Ministry of Science and Higher Education of the Russian Federation

ST. PETERSBURG
POLYTECHNICAL UNIVERSITY OF PETER THE GREAT

Yu.E. Adamyan, S.I. Krivosheev, S.G. Magazinov

High voltage technology. Laboratory course.

Textbook

Saint-Petersburg

2020

1
 The textbook corresponds to educational master program 13.03.02 Electrical
Engineering. Intended to implementation of the laboratory course "High voltage
technology" К.М.06.01. The main goal of the course is obtaining practical skills in
arrangement of experiments, measurements, processing and representation of the
experimental data in the specific environment of high voltage technique . For all
works the specialized laboratory installations are equipped, providing safe
operating conditions and variety of experimental abilities.

Some data needed for calculations are presented in the works descriptions, some,
that are commonly used, in the Appendix.

2
Оглавление
Оглавление................................................................................................................................................... 3
The list of laboratory works .......................................................................................................................... 4
Work 1. Breakdown of air gaps at DC voltage............................................................................................... 5
1.1. Work program .................................................................................................................................... 5
1.2. Work basics ........................................................................................................................................ 5
1.3. Determination of breakdown voltage dependence on the gap length for the gaps: pin – plane,
pin-pin at positive and negative pin polarity. .......................................................................................... 6
1.4 The procedure of breakdown voltages measurements: ..................................................................... 6
1.5 The report content. ............................................................................................................................ 8
Work 2. Breakdown of air gaps at AC voltage............................................................................................... 9
2.1 Work program ..................................................................................................................................... 9
2.2 Work basics ......................................................................................................................................... 9
2.3 Determination of AC breakdown voltage dependence on the gap length for the gaps pin – plane,
pin-pin, Rogowski electrodes and spherical electrodes. .........................................................................10
2.4 The report content. ...........................................................................................................................11
Work 3. Surface discharge in gas. ...............................................................................................................12
3.1. Work program ..................................................................................................................................12
3.2. Work description ..............................................................................................................................12
3.3 The experimental setup ....................................................................................................................13
3.4 The report content. ...........................................................................................................................14
Work 4. Characteristics of corona discharge on wires at DC voltage. ........................................................15
4.1. Work program ..................................................................................................................................15
4.2. Work basics ......................................................................................................................................15
4.3. Determination of the corona discharge starting voltage and discharge current dependence on
voltage at positive and negative wire voltage polarity. ..........................................................................15
4.4. Calculation of the corona discharge starting voltage. .....................................................................17
4.5. The report content. ..........................................................................................................................18
Work 5. Characteristics of corona discharge on wires at AC voltage. ........................................................19
5.1. Work program ..................................................................................................................................19
5.2. Work basics ......................................................................................................................................19
5.3. Measuring of the corona discharge critical voltage and obtaining dependence of energy loss on
voltage for coaxial wire system. ..............................................................................................................19
5.4 Calculation of the critical voltage for industrial frequency corona discharge. .................................20
5.5. The report content. ..........................................................................................................................21
Work 6. Voltage distribution on a string of insulators ................................................................................22

3
6.1 Work program ......................................................................................................................................22
6.2. Brief description of the work ...........................................................................................................22
6.3. Measuring of voltage distribution on a string of insulators .............................................................23
6.4. Determining of the “cap and pin” insulator breakdown voltage .....................................................24
6.5. The report content ...........................................................................................................................24
Work 7. High voltage pulse generator (Marx Generator). ..........................................................................25
7.1. Work basics ......................................................................................................................................25
7.2 Experimental setup. ..........................................................................................................................25
7.3. Measurements program. .................................................................................................................28
7.4. The lab report contents ...................................................................................................................30
Work 8. High voltage AC measurements. ...................................................................................................31
8.1. Work program ..................................................................................................................................31
8.2. Work basics ......................................................................................................................................31
8.3. Obtaining of the HV transformer graduation curve using test spark gap. .......................................31
8.4. High voltage measurement using the capacitive divider with electrostatic voltmeter. ..................33
8.5. High voltage measurement using the capacitive divider with microammeter. ...............................34
8.6. Estimation of the high voltage sinusoid quality. ..............................................................................34
8.7. The lab report content. ....................................................................................................................34
Appendix. ....................................................................................................................................................36

The list of laboratory works

The following laboratory works can be implemented on the basis of the

educational high voltage laboratory of the Institute of Energy:
1. Breakdown of air gaps at DC voltage
2. Breakdown of air gaps at AC voltage
3. Surface discharge in gas
4. Corona discharge at DC voltage
5. Corona discharge at AC voltage
6. String of insulators
7. High voltage pulse generator
8. High voltage AC measurements
4
Work 1. Breakdown of air gaps at DC voltage.

1.1. Work program

Obtain dependences of DC breakdown voltage on air gap length for the gap
geometries: pin – plane (at positive and negative pin polarity) and pin-pin.
Determine dependences of the average electric field strength on the gap length.

