Generation of Impulse Voltage

Overvoltages in electrical power system can be classified into two categories, viz.
a) Internal Overvoltage and b) External Overvoltage.
Internal overvoltages are generated due to several events such as switching
operations, load disturbances etc. But, external overvoltages are generated primarily due
to natural phenomenon such as lightning. Lightning overvoltages are again generated in
two different ways: i) Direct Stroke and ii) Indirect Stroke.

a) Direct Lightning Stroke:

The transmission line towers also carry a ground wire in addition to the live
conductors mainly for protection against lightning strikes, so that lightning cannot strike
a live wire directly. But, a lightning can strike the ground wire directly and then the very
high lightning current flows to ground through the tower. The tower has a finite
grounding resistance, even though it is low. Say, the grounding resistance is of the order
of 10 and the lightning is about 100kA. Then, the potential of tower becomes 1000kV
w.r.t. earth. This causes flashover of insulators from tower to conductor and is called
“Back-flashover”. But, it may be noted that direct strokes occur infrequently.

b) Indirect Lightning Stroke:

 When a charged cloud comes close to an overhead transmission line, charges of
polarity opposite to that of the cloud were induced on the section of the transmission
line near the cloud. These induced charges in effect charge the ground capacitances of
the line. When charge balancing between the cloud and the line is complete, the
charged ground capacitances discharge to other ground capacitances of the line lying
further away. This generates a travelling wave on the line moving in either direction.

Both direct and indirect lightning strokes generate transient voltages in overhead
transmission lines, the magnitude of which are many times greater than the normal
operating voltage.

Lightning Impulse Voltage:

A lightning impulse voltage has the following characteristics: i) Non-repetitive, ii)

Aperiodic and iii) it rises sharply to its peak value and then falls gradually. It is
characterized by – i) Polarity, ii) Peak value and iii) Waveshape represented by wave
front and wave tail times.
As per IS-2071, the waveshape of a standard lightning impulse voltage is – i)
wave front time = 1.2s ( 30%) and ii) wave tail time = 50s ( 20%).

Switching Impulse:

tf = 250s  20%
tt = 2500s  60%

Practically, it is very difficult to identify the actual origin O, because of imperfect

synchronization of measuring circuit and also due to ringing at a high frequency near the
origin. So, the waveshape is measured from the virtual origin O, which is the point on
the time axis, where the straight line joining the points of 30% and 90% of the peak
voltage cuts the time axis. Then

Wave-front time (tf) = OG

and Wave-tail time (tt) = OH, where H corresponds to the point on the voltage wave
when the voltage has fallen to 50% of its peak value.
Determination of Wave-front time:

In the Fig. 40, OFG and BDE are similar. So,

FG O G

DE BE
FG
or, O G   BE
DE
 Wave front time t f  
100
 t1  1.67t1
60
where, t1 = time taken to rise to 90% of peak value from 30%.

Principle of Impulse Voltage Generation:

Let, the condenser C1 is charged upto a voltage V. Now, if the switch is closed
then the condenser C2 of Fig.41 will be charged upto a steady voltage V1 (V1 < V)
depending upon the magnitudes of C1 and C2. If C1 = C2, then V1 = (V/2).

If two resistances are added in the circuit as shown in Fig.42, then time delays are
incorporated in charging and discharging of C2. As a result an impulse voltage wave is
obtained as output. The polarity of the output impulse can be simply reversed by
reversing the charging polarity of C1.

Fast-acting Switch:

For automatic and fast switching action, mechanical switches are not suitable and
hence, SCRs may be used as an alternative. But, the PIV of a single SCR is of the order
of 3kV and impulse voltages are about million volts in magnitude. So, large numbers of
SCRs are needed to be connected in series. In that case it becomes very difficult to obtain
exact synchronization of all the SCRs, which distorts the impulse waveform. Hence,
sphere gap is used as a switch, the breakdown time of which is about a microsecond only.
Basic Impulse Generator Circuit:

Cs = Source / Stage Capacitor Rd = Damping resistor

Cb = Load Capacitor Re = Discharge resistor
Rch = Charging resistor

Normally, Rd  0.1Re and Cb  0.1Cs.

i) Charge transfer from Cs to Cb

As Cb gets charged, the front of the impulse wave is generated. After the capacitor
Cb attains its peak voltage VP, both the capacitors starts discharging.

ii) Discharging of Capacitors Cs and Cb

Transform Circuit:

Vo Z2
so, V s   
s Z1  Z 2
1
Re 
1 sC b
where, Z 1  Rd  and Z 2 
sC s 1
Re 
sC b
By substitution one gets

Vo 1
V s    2
k s  as  b

1 1 1
where, a   
Rd C s Rd Cb Re Cb
1
b
Rd Re C s Cb
and k  Rd C b

In time domain it yields

V
V t   o
1
k k 2  k1 

e  k 1t  e  k 2 t 
2
a a
where, k1 , k 2     b
2 2

The output impulse voltage is therefore superposition of two exponential

functions of opposite polarity.
 Vo 1
Vo 1 e  k1t
 k k 2  k1 
k k 2  k1 

Vo 1
[ e  k 2 t ]
k k 2  k1 
Vo 1

k k 2  k1 

Wave shape 1/k1 1/k2

1.2/50s 68.2 0.405

250/2500s 2877 104

Voltage Efficiency:

VP
It is defined as v 
Vo
where, VP is the peak value of the output impulse wave.
In many respects the two basic circuits of single-stage impulse generator are
equivalent. But, they differ significantly in terms of voltage efficiency.

