With a more stable supply configuration I could start to experiment more with the speaker itself.
I started to try both the plasma tweeter type and corona wind arrangement. The supply and amplifier
circuitry was flexible enough to drive either type - plasma just needing higher currents and the corona a clean
This 5 pin brass screw arrangement was quite successful.
In this fuzzy image you can see the basic arrangement, 5 screws each with a ballast resistor on a board behind
pointing at a single ground side electrode. This is switched on in daylight and you can clearly see the
purple plasma glow in the centre even from the Polaroid. These are mounted on the top of 3 hot water
pipe insulation tubes - they were very useful to get the high voltage up and out of the way.
This is not far off the Hill Plasmatronic layout they used for their
plasma tweeter. I can't remember if I found out about their speaker and tried this or if this just came
about from experimenting. I was researching all the time whilst experimenting so difficult to tell
sometimes. To the left is an alternative star arrangement. The sound from both of these is pretty
much 360 degree and they are still fairly quiet - although if driven above 3kHz they could get a lot louder,
proving why the principle was more popular for use as a tweeter.
It's around this time I started to learn more about the general limitations and problems of plasmas. They
run hot, eat electrodes, generate toxic gases and are difficult to run stable.
I had limited myself in a way as I was using a DC supply and direct drive to a single ended class A
amplifier. The audiophile purist in me wouldn't allow me to use class A-B amplifiers, RF or transformers - I
liked the clean simplicity of the direct drive design and from my research could find no other designs trying to do
it that way. As this appealed I continued to try and find other ways around all the problems.
With the same equipment my trials started with corona wind. I
quickly found that many more pins were needed and that actually striking an arc was bad. The main idea
is to have the pins close enough to create a differential but not too close so the air doesn't break
down. The earliest successful attempt at this was the 25-pin design here. There is a piece of
breadboard at the back with a 5x5 grid of pins poking through. Each pin points to one of the ones
attached to the front wire grid. Each of the high voltage pins had a ballast resistor to balance the
current and after a bit of fine tuning they would all contribute to the sound.
This really started me off down the path of corona wind. This speaker had a much wider frequency response,
although much quieter in general and with more background hissing noise and crackle. This was the first
design I could actually sit back and listen to music on. One other benefit was that there was no risk of the
arc extinguishing because one was never struck (if aligned correctly and not over driven).
It appeared that the output limit was simply because of the
number of points and this led me in the direction of building ever more elaborate alternatives.
The one to the left was an attempt to get around the need for one resistor per pin. The black material is
antistatic foam - a conductive foam with roughly 100 ohm resistance. This proves to be too low - I was
usually using upper meg ohm values to balance them otherwise. This was a pretty design and a complete failure
to produce any sounds.
To the right is a very elaborate multi-point to plane corona
wind loudspeaker. I got up to around 50 pins with this design and although it worked it just couldn't
push much air. The end result was a bit like having one weak plasma.
This was all useful research - it's not like there was a book telling me the best way of doing these things and
I find the best thing to do is simply try an idea - plasma speakers are simple and many of the materials required
are cheap and easily available.