79

“Comparison Analysis of AC Voltage Controllers Based

on Experimental and Simulated Application Studies” Hamdy. A. Ashour and Rania. A. Ibrahim

Arab Academy for Science & Technology

Department of Electrical & Computer Control Engineering, 1029 Miami, Alexandria, Egypt

hashour@aast.edu, rania_assem@alexseeds.com

Abstract- This paper introduces a detailed comparison between possible connections of AC voltage controllers. For each

configuration, the experimental setup is implemented and the

corresponding simulation program is presented using Simulink

under Matlab. The simulated and experimental instantaneous

voltage and current waveforms in case of resistive and inductive

loads are matched well, validating the simulation comparison

for analysis. The comparison analysis includes the required

number of devices and isolated gate signals, which determines

the complexity and the size, hence the overall cost. Also

harmonic spectrum, total harmonic distortion, effective rms

value, dc offset and the control range are compared to specify

the performance. The implementation of a fixed- capacitor

thyristor controlled reactor (FC-TCR) and three phase

induction motor starters (SOFT STARTER) as two application

case studies of AC voltage regulators has been discussed.

Experimental and simulation results have been obtained and

well correlated, showing the effectiveness of such configurations

in the fields of control of reactive power flow and in the field of

controlling the starting performance of three phase induction

motor.

LIST OF SYMBOLS:

rmsoV : rms output load voltage

sV : rms supply voltage

n , m : Number of ON & OFF cycles

k : Duty cycle

α : Delay angle

β : Extinction angle

Θ : Load angle

δ : Conduction angle ( αβδ −= )

TCRFCI − , TCRFCV − : FC-TCR current and voltage

TCRFCB − : Compensator susceptance

cB : Capacitor susceptance

TCRB : Inductor susceptance as a function of

delay angle

maxB : Maximum inductor susceptance

ω : Angular frequency

C , L : Capacitance & inductance value

I. INTRODUCTION:

The AC voltage controller can be considered as a voltage

regulator device by which the root mean square load voltage

(rms), hence the power flow, can be set and maintained

constant at a certain desired value. The recent developments

achieved in the field of power electronics, control techniques

and microprocessors have introduced such AC voltage

controllers for the applications of power ranges from few

watts up to fractions of megawatts, such as light dimmers,

heating, melting, arc furnace, transformer tap changing,

cycloconverters, wind turbines, power factor improvement,

flexible transmission systems (FACTS), static switches, AC

motors control and operation [1-13]. The operation of the AC

voltage controllers have been explained in literatures [5-8].

This paper introduces a comparison between eleven possible

configurations of the AC voltage controllers based on

experimental and simulation analysis. Principle of operation

is reviewed, experimental setup and software simulation are

introduced, a detailed comparison has been carried out and

two different application case studies have been discussed for

practical validation.

II. AC VOLTAGE CONTROLLERS:

If a thyristor switch is connected between an AC supply

and load, the power flow can be controlled by varying the

rms value of the AC voltage applied to the load. This type of

power circuit is known as an AC voltage controller

(regulator). Since the input voltage is AC, the thyristor is line

commutated, so there is no need of extra commutation

circuitry and the circuits for AC voltage controllers are

simple and relatively inexpensive. AC switch can be

implemented either using a single triac with a single isolated

gate circuit but for lower power applications, using two back

to back thyristors with two isolated gate circuits, using two

diodes and two thyristors with a single isolated gate circuit,

or with four diodes and a single thyristor with a single

isolated gate circuit. The power flow to the load can be

controlled by the ON-OFF or phase angle control techniques.