1.2. Work basics

The high voltage level at which the air gap loses its electrical strength is called
breakdown voltage. Starting discharge voltage is the voltage at which self-
discharge conditions are reached in the gap. At the uniform and weakly
nonuniform field the breakdown and starting voltages are practically equal. In the
uniform field at the standard atmospheric conditions the average air electric
strength is about 30 kV/cm. In nonuniform fields, like pin-plane or pin-pin
geometries, the starting voltage is always lower than breakdown voltage. The
conditions at which the local field strength is higher than starting one are fulfilled
firstly in the vicinity of the electrode with small curative radius (pin).

At the positive polarity of the pin electrode, the excess positive space charge is
created near the electrode because of electrons drift in the electric field. The
electrons mobility is approximately two orders of magnitude higher than ions
mobility, so they are rapidly moving to the electrode where they are neutralized.
Being prolonged inside the gap this positive spatial charge increases the field
strength in outer zone of the developing discharge channel. Such field distribution
eases conditions of the breakdown development.

At the negative electrode polarity the space charge consists of positive ions in
narrow near-electrode zone being surrounded by electrons and negative ions. Such
distribution of the field reduces field out of the near-electrode zone and toughens
conditions of the breakdown development.

5
So, the breakdown voltages in nonuniform fields at negative polarity of the pin
electrode are substantially higher than at the pin positive polarity.

1.3. Determination of breakdown voltage dependence on the gap length for

the gaps: pin – plane, pin-pin at positive and negative pin polarity.

The DC experimental installation circuit diagram is displayed on fig. 1.1 The AC

voltage 220V 50Hz from the power network feeds the regulating transformer T1
through automatic circuit braker QF and the magnetic switch KM1. Magnetic
switch KM2 connects output T1 to the primary winding of the high voltage
transformer T2. The secondary (HV) winding of T2 is loaded by the rectifier VD,
C, where C is the filtering capacitor. The high voltage resistor R protects the circuit
at breakdown in the gap (TО). The gap voltage measurement is provided by
resistive-capacitive voltage divider R1,C1,R2,C2. The divider output voltage is
measured by low voltage voltmeter on the plant control panel.

Fig. 1.1 Experimental plant circuit diagram

Before starting measurements it is necessary to register the atmospheric pressure

and readings of the dry and wet thermometers of the laboratory psychrometer. The
difference between these temperatures allows calculation of the air humidity.

1.4 The procedure of breakdown voltages measurements:

Fix the electrodes on the supporting bench (plane electrode must be grounded). Set
the gap length 2cm.

Reconnect the circuit elements to get positive potential on the rectifier output.

Remove grounding rod from the high voltage supply output.

6
Switch on QF circuit breaker.

Push the button K1 on the control panel switching on the magnetic switch KM1.
Check for zero voltage on the secondary side of the voltage regulator T1 using
meter PV. If the voltage is not zero, set it by the counterclockwise rotation of the
regulator handle.

Push the button K2 on the control panel switching on the magnetic switch KM2.

Gradually raise the transformer T2 input voltage up to air gap breakdown (rise the
voltage approximately 30sec before breakdown).

Write down the readings of the meter on control panel (1µA corresponds to 1kV).
To reduce statistical error repeat measurement at least 3 times.

Measurements have to be performed at 5-7 different gap distances for each gap
geometry and polarity (For equipment safety keep maximal gap voltages not higher
than 80-90kV). Gap distance step should be opted for current experimental
conditions (the gap change step is approximately 1-2 cm).

Register atmospheric pressure and temperature before voltage measurements (it is

important for matching of tabulated spark gap discharge voltage with atmospheric
conditions).

Recalculate obtained breakdown voltages to the normal atmospheric conditions

using correcting coefficients
𝑈𝑚 𝐾𝛾
𝑈0 =
𝐾𝑝 𝐾𝑡

Here Um – measured voltage, K, Kp, Kt - correcting coefficients for the humidity,
pressure and temperature
𝑃 293
𝐾𝑝 = , 𝐾𝑡 = ,
𝑃0 273+𝑡

P – current pressure, P0 – normal pressure, t – current temperature in centigrades.

Absolute humidity correcting coefficient K have to be considered for DC

positive polarity and can be determined by graph fig A.1 in Appendix using
temperatures of the dry and wet thermometers (curve “a” on the left side of the
figure).

7
For the negative polarity humidity correcting coefficient is not used.

Average values of breakdown field strength can be determined by division of

breakdown voltages by the gap length.

1.5 The report content.

The report must contain:

a) Experimental plant circuit;

b) Tables of the results of measurements
c) Graphs representing dependences of breakdown voltages and average
electric field strength on the gap length for all gap geometries;
d) Results summary. Focus on the impact of the voltage polarity on the
breakdown voltages.

8
Work 2. Breakdown of air gaps at AC voltage.

2.1 Work program

Obtain dependence of AC breakdown voltage on air gap length for the following
gap geometries:

pin – plane, pin-pin, spherical electrodes, Rogowski electrodes.

