6.

0 AC–AC CONVERTERS
6.1 INTRODUCTION

A power electronic ac–ac converter, in generic form, accepts electric power from one system and
converts it for delivery to another ac system with waveforms of different amplitude, frequency,
and phase. They may be single-phase or three-phase types depending on their power ratings. The
ac–ac converters employed to vary the rms voltage across the load at constant frequency are
known as ac voltage controllers or ac regulators. The voltage control is accomplished either by
(1) phase control under natural commutation using pairs of silicon-controlled rectifiers (SCRs)
or triacs or (2) by on/off control under forced commutation/self-commutation using fully
controlled self-commutated switches. The ac–ac power converters in which ac power at one
frequency is directly converted to ac power at another frequency without any intermediate dc
conversion link (as in the case of inverters) are known as cycloconverters, the majority of which
use naturally commutated SCRs for their operation when the maximum output frequency is
limited to a fraction of the input frequency.

Although typical applications of ac voltage controllers include lighting and heating control,
online transformer tap changing, soft-starting, and speed control of pump and fan drives, the
cycloconverters are mainly used for high power, low-speed, large ac motor drives for application
in cement kilns, rolling mills, and ship propellers. The power circuits, control methods, and the
operation of the ac voltage controllers and cycloconverters are introduced in this chapter.

6.2 SINGLE PHASE AC-AC CONTROLLER

The basic power circuit of a single-phase ac–ac voltage controller, as shown in Fig. 6.1(a),
comprises a pair of SCRs connected back-to-back (also known as inverse-parallel or anti-parallel)
between the ac supply and the load. This connection provides a bidirectional full-wave
symmetrical control, and the SCR pair can be replaced by a triac (Fig. 6.1(b)) for low-power
applications. Alternate arrangements are shown in Fig. 6.1(c) with two diodes and two SCRs to
provide a common cathode connection for simplifying the gating circuit without requiring
isolation and in Fig. 6.1(d) with one SCR and four diodes to reduce the device cost but with
increased device conduction loss.

With phase control, the switches conduct the load current for a chosen period of each input cycle
of voltage and with on/off control, the switches connect the load either for a few cycles of input
voltage and disconnect it for the next few cycles (integral cycle control) or the switches are
turned on and off several times within alternate half cycles of input voltage (ac chopper or pulse
width modulated [PWM] ac voltage controller).

1
Figure 6.1 Single-phase ac voltage controllers: (a) full wave, two SCRs in inverse-parallel; (b)
full wave with triac; (c) full wave with two SCRs and two diodes; (d) full wave with four diodes
and one SCR.

6.2.1 Phase-Controlled Single-phase AC Voltage Controller

For a full-wave symmetrical phase control, the SCRs 𝑇1 and 𝑇2 in Fig. 6.1a are gated at 𝛼 and
𝜋 + 𝛼, respectively, from the zero crossing of the input voltage; by varying 𝛼 , the power flow to
the load is controlled through voltage control in alternate half cycles. As long as one SCR is
carrying current, the other SCR remains reverse-biased by the voltage drop across the conducting
SCR. The principle of operation in each half cycle is similar to that of the controlled half-wave
rectifier, and one can use the same approach for analysis of the circuit.

Operation with R-load: Figure 6.2 shows the typical voltage and current waveforms for the
single-phase bidirectional phase-controlled ac voltage controller of Fig. 6.1(a) with a resistive
load. The output voltage and current waveforms have half-wave symmetry and so no dc
component.

If 𝑣𝑠 = 2𝑉𝑠 𝑠𝑖𝑛𝜛𝑡 is the source voltage, the rms output voltage with 𝑇1 triggered at 𝛼 can be
found from the half-wave symmetry as

1 𝜋 0.5 𝛼 𝑠𝑖𝑛 2𝛼 0.5

𝑉𝑜 = 𝛼
2𝑉𝑠2 𝑠𝑖𝑛2 𝜛𝑡𝑑(𝜛𝑡) = 𝑉𝑠 1 − + (6.1)
𝜋 𝜋 2𝜋

Note that 𝑉𝑜 can be varied from 𝑉𝑠 to 0 by varying 𝛼 from 0 to 𝜋.