For the ON-OFF control, the rms output voltage for resistive

load can be expressed as [6]:

( )

( ) kVnm

nVttdV

mn

nV sssrmso =

+=

∫

+=

2

1

2

0

2sin22

πωω

π (1)

While for the phase control rms

oV can be expressed as:

( ) 2

1

2

1

2

0

22

2

2sin1sin2

2

2

+−=

∫=

ααπ

πωω

π

πssrmso VttdVV (2)

The ON-OFF type of control can be applied in applications

having mechanical inertia and high thermal time constant

such as industrial heating and speed control of small motors,

while the phase control can be utilized in many industrial

applications such as motor starters, transformer tap changing

and static VAR compensators. In case of inductive load, the

current will not be in phase with the controlled voltage. In

this case, in order to ensure full control of the AC voltage, a

single gate pulse should be replaced with continuous train of

pulses and the range of delay angle α is limited within the

range of:

80

παθ ≤≤ (3)

While (2) maybe then re- written as:

( ) 2

1

2

1

22

2

2sin

2

2sin1sin2

2

2

++=

∫=

βαδ

πωω

π

β

αssrmso VttdVV (4)

The AC voltage controllers can be configured to be used

either in the single phase low power domestic applications, or

in the three phase high power industrial applications.

Different possible configurations of the AC voltage

controllers are depicted in fig. 1 and will be compared

through the next sections.

III. EXPERIMENTAL SETUP AND SIMULATION SOFTWARE:

A general block diagram for the experimental setup is shown

in fig. 2-a while an example of practical connections is shown

in fig. 2-b. The setup is built in a modular form and consists

of a variable power supply, a variable RLC load bank, a

synchronizing and isolating firing gate signals, a controlled

set point (delay angle α) and a group of individual diodes and

thyristors power electronic devices. The setup is reliable and

flexible to be reconnected to get different configurations

shown in fig. 1.

The simulation analysis has been carried out using Simulink

under Matlab version 6.5 which provides strong power

electronics and analysis toolboxes. The system is simulated in

a modular form, as shown in fig. 3, typical to the

experimental block diagram for clear comparison and has an

additional block for harmonic analysis. This simulation block

calculates the Fourier coefficients, the total harmonic

distortion (THD), the effective rms value, the DC offset

component and then plots the harmonic spectrum as will be

demonstrated in the next sections.

IV. COMPARISON ANALYSIS:

For each configuration in fig. 1, the experimental setup is

reconnected and the experimental waveforms are obtained

using storage scope, then the simulation is reconfigured to

obtain the corresponding simulated waveforms for

comparison and analysis. An example of the experimental

connections and the corresponding Simulink simulated

program for one of the possible three phase configurations

are illustrated in fig. 2-b and (3-b) respectively. For the single

phase configuration, the gate signals (1, 2) are shifted by 180º

while for the three phase configurations the gate (1, 2, 3, 4, 5,

6) are shifted by 60º. For all configurations, the comparison is

carried out for load voltage waveforms at delay angle α=108º

and in two cases: unity power factor (resistive load) and 0.6

lagging power factor (resistive and inductive load).

The comparison depicted includes the followings:

A. Experimental and Simulation Waveforms:

These waveforms shown in fig. 4 are for clear comparison

and simulation validation. The scales of time, voltage and

current of the experimental waveforms are typical for these

shown in the corresponding simulation graphs. A good

agreement between the simulated and experimental

waveforms can be seen in fig. 4 for different configurations.

The slight differences noticed between waveforms,

particularly for inductive loads, could be due to the difference

between switching performance of the actual and simulated

devices.

B. Number of Devices:

Lower number reduces the cost; hence the cheapest

configuration is the single phase configuration (a) while the

cheaper one in the three phase is configuration (k). However,

the SCR current rating of configuration (k) should be 2

higher than the other three phase configurations. For three

phase, since the devices are connected at the phases for

1T

1DsV

oI+

−

Load

(a)

1T

2TsV

oI+

−

Load

(b)

1T

3T4T

5T6T

A

B

C

Load

Load

Load

ABV

BCV

CAV

AI

BI

CI

2T

(c) 1T

2D

3T4D

5T 6D

A

B

C

n

Load

Load

Load

(d)

1T

2T

3T

4T

5T6T

A

B

C

Load

Load

Load

(e)

