I have built myself a pair of electrostatic loudspeakers which differs from the standard flat planel.Sound quality is superb, however, I am using a pair of audio transformers from an old stereo amplifier. As a result the top end may not be as linear or extends high enough in the audio spectrum.
It is semi-circular and instead of the standard spaced diaphragm I use a monofilament line which is spaced half an inch apart on the back plate. The diaphragm is then stretched and laid over the line. A second run of line is run over the diaphragm between the first run of the monofilament. This causes the diaphragm to zig-zag between the perforated plates. The diaphragm's conducting side rests against the front perforated plate. Measurement is roughly 8 inches by 9 inches.
The high resistance part of the membrane, the front, is resting against the front plate at points separated approximately half an inch apart. My ESL is single mode operation, namely the back plate is at ground potential while the front plate is at 930 volts and in contact with the high resistance side of the membrane. I use antistatic spray on the front side of the membrane to develop the required high resistance. The membrane zig-zags against a monofilament line, thus separating the membrane from both plates. I am not sure of the freq. response, but sweeping the ESL with an audio frequency and using a cross-over of 350 Hz. I hear no dip in the response to the limit of my hearing. It obviously extends beyond 20kHz since the sound gives such definition to bells, picolos and other high frequency notes.
The best material I have found for the membrane is a turkey baking bag. I am not sure of its composition (could be some composition of mica), however, I found it to be very thin, does not stretch or shrink and withstands the 930 volts without arcing if adequate insulation in provided at the edges of the two plates. I use two layers of capton tape along the edges of the plates. The membrane is tensioned so that the resonance frequency is above the audio spectrum. /p>
Once the stators are formed it only takes about an hour to complete. The greatest difficulty is getting the membrane tensioned and fitted. I avoid using glue, instead I use double sided carpet tape. The membrane has a tendency to lose its tension so I run a layer of tuck tape over the edges of the membrane and fold the tuck tape over to the other side of the back stator.
If you look carefully at the ESL you will notice the front plate is held down by four small hold-down clamps and insulators. The back stator is larger than the front stator to allow for the clamps to be fastened with self-taping screws. This method of building the ESL was choosen to keep cost at a minimum. I am trying to find a better way to keep the membrane tensioned and to hold the two plates together without glue. Any suggestions?
The power supply is a transformerless symetrical voltage multiplier. This type of multiplier avoids taking one side of the AC line voltage to the electrostatic unit. In effect I end up with a +465 and -465 VDC with the +465 connected to the front stator and the -465 VDC to the back stator. Raising the voltage higher did not increase the output level of the electrostat. Besides I did not want to overstress the coupling capacitors. The 6 times voltage multiplier is for an AC input of 115 Volts.
The audio transformer is coupled to the front and back stators via two 0.01uF 400VDC polypropylene capacitors (would have preferred higher voltage caps but could not find any). The input to the audio transformer is terminated in a 10 ohm resistor. The transformer turns ratio is 1:50.
Assembling it all
The perforated stators are curved to an 8 inch diameter giving approximately 90 - 100 degrees sound dispersion, and are fitted with two mounting brackets angled at approximately 10 degrees. This allows for pointing up or down. The whole unit with the electronics is mounted on a piece of MDF and fitted in a box 17.5 inches high x 10.5 inches wide x 4 inches deep. (I had 2 pieces of precut MDF board which dictated the size).
I always read that electrostats are low in sensitivity, I was amazed at the output level. With a 50 watt amplifier the maximum level I can comfortably listen to is with the volume control set at the 12 o'clock position. Using CDs every nuance of the recording is audible. The staging and presence are very realistic.
Best regards, David Glean