2
The average value of output voltage is

1 𝜋 2𝑉𝑠
𝑉𝑑𝑐 = 2𝜋 𝛼
2𝑉𝑠 𝑠𝑖𝑛𝜛𝑡𝑑(𝜛𝑡) = 𝑐𝑜𝑠𝛼 + 1 (6.A)
2𝜋

Figure 6.2 Waveforms for single-phase ac full-wave voltage controller with R-load.

The rms value of load current, 𝐼𝑜 = 𝑉𝑜 𝑅 (6.2)

𝑃𝑜 𝑉0 𝛼 𝑠𝑖𝑛 2𝛼 0.5
The input power factor = = = 1− + (6.3)
𝑉𝐴 𝑉𝑠 𝜋 2𝛼

2𝑉𝑠
The average SCR current is 𝐼𝐴,𝑆𝐶𝑅 = 𝑐𝑜𝑠𝛼 + 1 (6.4)
2𝜋𝑅

Since each SCR carries half the line current, the rms current in each SCR is
𝐼𝑜
𝐼𝑂,𝑆𝐶𝑅 = (6.5)
2

3
Operation with RL Load: Figure 6.3 shows the voltage and current waveforms for the controller
in Fig. 6.1a with RL load. Due to the inductance, the current carried by the SCR 𝑇1 may not fall
to zero at 𝜛𝑡 = 𝜋 when the input voltage goes negative and may continue till 𝜛𝑡 = 𝛽, the
extinction angle, as shown.

The conduction angle,

𝜃 =𝛽−𝛼 (6.6)

of the SCR depends on the firing delay angle 𝛼 and the load impedance angle 𝜙

Figure 6.3 Typical waveforms of single-phase ac voltage controller with an RL load.

The expression for the load current 𝐼0 𝜔𝑡 , when conducting from α to β, can be derived in the
same way as that used for a phase-controlled rectifier in a discontinuous mode by solving the
relevant Kirchoff ’s voltage equation:

(6.7)

4
where 𝑍 = 𝑅 2 + 𝜔2 𝐿2 0.5
= load impedance and 𝜙 = load impedance angle = 𝑡𝑎𝑛−1 𝜛𝐿 𝑅 .

The angle 𝛽 , when the current 𝑖𝑜 falls to zero, can be determined from the following
transcendental equation resulted by using 𝑖𝑜 𝜛𝑡 = 𝛽 in Eq. (6.7).

(6.8)

From Eqs. (6.6) and (6.8), one can obtain a relationship between 𝜃 and 𝛼 for a given value of 𝜙,
as shown in Fig. 18.4, which shows that as 𝛼 is increased, the conduction angle 𝜃 decreases, and
thus the rms value of the current decreases.

Figure 6.4 θ versus α curves for single-phase ac voltage controller with RL load.

The rms output voltage

(6.9)

𝑉𝑜 can be evaluated for two possible extreme values of 𝜙 = 0 when 𝛽 = 𝜋, and 𝜙 = 𝜋/2 when
𝛽 = 2𝜋 − 𝛼.

5
The rms SCR current can be obtained from Eq. (6.7) as follows:

(6.10)

The rms load current, 𝐼𝑜 = 2𝐼𝑂,𝑆𝐶𝑅 (6.11)