Determine dependences of the average electric field strength on the gap length.

2.2 Work basics

The gap breakdown process development time is many times shorter than the
power network AC voltage period. At the alternating voltage applied to the
electrodes, the field distribution near the electrode with small curvative radius (pin)
are slightly different from DC voltage case because of residual charges at the
voltage oscillation.

So the gap breakdown takes place at positive voltage half-period on the electrode
with small curvative radius because the gap electric strength is lower at this
polarity.

In addition to the gap geometries used in the previous work the Rogowski
electrodes and spherical electrodes are used where the electric field is close to
uniform at gap length smaller than the electrode radius. For these geometries the
breakdown field strength is close to maximal value for the air Emax30 kV/cm.

9
 Fig. 2.1. Electric field of the
Rogowski electrodes. If the
electrodes are set precisely
parallel, the field in the
central space are close to
uniform.

2.3 Determination of AC breakdown voltage dependence on the gap length for

the gaps pin – plane, pin-pin, Rogowski electrodes and spherical electrodes.

For AC measurements exclude rectifier from the test circuit. Circuit diagram for
these measurements is displayed on fig 2.2.

Fig. 2.2. Test circuit diagram for AC measurements.

10
Measurements procedure is the same as in p.1.3. Note that the meter on the control
panel shows the amplitude value of the gap voltage.

Recalculate obtained breakdown voltages to the normal atmospheric conditions,

determine average breakdown field strength.

2.4 The report content.

The report contains:

a) Experimental plant circuit;

b) Measurements results tables;

c) Graphs representing dependences of breakdown voltages and average electric

field strengths on the gap length for all gap geometries;

d)Results summary. Focus on the impact of the electrodes shape on their

breakdown voltages;

11
Work 3. Surface discharge in gas.

3.1. Work program

1. Obtain dependence of the discharge voltage on the air gap size for gap laying
on glass surface at following conditions:
a) without grounded conductive sheet under glass.
b) with grounded conductive sheet under glass at two different values of the
glass thickness.
Present dependences of the discharge voltage and average electric field strength
on the air gap size for all cases at tabular and graph form.

3.2. Work description

A combination of current leading and insulating parts is inevitable at high

voltage equipment design. The stylized pictures of such compositions are presented
on fig. 3.

Fig. 3.1 Typical surface discharge geometries

In the first case (a), charged particles and ions of adsorbed water deform the
field in the gap. Inevitable presence of small gaps between electrodes and insulator
12
results in local field strength increasing. Thus, the discharge voltage decreases
slightly in comparison with the case of insulator absence inside the gap.

In the second case (b), the field between the electrodes is nonuniform. The
insulator on which the electrodes are placed does not actually influence on the
discharge voltage and this voltage is practically similar to the same geometry of
electrodes in air.

In the third case (c), where the insulator thickness is substantially less than
the gap length the electric field at the electrode edge is distinctly nonuniform due to
closeness of the conductive backing. Here the corona discharge starts at relatively
low voltage applied. At further voltage rise the diffused corona discharge
transforms into streamer stage and at the next stage – to arc (compare with pin-
plane discharge in the air).

Fig. 3.2. Test circuit diagram

The presence of the conductive backing dramatically reduces the discharge

voltage and this effect is more distinguished at thinner insulating layer.

3.3 The experimental setup

Operation of the experimental HV- installation is analogous to described in
work 2. It is possible to measure discharge voltage by the meter PV on the primary
side of the high-voltage transformer T2 with known transformation rate K=500. For
obtaining the voltage amplitude multiply the meter readings by 500*2.

There are two sheets of glass, metal sheet for backing, and the set of length-
measuring bars for assembling the discharge geometry. Measurements have to be
performed at 5-7 different gap distances for each gap geometry. Please, keep
maximal gap voltages not higher than [Link] measuring, try to distinguish
corona, streamer, and arc stages of the discharge registering corresponding
voltages.

13
3.4 The report content.

The report contains:

a) Experimental installation circuit diagram;

b) Measurements result tables;

c) Graphs representing starting voltages of corona, streamer and arc for all
studied configurations;

d) Results summary. Focus on the impact of conductive backing on all stages of

the discharge

14
Work 4. Characteristics of corona discharge on wires at DC voltage.

4.1. Work program

1. Determine starting voltage of the corona discharge and dependence of the

discharge current on voltage at positive and negative wire voltage polarity.

2. Calculate starting voltage of corona discharge for wires used.

4.2. Work basics

Corona discharge is self-discharge appearing in strongly non-uniform

electric field in the high field strength area. At avalanche corona discharge form
thickness of ionized gas is about few millimeters. In the rest of the discharge gap
having lower field strength, electron-ion impact ionization is absent but ions are
moving in the electric field. In collisions with gas molecules ions are transferring
to them a part of the kinetic energy obtained at the acceleration in the electric field.
As the result the gas is heated by the energy lost by ions. Besides the ionization the
molecules excitation process takes place. Molecules are retained in the excited
state for a very short time and then get back to the stationary state with photons
emission. That is why the corona discharge is accompanied with blue glow. Pulse
currents presenting in corona discharge have very high frequency that cases
electromagnetic influence for electronic devices.

