A Marx generator is a type of electrical circuit first described by Erwin Marx in 1924

whose purpose is to generate a high-voltage pulse. It is extensively used for simulating

the effects of lightning during high voltage and aviation equipment testing. A bank of 36
Marx generators is used by Sandia National Laboratories to generate X-rays in their Z
Machine. It can also be used as an ignition switch for thermonuclear devices.

Marx generator diagrams; Although the left capacitor has the greatest charge rate, the
generator is typically allowed to charge for a long period of time, and all capacitors
eventually reach the same charge voltage.

To deliver 5 ns rise time pulses the Marx generator is often built into a coaxial wave
guide. The spark gaps are placed as close as possible together for maximum UV light
exchange for minimum jitter. DC HV coming from bottom, pulsed HV leaving at top into
the coaxial line. The double line of spheres in the middle are the spark gaps, all other
spheres are to avoid corona discharge. Blue=water capacitor. Grey=solid metal. Black=
thin wire. The outer conductor also functions as a vessel, so that the gas and the pressure
can be optimized.

1
A number of capacitors are charged in parallel to a given voltage, V, and then connected
in series by spark gap switches, ideally producing a voltage of V multiplied by the
number, n, of capacitors (or stages). Due to various practical constraints, the output
voltage is usually somewhat less than n*V. Proper performance depends upon selection
of capacitor and the timing of the discharge. Switching times can be improved by doping
of the electrodes with radioactive isotopes caesium 137 or nickel 63, and by orienting the
spark gaps so that ultraviolet light from a firing spark gap switch illuminates the
remaining open spark gaps. Insulation of the high voltages produced is often
accomplished by immersing the Marx generator in transformer oil or a high pressure
electronegative gas such as sulfur hexafluoride (SF6).

Note that the closer a capacitor is to the charging power supply, the faster it will charge.
If the generator is allowed to charge long enough, all capacitors will attain the same
voltage.

In the ideal case, the closing of the switch closest to the charging power supply applies a
voltage 2*V to the second switch. This switch will then close, applying a voltage 3*V to
the third switch. This switch will then close, resulting in a cascade down the generator
(referred to as erection) that produces n*V at the generator output (again, only in the
ideal case).

The first switch may be allowed to spontaneously break down (sometimes called a self
break) during charging if the absolute timing of the output pulse is unimportant.
However, it is usually intentionally triggered by mechanical means (reducing the gap
distance), triggered electrically, triggered via a pulsed laser, or by reducing the air
pressure within the gap after all the capacitors in Marx bank have reached full charge.

The charging resistors, Rc, need to be properly sized for both charging and discharging.
These resistors are sometimes replaced with inductors for improved efficiency and faster
charging. In many generators the resistors are made from plastic or glass tubing filled
with dilute copper sulphate solution. These liquid resistors overcome many of the
problems experienced by the more normal solid resistive materials which have a tendency
to lower their resistance over time under high voltage conditions.

Contents
[hide]

• 1 Short pulses
• 2 Stage variants
• 3 Applications
• 4 Further reading
• 5 Patents

• 6 External articles

2
 [edit] Short pulses
The Marx generator is also used to generate short high power pulses for Pockels cells,
driving a TEA laser, ignition of the conventional explosive of a nuclear weapon, and
radar pulses. Short is relative, as even high speed versions stay below 1 GHz frequency
and thus almost all low power electrical components are faster. In high speed devices
electrodynamics is important and the Marx generator regards it insofar as it uses short
thick leads between its components, but the main design is purely an electrostatic one.
Using electrodynamic language one says if the first stage breaks down it creates a
spherical wave, with an electric field vector opposed to the static high voltage and a
circular magnetic field. This field has the wrong orientation to trigger the next stage and
the wave points in the wrong direction, it may even reach the load and in many switching
application noise in front of the edge is not wanted. If the generator is inside a tube - lets
say with 1 m diameter - one would guess after 10 reflections of radio waves the field is
settled, and the electrostatic equations are fulfilled. This would limit useful operation
above 30 ns pulse leading edge width. Smaller devices are faster. When the first gap
breaks down, electrostatic tells us, that the voltage across all stages rises. More precisely,
all stages couple capacitively to ground and tend to screen such a voltage rise, and thus
the very next stage feels the largest voltage rise. Also electrodynamic prefers the very
next spark gap and the stages can be sorted by their sensitivity. From stage to stage the
voltage rise rises, which speeds up their operation. Thus a voltage rise, which is fed into
the first stage, will get amplified and steepened at the same time.

Looking into details, the speed of a switch is determined by the speed of the charge
carriers, which gets higher with higher voltage and the ration between current and
parasitic capacities. And a high voltage leads to high current in avalanche devices.
Because the high voltage is applied only for a short time, the solid state parts of the
device will not heat up until they evaporate. And as a compensation the later stages have
to carry a lower overall charge. Stage cool down and capacitor recharging goes well
together.

[edit] Stage variants

Avalanche diodes can replace the spark gap for voltages around 500 volt. The charge
carriers easily leave the electrodes, so no extra ionization is needed and jitter is low, also
the diodes live longer.

A speedy switching device is a NPN avalanche transistor outfitted with coil between
emitter and base, so that the transistor is closed, at least at the beginning. Emitter and
collector are externally connected. About 300 Volts fall of across the base collector
junction. The voltage is so high that a charge carrier in this region can create more
carriers by impact ionization, but the probability for this is just to low to form a big
avalanche, rather a noisy leakage current flows. If a preceding stage breaks, electrostatics
tells us that in the next stage the emitter base junction is pushed into forward bias and the
base collector junction enters avalanche mode. That means charge carriers are injected

3
into the base collector region and then are multiplied. After the Marx generator has fired,
voltage is generally lower, avalanches stop, the matched coil pushes the emitter base
junction into reverse bias and the low static field sucks remaining charge carriers out of
the base collector junction.

At 5 V for signal processing tunnel diodes may be useful. As stated under 'short pulses',
for low voltages better look into step recovery diode, sampling bridge, transmission line,
mode locking.

[edit] Applications
An application is boxcar (figurative name for an electric signal) switching of a Pockels
cell. 4 Marx generators are used. Each electrode of the Pockels cell is connected to a
positive pulse generator and a negative pulse generator. First two generators are fired to
charge the Pockels cell into one polarity, this will also partly charge the other generators,
but not trigger them, because they have been charged only partly before. The leakage
through the resistors needs to be compensated by small bias current through the
generator. At the trailing edge of the boxcar, the other generators are fired.

4
5