1 𝛽
The average value of SCR current, 𝐼𝐴,𝑆𝐶𝑅 = 𝑖 𝑑(𝜛𝑡) (6.12)
2𝜋 𝛼 𝑜

Gating Signal Requirements: For the inverse-parallel SCRs as shown in Fig. 6.1(a), the gating
signals of SCRs must be isolated from one another because there is no common cathode. For R-
load, each SCR stops conducting at the end of each half cycle and under this condition, single
short pulses may be used for gating as shown in Fig. 6.2. With RL load, however, this single
short pulse gating is not suitable as shown in Fig. 18.6. When SCR 𝑇2 is triggered at 𝜛𝑡 = 𝜋 + 𝛼,
SCR 𝑇1 is still conducting due to the load inductance. By the time the SCR 𝑇1 stops conducting
at β, the gate pulse for SCR 𝑇2 has already ceased and 𝑇2 will fail to turn on resulting the
converter to operate as a single-phase rectifier with conduction of 𝑇1 only which increases the
dissipation in SCR gate circuit and a large. This necessitates the application of a sustained gate
pulse either in the form of a continuous signal for the half-cycle period, isolating pulse
transformer or better a train of pulses (carrier frequency gating) to overcome these difficulties.

Figure 6.5 Single-phase full-wave controller with RL load: gate pulse requirements
6
6.2.2 Single-Phase AC–AC Voltage Controller with On/Off Control
Integral Cycle Control: As an alternative to the phase control, the method of integral cycle
control or burst-firing is used for heating loads. Here, the switch is turned on for a time 𝑡𝑛 with n
integral cycles and turned off for a time 𝑡𝑚 with m integral cycles (Fig. 6.6). As the SCRs or
triacs used here are turned on at the zero crossing of the input voltage and turn off occurs at zero
current, supply harmonics and radio frequency interference are very low.

However, subharmonic frequency components may be generated, which are undesirable because
they may set up subharmonic resonance in the power supply system, cause lamp flicker, and may
interfere with the natural frequencies of motor loads causing shaft oscillations.

Figure 6.6 Integral cycle control: (a) typical load voltage waveforms and (b) power factor with
the duty cycle k.

For sinusoidal input voltage, 𝑣 = 2𝑉𝑠 𝑠𝑖𝑛𝜛𝑡, the rms output voltage,

𝑉𝑜 = 𝑉𝑠 𝑘 (6.13)

where 𝑘 = 𝑛/ 𝑛 + 𝑚 = 𝑑𝑢𝑡𝑦 𝑐𝑦𝑐𝑙𝑒

and 𝑉𝑠 = 𝑟𝑚𝑠 𝑝𝑕𝑎𝑠𝑒 𝑣𝑜𝑙𝑡𝑎𝑔𝑒

The power factor = 𝑘 which is poorer for lower values of the duty cycle k.

6.3 THREE-PHASE FULL-WAVE CONTROLLERS

The unidirectional controllers, which contain dc input current and higher harmonic content due
to the asymmetric nature of the output voltage waveform, are not normally used in ac motor
drives; a three-phase bidirectional control is commonly used. The circuit diagram of a three-
phase full-wave (or bidirectional) controller is shown in Fig. 6.7 with a Y-connected resistive
load. The firing sequence of the thyristors is 𝑇1 , 𝑇2 , 𝑇3 , 𝑇4 , 𝑇5 , 𝑇6

7
 Figure 6.7 Three-phase bidirectional controller

If the instantaneous input phase voltages are defined as

𝑣𝐴𝑁 = 2𝑉𝑠 𝑠𝑖𝑛𝜔𝑡

2𝜋
𝑣𝐵𝑁 = 2𝑉𝑠 𝑠𝑖𝑛 𝜔𝑡 −
3

4𝜋
𝑣𝐶𝑁 = 2𝑉𝑠 𝑠𝑖𝑛 𝜔𝑡 −
3

The instantaneous input line voltages are

𝜋
𝑣𝐴𝐵 = 6𝑉𝑠 𝑠𝑖𝑛 𝜔𝑡 +
6
𝜋
𝑣𝐵𝐶 = 6𝑉𝑠 𝑠𝑖𝑛 𝜔𝑡 −
2
7𝜋
𝑣𝐶𝐴 = 6𝑉𝑠 𝑠𝑖𝑛 𝜔𝑡 −
6

The waveforms for the input voltages, conduction angles of thyristors, and output phase voltages
are shown in Fig. 6.8 for 𝛼 = 60° and 𝛼 = 120° .