4.3. Determination of the corona discharge starting voltage and discharge

current dependence on voltage at positive and negative wire voltage polarity.

Determination of the voltage of the corona discharge beginning and discharge

current dependence on voltage are carried out on the special high voltage plant.
The circuit diagram is displayed on fig. 4.1.

Uninsulated wire is located coaxially in metallic cylinder 1 having 600mm

diameter and active length 1.2 m. The cylinder edges are rounded to avoid edge
effects. The cylinder is insulated from the ground and surrounded by the metallic
screen for measuring circuits protection from electromagnetic influences. For
15
measuring of the corona current the microammeter PA is connected between the
cylinder and the ground. The PA scale has zero point in its middle that allows
conducting measurements at positive either at negative wire voltage polarity.

The wire voltage measurement is provided by resistive-capacitive voltage divider

R1,C1,R2,C2. The divider output voltage is measured by low voltage voltmeter on
the plant control panel. Measuring devices and magnetic switches control buttons
are mounted on the control panel.

Fig.4.1. The schematic of the for corona discharge installation

Before starting measurements it is necessary to register the atmospheric pressure

and reading of the dry thermometer.

Measurements procedure:

Fix the 1.2mm diameter wire in the cylinder 1.

Remove shorting connector SW1 from the HV diode VD, set VD anode to the
transformer output, connect diode cathode with filtering capacitor C. Remove the
grounding rod from the capacitor HV electrode and close the fence door.

Set position “x1” of the switch S3 on the control panel, switch on S1 and S2.

Switch on QF circuit breaker.

16
Push the button K1 on the control panel switching on the magnetic switch KM1.
Check for zero voltage on the secondary side of the voltage regulator T1 using
meter PV.

Push the button K2 on the control panel switching on the magnetic switch KM2.

Gradually raise the transformer T2 input voltage up to appearing of the corona

current through the microammeter PA.

Write down the reading of the HV voltage meter (1µA corresponds to 1kV).
Voltage obtained is the measured voltage of the corona discharge beginning.

To reduce statistical error repeat measurement at least 3 times.

Gradually raise the wire voltage to get corona voltage-current dependence.

Measurements have to be performed at 8-10 voltage values. When the current
reaches 80-100 µA set S3 to the position “x5”.

Switch off the installation, set regulating transformer to zero, open the door and put
the grounding on the HV transformer output.

Rotate the HV diode VD at 180 degrees to change the wire voltage polarity.

Repeat measurements at negative wire voltage.

To study the influence of the wire diameter conduct measurements at 2mm and
3mm wire.

Match measured voltage values to the normal atmospheric conditions.

There is no need to insert the humidity correcting coefficient in the corona

measurements. Determine the atmospheric conditions correcting coefficient
𝑃 ∗ 293
𝛿=
𝑃0 (273 + 𝑡)

Voltage matched value U0=Um/, here Um is the measured voltage, P and P0 –

current and normal atmospheric conditions.

4.4. Calculation of the corona discharge starting voltage.

For coaxial wire system the corona discharge starting voltage can be calculated by
formula Uv=Evr1ln(r2/r1). Here r1 and r2 – radii of the wire and the cylinder ;
17
Ev – field strength of the corona discharge start, which can be calculated for the
0.298
wire radius less than 1cm as 𝐸𝑣 = 𝐴 ∙ 𝛿 ∙ (1 + ) (Peek's law, index “v”
√𝛿∙𝑟1/𝑟
means "visual critical corona voltage" ). A=3.03106 V/m,  - relative air density,
r1 – wire radius, r=0.01m. Calculate this value for all wire diameters used.

4.5. The report content.

The report must contain:

e) Experimental plant circuit;

f) Measurements result tables
g) Graphs representing corona voltage-current characteristics ;
h) Experimental and calculated values of the voltages of the corona discharge
beginning.
i) Results summary.

18
Work 5. Characteristics of corona discharge on wires at AC voltage.

5.1. Work program

1. Determine critical voltage of the AC corona discharge and obtain power loss
dependence on voltage for coaxial wire system.

2. Calculate critical voltage of AC corona discharge.

5.2. Work basics

In the electric field two types of ions (positive and negative) can appear in
air media. If the ion polarity is coincided with wire voltage polarity it moves
outside from the wire. Ions of the opposite polarity are in the majority neutralized
on the wire surface. Therefore, the spatial discharge near the wire consists mostly
of ions with polarity corresponding to the wire polarity. At AC voltage interaction
between charged particles generated in the gas at previous half period leads to
partial neutralization of the spatial discharge. It eases the corona triggering and
intensifies corona discharge. At AC corona minimal voltage and field strength at
that the corona discharge exists are called “critical”. Corona discharge is
accompanied with energy losses. At AC voltage losses during one voltage period
can be measured as square of Volt-Coulomb characteristics of the discharge gap.
For obtaining of the transferred charge signal, the corona current signal is
integrated by capacitor.