1T

3T

5T

A

C

4D

6D

2D

B

Load

LoadLoad

(f)

1T

3T5T

A

C

B

4T

6T 2T

Load

Load

Load

(g)

1T4D

3T

6D

5T

2D

A

C

B

Load

Load

Load

(h)

Figure 1: Different circuit configurations for the implementation of AC voltage controllers

1T 4T

3T

6T

2T5T

A

B

C

Load

Load

Load

(i)

1T

3T

4T

6T

5T

A

C

B

2T

Load

Load

Load

(j)

6T1T

3T

A

C

B

Load

Load

Load

(k)

(a) General block diagram (b) Example of actual connections (conf. c)

Figure 2: Experimental set-up

(b) Example of simulation program (Conf. c)

Figure 3: Simulation software

(a) General block diagram

81

configurations (i, j, k), and not at the lines like others, they

could have lower current and higher voltage ratings.

C. Number of Isolated Gates:

Increasing this number complicates the circuits, increases

the size and the overall cost. Configuration (a) is the best

from this point of view as it needs only one gate signal, while

the three phase configurations (d, f, h, k) require only three

isolated signals rather than six required by the rest.

D. Number of Load Terminals:

For the three phase configurations (g, h, i, j, k), the six

terminals of the load should be available and not connected

as star or delta. This could limit the applications of these

configurations according to the nature of the available loads.

It is not the case for other configurations which require loads

with only three terminals and are suitable for most three

phase loads (connected or not connected as star or delta).

E. Effective rms Value and the DC Offset:

These values are calculated in per unit for the supply

voltage taken as a base voltage and for a certain delay angle

α=108º and the load impedance is the base for current

calculation. Due to the presence of diodes in configurations

(a, d, f, h) or the possible forced path through the ON SCR in

configuration (k), the output voltage and input current are

asymmetrical containing a DC component. This is very clear

in the single phase configuration (a) which also has a limited

range of control as the Vorms can be only varied from 0.7 p.u.

to 1 p.u. If there is a magnetic element, such as a transformer,

such DC component may cause saturation problems. For

these reasons, these configurations, named as unidirectional,

are more suitable for resistive loads, such as heating and

lighting applications. Configurations (b, c, e, g, i, j) are

bidirectional control and the waveforms are symmetrical

along the x-axis, containing no DC component. These

configurations are most suitable for AC motor controls and

power system applications.

F. Harmonic Spectrum and Total Harmonic Distortion

(THD):

Harmonics may cause problems particularly for motors

(negative torque) and power systems (resonance and noise

interference). Harmonics may be useful in some applications,

such as heating, since the effective rms may be increased.

Configurations (c, g) and (d, h) produce similar waveforms

respectively as seen from fig. 4-c & 4-d. From fig. 4-a & 4-b

it can be seen that the unidirectional configurations introduce

even and odd harmonics while the bidirectional

configurations introduce only odd harmonics due to the

symmetrical positive and negative parts of the waveforms.

Even harmonics may cause problems in motor applications. It

can be also seen that the inductive load increases the value of

harmonic components and THD values due to the distortion

in the waveforms. The triplen harmonics will be disappeared

in the line values for the delta connected loads

(configurations e, f, i). It should be noted that the control

range of the delay angle may change the voltage wave shapes

and fig. 6 depicts the different voltage wave shapes at

different delay angles using resistive load for clarity.

Reducing the delay angle α reduces the value of harmonic

components and the THD, since the output waveform tends

to be sine wave.

