Materials for each speaker cost approximately $6.50, and labor for machining cost $35, for 
a total cost of $41.50 in 1966. 
The Mylar diaphragm is placed with the metalized side in contact with the body and the 
holes are alined. The sintered brass filter serves as the backplate for the speaker; the com- 
mercially available item we used requires only sanding with fine sandpaper to remove the sharp 
edges (the normally textured surface provides the minute and rounded projections necessary 
for operation of the electrostatic speaker); then, after it is cleaned with 100 percent ethyl 
alcohol, it is placed with the smoothed side toward the diaphragm. The Plexiglas backplate is 
placed against the brass backplate so that it exerts a slight pressure, and the Plexiglas cup 
is positioned on the body and bolted in place. The tension on the diaphragm is adjusted with the 
adjustment bolt until wrinkles in the diaphragm are eliminated. Although tension on the diaphragm 
is not critical for sound-level output, the outputcan sometimes be increased by 2 to 3 db by care- 
ful adjustment. 
INSTRUMENTATION 
The speakers were calibrated with the equipment shown in the photograph in figure 3 and in 
the diagram in figure 4. RCA signal generator model WA-44C produced the signal to drive the 
driver amplifier. The output signal was monitored onRCA vacuum tube voltmeter (VTVM) model 
WV98B, Tektronix oscilloscope model 502, and Hewlett Packard electronic counter model 523D. 
The driver amplifier was the type described by McCue’ with a wide frequency range and an 
output impedance capable of driving our electrostatic speaker. It drove the speaker directly and 
was also equipped with a 220-volt d.c. power supply for the necessary polarization voltage. 
Bruel and Kjaer (B&K) 1/4-inch condenser microphone type 4135 with a cathode follower 
was used as the transducer for B&K microphone amplifier type 2604. The output of the micro- 
phone amplifier read directly in decibels, and the waveform, voltage, and frequency from the 
output were monitored on the oscilloscope, the VTVM, and the frequency counter. 
CALIBRATION PROCEDURES 
A calibration chamber was designed and constructed so that stray or reflected signals could 
be attenuated or eliminated (fig. 3), and a microphone mount was installed in the rear of the 
chamber so that only the active portion of the microphone was contained in the chamber. A small 
lamp was installed 1 inch above the center of the microphone and focused on the Mylar diaphragm 
at the front of the speaker; when the light beam was reflected into the source, the speaker was 
assumed to be properly alined with the microphone. The speaker mount, equipped with adjustments 
for vertical and horizontal alinement and built to be rotated 360° in the horizontal plane, held 
the speaker 1 meter from the front plate of the microphone. 
The driver amplifier was calibrated with a constant 2-volt peak-to-peak (p-p) sine wave 
signal from the oscillator that was monitored by a VTVM. The waveforms and frequency were 
monitored at the oscillator on the oscilloscope and electronic counter, respectively. The signals 
were also monitored at the output of the driver amplifier and the microphone ‘amplifier. The 
waveforms were compared for distortion at all three points (fig. 5). Frequency response of the 
driver-amplifier is shown in figure 6, 
The condenser speaker was calibrated by using a constant 100-volt p-p signal from the out- 
put of the driver amplifier, and the frequency response of the speaker was measured at fre- 
quencies from 10 to 100 KHz (fig. 7). The directional characteristics of the output of the speaker 
were determined in the horizontal .plane at angles between 0° to 90° in relation to the plane 
of the microphone; the data (fig. 8) show a typical response curve for this type of speaker and 
TSee footnote 6, 
