LECTURE-6 (B) Generation of HV

High Voltage Engineering
Generation of
High-Voltages
LECTURE- 6 (B)
Prof Dr. Suhail A. Qureshi.

Elect. Engg. Deptt.
1
Tripping of an Impulse Generator
In large impulse generators, the spark gaps generally

consist of spherical electrodes of such size that the
spacing for the maximum voltage does not exceed a
sphere diameter. The gaps are so arranged that each gap
can "see" the gap below so that the consistency in the
breakdown of a gap can be improved as a result of the
ultraviolet light arising from the spark of another gap.
"Prof Dr. Suhail Aftab Qureshi" 2

A simple method for tripping an impulse generator is to

set the lowest spark gap to break down at a
predetermined value. As soon as the charging voltage is
sufficient to cause this gap to break down, the generator
as a whole operates and a voltage impulse is produced.
This method of self tripping depends upon the reliability
with which the first gap breaks down and has the
disadvantage that the instant of discharge is not
accurately predictable. The latter objection also applies to
mechanical closure of the first spark gap to initiate the
discharge of the generator.

A simple method of controlled tripping utilizes a three -

electrode gap as the first stage spark gap. It is connected
in such a way that the charging voltage is applied across
the outer electrodes. The central electrode may be
earthed through a high resistance or may be biased at
some voltage intermediate between those of the outer
electrodes.

The electrode system is so adjusted that it withstands the

charging voltage by a small margin and the tripping of the
generator is effected by applying a short-duration voltage
pulse to the central electrode. With this arrangement, the
timing of the generator tripping is accurately controlled by
the tripping pulse. The voltage unbalance may also be
affected by applying an alternating voltage to the central
electrode, when tripping will occur near the crest of the
alternating voltage wave.

The tripping circuit using a three-electrode gap suggested

by White is presented in Fig.12. The centre sphere is
insulated from earth and is mounted on a lever system
which ensures that as the main sphere is moved the air
gaps remain equal. The voltage between the outer
spheres is equally divided between the two gaps by
connecting a high resistance R1 between the outer
spheres and joining the centre point of R1 to the centre
sphere.

Fig: 12.

The tripping is initiated by applying an impulse of either

polarity of a peak voltage not less than one-fifth that of
the charging voltage. This impulse is applied through a
small capacitor C1 which is in the form of a porcelain pin-
type insulator. The pin of the insulator is connected to the
anode of a thyratron G1 through a resistor R2. During the
charging period of the generator, the anode of G1 is
maintained at a positive potential of 20 kV and conduction
is prevented by applying an appropriate negative bias to
the grid by means of a battery B1. If the potential on the
grid is momentarily neutralized by closing the switch S1,
the thyratron becomes conductive and a negative impulse
of 20 kV peak is generated at its anode.

The oscillograph sweep circuit is also initiated by closing

the switch S1 which applies an impulse of a few hundred
volts to the tripping circuit of the oscilloscope. The
inherent time lag in the operation of the thyratron ensures
that the sweep circuit begins to operate before the start of
the high voltage impulse. A further delay can be
introduced, by adding a simple resistance capacitance
circuit in order to flatten the wave front of the pulse
applied to the thyratron grid. The resistance R2 attenuates
high-frequency voltages which may be produced by the
breakdown of the sphere gaps and which might be
transmitted to the oscilloscope through the common
tripping circuit.

The output voltage of the impulse generator is controlled

manually by operating the switch S1 when the voltage of the
first stage has reached a predetermined value. This voltage
is read on an electrostatic voltmeter V connected across a
tapping on the resistor R1 at its earthy end. Alternatively, an
automatic control of the output voltage can be obtained by
replacing the hand operated switch 81 by a small thyratron.
In the circuit arrangement of Fig. 13 suggested by White,
the grid of the thyratron G2 is connected through a battery B2
to a tapping near the earthed end on the resistor R1.

Fig: 13.

Assuming the charging voltage to be positive, the grid

voltage rises as the charging proceeds and at a critical
voltage the thyratron G2 becomes conducting and a pulse
is applied to the grid circuit of the thyratron G1. The output
voltage of the generator is controlled by the position of
the tapping on R1 and the voltage of the battery B2. This
method is only applicable when the charging voltage is
positive.

A method for automatic control of voltage, suitable for either

polarity of the charging voltage, is shown in Fig.14. The
auxiliary spark gap is charged from the input voltage of the
generator through a resistor R3. The gap is adjusted so that
it breaks down at a voltage slightly below the spark-over
voltage of the generator gaps. When the charging voltage is
positive, the spark-over of the auxiliary gap produces a
positive impulse at the point P.

