ECE 8830 - Electric Drives

Topic 10:

Cycloconverters

Spring 2004

Introduction
Cycloconverters directly convert ac signals of
one frequency (usually line frequency) to ac
signals of variable frequency. These variable
frequency ac signals can then be used to
directly control the speed of ac motors.
Thyristor-based cycloconverters are typically
used in low speed, high power (multi-MW)
applications for driving induction and wound
field synchronous motors.

Phase-Controlled Cycloconverters
The basic principle of cycloconversion is
illustrated by the single phase-to-single
phase converter shown below.

Phase-Controlled Cycloconverters
(contd)
A positive center-tap thyristor converter
is connected in anti-parallel with a
negative converter of the same type.
This allows current/voltage of either
polarity to be controlled in the load.
The waveforms are shown on the next
slide.

Phase-Controlled Cycloconverters
(contd)

Phase-Controlled Cycloconverters
(contd)
An integral half-cycle output wave is created which
has a fundamental frequency
f 0=(1/n) fi where n
is the number of input half-cycles per half-cycle of
the output. The thyristor firing angle can be set to
control the fundamental component of the output
signal. Step-up frequency conversion can be
achieved by alternately switching high frequency
switching devices (e.g. IGBTs, instead of thyristors)
between positive and negative limits at high
frequency to generate carrier-frequency modulated
output.

Phase-Controlled Cycloconverters
(contd)
3 to single phase conversion can be
achieved using either of the dual
converter circuit topologies shown below:

Phase-Controlled Cycloconverters
(contd)
A Thevenin equivalent circuit for the dual
converter is shown below:

Phase-Controlled Cycloconverters
(contd)
The input and output voltages are adjusted to be
equal and the load current can flow in either
direction. Thus,

V0 Vd Vd 0 cos p Vd 0 cos n
where Vd0 is the dc output voltage of each
converter at zero firing angle and p and N are
the input and output firing angles. For a 3 halfwave converter Vd0 =0.675VL and Vd0 = 1.35VL for
the bridge converter (VL is the rms line voltage).

Phase-Controlled Cycloconverters
(contd)
Voltage-tracking between the input and
output voltages is achieved by setting the
sum of the firing angles to . Positive or
negative voltage polarity can be achieved
as shown below:

Phase-Controlled Cycloconverters
(contd)
A 3 to 3 cycloconverter can be
implemented using 18 thyristors as
shown below:

Phase-Controlled Cycloconverters
(contd)
Each phase group functions as a dual
converter but the firing angle of each
group is modulated sinusoidally with 2/3
phase angle shift -> 3 balanced voltage
at the motor terminal. An inter-group
reactor (IGR) is connected to each phase
to restrict circulating current.

Phase-Controlled Cycloconverters
(contd)
An output phase wave is achieved by
sinusoidal modulation of the thyristor
firing angles.

Phase-Controlled Cycloconverters
(contd)
A variable voltage, variable frequency
motor drive signal can be achieved by
adjusting the modulation depth and output
frequency of the converter.
The synthesized output voltage wave
contains complex harmonics which can be
adequately filtered out by the machines
leakage inductance.

Phase-Controlled Cycloconverters
(contd)
A 3 to 3 bridge cycloconverter
(widely used in multi-MW applications)
can be implemented using 36 thyristors
as shown below:

Phase-Controlled Cycloconverters
(contd)
The output phase voltage v0 can be written as:

v0 2V0 sin 0t

where V0 is the rms output voltage and 0 is

the output angular frequency. We can also
write:

v0 Vd 0 cos p Vd 0 cos n m f Vd 0 sin 0t

where the modulation factor, mf is given by:

m f 2V0 / Vd 0

Phase-Controlled Cycloconverters
(contd)
From these equations, we can write:

p cos 1[m f sin 0and

t]

N P

Thus for zero output voltage, mf=0 and

P= N= /2. For max. phase voltage,

mf=1 => P=0, N= . See below figure

for P and N values for mf=0.5 and 1.

