The power supply used in the pictures shown on this page functions by dumping the energy stored in 3 uF of capacitance (which is charged to about 250 volts) through a common automotive ignition coil 120 times a second via a common quadrac. This creates 120 high voltage pulses a second of alternating polarity. Peak voltage (with respect to ground) is very roughly 75,000 volts. The voltage was estimated from the spark length between two pin points. Note that the pulse width is about 10 uSec.
An interesting thing to note is that while the ignition coil was capable of generating 75,000 volts, its insulation was not capable of preventing that much voltage from flashing over the outside of the coil to ground. This made it necessary to repot the coil. To repot the coil, the coil was removed from its steel and phenolic (?) housing and placed in a plastic soda pop bottle full of paint thinner. All of the leads to the coil were replaced with high voltage wire, with silicon-insulated HV wire emplyed for the HV output lead. While it would probably be better to use transformer oil in this application, paint thinner works just fine, and paint thinner is readily available and dirt cheap...
Click on any of the images to get a much higher
resolution version...
These two pictures were created by setting the dime on a
4 x 5 inch piece of print paper, setting the print paper
on 8 sheets of 3 mil polyethylene plastic (for insulation),
and then setting that on a sheet of grounded aluminum foil.
The power supply was turned on for about 1/4 of a second
(I turned the power on and then immediately turned it
off again.) The print was developed for about 5-10 seconds.
(The "correct" development time for the paper and developer
is 60 seconds. The arcs grossly over-exposed
the paper, however, so I had to pull it out of the developer
quickly before the paper turned pure black.)
This picture was created by setting the dime on the
print paper as before, but without a ground plane.
Without the ground plane, the sparks were much less intense.
(Instead of hot, white sparks, there was a lot of
ozone-generating corona.) This made it possible
to leave the power
supply on for about 3/4 of a second without over-exposing
the paper, which was developed
for a full 60 seconds.
This picture was created in exactly the same manner as the picture
above it, except that instead of setting a dime on the
print paper, the end of a 0.019" copper wire was pressed
against the center of the paper.