Maximum Stored Energy:

A very important parameter of impulse generator is the maximum stored energy,

which is given by

1
C s Vo max 
2
W 
2

This energy primarily determines the size and the cost of a generator.

Significance of Rd:

In practice some inductance, however, small it may be, is always present in all the
branches of the impulse generator circuit. Combining all the inductances into a lumped
inductor, the charging circuit of Cb can be represented as
For this circuit to be non-oscillatory.

L CC
Rd  2 , where, C  s b
C C s  Cb
The discharge circuit of Cs and Cb is always non-oscillatory due to the presence of Re,
which increases the resistance beyond the critical resistance.

Multi-Stage Impulse Generator Circuit:

Following problems are generally encountered, if peak voltages higher than

200kV are to be generated using single-stage impulse generator circuit.

i) increase of the physical size of the circuit elements

ii) difficulty in obtaining high dc voltages for charging Cs
iii) difficulty in using spark gaps for switching at very high voltages and
iv) difficulties in suppressing corona discharges from the structure and leads.

In order to over come these difficulties, in 1923 Marx proposed a circuit where a
number of source capacitors are charged in parallel and then discharged in series through
spark gaps.

Marx Generator Circuit:

A multi-stage Marx generator circuit can be divided into three sections.

i) D.C. generating section
ii) Voltage multiplying section.
iii) Wave shaping section.
i) D.C. Generating Section:

Conventional Cockcroft-Walton Voltage doubler circuit is used in this section that

utilises power frequency input at 50Hz. The output DC voltage is normally kept
between 100kV – 200kV.

ii) Voltage Multiplier Circuit:

During charging all the source capacitances Cs1, Cs2, Cs3 and Cs4 are charged to a
voltage V over the resistances R, Rch and Re1. The resistances are incorporated in
the circuit to limit the charging current drawn by the capacitors. As the sphere gap
S4 is not in conducting mode, the load capacitor Cb is not charged. The gap
distances of S1, S2, S3 and S4 are kept in ascending order such that S1 < S2 < S3 <
S4. The gap of S1 is so set that it does not break down at the voltage V. When all
the source capacitors are charged to the voltage V, the sphere gap S 1 is triggered
to break down. With the break down of S1, VB = VA = V and hence VC = 2V as
Cs2 is charged to a potential of V. The potential difference across the sphere gap
S2 is therefore 2V, which results in the break down of S2. Then, VD = 2V and VE
= 3V. Thus the voltage appearing across sphere gap S3 is 3V and it breaks down.
Consequently, VF = 3V and VG = 4V which causes breakdown of sphere gap S4. It
leads to charging of Cb followed by discharging of source and load capacitors.
Thus the required impulse voltage wave is generated across Cb.

iii) Wave Shaping Circuit:

To achieve standard impulse waveshape, the resistances Rd2 and Re2 are
commonly changed. The equivalent Rd and Re is partially distributed within the
impulse generator as Rd1 and Re1. The rest are placed as lumped parameters as Rd2
and Re2 externally. While testing objects of low surge impedance, standard
waveshapes are difficult to achieve. In that case along with lumped Rd2 and Re2,
Re1 and Rd1 might also be changed.

Representation of Multistage Generator by an Equivalent Single-Stage Generator:

When all the spark-gaps, viz. S1, S2 and S3 break down, then all the stage capacitors are
connected in series.
Let, Cs1 = Cs2 = Cs3 = Cs4 = Cs.

Cs
Hence, C eq 
4
Cs
 In general, C eq 
n

If this circuit is compared to a single – stage circuit, then it may be said that both
Re and Rd have been divided into two groups: One placed within the generator and the
other in the wave shaping section.
Problem:

Design an eight-stage impulse generator to generate 1400kV with voltage

efficiency of 96%. The energy stored in the impulse generator is 16kJ and the input dc
voltage is 175kV.

Solution:

When the sphere-gap Sn breaks down, then the capacitors Ceq and Cb are
connected in series. Neglecting Rd1 and Rd2 the output voltage

C eq
V2   V1
C eq  Cb
V2
Given, Voltage efficiency  v    0.96
V1
C eq
or,  0.96
Again, energy stored in an impulse generator C eq  Cb
= Energy stored in the stage capacitors.