(d) Conf. d &h

(e) Conf. e

(f) Conf. f

(g) Conf. i

(a) Conf. a

(c) Conf. c &g

(b) Conf. b

82

α =40º

α=80º

α=120º

α=40º

α=80º

α=120º

α=40

α=80

α=120

V. APPLICATION CASE STUDIES:

Two different application case studies using AC voltage

controllers will be discussed through the following sections

(i) Conf. k Figure 4: Simulation and experimental load phase voltage and current waveforms

for different configurations. Graphs in sequence are: Left graphs: SIMULATION Right graphs: EXPERIMENTAL

1st = Voltage, R -load 1st = Voltage, R -load

2nd = Current, R- load 2nd = Current, R- load 3rd= Voltage, RL- load 3rd= Voltage, RL- load

4th= Current, RL- load 4th= Current, RL- load

Scales: Voltage: 50 V / div, Current: 1 A / div, Time: 0.01 sec/div

(Conf. c & g) (Conf. a) (Conf. b)

(Conf. d & h) (Conf. e) (Conf. f)

Figure 6: Effect of varying delay angle on output voltage for different AC

voltage controllers on R-load.

(Conf. j) (Conf. i) (Conf. k)

(h) Conf. j

(a) Conf. a

THD = 67.68%

RMS= 0.8 p.u

DC= -0.3 p.u

THD = 69.89%

RMS= 0.86 p.u

DC= -0.21 p.u

(b) Conf. b

THD = 87.2%

RMS=0.54 p.u

DC=0 p.u

THD = 129.1%

RMS= 0.67 p.u

DC= 0 p.u

(c) Conf. c & g

THD = 104.9%

RMS= 0.33 p.u

DC= 0 p.u

THD = 266.9%

RMS= 0.69 p.u

DC= 0 p.u

(d) Conf. d & h

THD = 82.82%

RMS= 0.68 p.u

DC= -1e-3 p.u

THD = 105%

RMS= 0.74 p.u

DC= -1e-3 p.u

(e) Conf. e

THD = 100.2%

RMS= 0.57 p.u

DC= 0 p.u

THD = 254%

RMS= 0.96 p.u

DC= 0 p.u

(f) Conf. f

THD = 82.95%

RMS= 82.95 p.u

DC= -7e-4 p.u

THD = 96.43%

RMS= 1.25 p.u

DC= -2e-3 p.u

(g) Conf. i

THD = 147.6%

RMS= 0.47 p.u

DC= 0 p.u

THD = 206.4%

RMS= 0.97p.u

DC= 0 p.u

(h) Conf. j

THD = 105.1%

RMS= 0.33 p.u

DC= 0 p.u

THD = 186.8%

RMS= 0.48 p.u

DC= 0 p.u

Figure 5: Harmonic analysis for output voltage for different configurations

Left: R- load Right: RL- load

(i) Conf. k

THD = 82.85%

RMS= 0.68 p.u

DC= -1e-3 p.u

THD = 99.8%

RMS= 0.76 p.u

DC= -2e-3 p.u

83

α= 125º (lagging) α= 160º (leading) α= 133º ≈ in phase

(b) Simulation Results

Figure 8: Effect of changing delay angle on the FC-TCR voltage and current waveforms

CH1 voltage: 20 V / div, CH2 current: 0.2 A / div, Time: 0.01 sec/div

(a) Experimental results α= 133º ≈ in phase

α= 160º (leading) α= 125º (lagging)

A. Static Power Factor Improvement (FC-TCR):

Recently, the AC voltage controllers have been utilized in

the field of power system quality and flexible AC

transmission FACTS [9-11]. Unlike traditional shunt reactive

elements, a fixed capacitor– thyristor- controlled reactor (FC-

TCR) is able to rapidly and smoothly supply or absorb

reactive power by controlling the firing delay angles of

thyristors. As shown in fig. 7, each branch has a fixed

capacitor and two anti- parallel thyristors controlled in series

with an inductor. For such configuration, (5) can be written

from [10].

( )L

BBB

CBBBB

jBVI

TCR

CTCRCTCRFC

TCRFCTCRFC

ωα

ππα

α

ω

1,2sin

122

,

maxmax =

−−=

=−=

=

−

−−

(5)

Fig. 7-c shows the operating characteristics and the

susceptance ( TCRFCB − ) of this type of compensator based on

(5) and it can be seen that VAR (reactive power) production

as well as VAR absorption is possible by varying the delay

angle of thyristors; hence the power factor changes from

leading to lagging. The firing gates of the thyristors are

synchronized with the capacitance voltage and can be varied

from 90º to 180º.