Fig: 14.

With negative polarity charging voltage, the breakdown of

this gap is followed by a positive polarity impulse
appearing at the point Q. A suitable fraction of this
impulse can be applied to the grid of thyratron G1 and to
the tripping circuit of the oscillograph sweep, as shown in
Fig.14. The output voltage is then determined by the
length of the auxiliary gap.

Another form of a three-electrode trigger gap is the

trigatron of the type shown in Fig.15 where the main
electrodes are spheres of a suitable size. The earthed
electrode is provided with a small hole in which projects a
metal rod with an annular clearance of about 1 mm. The
best position of the rod is with its face flush with the
outside surface of the sphere. Tripping of the generator is
effected by applying a tripping pulse which produces a
spark between the tip of the rod and the earthed sphere.

Fig: 15.

The space charges of the auxiliary spark distort the

electrostatic field in the main gap and cause it to break
down at a voltage much lower than that required in the
absence of the trigger pulse. The trigatron is polarity
sensitive, and if the generator is required to deliver both
positive and negative polarity voltages a polarity change-
over switch is necessary in the trip circuit.

A circuit operating a trigatron incorporated in the first

stage of an impulse generator is shown in Fig.16. The
d.c. supply charges the capacitor C1 through a high
resistance R1 and upon closing the remotely controlled
switch S a pulse is applied to the CRO trip circuit through
the capacitor C2. At the same time the capacitor C1 a is
charged up and a triggering pulse is applied to the trigger
electrode.

Fig: 16.

The tripping potential on the electrode can be adjusted by

adjusting R3 or C3. The R3C3 circuit also provides a way
for delaying the operation of the trigatron and hence that
of the impulse generator spark gaps. The high resistance
R2 provides a discharge circuit for the residual charges on
C3.
The breakdown mechanism of triggered spark gaps have

been studied by several investigators and the possible
theories have been suggested by Broadbent and
Broadbent and Shalash.

The application of triggered spark gaps for controlled

tripping circuits has been described by Husbands and
Higham. A circuit developed by these authors is given in
Fig.17. Where the tripping pulse is obtained from the time
sweep circuit of the oscillograph. The oscillograph used
with the impulse generator is a 50 kV continuously
evacuated type. Its time-sweep circuit incorporates a
three-sphere gap operating at 6 kV.

Fig: 17.

With the initiation of the time sweep, the voltage on one

sphere falls rapidly from 6 kV to zero and this voltage
change provides a suitable pulse for tripping of the
impulse generator. The application of this pulse directly to
the first gap of the main impulse generator could cause
oscillations in the time-sweep circuit because of the large
impulse currents. To avoid this an Intermediate single-
stage pulse generator was introduced.

This pulse generator was charged to 6 kV by the

oscillograph time-sweep supply and was tripped through
the gap G2 by the illumination from the three-sphere gap
of the time-sweep circuit. The gap G2 was positioned
about t in. from the gap G1. It was found that the time lag
inherent in the breakdown of various gaps provided
insufficient delay in the tripping of the impulse generator,
therefore a delay cable was introduced.

With the circuit shown in Fig.17 the delay in the start of

the wave front on the oscillograms was nearly 0.5 μsec.
The circuit has been operated successfully with impulse
generators having charging voltages up to 167 kV of
either polarity. It has the further advantage that the main
gap functions satisfactorily with a tripping voltage pulse of
only 5 kV.

Broadbent and Cooper have used a beam of light for

tripping a single-stage 1-MV impulse generator. Their
method is shown in Fig. 18(a) and the tripping sequence
is illustrated "in Fig. 18(b). A push-button switch triggers
both the oscillograph time base and a xenon flashtube
which gives a light pulse of few microseconds duration.

Fig: 18.

Fig: 18. (b)

The main spark gap of the generator is a trigatron gap

with the high-voltage sphere housing the triggering
equipment which consists of a photocell, amplifier, and 10
kV pulse generator. The ftashtube is placed in such a
position that light illuminates the sphere and falls on to
the cathode of the photocell and so activates the system
which follows the sequence shown in Fig. 18(b).

The mean time lag between the start of the time base and
breakdown of the trigatron annular gap was 1.3 μsec. The
authors have presented curves relating the time lag
between initiation of the time base and breakdown of the
main gap.
A new technique of triggering a spark gap has recently

been described by Broadbent. It is in a way similar to the
trigatron but the triggering spark is produced by breaking
the current in a low-voltage inductive circuit. This method
eliminates the necessity to produce a 10 kV trigger pulse
which is required for the operation of the conventional
trigatron gap.

