Generation of High DC Voltage

High DC voltages are extensively used for scientific research work,

electromedical equipment, electrostatic precipitators, electrostatic painting etc. It is also
used for testing equipment related to HVDC transmission lines. Another major use is in
testing HVDC power cables of long lengths to circumvent the problem of high charging
current to be fed to the cables under test.
Earlier HVDC was generated with the help of electrostatic generators. But now-a-
days HVDC is generated by means of rectifying circuits in conjunction with a.c. voltage
source. There are two rectification schemes in use, viz. a) Asymmetric and b) Symmetric.

Asymmetric Rectifier Circuit:

Neglecting leakage reactance of the transformer and the forward drop of the
diode, the capacitor C is charged to +Vm of the a.c. voltage when the diode D conducts. If
the output load current is zero (RL  ), then the potential of ‘b’ remains constant at
+Vm. But, the potential of “’’ oscillates between  Vm. So, the diode D must be able to
withstand a PIV of 2Vm.

22
 The output voltage V does not remain constant if the circuit is loaded. After the
condenser is charged upto +Vm, the potential of ‘b’ becomes less than +Vm during the
period t2 so that the diode D gets reverse biased. In this period the capacitor C discharges
some amount of charge to the load and the output voltage gradually decreases depending
upon the time constant of the load circuit in series with capacitor C. During the period t 1,
the diode again becomes forward biased and the capacitor again gets charged to +Vm. The
charge lost by the capacitor C during t2 is gained during t1. Since t1 << t2, the current
supplied by the transformer is pulsed in nature. The output ripple is given by

V  Vmin
V  m
2

Disadvantages:

i) Since half-wave rectification is employed, the voltage generation is asymmetric.

ii) The rectifier and the HT transformer need to be robust and hence expensive.
iii) The current supplied by the transformer to charge the capacitor C introduces a dc
component on the secondary side. Hence, no mmf balance takes place between
the primary and secondary. This may cause saturation of the transformer core
with the increase of load current.
To avoid core saturation following steps could be taken:
a) Core cross-section is to be increased.
b) Special transformers are to be used.
c) The existing transformers have to be operated at a relatively low voltage
compared to its rating. So that the operating point lies will below the knee
point.

One way of solving all these problems is to use single-phase full-wave rectifier circuits.

However, such full-wave rectifier circuits can only be used for the cases where
the HV winding of the transformer can be earthed at its midpoint.
Hence, more commonly used method is single-phase voltage doubler.

23
Single – Stage Voltage Doubler Circuit:

Such voltage doubler circuits are also known as Cockcroft – Walton Voltage
doubler circuits (1932).

Open – Circuited HV Output: (I = 0)

In the loop 0–A–B–C, the capacitor C1 charges upto a voltage of +Vm, when v(t)
reaches the lowest potential –Vm. If C2 is uncharged, then the diode D2 conducts as soon
as v(t) increases from –Vm and D1 gets reverse biased. During the period, when (dv/dt) is
positive, the potential across C1 remains constant at +Vm as it has no path to discharge.
Thus, the potential across diode D1 during this period is VBA = VBC + VCO and is an
oscillating voltage varying between 0 to +2Vm. But VBA is unidirectional in nature.
During this period when D2 conducts, capacitor C2 gets charged upto the maximum value
of VBA, i.e. 2Vm. Hence, a steady dc voltage whose value is double the peak value of
transformer secondary voltage is obtained across the capacitor C2 as output.

HV Output when loaded: (I > 0)

If the generator supplies a load current I, then the output voltage will never reach
the value 2Vm and there will also be a ripple on the output voltage.

The peak value of VD is reached when VC is at +Vm and D2 just stopped

conduction. After that the current I continuously discharges C2. There is a sudden drop in
VD during the period when D1 conducts to charge C1. VD reaches its minimum value at
the instant when D2 starts to conduct and C2 gets charged. During the period of
conduction of D2, C2 again gets charged upto VDmax. During the period t1, q amount of
charge is transferred to load from C2 and during the period t2 the same amount of charge
q is replenished by C1 to C2.

24
Voltage Regulation:

VD max = Maximum output voltage

VDmin = Minimum output voltage
V  VD min
Average DC output voltage  VDC  D max
2

a) Let, I be the load current. Then the amount of energy transferred to the load is as
follows.