8
 Figure 6.8 Waveforms for three-phase bidirectional controller

For 0 ≤ 𝛼 < 60° , immediately before the firing of 𝑇1 , two thyristors conduct. Once 𝑇1 is fired,
three thyristors conduct. A thyristor turns off when its current attempts to reverse. The conditions
alternate between two and three conducting thyristors.

For 60° ≤ 𝛼 < 90° , only two thyristors conduct at any time. For 90° ≤ 𝛼 < 150° , although two
thyristors conduct at any time, there are periods when no thyristors are on. For 𝛼 ≥ 150° , there is
no period for two conducting thyristors and the output voltage becomes zero at 𝛼 = 150° . The
range of delay angle is
0 ≤ 𝛼 ≤ 150°
Similar to half-wave controllers, the expression for rms output phase voltage depends on the
range of delay angles. The rms output voltage for a Y-connected load can be found as follows.
For 0 ≤ 𝛼 < 60° :

9
The power devices of a three-phase bidirectional controller can be connected together, as shown
in Fig. 6.9. This arrangement is also known as tie control and allows assembly of all thyristors as
one unit. However, this arrangement is not possible for motor control because the terminals of
the motor windings are not normally accessible.

Figure 6.9 Arrangement for three-phase bidirectional tie-control

10
Gating Sequence. The gating sequence is as follows:

1. Generate a pulse signal at the positive zero crossing of the supply phase voltage 𝑣𝑎𝑛 .
2. Delay the pulse by angles 𝛼 , 𝛼 + 2𝜋 3 , and 𝛼 + 4𝜋 3 for gating 𝑇1 , 𝑇3 and 𝑇5
respectively through gate-isolating circuits.
3. Similarly, generate pulses with delay angles 𝛼 + 𝜋, 𝛼 + 5𝜋 3, and 𝛼 + 7𝜋 3 for gating
𝑇2 , 𝑇4 and 𝑇6 respectively.

6.4 THREE-PHASE BIDIRECTIONAL DELTA-CONNECTED CONTROLLERS

If the terminals of a three-phase system are accessible, the control elements (or power devices)
and load may be connected in delta, as shown in Fig. 6.10. Because the phase current in a normal
three-phase system is only 1 3 of the line current, the current ratings of thyristors would be
less than that if thyristors would be less than that if thyristors (or control elements) were placed
in the line.

Figure 6.10 Delta-connected three-phase controller

Assuming that the instantaneous line-to-line voltages are

𝑣𝐴𝐵 = 𝑣𝑎𝑏 = 2𝑉𝑠 𝑠𝑖𝑛𝜔𝑡

2𝜋
𝑣𝐵𝐶 = 𝑣𝑏𝑐 = 2𝑉𝑠 𝑠𝑖𝑛 𝜔𝑡 −
3

4𝜋
𝑣𝐶𝐴 = 𝑣𝑐𝑎 = 2𝑉𝑠 𝑠𝑖𝑛 𝜔𝑡 −
3

The input voltages, phase and line currents, and thyristors gating signals are shown in Fig. 6.11
for 𝛼 = 120° and a resistive load.

11
Figure 6.11 Waveforms for delta-connected controller

12
For the resistive loads, the rms output phase voltage can be determined from

2𝜋 1/2 𝜋 1/2
1 2
2
𝑉𝑜 = 𝑣𝑎𝑏 𝑑 𝜔𝑡 = 2𝑉𝑠2 𝑠𝑖𝑛𝜔𝑡𝑑 𝜔𝑡
2𝜋 2𝜋
𝛼 𝛼
1/2
1 𝑠𝑖𝑛2𝛼
𝑉𝑜 = 𝑉𝑠 𝜋−𝛼+
𝜋 2

The maximum output voltage would be obtained when 𝛼 = 0, and the control range of the delay
angle is