Phase-Controlled Cycloconverters
(contd)
The phase group of a cycloconverter can be operated in
two modes:
1) Circulating current mode
2) Non-circulating current (blocking) mode
In the circulating current mode, the current continuously
circulates between the +ve and -ve converters. Although
the fundamental output voltage waves of the individual
converters are equal, the harmonics will cause potential
difference which will result in short-circuits without an
IGR.

Phase-Controlled Cycloconverters
(contd)
The equivalent circuit of a phase group with an IGR is
shown below.

The inclusion of an IGR leads to self-induced

circulating current as illustrated in the next slide.

Phase-Controlled Cycloconverters
(contd)

Phase-Controlled Cycloconverters
(contd)
At t=0, +ve load current is taken by the +ve converter
only (iP=i0).
From 0->/2, rising +ve load current will create a +ve
voltage drop (vL=Ldi0/dt) in the primary winding of the
IGR. This creates a -ve voltage drop in the secondary
winding of the IGR
-> D N reverse biased. no
current flow in the -ve converter.
At /2, i0 peaks at Im-> vL=0. After this vL tends to
reverse polarity inducing current in the -ve converter.
Voltage across IGR becomes clamped to 0 -> self-induced
circulation current between +ve and -ve converters.

Phase-Controlled Cycloconverters
(contd)
The +ve and -ve converter currents can be
expressed as:

iP 0.5 I m 0.5I m sin 0t

iN 0.5 I m 0.5 I m sin 0t

The self-induced circulating current is simply iP-iN.

In practice, the waves will not be pure sine waves
but include a ripple current. Practical waveforms
are shown on the next slide.

Phase-Controlled Cycloconverters
(contd)

Phase-Controlled Cycloconverters
(contd)

Advantages of circulating current mode of

operation, over blocking mode include:
Output phase voltage wave has lower harmonic
content than in blocking mode.
Output frequency range is higher.
Control is simple.
Disadvantages
Bulky IGR increases cost and losses.
Circulating current increases losses in thyristors.
Over-design increases cost.

Phase-Controlled Cycloconverters
(contd)
In the blocking mode of operation, no IGR
is used and only one converter is
conducting at any time.
Zero current crossing detection can be
used to select +ve or -ve converter
conduction as shown below:

Phase-Controlled Cycloconverters
(contd)
Since the cycloconverter is usually connected
directly to a motor, the harmonics from the
converter will induce torque pulsations and
machine heating resulting in increased machine
losses. Also, since the cycloconverter is
essentially a matrix of switches without energy
storage (neglecting IGR) Pin=Pout . Thus
distortions in the output voltage waveform
reflect back into the line input. See text for a
discussion of the load voltage and line
harmonics.

Phase-Controlled Cycloconverters
(contd)
A major disadvantage of cycloconverters
is poor DPF (displacement power factor).
To calculate DPF, consider a phase group
of an 18-thyristor cycloconverter shown
below.

Phase-Controlled Cycloconverters
(contd)
Assume the +ve converter is operating in
continuous conduction and is connected to a
high inductance load and assume that the
cycloconverter is operating at low frequency.
Segments of the output current and voltage
waves are as shown below:

Phase-Controlled Cycloconverters
(contd)
The Fourier series of the line current is
given by:
i

i0
3
1
1
1

i0 [sin( t P ) cos 2( t P ) cos 4( t P ) sin 5( t P ) ...]

3
2
4
5

where i0 is the load current and is the

supply frequency. The current wave has a
dc component and a fundamental
component with a lagging phase angle, P.