From (5) and for L= 448 mH, C= 18 µF, ω =314.15 rad/sec,

then:

if α=90º BFC-TCR = Bc - Bmax= (1.45 e-3

) mho.

or if α=180º BFC-TCR= Bc = (5.65e-3

) mho.

Such configuration of the FC-TCR has been experimentally

connected and simulated using Matlab. Fig. 8 shows the

experimental and the corresponding simulation waveforms of

the FC-TCR voltage and current for different delay angles. It

can be seen that experimental and simulation waveforms are

matched and show the validity of controlling the flow of

reactive power. This configuration could be utilized to

replace the traditional bank of capacitors in many

applications such as power factor improvement, power

system voltage and reactive power control, voltage control of

induction generator and performance optimization of three

phase induction motor operated from a single phase supply.

B. Three Phase Induction Motor Starters (Soft Starters):

Controlling the starting performance of three phase

induction motor has become one of the major concerns in

industrial fields [12 -13]. The purpose is to control the

starting voltage, current and torque as desired. Configurations

(e, f) in fig. 1 were utilized to examine the starting

performance on a three phase delta connected induction

motor of 0.75 kW, 75 V, 1.5 A, 50Hz, both experimentally

and using simulation. As seen from the results depicted in fig.

9 & 10, configuration (e) produces symmetrical wave forms

for both voltage and currents, unlike configuration (f) which

produces unsymmetrical voltage and current waveforms. The

unsymmetrical waveforms caused by the usage of only 3

SCRs is the reason for the appearance of odd and even

harmonics, DC voltage and current, torque pulsations and a

longer time to reach the desired speed as can be seen from

fig. 9. Heat loss in configuration (f) is higher than

configuration (e) due to the higher rms value of current. Fig.

11 & 12 depict the experimental waveforms obtained for the

practical implemented setup for both the 3 SCRs and 6 SCRs

corresponding to those simulated waveforms in fig. 9 & 10 at

no-load. The differences between the experimental and

simulation waveforms are due to the increase of α was done

manually unlike the simulation.

Voltage Voltage Voltage

Current Current Current

(b) Line current

(c) Motor speed

(a) Line voltage

(d) Average heat losses

Losses=

5 watt Losses=

5.4 watt

Figure 7: Configuration and characteristics of the FC-TCR

(b) Per phase

connection

(c) Characteristic curves (a) Three phase

connection

84

Figure 9: Simulation of three phase induction motor using 6 and 3 SCRS with alpha ramp from 0 to 220 V at 1 sec (0.5 N.m. loading)

Left: 6 SCRs Right: 3 SCRs

(e) Developed torque

VI. CONCLUSION:

A comparison study for different configurations of AC

voltage controllers has been introduced through this work.

Experimental and simulation waveforms are matched and

validated for the configurations. The analysis showed that

unidirectional configurations, having a combination of diodes

and thyristors, produced even and odd harmonics and also

contained a DC offset hence they are most suitable for

heating, melting and welding applications, while the

bidirectional configurations are suitable for AC motors,

power systems and electrical drives applications due to the

waveforms symmetry of the controlled voltages. This paper

also validates the analysis of each configuration for any value

of delay angle and control range. Using the AC voltage

controllers in reactive power control and controlling the

starting performance of three phase induction motor through

FC-TCR and SOFT STARTERS respectively as two

application case studies of AC regulators are demonstrated by

the aid of experimental and simulation results.

REFERENCES:

[1] K. H. Sueker, S.D. Hummel and R.D. Argent, “Power Factor

Correction and Harmonic Mitigation in a Thyristor Controlled Glass

Melter,” IEEE Trans. Industry Applications, Vol. 25, No. 6, pp. 972-975, November/December 1989.