5.3. Measuring of the corona discharge critical voltage and obtaining

dependence of energy loss on voltage for coaxial wire system.

Measuring of the corona discharge critical voltage and obtaining dependence of

energy loss on voltage is performed on the same plant as for DC corona (fig.4.1).
The plant circuit should be modified as follows:

Switch off S2 on the control panel. After grounding of the installation:

HV diode VD must be shorted by connector SW1.

19
Filtering capacitor C must be disconnected.

Fix the 1.2mm diameter wire in the cylinder 1.

Volt-Coulomb characteristics are measured by the oscilloscope in XY-mode.

Connect the voltage signal output from the measuring divider with oscilloscope Y
channel.

Connect the charge signal from the capacitor C3 with oscilloscope X channel.

Switch the equipment on in the same way as in 4.3.

If the corona discharge is absent, figure on the oscilloscope screen looks like
straight sloping line. As corona discharge appears the line transforms to loop.
Register this moment for determination of the critical voltage.

Voltage increase leads to increase of the loop area. Make 5-7 oscillograms of
Volt-Coulomb characteristics at different voltage amplitudes from the critical
voltage to 80-90kV.

Determine power loss as P=SKuKqf

Here S – square of a Volt-Coulomb characteristic (div2), Ku and Kq – voltage and

charge scales, f – power net frequency (50Hz).

Ku=U/hy. Here U – voltage amplitude on the wire, hy – maximal vertical

oscilloscope beam displacement.

Respectively Kq=hxC3. Here hx – oscilloscope horizontal sensitivity 0.2 or 1.0

V/div depending on the oscilloscope settings; C3 – 0.4710-6F.

Repeat measurements at wires diameters 2 and 3 mm.

Match the results obtained to normal atmospheric conditions as it described at 4.3

5.4 Calculation of the critical voltage for industrial frequency corona

discharge.

For coaxial wire system the critical voltage of the corona discharge can be
calculated by formula Uc=Ecr1ln(r2/r1). Here r1 and r2 – radii of the wire and the
cylinder ;Ec –critical field strength of the corona discharge which can be
calculated for the industrial frequency as
20
 0.613
𝐸𝑐 = 24.5 ∙ 𝛿 ∙ (1 + ) kV/cm
(𝛿∙𝑟1)0.4

Here,  - relative air density, r1 – wire radius. Calculate this value for all wire
diameters used.

5.5. The report content.

The report must contain:

a) Experimental plant circuit;

b) Measurements result tables
c) Oscillograms of Volt- Coulomb characteristics;
d) Graphs representing dependences of corona loss on wire voltage
e) Experimental and calculated values of the voltages of the corona critical
voltages.
f) Results summary.

21
Work 6. Voltage distribution on a string of insulators

6.1 Work program

Determine: a) Voltage distribution on a string of insulators;

b) Break-down voltage of a “cap and pin” insulator .

6.2. Brief description of the work

Cap and pin insulators are widely used on power lines 35 kV or higher.
Porcelain, glass or plastic are typical materials for this class of insulators.

For AC voltage a string of insulators can be represented by circuit,

consisting of capacitors (see the fig. 6.1), that equivalents constructive
capacitances of insulators and between insulator pins, wire and grounded elements
of power line.

Fig. 6.1. Equivalent circuit of the string

of insulators
Herein, С0 = 50–70 pF – the capacity
between pins of insulators, Сп = 0.5–1
pF – capacity of separate insulators to
HV-wire, and Сз = 4–5 pF capacity to
grounded parts of a transmission line
tower. Currents, flowing through the
capacities of Сз and Сп, substantially
disturb the linear distribution of the
voltage along the string, formed by
capacities of С0 that are connected in
series. As Сз>Cп, the highest voltage is
applied to the first insulator which is
closest to the HV wire.

22
 In accordance with technical regulations, the string must be designed so that
the highest voltage level on this insulator does not exceed the allowed value,
corresponding to the acceptable level of electromagnetic influences (noises). For
lowering of the voltage value on the first insulator, electric shields (e.g. of toroidal
shape) can be used.

6.3. Measuring of voltage distribution on a string of insulators

Measuring of voltage distribution on a string of insulators can be

implemented on the high voltage installation as it is displayed on fig. 6.2.

Special device is used which consists of insulating handle MR and small

spark gap (1-2mm). Thin metal wires (antennae) are intended for connection of
the gap with electrodes. For the safety, the handle has grounding wire which is
connected with the grounding circuit.

Switch on the QF and KM1. Set regulating transformer T1 to zero output

voltage. Switch on KM2.