1 1
C 2 VD2 max  C 2 VD2 min  VDC  I  t 2
2 2

V  VD min 
or, C 2   D max  VD max  VD min   VDC  I  t 2
 2 
or, C 2 VDC  V2  VDC  I  t 2
I .t 2
or, V2 
C2

Again, T  t1  t 2 and t 2  t1
So, T  t 2
I T I
 V2  
C 2 fC 2

25
b) C1 supplies the charge to C2 when D2 conducts. Moreover, some charge is drained
from C1 to C2 through D2 even when it is turned off. As a result, VB is decreased
from 2Vm by an amount V1.

dq dv V
so, I   C1  C1 1
dt dt T
IT I
or, V1  
C1 fC1

VD max  VD min
so, Av DC output voltage VDC  
2
V2
 2Vm  V1 
2
I I
 2Vm  
fC1 2 fC 2
I  1 1 
 2Vm    
f  C1 2C 2 

Thus the output voltage depends on – i) the load current, ii) frequency and iii) stage
capacitances. Higher the frequency, lesser the voltage drops. Hence, often Cockcroft–
Walton voltage doubler circuits are fed at a higher frequency from oscillators.

Problem:

A Cockcroft – Walton voltage doubler circuit is used to test a cable at 170kV. The
insulation resistance of the cable is 6  109 /m and the length of the cable is 12m. Stage
capacitances are both 0.1F. The generator is supplied from a 230V/250kV testing
transformer. Calculate the voltage to be applied to the input of the transformer at 50Hz.

Solution:

Given : C1  C 2  0.1F
RL  6  10 9  / m
6  10 9
 Total RL   5  10 8 
12
Test Voltage  170kV
170  10 3
 Load current (I)   3.4  10  4 A
8
5  10

26
  1
I 1 
Now : VDC  2Vm    
f C1 2C 2 
3.4  10  4  1 1 
or, 2Vm  170  10 
3
 6
 6 
50  0.1  10 2  0.1  10 
 170.102 kV
 Vm  85.051 kV

85.051
 Transformer primary Vmax   230  78.25 V
250
78.25
 Transformer primary Vrms   55.33 V (Ans)
2

Multistage Cockcroft – Walton Voltage Multiplier Circuit:

27
 Terminal Potential w.r.t earth Nature of waveform

A 0 – 2Vm Oscillatory
B 2Vm DC
C 2Vm – 4Vm Oscillatory
D 4Vm DC
E 4Vm – 6Vm Oscillatory
F 6Vm DC
G 6Vm – 8Vm Oscillatory
H 8Vm DC

Let, Vm be the maximum value of the transformer secondary voltage. When the diode D1
conducts, C1 is charged to + Vm. When D1 is turned off, the potential of A oscillates
between 0 to 2Vm as transformer secondary voltage varies from –Vm to +Vm. When D1 is
off, D2 will turn on and C2 will be charged upto +2Vm. When D3 is on, then D1 is on and
D2 is off. Then the potential of A is zero and thus C3 gets charged to 2Vm from C2. When
D1 and D3 are off and D2 and D4 are on. Then the potential of C oscillates between 2Vm
to 4Vm and C4 gets charged upto 2Vm. Since, the potential of B is 2Vm w.r.t. earth, hence,
the potential of D is 4Vm w.r.t earth. In this way C6 and C8 are also charged to 2Vm each
and so 8Vm may be obtained as the output DC voltage.
Practically, number of stage more than 4 is not used because (i) very high
insulation is required for higher stages and (ii) the diodes are not ideal and identical
which causes many problems.

Symmetric Voltage Doubler Circuit:

28
 The symmetric voltage doubler circuit was proposed by Allibone in 1934. The
HV transformer feeds two half-wave rectifiers and two storage capacitors are connected
in series. When the diode D1 conducts in one half-cycle, then C1 is charged to +Vm w.r.t.
earth, where Vm is the peak value of the transformer secondary voltage. In the other half
cycle, D2 conducts and D1 is off. Then C2 is charged to +Vm, so that +2Vm is obtained as
output. When the diodes are turned off, then potential across the diodes vary from 0 to
2Vm. Hence, the PIV of the diodes must be 2Vm. As the diodes conduct identically in the
two half cycles it is called symmetric voltage doubler circuit and there is no problem of
transformer core saturation. In this circuit neither terminal of the transformer secondary
could be earthed and the potential of one terminal of transformer secondary is a steady
DC of magnitude +Vm. Hence, the insulation between the primary and the secondary of
the transformer should be for a dc voltage +Vm.

Multi-stage Circuit:

In one half cycle, D1 and D3 conduct and C1 and C3 are charged to +Vm each. In
the other half cycle, D2 and D4 conduct and C2 and C4 are charged to +Vm each. Since, all
the four storage capacitors are connected in series, +4Vm is obtained as output. Since, one
terminal of the secondary of transformer T2 is at a steady DC potential +3Vm, insulation
for 3Vm need to be provided between the primary and secondary of T2. As it would be
difficult to provide such high insulation within a single transformer, an isolating
transformer T3 is used that feeds T2. Thus, this need for additional isolating transformers
makes the use of more than two stages uneconomical.