[2] V. V. Sastry, M.R.Prasad and T.V. Sivakumar, “ Optimal Soft Starting

of Voltage- Controller Fed IM Drive Based on Voltage Across Thyristor,” IEEE Trans. Power Electronics, Vol. 12, No. 6, pp. 1041-

1051, November 1997.

[3] A. García- Cerrada, P. García- González, R. Collantes, T. Gőmez and J. Anzola, “ Comparison of Thyristor- Controlled Reactors and Voltage –

Source Inverters for Compensation of Flicker Caused by Arc

Furnaces,” IEEE Trans. Power Delivery, Vol. 15, No. 4, pp. 1225-1231, October 2000.

[4] A-R. A. M. Makky, F. N. Abd- Elbar and E. E. Mobarak, “ Speed

Control of 3- Phase Induction Motor Using Delta- Connected Thyristors Switch,” The Ninth International Middle- East Power

Systems Conference MEPCON’, pp. 827-833, December 2003.

[5] P.C. Sen, “Power Electronics”, TATA McGraw-Hill, 1987, ch. 8, pp. 588-689.

[6] Muhammad H. Rashid, “Power Electronics Circuits, Devices and

Applications”, Prentice – Hall International, Inc, 1993, ch. 6, pp. 190-222.

[7] A. Ginart, R. Esteller, A. Maduro, R. Pifiero and R. Moncada, “ High

Starting Torque for AC SCR Controller,” IEEE Trans. Energy Conversion, Vol. 14, No. 3, pp. 553-559, September 1999.

[8] M.S. Jamil Asghar, “Power Electronics”, Prentice-Hall of India, 2004,

ch.8 , pp.158-170 [9] L. J. Bohmann and R. H. Lasseter, “Stability and Harmonics in

Thyristor Controlled Reactors,” IEEE Trans. Power Delivery, Vol. 5,

No. 2, pp. 1175-1181, April 1990. [10] J. H. Chen, W. J, Lee and M. Chen, “ Using a Static Var Compensator

to Balance a Distribution System,” IEEE Trans. Industry Applications,

Vol. 35, No. 2, pp. 298-304, March/April 1999. [11] C. Champoo-inwai, C. Yingvivatanapong, K. Methaprayoon and W.

Lee, “ Reactive Compensation Techniques to Improve the Ride –

Through Capability of Wind Turbine During Disturbance,” IEEE Trans. Industry Applications, Vol. 41, No. 3, pp. 666-672, May/June

2005. [12] Frede Blaaberg, John K. Pedersen, Soren Rise, Hans- Henrik Hansen

and Andrzej M. Trynadlowski, “Can Soft- Starters Help Save

Energy?,” IEEE Industry Applications Magazine, pp. 56-66, September/ October 1997.

[13] Adel Gastli and Mohamed Magdy Ahmed, “ANN- Based Soft Starting

of Voltage- Controlled- Fed IM Drive System,” IEEE Trans. Energy Conversion, pp. 1- 7, 2005.

Figure 10: Simulation waveforms of three phase induction motor at α=130º Left: 6 SCRs Right: 3 SCRs

(d) Harmonic Analysis

(a) Phase voltage

(c) Phase current

(b) Line current

Figure 12: Experimental waveforms of three phase induction motor α=130º

CH1 voltage: 50 V / div, CH2 current: 1 A / div, Time: 0.01 sec/div

Left: 6 SCRs Right: 3 SCRs

(a) Phase voltage and line current

(b) Phase voltage and phase current

(a) Phase voltage and phase current

(b) Line voltage and line current

Figure 11: Experimental Waveforms of Three Phase Induction motor

CH1 voltage: 50 V / div, CH2 current: 0.75 A / div, Time: 0.01 sec/div

Left: 6 SCRs Right: 3 SCRs

THD= 241.5%

DC=0

THD= 226.5%

DC=0.05