After connecting of the measuring gap with electrodes, gradually rise the
voltage applied to the HV wire. Write down the voltage Ui, corresponding to the
break-down of the gap at insulator #i.

Fig. 6.2. Measuring circuit

Set regulating transformer T1 to zero output voltage. Repeat measurement
on the next insulator.

After having all voltages measured, i – quote of each insulator, can be

calculated as

1
Ui
i  ,
in1 1
Ui

23
here n – number of insulators in the string; i –index number of the insulator,
counted from HV-wire site.

Install electric shields on top and bottom of the string. Repeat measurements in the
shielded geometry.

6.4. Determining of the “cap and pin” insulator breakdown voltage

For determining of the break-down voltage of the “cap and pin” insulator,
remove electric shields, short with the special wire all insulators of the string
except one which is closest to HV wire.

Switch on the installation as it is described in 6.3. Gradually rise applied voltage

until break-down is observed. Immediately after breakdown switch the HV off
pushing red button on the control panel.

To get actual RMS value of the break-down voltage multiply the reading of
meter PV by 500. That number is the transformation rate of the HV transformer
T2.

6.5. The report content

a) Experimental plant circuit;

b) Measurements result tables;

c) Graphs representing the voltage distributions along the string of “cap and
pin” insulators for geometries with and without shields

d) Conclusion, containing information about voltage distribution nonuniformity

and effect of shields.

24
Work 7. High voltage pulse generator (Marx Generator).

7.1. Work basics

Marx generator is intended for generation of single HV pulses with
amplitudes higher than 100 kV (typically 200-1000kV). Mostly the pulse shape is
bi-exponential and can be characterized by two time constants – rise time (r) and
fall time (f). Standard values of time constants for lightning modeling are 1.2 µs
and 50µs respectively.

The generator operating principle is based on storing energy in set of capacitors at

relatively low voltage level (up to 100kV) and discharging them through spark
gaps in combination, providing series connection of the capacitors for the
discharge time.

7.2 Experimental setup.

The generator schematic is presented on fig.7.1.

The HV transformer is supplied by AC voltage from regulating low voltage

transformer (not shown on the schematic diagram). The transformer output is
connected with two voltage doubling rectifiers producing DC high voltages of
different polarities ( fig.7.1. "+" and "-"). The rectifiers output voltages (indicated
on the control panel) can be connected with capacitors by two ways, changing
pulse polarity.

For the presented kind of schematic the generator output pulse amplitude roughly
can be determined as capacitor charging voltage multiplied by doubled sections
number. There are 4 sections in the laboratory generator, so the voltage
multiplication rate is 8.

The generator main technical characteristics claimed by manufacturer are:

1 Output voltage amplitude kV 80-600

2 Voltage pulse shape (rise time/fall time) µs 1.2/50
3 Maximal energy stored kJ 9
4 Maximal value of short circuit current A >1000
5 Maximal charging voltage kV 80
25
The pulse front time is determined by front capacitor Cf and by discharge and
cross-section resistors Rd,Rcs. The fall time is determined by the equivalent
capacity of the stage Cs/2 and cross-section resistors Rcs.

Before pulse generation the capacitors Cs must be charged with certain level of
uniformity. It requires long enough time because of their different charging time
(being charged through bigger resistance capacitors of higher stages have longer
charging time). To provide reliable operation, the spark-gap switches S1-S4 have
the electrically driven gap adjustment system. Switch S1 is trigatron - type, so it
can be triggered by high voltage pulse from control panel. After breakdown of the
S1, transient process in the discharge circuit leads to breakdown of the resting
switches.

For each level of output voltage the proper gap distance for spark gaps S1-S4
must be set according to fig.2. If the gap value is set inside working interval for the
given charging voltage, the triggering pulse being applied to S1 after charging
results in breakdown of all the spark gaps and forming of the output pulse.

Increasing spark gap needs to be accompanied with corresponding rise of the HV

supply voltage. The criterion of normal operation is charging time before
breakdown abut 20s.

The capacitive voltage divider Cf is used for a pulse shape registration by

oscilloscope. Its transfer coefficient is 1:10000. Simultaneously it operates as a
pulse front correcting capacitor. It does not fit well with the measurement functions
because of resonant properties of the circuit branch formed by Cf and inductance of
its connection to the generator output. So, it is preferable to use additional
resistive-capacitive divider HVD for more precise pulse shape registration. In this
case it needs special calibration (see below).

26
Fig . 7.1. The generator schematic

27
 Spark gap mm

Fig.7.2. To the spark gap determination

The installation load part is presented on the right part of fig.7.1. Voltage pulse is
formed in the point Mline and transferred to the point M by connecting wire Lcon.
Objects of measurements can be connected with this point. E.g. it can be nonlinear
overvoltage arrester (1) with the equivalent circuit of grounding (2), pin-plane
spark gap (3), ball spark gap(4) or string of insulators (5).

Ball spark gap is a standard device for HV measurements (see Table A.1).

