You are quite correct with your questions. Arkansas has a very high
humidity most of the year and our window for optimally working with
electrostatics is limited.
I remember that the day scheduled for the test was not a good one and I
was worried that we would get poor results. The Physics building
however seemed dryer than outside and we drew about three inch sparks
from a toy one foot world globe which was sitting on a two foot stalk of
One pulley was lathed from wood and the other from plastic. The belt
was made from an old bicycle innertube. The brushes were made from
eight sewing needles arranged in line with their points aimed at the
belt. (I was amazed that the needles could be moved as far as 1/2 inch
from the belt without having much effect on the sparks that were
The whole thing was mounted on a 12in. X 24in. X 3/4 in. board. The
motor/belt/pulley arrangement was on one end of the board and the dome
with its support was on the other. There was a connecting wire
(insulated with thick Teflon) running from the upper brushes over to the
globe. All of the open connections were coated with corona dope.
The thing produced sparks the first time it was tried, but I was unhappy
with all of the manufacturers blue writing on the PVC pipes, so I
disassembled everything and carefully cleaned the pipes until they were
gleaming white. Washed them in denatured alcohol and polished them with
no lint paper towels. For almost a week the machine refused to make any
sparks at all.
The date for the showing was fast approaching and I was going a little
nuts. finally I took the PVC pipes off and replaced them with new ones
with the markings still in place. Glory !! The sparks came back. I
carefully sanded the original tubes and while wearing white
photographers gloves and replaced them in the machine. The sparks were
bigger and longer than ever, and I still had about 24 hours to
experiment with getting maximum bang for the buck before show time.
There were no quantitative evaluations made. The thing made sparks and
whether it would ever do so as well as an inside belt machine was not
the question of the moment.
Instead of being like a tub of water which is being filled one spoonful
at a time, (which has been my assumption for years) this would imply
that the "belt pulley friction arrangement" was putting enough charge on
the belt that it was always higher, by some appreciable amount, than the
charge on the dome that was throwing the sparks. That thought bothers
me almost as much as the experiment of moving the brushes away from the
I was amazed that a bicycle innertube that was black as could be would
not leak charge like crazy. It was what I had on hand and I took a
flyer. I tried many different ways of joining the belt ends including
vulcanizing them, but the best method I found was to scarf the edges to
a very thin edge and carefully glue the two, relatively large surface
areas thus produced, with some common old super glue. It has held up
for three years and survived numerous demonstrations.
Some how I seem to get caught up on those damn needles when ever I get
near the machine. I guess that I must be positively attracted to them.
P.S. I don't have much time for tinkering these days, I am to busy
talking on the web. (computers Bah) but if you have an experiment you
would like for me to run, I will try and see if it will work. NOTE: The
relative humidify in Arkansas today is about 90% so we may have to
postpone it for a while.