The principle of operation is shown in Fig.19. The spark

gap consists of two spherical electrodes with a pair of
contacts placed at the sparking surface of one sphere.
The contacts are connected to a low-voltage battery
through an inductance in series. When the contacts are
made to open, the energy stored in the inductance
appears in the form of a spark across the contacts and
the breakdown of the main gap follows.

Fig: 19.

The device was tested with 15-cm diameter spheres

using a 9-V battery in series with a 230 mH inductor. The
operating range of voltage was about the same as that for
the trigatron, for the same polarity and gap length, and it
was hardly affected by the value of either the voltage or
the inductance.
The contacts can be operated by a remote control device

using a photosensitive cell and a relay.

An advantage of this arrangement is that the sphere

containing the triggering device can be situated at any
point in the high-voltage circuit, whether insulated from
the earth or not. In the usual form of the trigatron gap, the
sphere containing the trigger rod must be at earth
potential, otherwise insulation difficulties occur in
providing the high-voltage pulse to the rod. This restricts
the application of the trigatron as the spark switch in
impulse generator circuits where one end of the
generator capacitance is earthed. In the new triggering
device these difficulties are easily overcome, as the
triggering circuit operates at low voltage.

When tests are carried out with impulse voltages, it is

often necessary to record the voltage wave shape on an
oscilloscope operating in conjunction with a potential
divider. It becomes, therefore, necessary to ensure that
the time base of the oscilloscope must be triggered in
synchronism with the impulse generator. This can be
achieved by first tripping the oscilloscope time-base
circuit which in turn causes a pulse to be applied to the
trip circuit of the impulse generator.

A short delay is introduced between the initiation of the

time base and the application of the pulse to the
generator tripping gap. Thus with this arrangement, the
beam of the oscilloscope will have already traversed a
small part of the screen before the impulse wave reaches
the deflection plates. Alternatively, the generator is first
tripped and the oscilloscope time base is initiated by
means of a pick-up aerial from the electrical disturbance
caused by the generator tripping.

(3). Direct Voltages (HVDC)
Direct current voltages are chiefly used for pure scientific
research work. In industry the main application of direct
voltage is in tests on cables with a relatively large
capacitance which take a very large current if tested with
alternating current voltages. It may be pointed out that
although the direct voltage tests are more economical
and convenient, the experimentally obtained stress
distribution may differ from the normal working conditions
where the cable is transmitting power at low-frequency
alternating voltages. With the growing interest in high-
voltage direct current transmission, an increasing number
of high-voltage laboratories is being equipped with
sources producing direct current at high voltages.

a) Voltage Doubler and Cascade Circuits
Direct current supplies for high-voltage testing purposes

are generally obtained by rectification of alternating
current. For voltages of up to about 100 kV with a
relatively low current output the size of the high-voltage
transformer may be considerably reduced by rectifying
high-frequency currents of about 30 to 100 kc/s produced
by an electronic oscillator. This arrangement, however,
introduces extra complications and most d.c. testing sets
are supplied at lower frequency.

In addition to the half-wave and full-wave rectifier circuits,
the two most common methods of obtaining d.c. at high
voltages are:
(a) voltage-doubler circuit, and

(b) cascade-circuit.
These circuits are discussed below.
A simple circuit for half-wave rectification is shown in Fig.

20(a). When the rectifier conducts, the load capacitance
CL is charged to the maximum value, V max' of the
output voltage of the h.t. transformer.

Fig: 20 (a).

In the next half cycle of the alternating voltage, the

voltage across CL remains the same while the voltage of
the h.t. terminal of the transformer rises to a V max' The
rectifier must be, therefore, designed to withstand a
voltage of 2Vmax The rectifier is protected against
excessive currents by the high resistance R inserted in
the h.t. circuit.


Full-wave rectification can be obtained by using two
rectifiers arranged in a manner shown in Fig. 20(b).
During the half-cycle when the valve II conducts, the
auxiliary capacitor C1 is charged to a voltage V max with
polarities as marked in the Figure. During the next half-
cycle, current flows through the valve I and the capacitor
C1 is charged to Vmax with polarities as marked. The
rectifier has to withstand a voltage of 2Vmax.

Fig: 20 (b).

Figure 20 (c) shows a Cockcroft-Walton type voltage

doubler. When the point A is negative the capacitor C1
charges to a voltage V. As A rises to zero the point B
rises with it. Thus B rises to 2V.