7.3. Measurements program.

1. Voltage divider calibration

28
These measurements need the pulse registration by oscilloscope.

Set the ball spark gap distance 50mm.

Gradually rise pulse amplitude, registering pulses shape by the oscilloscope.

Determine oscilloscope voltage amplitude corresponding to 50% probability of the
ball spark gap breakdown. From 50% breakdown oscilloscope voltage determine
transfer coefficient of the HV divider using Table A. 1.

2. Obtaining voltage-time characteristics of ball spark gaps

Rising pulse amplitude above 50% breakdown level, obtain dependence of

breakdown voltage on breakdown time interval from the pulse start. On fig. 7.3
two typical oscillograms of pulse voltages on air gap are presented. Here, U1, U2 -
breakdown voltages, t1, t2 - breakdown times. Make at least 3 measurements on
each level of charging voltage. With respect to the statistical deviations, collect 10-
15 points, changing gap distance of the generator switches, and, correspondingly,
HV supply output voltage.

Fig. 7.3. Typical oscillograms of pulse voltages on air gap.

3. Voltage-time characteristics of HV line insulator

29
Determine 50% breakdown voltage for standard pulse as above.

Gradually rising pulse amplitude obtain dependence of breakdown voltage on

breakdown time. Make at least 3 measurements on each level of charging voltage

Change pulse polarity and repeat measurements.

7.4. The lab report contents

please present results of measurements as tables and graphs.

Typical voltage –time characteristics are presented on fig.7.4.

Fig.7.4.

30
Work 8. High voltage AC measurements.

8.1. Work program

[Link] graduating curve of a HV transformer using standard air gap.

2. Making HV measurements using capacitive divider

a) by static voltage meter

b) by microammeter

3. Evaluating voltage sinusoid quality.

8.2. Work basics

The simplest way of AC measurements on a transformer's high voltage side is to

measure voltage on low voltage side and recalculate it to HV using the transformer
rate. But it is necessary to take into account that the relation between voltages on
high and low sides will differ depending on voltage level and the transformer load.

It is caused by the transformer impedance, magnetic core nonlinearity and corona

discharge losses on the HV side. That is why for precise recalculation of the
measured low voltage to HV side it is necessary to have the transformer graduating
curve specific for its actual load.

The another way, based on a capacitive divider application, requires additional

equipment. Low voltage measuring devices have to be connected in parallel with
the divider lower part. Measured high voltage is determined as low voltage meter
count multiplied by divider rate. Measurement accuracy can be satisfactory if the
meter impedance is much higher than the capacitive resistance of the divider lower
part. DC voltage cannot be measured by this method.

8.3. Obtaining of the HV transformer graduation curve using test spark gap.

The circuit diagram is displayed on fig. 8.1

31
 Fig. 8.1 Experimental plant circuit diagram

Before starting measurements register atmospheric pressure and temperature . It is

important for matching of tabulated spark gap discharge voltage with atmospheric
conditions.

To get one point of the graduation curve:

set the 0.5 cm distance between 12.5cm measuring spherical electrodes;

read from Table A.1 breakdown voltage for this gap length and voltage type;

remove grounding bar from HV electrode of the spark gap;

switch on QF circuit breaker.

Push the button K1 on the control panel switching on the magnetic switch KM1.
Check for zero voltage on the secondary side of the voltage regulator T1 using
meter PV. If the voltage is not zero, set it by the counterclockwise rotation of the
regulator handle.

Push the button K2 on the control panel switching on the magnetic switch KM2.

Gradually raise the transformer T2 input voltage up to air gap breakdown (rise the
voltage approximately 30sec before breakdown). Register the voltage value on PV
just before breakdown (Ub). Write down the value obtained in result table.
Measured value is RMS (repetitive medium square).Repeat the measurement two
more times to get average value of the primary breakdown voltage.

Determine the atmospheric conditions correcting coefficient

𝑃 ∙ 293
𝛿=
𝑃0 (273 + 𝑡)

32
Get the amplitude value of the spark gap breakdown voltage for current conditions

Um=U0. It is recommended to set X-axis as PV voltage, Y-axis as Ub.

For the graduating curve forming repeat measurements at distances between

measuring electrodes consequently 1.0, 1,5 and 2.0 cm.

The measurement error by graduating curve can be estimated as

∆𝑈𝑚 ∆𝑆 ∆𝑃 2 ∆𝑈𝑝𝑣 2 ∆𝑡
= √( )2 + ( ) +( ) +( )2 + 0.032 , where estimated errors of
𝑈𝑚 𝑆 𝑃 𝑈𝑝𝑣 273+𝑡
the gap length, pressure, temperature and primary voltage measurements are
considered as well as the error of spark gap standard measurements (0.03 are
considered.)