0≤𝛼≤𝜋

The line currents, which can be determined from the phase currents, are

𝑖𝑎 = 𝑖𝑎𝑏 − 𝑖𝑐𝑎
𝑖𝑏 = 𝑖𝑏𝑐 − 𝑖𝑎𝑏
𝑖𝑐 = 𝑖𝑐𝑎 − 𝑖𝑏𝑐

Notice from Fig. 6.11 that the line currents depend on the delay angle and may be discontinuous.
The rms value of line and phase currents for the load circuits can be determined by numerical
solution or Fourier analysis. If 𝐼𝑛 is the rms value of the nth harmonic component of a phase
current, the rms value of phase current can be found from

𝐼𝑎𝑏 = 𝐼12 + 𝐼32 + 𝐼52 + 𝐼72 + 𝐼92 + 𝐼11

2
+ ⋯ + 𝐼𝑛2 1 2

Due to the delta connection, triplen harmonic components (i.e those of order 𝑛 = 3𝑚, where m is
an odd integer) of the phase currents would flow around the delta and would not appear in the
line. This is due to the fact that the zero-srquence harmonics are in phase in all three phases of
load. The rms line current becomes

𝐼𝑎 = 3 𝐼12 + 𝐼52 + 𝐼72 + 𝐼11

2
+ ⋯ + 𝐼𝑛2 1 2

As a result, the rms value of line current would not follow the normal relationship of a three-
phase system such that

𝐼𝑎 < 3𝐼𝑎𝑏

An alternative form of delta-connected controllers that requires only three thyristors and
simplifies the control circuitry is shown in Fig.6.12. This arrangement is also known as a neutral-
point controller.

13
 Figure 6.12 Three-phase three-thyristor controller

6.5 DESIGN OF AC VOLTAGE-CONTROLLER CIRCUITS

The ratings of power devices must be designed for the worst-case condition, which occurs when
the converter delivers the maximum rms value of output voltage 𝑉𝑜 . The input and output filters
must also be designed for worst-case conditions. The output of power controller contains
harmonics, and the delay angle for the worst-case condition of a particular circuit arrangement
should be determined. The steps involved in designing the power circuits and filters are similar
to those of rectifier circuit design.

6.6 CYCLOCONVERTERS
The ac voltage controllers provide a variable output voltage, but the frequency of the output
voltage is fixed and in addition the harmonic content is high, especially at a low output voltage
range. A variable output voltage at variable frequency can be obtained from two-stage
conversions: fixed ac to variable dc (e.g. controlled rectifiers) and variable dc to variable ac at
variable frequency (e.g. inverters). However, cycloconverters can eliminate the need of one or
more intermediate converters. A cycloconverter is a direct-frequency changer that converts as
power at one frequency to ac power at another frequency by ac-ac conversion, without an
intermediate conversion link.

The majority of cycloconverters are naturally commutated and the maximum output frequency is
limited to a value that is only a fraction of the source frequency. As a result the major
applications of cycloconverters are low-speed ac motor drives ranging from 15,000 kW with
frequencies from 0 to 20 Hz.

6.7 SINGLE-PHASE CYCLOCONVERTERS

The principle of operation of single-phase/single-phase cycloconverters can be explained with
the help of Fig. 6.13(a). The two single-phase controlled converters are operated as bridge
rectifiers. However, their delay angles are such that the output voltage of one converter is equal
and opposite to that of the other converter. If converter P is operating alone, the average output
voltage is positive and if converter N is operating alone, the output voltage is negative. Fig.
6.13(b) shows the simplified equivalent circuit of the dual converter. Fig. 6.13(c) shows the
waveforms for output voltage and gating signals of positive and negative converters, with the
positive converter on for time 𝑇0 2 and the negative converter operating for time 𝑇0 2 . The
frequency of the output voltage is 𝑓𝑜 = 1 𝑇0 .

14
Figure 6.13 Single-phase/single-phase cycloconverter

15
If 𝛼𝑝 is the delay angle of the positive converter, the delay angle of the negative converter is
𝛼𝑛 = 𝜋 − 𝛼𝑝 . The average output voltage of the positive converter is equal and opposite to that
of the negative converter.