For the eight stages, there are eight stage capacitors, each of which is charged to
175kV.
So,
Again,
1
2
 2

 C s  175  10 3  8  16  10 3 C eq
 0.96
or, C s  0.1306 F . C eq  C b
Cs or, C b  0.68nF .
 C eq   16.32nF .
8

Now, the wave-front time is determined by the charging time of Cb, when all the Cs are in
series.
This is because of the following:
6
3.243C eq  Cb  t
1.2  10   Rd 
C eq  Cb  v  V (1  e 
)
or, Rd  566.8 1 1
So, t 30   ln ( ) and t 90   ln ( )
1  0.3 1  0.9
t 90  t 30   1 1 
tf   ln ( )  ln ( )
0.9  0.3 0.6  1  0.9 1  0.3 
 3.243 
The tail time is determined by the discharging time of Cb through both Re and Rd. During
this discharge, Ceq and Cb are connected in parallel.

 
 50  10 6  1.2  10 6  0.693C eq  Cb Re  Rd 
or Re  Rd  4142.2 t
 1 1
or, Re  3575.3  v V e 
, So, t 50   ln   ln  0.693 
v /V 0.5

Triggering of Impulse Generator

The following problems are generally associated with the breakdown of the 1st
stage sphere gap:
i) If the gap is too large, then no breakdown occurs.
ii) If the gap is too small, then breakdown occurs before the capacitors are
charged to the desired voltage.
So, for proper operation of the impulse generator at a desired voltage, it is
required that the sphere gap in the 1st stage breaks down at the desired voltage and
desired instant. The mechanism by which the initiation of the breakdown of the 1 st stage
gap occurs is called triggering.
Triggering may be manual or electronic. In manual triggering, the spark gaps are
so set that they do not breakdown at the voltage to which the stage capacitors are
charged. When all the capacitors are charged to the desired voltage level, the sphere gap
spacings are reduced by mechanical means. The reduction of distance between the
spheres results in breakdown and thus triggering is achieved. However, for impulse
generators with higher number of stages, mechanical triggering is very inefficient and
hence is not used. So in practice electronic triggering is employed.
 For uniform field, e.g. parallel plate capacitor, air breakdown occurs at 30kVp/cm
at normal temperature and pressure that is same everywhere.
For non-uniform field, the potential distribution is non-linear and hence electric
field intensity is not same everywhere. Hence, for the same spacing non-uniform field
will breakdown at an applied voltage that is much lower than that required to break
uniform field gap.
For electronic triggering the concept of breakdown in non-uniform field is
utilized.

Trigatron Gap

The three-electrode sphere gap arrangement used for triggering is known as

Trigatron Gap and is used in the 1st stage of the impulse generator circuit. When
triggering is required, a pulse of about 10-12 kV is applied at the 3rd electrode causing
breakdown of the annular gap between the earthed main electrode and the 3rd electrode.
The charge particles thus produced accumulate in the main gap and distorts the main field
between the high voltage and earthed spheres. In this process the field in the main gap
becomes non-uniform and as a result field intensity increases. This causes breakdown of
the main gap and triggering is achieved.
The 3rd electrode is always housed in the earthed sphere of the 1st stage and is
insulated from the earthed sphere. It is normally made of brass or copper.

Generation of Trigger Pulse

 The transformer having a high turns ratio is connected to a dc source through a
polarity reversal switch. The switches S1 and S2 operate simultaneously. When S1 is
closed S2 is also closed and hence, the secondary is short-circuited by the capacitor C. A
dc current is established on the primary side that is governed by the resistance R. When
S1 and S2 are opened, then due to breaking of current on the primary side, a voltage is
induced in the primary winding due to L(di/dt), which is reflected to the secondary side
with amplified magnitude. This high voltage pulse is applied between the third electrode
and the earthed main electrode to achieve triggering.

Delay Cable:

While recording an impulse signal by CRO, it has to be ensured that the CRO is
properly synchronized with the impulse signal. Otherwise the entire impulse signal will
not be recorded. For this purpose, the time-base of the CRO is activated first and the
impulse signal is made to arrive at the vertical plates with a small delay. This is achieved
with the help of delay cable as shown in the Fig.56.

The antenna picks up the signal generated due to discharge of the sphere-gap
within the impulse generator and fires the time-base of CRO, as soon as the impulse
generator is triggered. The impulse signal arrives at the vertical plates after passing
through a delay cable, which introduces a delay of about 0.1 – 0.5s depending upon the
insulating medium and length of the delay cable. This ensures that the complete impulse
signal is recorded by the CRO.

The propagation velocity of a signal along a transmission line is

c
v
 r r

For normal signal cables r = 1, so that the above equation can be simplified to

c
v
r
 The relative velocity is defined as the ratio of the actual velocity in the cable to
the velocity of light

v
v rel   100%
c

In standard signal cables using solid polyethylene or Teflon as a dielectric, the

propagation velocity is about 60 – 70% of the velocity of light.

The reciprocal of propagation velocity is the transit time per unit length

1
T ,
v

which is normally given in s/cm. The absolute transition time and the appropriate cable
length necessary to delay a signal by a specified amount can be calculated from

L
  T .L  r
c

Trigger Pulse Generation for multi-purpose use:

CBL  100pF