DEVELOPMENT OF A HIGH VOLTAGE ARBITRARY WAVEFORM GENERATOR

CAPABLE OF SIMULATING THE NATURAL ELECTRIC FIELDS ARISING FROM

STEPPED DOWNLEADERS

J.R. Gumley, F. D’Alessandro & C.J. Kossmann

ERICO Lightning Technologies, AUSTRALIA.

Abstract: In this paper we describe the main rapidly escalating waveform for a nearby lightning
features of a new prototype, high voltage, arbitrary strike [11] which cannot be reproduced by the
waveform generator. The generator is capable of present generators. Researchers have attempted to
precisely simulating the rapidly escalating electric circumvent this problem by using switching impulses
field due to a lightning downleader, including the and a peak voltage for a given terminal-plate gap
well-known steps or pauses as the leader progresses such that the field risetimes are similar to those
toward its point of attachment near the ground. It observed in nature, namely ~ 1 kV/m/µs [6].
utilises a computer controlled interface that drives a However, this practice disregards the importance of
set of series-stacked flyback transformers. The the build-up to breakdown and the role it plays in air
resulting output is any desired (montonically) terminal performance. This is particularly so in non-
increasing waveform, downloaded from the conventional air terminals that rely on capacitive
computer. coupling to a downleader in order to gain the
necessary energy for their operation. A comparison
of the basic waveforms is shown in Figure 1.
1. Introduction

Research into direct-strike lightning protection 1000

design and, specifically, comparative air terminal 250/2500µs switching impulse
performance, is usually conducted by one of three
Electric field (arbitrary units)

means: (i) “natural” field experiments where air 800

terminals are placed in locations of high lightning
incidence [1, 2], (ii) “artificial” field experiments
where air terminals are exposed to lightning 600
triggered by rockets [3, 4], and (iii) laboratory
experiments using high voltage impulse generators
400
[5, 6, 7, 8].

The last of these is an attractive option because 200 Natural e-field variation
testing of air terminals on demand can provide
results much more quickly than having to rely on the
vagaries of field testing. However, progress in the 0
0 1000 2000 3000
laboratory has been limited for a number of reasons.
Time (µs)
One problem relates to the scaling of results from the
laboratory to the field, although some aspects of this
limitation can be overcome by performing a large Figure 1: Comparison of waveform obtained from
number of systematic experiments. A more difficult Marx-style generator with that observed in nature
problem derives from the fact that the Marx-style from a progressing lightning downleader.
impulse generator [9, 10], which is used in most high
voltage laboratories around the world, produces an
RC-type waveshape. Hence, these generators are Figure 2 shows an active air terminal configuration
unable to simulate the temporal electric field designed to operate like the “Trigatron” in the firing
waveforms evident in natural lightning phenomena. mechanism of a Marx generator. The main difference
The natural field at ground level has two in this configuration is that energy is derived from
components: a “permanent” (or DC) and an the approaching leader and not from a laboratory
“impulse” component. The latter component is a power supply. In the pre-stroke period with electric
fields around 10-20 kV/m, the floating sphere will Conversely, with the natural lightning waveform,
collect random ions. These are passed to ground trigger pulses hold off until dV/dt rises to an
through the impedance which couples the sphere and adequate value. Thereafter, pulse rate actually
earth rod. Under these conditions, and even in the increases with the increasing field strength, as shown
early stages of leader approach, the sphere remains in Figure 3(b). Streamer generation is actually
effectively grounded and presents a spherical surface retarded until the near field has adequate strength to
of low field intensification. This acts to preclude support streamer formation. It is readily seen that the
corona and space charge formation. Marx generator waveform is totally unsuited for
testing this type of terminal.

MARX WAVEFORM

CURRENT PULSES

Figure 2: Basic design of a corona reducing,

capacitively coupled air terminal.

(a)
However when the field is increasing at a rate
approaching 1 kV/m/µs due to an approaching
downleader, the capacitive reactance due to the
coupling decreases and current attempts to increase.
However, the presence of the impedance Z restricts
the flow of displacement currents. This causes the
sphere to rise in potential until a triggering arc is
created across the gap between the sphere and the NATURAL WAVEFORM
earthed rod.

Should triggering occur too early, a failed streamer

will leave space charge above the terminal, and this
will act to reduce the local field strength. Testing
with a Marx generator can give very misleading
results because the highest dV/dt occurs at the
commencement of the pulse and there is no trigger
when peak voltage is being approached since dV/dt CURRENT PULSES
approaches zero.

