large 'indirect' sound
source; there are so many different sound path lengths to the listener's
ear that the sound cannot help but be blurred. Thus, large apparent source
area has been achieved at the expense of a moderately muddied sound. By
keeping the radiating area large at high frequencies and by having the
radiating elements (in this case, the cells) tangent to the surface of a sphere, a virtual point source is produced, but it is well behind the speaker. This places the apparent source of the sound (stereo image effects ignored for the moment) at a realistic distance from the listener; the actual sound gradients in the listening area are lower, and therefore the listening area is deeper. The direct/reflected sound ratio is higher, and more coherent, and a better defined transient response is a result. Details can be heard that normally would have been overlaid with other sounds, and listening fatigue is reduced.
It is important to ensure that the units are used with sources and equipment of only the highest calibre; although the
low distortion and excellent definition will reveal many musical realities, it will also reveal any system deficiencies.
High order distortion produces which were veiled by other speakers will now become apparent, harshness can result, . it inferior or ill-matched equipment is used. However, very high amplifier power can be tolerated. We
have numerous customers who have used as much as 700 Watts per speaker - they find that at high listening levels
the speaker will show up clipping in smaller power amplifiers which they can no longer tolerate. On the other hand, we have many customers who are quite content with 50 Watts/channel.
USE OF THE SPEAKERS IN ARRAYS
Through the use of the interlocking I section splines, the Mark 3 cabinets
can be stacked in multiple arrays for sound reinforcement use, as shown
in the drawing. Work is under way on hardware for these applications as
Remember, because of the hermetically sealed construction, the cells are not damaged by moisture or dust.
RESPONSE AND DISPERSION CHARACTERISTICS
The 'Bin' in the Mk Ill has been redesigned to effectively tilt back the speaker internally. If two speakers were stacked - the lower one inverted - the polar response characteristics would be complimentary. By utilization of a carefully developed in-plant ageing program, the initial efficiency of the XG-8 Mk Ill Speakers is considerably higher than the Mk II or Mk I units.
Basic Unit 1 Meter x 1 Meter x 24.2 cm deep
39.365" x 39.365" x 9.5"
Weight: 56 lbs./Speaker
ST-300 A 112 lbs. Approx.
Finish: Black Anodized Pebblegrain
Frequency Response: 32 Hz to 24 kHz +/- 4 db.
Distortion: Typical T.H.D. @ 1 kc 96 clB under .21%
Speaker alone: 80,000 ohm nominal with ST-300A 4 ohms nominal 2 pair of XG-8 Mk I I I's may run off 1 ST-300 providing adaptor cords are used.
Minumum Power Required: 60 Watts per channel
Maximum Power Recommended: 4500 Watts channel 39.37 inches 9.5 inche-1
XG-8 Mk 3 STYLE SPEAKER
A modified form of the XG-8 Mk 3 Speaker is now being manufactured retaining the same bin, cells and electricals as the Mk 3 Aluminum unit. It uses an internal welded steel cabinet as the sealed enclosure, the exterior being sheathed in wood. The grill cloth is wrapped around the unit, which is capped top and bottom with oiled walnut or rosewood (Specify finish when ordering).
A black detail line follows the circumference of the cabinet breaking the side and top faces and 'lightening' the scale of the unit. Because of the construction, the units are heavier than the Aluminum cabinet type.
Size: Basic Unit 42" high x 39" wide x 9-1/2" deep. Weight: Approx.
80 lbs/speaker. ST-300A approx. 112 lbs. Finish: Black grill cloth - Walnut
or Rosewood caps. Frequency Response: 32 Hz to 24 kHz +/- 4 db Distortion:
Typical T. H. D. @ 1 k Hz, 96 db < .21 % I mpedence: Speaker 80,000
ohm Nominal with ST-300A 4 ohm nominal Minimum Power Required: 60 Watts/channel
Maximum Power Recommended: <500 Watts channel
DAYTON WRIGHT ASSOCIATES
P.O. BOX 419 THORNHILL ONTARIO CANADA Ph (416) 884-3422
NOTE THAT THIS INFORMATION IS NOT CURRENT - IT IS PROVIDED FOR HISTORICAL ACCURACY
(c) 1975 Dayton Wright Associates Limited.
© 1980, 1999 Wright Electroacoustics
to Home Page