8.4. High voltage measurement using the capacitive divider with electrostatic
voltmeter.

For implementation of voltage measurement using the capacitive divider with

electrostatic voltmeter it is necessary to turn on the installation as it is described in
8.3, set the voltage on the PV voltmeter to 60-70 V and write down the readings of
the electrostatic voltmeter PV1. The amplitude of the transformer output voltage
U=KUpv1. K –division rate of the capacitive divider C1,C2 .
𝐶1+𝐶2
𝐾= , C1=350 nF,C2=0.325µF. Note: due to action of rectifier VD1 and
𝐶1
filter capacitor C3, Upv1 is the amplitude of the divider output.

33
The voltage measurement error can be estimated as

∆𝑈𝑚 ∆𝐾 ∆𝑈𝑝𝑣1 2
= √( )2 + ( ) + 0.022
𝑈𝑚 𝐾 𝑈𝑝𝑣1

∆𝐾
Here = 0.02 - relative error of the division rate measurement ∆𝑈𝑝𝑣1 –
𝐾
absolute error of the electrostatic voltmeter, 0.02 – relative error of the
measurement circuit, determined by capacitors C3=C4=1µF and resistors
R1=R2=910 kOhm;

8.5. High voltage measurement using the capacitive divider with

microammeter.

Another option is measuring of the divider secondary voltage using resistor R1 and
microammeter PA. The high voltage amplitude is U=KR1I, here I is the
microammeter reading.

The voltage measurement error can be estimated as

∆𝑈𝑚 ∆𝐾 2 ∆𝐼 ∆𝑅1 2
= √( ) + ( )2 + ( ) + 0.022
𝑈𝑚 𝐾 𝐼 𝑅1

∆𝑅1
Here ∆𝐼 – absolute error of the current measurement by PA, = 0.01 relative
𝑅1
error of the R1 value.

8.6. Estimation of the high voltage sinusoid quality.

For the estimation of the sinusoid quality set the transformer primary voltage as it
described in 8.4 and register the voltage amplitude Um by the electrostatic
voltmeter PV1. Then disconnect PV1 with no switching off the plant and no
changing voltage and switch it to the point “a”. Now PV1 will measure
effective(RMS) voltage on the transformer output U.
𝑈𝑚
If the voltage is practically sinusoidal = √2 ± 0.07
𝑈

8.7. The lab report content.

The report must contain:

34
a) Experimental plant circuit;

b)Measurements result tables

c)The transformer graduation curve;

d)Results summary.

35
Appendix.

Fig. A.1. To determination of the humidity correcting coefficient

Curve a –DC and AC voltages, b – pulse voltages

36
Table A.1

Breakdown voltages for alternating sinusoidal voltages, DC voltages of both

polarities, full standard and longer pulsed voltages of negative polarity (50%
discharge) - a) and full standard and longer pulsed voltages of positive polarity
(50% discharge) - b), kV

Distance Spheres diameter , cm

between

spheres, cm 6.25 12.5 25 50

a b a b a b a b

0.4 14.2 14.2

0.5 17.2 17.2 16.8 16.8

0.6 20.2 20.2 19.9 19.9

0.7 23.2 23.2 23.0 23.0

0.8 26.2 26.2 26.0 26.0

0.9 29.1 29.1 28.9 28.9

1.0 31.9 31.9 31.7 31.7 31.7 31.7

1.2 37.5 37.6 37.4 37.4 37.4 37.4

1.4 42.9 43.2 42.9 42.9 42.9 42.9

1.5 45.5 45.9 45.5 45.5 45.5 45.5

1.6 48.1 48.6 48.1 48.1 48.1 48.1

1.8 53.5 54.0 53.5 53.5 53.5 53.5

2.0 58.5 59.0 59.0 59.0 59.0 59.0 59.0 59.0

2.2 63.0 64.0 64.5 64.5 64.5 64.5 64.5 64.5

2.4 67.5 69.0 70.0 70.0 70.0 70.0 70.0 70.0

37
2.6 72.0 73.5 75.5 75.5 75.5 75.5 75.5 75.5

2.8 76.0 78.0 80.0 80.5 81.0 81.0 81.0 81.0

3.0 79.5 82.0 85.0 85.5 86.0 86.0 86.0 86.0

3.5 (87.5) (91.5) 97.0 98.0 99.0 99.0 99.0 99.0

4.0 (95.0) (101) 108 110 112 112 112 112

4.5 (101) (108) 119 122 125 125 125 125

5.0 (107) (115) 129 134 137 138 138 138

5.5 138 145 149 151 151 151

6.0 146 155 161 163 164 164

6.5 (154) (164) 173 175 177 177

7.0 (161) (173) 184 187 189 189

7.5 (168) (181) 195 199 202 202

8.0 (174) (189) 206 211 214 214

9.0 (185) (203) 226 233 239 239

10.0 (195) (215) 244 254 263 263

11.0 261 273 286 287

12.0 275 291 309 311

13.0 (289) (308) 331 334

14.0 (302) (323) 353 357

15.0 (314) (337) 373 380

16.0 (325) (350) 392 402

17.0 (336) (362) 411 422

18.0 (347) (374) 429 442

38