29
Improved Symmetric Voltage Multiplier Circuit:

An essential improvement is made when the different stages are excited by

specially designed cascaded transformer. Every transformer per stage consists of a LV
primary (1), HV secondary (2) and LV tertiary (3) winding. The tertiary winding feeds
the primary of the transformer in the next higher stage. The necessary dc insulation
within each transformer T1, T2 etc can be subdivided within the transformers. However, a
problem with this circuit is that the lower stage transformers have to supply the energy to
the upper stage units.

Types of Rectifiers:

Two types of rectifiers are used in practice, i) vacuum tube diode and ii) solid-
state rectifiers.

Vacuum Tube Diode:

In vacuum tube filament is used to raise the temperature of cathode so that

electrons are emitted and are available for conduction when a forward voltage is applied
between A and C. Normally, C1 is charged with +Vm, where Vm is the peak value of the
HT transformer secondary voltage. Now, F and C are equivalent to two electrodes with a
capacitance CCF between them. Again, there is a capacitance CFT between the primary
and secondary of the filament transformer, whose one terminal N is at close to earth
potential.

30
Thus, the C1–C–F–FT equivalent circuit becomes

Hence, maximum potential drop occurs across CCF. Now if the insulation between C and
F is insufficient to withstand this voltage, then breakdown starts from C to F. That is why
C and F are connected electrically. In that case the total voltage appears across CFT.
Hence, the insulation between primary and secondary of Filament Transformer must be
designed to withstand this high voltage and hence becomes expensive.
If during testing the test object fails, then the full voltage appears across the
rectifier. But, vacuum tube is capable of withstanding this as the current through the tube
saturates after a certain voltage. This is known as self-protecting property of vacuum
tubes.

Solid State Rectifiers:

Two types of solid state diodes are used – i) Silicon and ii) Selenium diodes.
Solid state diodes have PIV of about 1kV (continuous) and exhibits surge rating
which is about 4 to 10 times the continuous rating.
To withstand the large PIV across the diodes during testing, large no. of solid
state diodes are connected in series.

V
No. of diodes in series   1.2
PIV

31
 The multiplying factor 1.2 is needed because of the following reason. Even when
the diodes are of the same batch, their reverse bias voltage rating vary, though they match
in forward biased condition. Hence, the potential distribution will not be uniform in
reverse biased condition and so a safety margin of 20% is required for reliable operation.

The time required to change from forward biased to reverse biased condition is
called clearing time. When a large no. of solid state diodes are connected in series, their
clearing times are never identical. So, in reverse biased condition some diodes may
remain on for some time when the others are off. During this time the entire PIV is to be
supported by those diodes, which are off. This often leads to flashover of diodes.

However, when a vacuum tube diode fails, it has to be replaced in full and hence
it is costly. But, in the case of solid state rectifier bank, the entire bank is never damaged.
Generally, 30 – 40% of the diodes get damaged. Hence, replacement cost is less.

Comparison between Vacuum Tube Diode (VTD) and Solid State Diode (SSD):

i) A VTD requires filament for electron emission, which is not needed for a
SSD.

ii) A VTD has large PIV, whereas a SSD has a low PIV. Typically 30 – 50V
for Se and 1 – 2 kV for Si. Hence, a large number of SSDs are connected
in series.

iii) Failure of a VTD results in the collapse of the test circuit. But, the failure
of a SSD does not result in total collapse, as there are many SSDs in
series.

iv) A VTD can sustain overload, but a SSD cannot do so.

v) In a VTD, the current saturates after a certain value of voltage. So, a VTD
has self-current limiting characteristics. But, for a SSD current increases
exponentially with voltage. Thus, a current limiting element must be used
along with SSD bank.

vi) SSD bank is susceptible to failure under transient voltage due to non-
uniform potential distribution along with bank. No such problem arises
with a VTD.

vii) A SSD has a lower forward voltage drop than a VTD.

viii) Temperature rise of SSD is lower. This along with the reliability of SSD,
allows SSD to be used in sealed oil pack. A SSD bank in oil requires less
space than VTD.

32
Comparison between Selenium and Silicon Rectifiers:

A Si rectifier has higher PIV and current rating. But the inverse characteristics
and clearing time are not identical for all Si diodes. So, in the case of Si diode banks it
may so happen that some are off and some are on during clearing. Then the full PIV is
applied across those diodes that are off resulting in flashover.

Se diodes have lower PIV and current rating. But, their inverse characteristics are
identical and their clearing times are also equal. That is why Selenium diodes are often
preferred over Silicon diodes for high voltage DC generation.

33