Marx Generator Measurements
This picture shows the set-up for measuring the
output of my Marx generator, using my high
voltage probe and oscilloscope. The
probe was mounted on a camera tripod with an adapter made from a piece
of half inch thick polycarbonate. The probe tip is in contact with
the toroid. The two thick wires hanging from the cieling in the background
are grounds for the base of the generator and the spark terminus.
To collect the following wave forms, the oscilloscope was run in "single
shot" mode with a trigger voltage of 8 volts and the input set-up for a
1:1 probe. Since my probe has a 10,000:1 division ratio, all readings
should be multiplied by 10,000.
This shows one of the nicer wave forms that I
captured during testing. I had initially expected a peak voltage
of ~300 KV (the gaps were set for firing at a charging voltage of ~30 KV).
This particular shot shows a positive peak of 1.36 MV followed by a negative
peak of -1.38 MV! I knew when I chose to use inductors for charging
the capacitors that the output would be the sum of the output from ten
LC oscillators and would not be a simple square wave, but I was very surprised
by the complexity of the output.
This wave form was collected with the Marx generator
gaps opened up for a charging voltage of ~50 KV. The initial peak
is at a potential of 2.68 MV. A later negative peak, not captured
on screen but recorded in the data, had a potential of -4.16 MV.
I tried some shots with a longer time base, but the scope tended to miss
the trigger. The scope's specifications claim that it samples at
a rate 5 GHz in repetitive mode, but only 25 MHz in single shot mode and
maybe less when using longer time bases. Here is one that I did get
a good record of (until the negative peak tripped the input overload protection).
This shot should give you some idea of the variability
of the output from shot to shot. The maximum potential is 3.56 MV,
while the minimum recorded is -4.84 MV (this was chopped off, the
actual value was more negative). I'm not sure if these massive negative
peaks are really part of the Marx generator's output, or if they are indicative
of poor grounding. During firing, I noticed that a spark was jumping
between a valve handle on the water pipe that I was using as a ground (for
both the base of the generator and the spark terminus) and a nearby (~1/4"
separation) gas pipe. Since the scope was grounded to the water pipe,
I decided to try switching the other grounds to the gas pipe, near where
it enters, on the side of my basement opposite to where the water pipe
With the Marx generator grounded to the gas pipe, the negative peaks
were much reduced.
The oscilloscope worked OK during most of this testing. The only
problems I noticed (scary at the time) was that the scope occasionally
turned itself off during a shot. This did not seem to cause any damage.
I will probably try to round up all of the pieces of my Tesla coil and
see if the scope needs more shielding or not.
Back to Oscilloscope