𝑉𝑑𝑐 2 = −𝑉𝑑𝑐 1

Gating sequence. The gating sequence is as follows:

1. During the first half period of the output frequency 𝑇0 2, operate converter P as a normal
controlled rectifier with a delay angle of 𝛼𝑝 = 𝛼, that is, by gating 𝑇1 and 𝑇2 at 𝛼 and
gating 𝑇3 and 𝑇4 at 𝜋 + 𝛼.
2. During the second half period of the output frequency 𝑇0 2, operate converter N as a
normal controlled rectifier with a delay angle of 𝛼𝑛 = 𝜋 − 𝛼, that is, by gating 𝑇1′ and 𝑇2′
at 𝜋 − 𝛼 and gating 𝑇3′ and 𝑇4′ at 2𝜋 − 𝛼

6.8 THREE-PHASE CYCLOCONVERTERS

The circuit diagram of a three-phase/single-phase cycloconverter is shown in Fig. 6.14(a). The
two ac-dc converters are three-phase controlled rectifiers. The synthesis of output waveform for
an output frequency of 12 Hz is shown in Fig. 6.14(b). The positive converter operates for half
the period of output frequency and the negative converter operates for the other half period. The
analysis of this cycloconverter is similar to that of single-phase/single-phase cycloconverters.

The control of ac motors requires a three-phase voltage at a variable frequency. The

cycloconverter in Fig. 6.14(a) can be extended to provide three-phase output by having 6 three-
phase converters, as shown in Fig. 6.15(a). Each phase consists of 6 thyristors, as shown in Fig.
6.15(b), and a total of 18 thyristors are required. If six full-wave three-phase converters are used,
36 thyristors would be required.

Gating sequence. The gating sequence is as follows:

1. During the first half period of the output frequency 𝑇0 2, operate converter P as a normal
three-phase controlled rectifier with a delay angle of 𝛼𝑝 = 𝛼.
2. During the second half period of the output frequency 𝑇0 2, operate converter N as a
normal controlled three-phase rectifier with a delay angle of 𝛼𝑛 = 𝜋 − 𝛼.

16
Figure 6.14 Three-phase/single-phase cycloconverter

17
 Figure 6.15 Three-phase/three-phase cycloconverter

6.9 REDUCTION OF OUTPUT HARMONICS

From Figures 6.13(c) and 6.14(b) it can be noticed that the output voltage is not purely sinusoidal,
and as a result the output voltage contains harmonics. The power factor depends on the delay
angle of thyristors and is poor, especially at low output voltage range.

The output voltage of cycloconverters is basically made up of segments of input voltages and the
average value of a segment depends on the delay angle for that segment. If the delay angles of
segments were varied in such a way that the average values of segments correspond as closely as
possible to the variations of desired sinusoidal output voltage, the harmonics on the output
voltage can be minimzed. Equation (6.A) indicates that the average output voltage of a segment
is a cosine function of delay angle. The delay angle for the segments can be generated by
comparing a cosine signal at source frequency 𝑣𝑐 = 2𝑉𝑠 𝑐𝑜𝑠𝜔𝑠 𝑡 with an ideal sinusoidal
reference voltage at the output frequency 𝑣𝑟 = 2𝑉𝑠 𝑠𝑖𝑛𝜔0 𝑡 . Fig. 6.16 shows the generation of
gating signals for the thyristors of the cycloconverters in Fig. 6.14(a).

18
 Figure 6.16 Generation of thyristor gating signals

The maximum average voltage of a segment (which occurs for 𝛼𝑝 = 0) should be equal to the
peak value of output voltage; for example,

2 2𝑉𝑠
𝑉𝑝 = = 2𝑉𝑜
𝜋

which gives the rms value of the output voltage as

2𝑉𝑠 2𝑉𝑝
𝑉𝑜 = =
𝜋 𝜋

19