Phase-Controlled Cycloconverters
(contd)
Since the supplys active and reactive power
components are contributed only by the fundamental
current, the instantaneous active P i and reactive
power Qi for the positive converter is given by:

Pi 3Vs
'

3i0
cos P
2

3i0
where Vs =rms' line voltage.
Qi 3Vs

sin P
2

Phase-Controlled Cycloconverters
(contd)
These equations can be rewritten as:

Pi ' (1.17Vs cos P )i0 Vd 0 cos P i0 v0i0

Qi' (1.17Vs sin P )i0 Vd 0 sin P i0

If the firing angle is constant, the converter acts as a

rectifier and Vd=Vd0cosP and i0=Id.
In a cycloconverter P and i0 vary sinusoidally and so
Pi and Qi are also modulated. We need to average
these parameters to determine loading on the source.

Phase-Controlled Cycloconverters
(contd)

Phase-Controlled Cycloconverters
(contd)
The expression for the average reactive power contributed by
the line, Qi, is given by:

1
Qi

1
cycle
2

(1.17Vs sin P )i0 d 0t

1
where
Vs cos
0t )(power
I m sin( 0factor
t )) dangle.
load
(1.17=
0t ( 1.17Vs cos 0 t )( I m sin( 0t )) d 0t

/2
Performing the integration above
yields:

/2

where P0, Q0 are the real and reactive output power per phase,
2 P0
2Q0
1

2
respectively.
Qi
cos

sin
2

Phase-Controlled Cycloconverters
(contd)
P0 and Q0 are given by:

P0 V0 I 0 cosand

Q0 V0 I 0 sin

Since the real output power = real input power, we

can write:

2
1

2
Pi jQi Pi j Pi cos Q0 sin 2
The input
DPF canbe

expressed as: 2
DPF = cosi =

Pi
Pi jQi

Phase-Controlled Cycloconverters
(contd)
DPF =

1
Q0
2
1

2
1 j cos sin 2

Pi
2

1
1 j

2
1

2
cos

tan

sin
2

1
1 j

2
(1 tan )

where tan = Q0/Pi (=Q0/P0)

Phase-Controlled Cycloconverters
(contd)
This equation for DPF applies when additional
phase groups are added or if a 36-thyristor
implementation is considered.
mf=1 was assumed in this derivation.
For
mf 1:
DPF

mf
2
1 j (1 tan )

The maximum value of line DPF is 0.843.

Phase-Controlled Cycloconverters
(contd)

See Bose text pp. 180-184 for methods

to improve DPF.

Phase-Controlled Cycloconverters
(contd)
The control of a cycloconverter is very
complex. A typical variable speed constant
frequency (VCSF) system is shown below:

Phase-Controlled Cycloconverters
(contd)
Generator bus with regulated voltage but
variable frequency (1333-2666 Hz) is fed to
the cycloconverter phase groups. (A generator
speed variation of 2:1 is assumed). The dual
converter in each phase group uses a lowpass filter to generate a sinusoidal signal.
modulator receives biased cosine waves
from generator bus voltage and sinusoidal
control signal voltages to generate thyristor
firing angles.

Phase-Controlled Cycloconverters
(contd)
3 sinusoidal control signals are generated
through the vector rotator. The feedback
voltage Vs is generated from the output
phase voltages.

Phase-Controlled Cycloconverters
(contd)
Details of modulator are shown below:

Matrix Converters
These types of cycloconverters use highfrequency, self-controlled ac switches (e.g.
IGBTs). A 3 to 3 converter is shown
below:

Matrix Converters (contd)

A matrix of nine switches where any input
phase can be connected to any output
phase. The switches are controlled by
PWM to fabricate an output fundamental
voltage whose amplitude and frequency
can be varied to control an ac motor.
The output waveform synthesis is shown
on the next slide.

Matrix Converters (contd)

Matrix Converters (contd)

Matrix converters offer the advantage
over thyristor cycloconverters of being
able to produce unity PF line current.
However, compared to PWM voltage-fed
converters, the parts count is
significantly higher.

High-Frequency Cycloconverters

See Bose text pp. 186-189