Fig: 20. ( c )

Conduction in the outer circuit takes place as soon as the
point B is above the output voltage, i.e. when B is more
positive than D and charge from C1 flows into C2 making
the no load voltage across C2 equal to 2V. Such circuit
can be connected again in series to produce a further
voltage doubling and a practical arrangement presented
by Allibone is illustrated in Fig. 21. The rectifiers R1 to R4
are all in series and the potential distribution across them
is uniform since capacitors C1 to C4 are all equal. The
second h.t. transformer T2 is supplied through an isolating
transformer T.

Fig: 21.
The principle of cascade generator using the Cockcroft
Walton circuit is illustrated in Fig:22.
The portion FTE is a single-phase half-wave rectifier

circuit and when the rectifier I conducts the condenser C1
charges up to a voltage of + V max. The potential of the
point F with respect to earth therefore oscillates between
zero and +2Vmax and the condenser C2 is charged to
+2V max through the rectifier 2. The point D then attains
a steady potential of + 2V maV and the voltage applied to
C3 rectifier 3 therefore varies between +2Vmax and zero.
Thus C3is charged up to +2V max.

The potential of G therefore oscillates between + 2V max
and +4V rnax and the condenser C4 is charged through
the rectifier 4 to a voltage of +4V max. The potential of C
with respect to earth, therefore, attains a value of +4V
max' The whole circuit is a cascade arrangement and the
points D, C, B and A are successively raised to potentials
of 2V max 4V max 6V max and 8V max. The use of
several stages arranged in this manner enables very high
voltages to be obtained. Each capacitor and rectifier has
to withstand only twice the transformer voltage
irrespective of the output voltage.

Fig: 22. Cascade Circuit.

Fig: 23.

The horizontal rectifier arms are permanently connected

to the capacitor junction points on the feed capacitor
column and the other ends are supported by insulated
columns, so that various series-parallel arrangements of
connections can be made by means of links. These
insulated columns are also used to direct the flow of oil
through the horizontal rectifier arms to form a closed
cooling loop of rectifiers, pump and radiator.

The output voltage of the generator is measured by a

movingcoil instrument connected in series with an oil-
cooled resistance unit. The output voltage is taken to the
load through a protective resistor of sufficient length to
withstand the full generator voltage which appears across
it when the test object flashes over. The h.v. output
current of the generator is measured by means of a
moving-coil milliammeter connected in the earthy end of
the generator output.

The 1.2-MV generator gives an output of 900 kV when

supplying a continuous load current of 30 mA. Several
types of power rectifiers have been commonly employed
in high-voltage rectifier circuits. These include hot
cathode vacuum valves, hot cathode gas-filled valves,
mercury pool rectifiers and solid state rectifiers.

The particular choice depends on many properties which

are of importance, such as the ability of the rectifier to
withstand overload current and inverse voltage surges;
the life and reliability of the rectifier, the cost of the
rectifier and the auxiliary equipment, cook has
summarized the important secondary properties of the
rectifiers and the present trend is to use the silicon type
semiconductor rectifier in the high voltage field.

An analysis of half-wave and full wave rectifier circuit has

been made by Mitchell and expressions have been
derived for obtaining the circuit data in terms of the
known constants of the rectifier. The treatment was later
extended to cover voltage-doubler circuits and
subsequently Mitchell presented a detailed analysis of the
basic circuit of cascade generators from which the
performance of such circuits may be obtained.

b) Electrostatic Machines
The electrostatic belt-driven generator developed by Van
de Graff is in common use in several nuclear physics
research laboratories. Fig:24. shows the principle of
operation of such a generator. Charge is sprayed on to an
insulating moving belt by means of corona discharge
points which are at about 10-100 kV from earth potential.
The belt (about 20 in wide) is driven at about 3000-6000
ft/min by means of a motor and the charge is conveyed to
the upper end where it is removed from the belt by
discharging points connected to the inside of an insulated
metal electrode through which the belt passes. The entire
equipment is enclosed in an earthed metal tank.

Fig: 24.

The potential of the h.v. terminal at any instant is V = Q/C
above earth, where Q is the charge stored and C is the
capacitance of the electrode to earth. The potential of the
terminal rises at a rate given by dv/dt = d/dt(Q/C) = l/C,
where I is the net charging current to the terminal. In
practice dv/dt may reach a value of 1 MV/s and it appears
that the final potential of the h.v. electrode would be
infinite in the absence of any mechanism of charge loss.
Equilibrium is established at a potential such that the
charging current equals the discharge current which
includes leakage current, corona losses and load currents
due to voltage divider, X-ray or ion tubes.