Figure 3(a) shows how testing with a Marx generator

produces very early triggers when voltage is (b)
impossibly low to form a streamer. The figure shows
how the rate of pulse repetition progressively reduces Figure 3: (a) Early triggering due to a typical Marx-
with time as the dV/dt reduces. At the current peak, style switching impulse waveform; (b) Delaying
dV/dt is zero and no triggering arc can occur. triggering that would actually occur under natural
conditions.
 which is a pulse width modulated (PWM)
representation of the rate of rise of a point on the
In the remainder of this paper, we describe the main desired waveform. Typically, the PWM frequency is
features of a new prototype, high voltage, arbitrary 100 kHz and the data rate is 2 MHz. This enables a
waveform generator. The generator is capable of PWM resolution of 5 %. Hence, it allows delays to be
precisely simulating the electric field due to a inserted between each bit in the data stream in 5 %
lightning downleader, including the well-known steps in order to create a ripple effect. A simple
steps or pauses as the leader progresses toward its example of this “interleaving” principle is illustrated
point of attachment near the ground (Figure 4). in Figure 6.

2000

SLOPE = 0

1500
Electric field (kV/m)

1000
SLOPE = 2x

500

0
0 200 400 600
TIme (µs)
SLOPE = x

Figure 4: Simulated electric field waveform at the SLOPE = 0

ground from a stepped downleader. y 2y

O/P 1
2. Generator Design O/P 2
O/P 3
A block diagram of the mechanical design of our 10- O/P 4
stage, 200 kV prototype system is shown in Figure 5. O/P 5
The key to the whole concept lies in the combined O/P 6
effect of series stacking a number of specially O/P 7
tailored transformers. The generator is capable of O/P 8
accurately simulating any monotonically increasing O/P 9
waveform, such as the electric field due to a stepped O/P 10
lightning downleader.
Figure 6: Simplified example of the interleaving
principle of the generator. The waveform slope is
approximately proportional to the duty cycle.

Each delayed bit of the output data is passed into one

. X ORPPA m8. 1

- 200 kV PEAK
of ten opto-driven cables as shown in Figure 7. Each
of these fibre optic signals is then passed into an
isolated switched-mode power supply (SMPS) with
its own floating DC power supply. The fibre optic
signals are converted back to electrical form inside
the SMPS’s.

The output stage of the generator uses transformers

Figure 5: Mechanical design of the new high voltage configured in a flyback topology to eliminate the
generator. need for output inductors. When the transformer
primary winding switches are activated by the
amplified SMPS outputs, the primary side of each
A personal computer installed with a high speed transformer acts as an inductor due to the blocking
digital I/O card is used to output a 10-bit data word action of the output diode. When the switches are
deactivated, the voltage reverses and the inductive time with respect to absolute field strength and
energy stored in the primary is released through the rate of rise of the electric field
secondary winding. The output diode then conducts • compare streamer generation from active
so that a negative voltage appears on each output. terminals and passive sharp/blunt rods.

Risetimes > 1 kV/µs are achievable with this system. The authors believe that a true streamer-to-leader
The advantage of series stacking the modules transition will not be observed with a generator
comprising the generator is that each module only producing less than 1 MV. Of course, the voltage
needs to be able to output a voltage of Vout/n and, required to observe this transition is not known with
more importantly, output it at a rate of only 1/n of the any certainty. All past testing with Marx generators
required slew rate, where n is the number of has a progressively reducing rate of rise of electric
modules. An additional benefit of increasing the field from t = 0. At the time the generator has
number of modules is that the ripple effect from the reached the general area of the critical breakdown
interleaving is smoothed even further. voltage, the dV/dt is considerably reduced.

Each power unit in the prototype can produce up to On the other hand, this generator produces the
20 kV with a series stack of 10 units reaching an typical waveform observed in nature and hence
output voltage of 200 kV after additional filtering. causes a streamer to be launched into a sustainable
Also, each unit can maintain a constant voltage electric field strength, at a time when the rate of rise
output, thus allowing the waveform to rise from a is rapidly escalating. It could quite easily be observed
predetermined static level. At this stage, no attempt that no change in the time to breakdown is recorded,
has been made to control the current waveform of but systematic tests are required to prove this.
any subsequent air discharge. The prototype has been
designed to simulate only a millisecond or so
immediately prior to the return stroke of a discharge. 4. Conclusions

Other advantages of this system include: (i) the test This paper has described a new design of high
waveform can be changed from concave, to linear, to voltage generator that is capable of producing a
convex in a relatively short period of time (of the waveform that faithfully replicates the type of
order of minutes) so that empirical corrections for lightning electric field waveform observed in nature.
variations in temperature, pressure and humidity are The generator tests presently in progress are part of a
not needed; (ii) computer control means that the broad program of lightning protection research also
waveshapes can be stored and recalled at any time to involving computer modelling and field testing. The
repeat a test. results will be published as soon as they become
available.