The shape of the h.v. electrode should be such that local

discharges are eliminated from its surface and the field
distribution between the electrode and earth is made as
uniform as possible by means of suitably placed grading
rings.
The lower spray unit, shown in Fig.24, consists of a

number of needles connected to a d.c. source so that the
discharge between the points and the belt is maintained.
The collector needle system is placed near the point
where the belt enters the h.v. terminal.

A self-inducing arrangement is commonly used for

spraying on the down-going belt charges of polarity
opposite to that of the h.v. terminal. The rate of charging
of the terminal, for a given speed of the belt, is therefore
doubled. To obtain a selfcharging system, the upper
pulley is connected to the collector needle and is
therefore maintained at a potential higher than that of the
h.v. terminal.

The device includes another system of points (shown as

upper spray points in Fig. 24 which is connected to the
inside of the h.v. terminal and is directed towards the
pulley at the position shown. As the pulley is at a higher
positive potential the negative charges of the corona at
the upper spray points are collected by the belt. This
neutralizes any remaining positive charges on the belt
and leaves any excess negative charges which travel
down with it and are neutralized at the lower spray points.

LECTURE-6 (B) Generation of HV

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

LECTURE-6 (B) Generation of HV

Uploaded by

Copyright:

Available Formats

High Voltage Engineering

Prof Dr. Suhail A. Qureshi.

In large impulse generators, the spark gaps generally

"Prof Dr. Suhail Aftab Qureshi" 2

A simple method for tripping an impulse generator is to

"Prof Dr. Suhail Aftab Qureshi" 3

A simple method of controlled tripping utilizes a three -

"Prof Dr. Suhail Aftab Qureshi" 4

The electrode system is so adjusted that it withstands the

"Prof Dr. Suhail Aftab Qureshi" 5

The tripping circuit using a three-electrode gap suggested

"Prof Dr. Suhail Aftab Qureshi" 6

"Prof Dr. Suhail Aftab Qureshi" 7

The tripping is initiated by applying an impulse of either

"Prof Dr. Suhail Aftab Qureshi" 8

The oscillograph sweep circuit is also initiated by closing

"Prof Dr. Suhail Aftab Qureshi" 9

The output voltage of the impulse generator is controlled

"Prof Dr. Suhail Aftab Qureshi" 10

"Prof Dr. Suhail Aftab Qureshi" 11

Assuming the charging voltage to be positive, the grid

"Prof Dr. Suhail Aftab Qureshi" 12

A method for automatic control of voltage, suitable for either

"Prof Dr. Suhail Aftab Qureshi" 13

"Prof Dr. Suhail Aftab Qureshi" 14

With negative polarity charging voltage, the breakdown of

"Prof Dr. Suhail Aftab Qureshi" 15

Another form of a three-electrode trigger gap is the

"Prof Dr. Suhail Aftab Qureshi" 16

"Prof Dr. Suhail Aftab Qureshi" 17

The space charges of the auxiliary spark distort the

"Prof Dr. Suhail Aftab Qureshi" 18

A circuit operating a trigatron incorporated in the first

"Prof Dr. Suhail Aftab Qureshi" 19

"Prof Dr. Suhail Aftab Qureshi" 20

The tripping potential on the electrode can be adjusted by

The breakdown mechanism of triggered spark gaps have

"Prof Dr. Suhail Aftab Qureshi" 21

The application of triggered spark gaps for controlled

"Prof Dr. Suhail Aftab Qureshi" 22

"Prof Dr. Suhail Aftab Qureshi" 23

With the initiation of the time sweep, the voltage on one

"Prof Dr. Suhail Aftab Qureshi" 24

This pulse generator was charged to 6 kV by the

"Prof Dr. Suhail Aftab Qureshi" 25

With the circuit shown in Fig.17 the delay in the start of

"Prof Dr. Suhail Aftab Qureshi" 26

Broadbent and Cooper have used a beam of light for

"Prof Dr. Suhail Aftab Qureshi" 27

"Prof Dr. Suhail Aftab Qureshi" 28

Fig: 18. (b)

"Prof Dr. Suhail Aftab Qureshi" 29

The main spark gap of the generator is a trigatron gap

"Prof Dr. Suhail Aftab Qureshi" 30

A new technique of triggering a spark gap has recently

"Prof Dr. Suhail Aftab Qureshi" 31

The principle of operation is shown in Fig.19. The spark

"Prof Dr. Suhail Aftab Qureshi" 32

"Prof Dr. Suhail Aftab Qureshi" 33

The device was tested with 15-cm diameter spheres

The contacts can be operated by a remote control device

"Prof Dr. Suhail Aftab Qureshi" 34

An advantage of this arrangement is that the sphere