3. Preliminary tests We aim to increase the capacity of the generator to

≥ 1 MV. If this is achievable, it will revolutionise
The prototype generator has a design objective of lightning downleader simulation experiments in
producing a variable, programmable wavefront with general and, specifically, comparative air terminal
200 kV peak voltage. Accordingly, the authors have testing in the high voltage laboratory.
no expectation of creating an upward leader in the
laboratory. This may come at a later time using a full
size generator based on our experience with this References
prototype. We do, however, expect to generate
streamers and to break down a gap with these [1] Gumley, J.R.: “Lightning interception
streamers. techniques”, 20th International Conference on
Lightning Protection, Interlaken, Switzerland,
Notwithstanding these limitations, we expect some paper 2.8, 1990.
useful testing to be performed. We currently propose
to: [2] Gumley, J.R.: “Lightning interception and the
• compare sharp and blunt rods with and without upleader”, 22nd International Conference on
prior corona producing fields Lightning Protection, Budapest, Hungary, paper
• test sharp and blunt rods with both concave and R 2-11, 1994.
convex waveshapes
• test capacitively coupled air terminals under
different wavefronts to optimise their triggering
[3] Moore, C.B., Rison, W., Mathis, J. & Paterson, blunt lightning rods”, preprint, 1997.
L.: “Report on a competition between sharp and
[4] Uman, M.A., et al: “1995 Triggered Lightning Conference on Gas Discharges and their
Experiment in Florida”, 10th International Applications, Tokyo, Japan, 1995.
Conference on Atmospheric Electricity, Osaka,
Japan, pp. 644-7, 1996. [9] Marx, E.: Deutsches Reichspatent no. 455933,
1923.
[5] Gary, C., et al: “Laboratory aspects regarding
the upward positive discharge due to negative [10] Kuffel, E. & Zaengl, W.S.: “High Voltage
lightning”, Rev. Roum. Sci. Techn. - Engineering”, Pergamon Press, Oxford, 1984.
Electrotechn. et Energ., Vol. 34, pp. 363-377,
1989. [11] Beasley, W.H., et al: “Electric fields preceding
cloud-to-ground lightning flashes”, J. Geophys.
[6] Berger, G.: “The early streamer emission Res., Vol. 87, pp. 4883-4902, 1982.
lightning rod: laboratory simulation of the
connecting discharge from a lightning rod
conductor”, 15th International Aerospace and Address of Authors
Ground Conference on Lightning and Static
Electricity, Atlantic City, published by US Dept. J.R. (Rick) Gumley (Chief Research Engineer)
of Transportation, pp. 38-1 to 38-9, 1992. Dr. F. D’Alessandro (Research Physicist)
C.J. Kossmann (Electrical Engineer)
[7] Berger, G.: “Formation of the positive leader of
long air sparks for various types of rod ERICO Lightning Technologies
conductor”, 22nd International Conference on Technopark, Dowsings Point, TAS. 7010
Lightning Protection, Budapest, Hungary, paper GPO Box 536, Hobart, Tasmania. 7001
R 2-01, 1994. AUSTRALIA.
Tel: +61 3 62 373 200
[8] Berger, G.: “Inception electric field of the Fax: +61 3 62 730 399
lightning upward leader initiated from a Email: rgumley@erico.com
Franklin rod in laboratory”, 11th International fdalessandro@erico.com
ckossmann@erico.com

SW ITCH
DELAY
D RIVER

PC

HIGH SPEED
DIGITAL
I/O CARD

MULTIPLE
OUTPUT -200 kV
STAGES OUTPUT

SW ITCH
DELAY
D RIVER

OPTO FIBRE FIBRE

DRIVER OPTIC OPTIC
OUTPUT RECEIVER

INTERLEAVED
DELAY
D.C.
SOURCE

Figure 7: Diagram of the prototype 200 kV high voltage arbitrary waveform generator.