SEMINARIO DE TESIS (MAYO-AGOSTO 2014)

Electrical Signal Acquisition and Processing for

Electric Power Quality Applications
AbstractIn this work a study of electric power quality was
done. Several documentation was made and the state of the art
in electric power quality was established for the thesis work.
Also, a review of the basic concepts of the three-phase power
was done, in the look of a possible target of measurement and
instrumentation (a motor for example). At last, the recent work of
the computation of an 8-Point Fast Fourier Transform in VHDL
is shown as the core of a more complex FFT to obtain key features
of an electric signal for electric power quality.
Index TermsSurvey, Electric Power Quality, EPQ, Harmonics, Three-phase power, FFT, Fourier transform, VHDL, FPGA.

I. I NTRODUCTION
LTHOUGH Electric Power Quality (EPQ) is defined by
the IEEE as the concept of powering and grounding
sensitive electronics equipment in a manner suitable for the
equipment, all electrical devices are susceptible to failure or
malfunction when they are exhibited to one or more power
quality problems. This device might be an electric motor,
a transformer, a generator, a computer, telecommunications
equipment, etc. From a simple alarm clock to complex industrial controllers, electric devices are susceptible to failure due
to a poor electric power quality.
Many other definition to EPQ exist. One example is A
set of electrical boundaries that allows a piece of equipment
to function in its intended manner without significant loss of
performance of life expectancy [9]. Another definition might
state The term that refers to maintaining the near sinusoidal
waveform of power distribution bus voltages and currents at
rated magnitude and frequency [5].
Poor quality is normally caused by power line disturbances
such as impulses, notches, voltage sag and swell, voltage
and current unbalances, momentary interruption and harmonic
distortions.
Other major contributors to poor power quality are
harmonics and reactive power.

The utilization of high-speed switching solid state control

of AC power are the main source of harmonics
Different non-linear loads contribute to excessive drawl
of reactive power from supply.

Poor energy quality has detrimental effects on health of

different equipment and systems, besides of the power system
stability, continuity and reliability fall with the degradation of
quality power [1].
II. E LECTRIC P OWER Q UALITY
Now that a more accurate definition of Electric Power
Quality was given, the characteristics of EPQ must be dis-

cussed.The quality in electric power can be classified as

follows:
Voltage quality
Current quality
Quality of power supply
Reliability of service
The study on EPQ may be divided into the following stages:
1) Fundamental concepts: identifies the parameters and
their degree of variation with respect to their rated
magnitude which are the base reason for degradation
of quality of electric power.
2) Sources: Are regions or location or events which causes
the unwanted variation of parameters.
3) Effects: The effects faced by the system and consumer
equipment.
4) Modeling and Analysis: Attempts are taken to configure the disturbance, its occurrence, sources and effect,
mainly based on the mathematical background.
5) Instrumentation: For monitoring EPQ, constant measurement and instrumentation of electric parameters
are necesary.
6) Solutions: Complete solution (delivery of pure power
to the consumerside is practically impossible. Our
target is to minimize the probability of occurrence of
disturbances and to reduce the effects of EPQ problems.)
EPQ describes the variation of voltage, current and frequency in a power system. Recent equipment is not resistant
to wide variations of voltage, current and frequency.
The sources for Electric Power Quality deterioration in a
power system can be divided into two groups:
1) Actual loads, equipment and components
2) Subsystems of transmission and distribution system

A. Classification of Power System Disturbances

Most EPQ disturbances depend on amplitude or frequency
or on both frequency and amplitude [4]. Events can be divided
into short, medium or long period.
Interruption / Under voltage / Over voltage:
Interruption: Is when the voltage goes down to zero
(short, medium or long period)
Under voltage / over voltage: These are fall and
rise of voltage levels of a particular bus with respect
to standard bus voltage. Sometimes under and over
voltages of little percentage are allowable; but when
they cross the limit of desired voltage level, they are
treated as disturbances.

SEMINARIO DE TESIS (MAYO-AGOSTO 2014)

Fig. 1. Under / Over voltage example

Fig. 3. Oscillatory Transient

Fig. 4. Sags/Swells
Fig. 2. Harmonic example

Voltage / Current unbalance: Voltage and current unbalance may occur due to the unbalance in drop in the
generating system or transmission system and unbalanced
loading. During unbalance, negative sequence components appear.
Harmonics: Are the alternating components having frequencies other than fundamental present in voltage and
current signals. Caused by non linearity, excessive use of
semiconductor based switching devices, different design
constraints, etc. (Harmonic component of current of order
n 1).
in = In sin 2nf

(1)

Harmonics are classified as integer harmonics (integer

multiple of fundamental frequency), sub harmonics (frequencies smaller than fundamental frequencies) and inter
harmonics (Greater than fundamental frequency). Other
classification for harmonics is Time harmonics, and spatial (space) harmonics.
Transients: May generate in the system itself or may
come from the other system. Transients are classified
into two categories: dc transient and ac transient. AC
transients are further divided into two categories: single
cycle and multiple cycles.
Voltage sag: It is a short duration disturbance. During
voltage sag, R.M.S. voltage falls to a very low level for
short periods of time.
Voltage swell: Short duration disturbance. R.M.S. voltage
increases to a very high level for short period of time.
Flicker: Undesired variation of system frequency.
Ringing waves: Oscillatory disturbances of decaying
magnitude for short period of time. May be called a
special type of transient.

Outage: Special type of interruption where power cut has

occurred for not more than 60s.
III. T HREE - PHASE POWER

Three-phase electric power is a common method of

alternating-current electric power generation, transmission,
and distribution. It is a type of polyphase system and is the
most common method used by electrical grids worldwide to
transfer power. It is also used to power large motors and other
heavy loads. A three-phase system is usually more economical than an equivalent single-phase or two-phase system at
the same voltage because it uses less conductor material to
transmit electrical power.
In a three-phase power supply system, three conductors each
carry an alternating current (of the same frequency) but the
phase of the voltage on each conductor is displaced from
each of the other conductors by 120 degrees. (One third of
a 360 degreecycle.) Hence, the voltage on any conductor
reaches its peak at one third of a cycle after one of the other
conductors and one third of a cycle before the third conductor.
Using the voltage on one conductor as the reference, the peak
voltage on the other two conductors is delayed by one third
and two thirds of one cycle respectively. This phase delay
gives constant power transfer over each cycle. It also makes
it possible to produce a rotating magnetic field in an electric
motor [6].
With a three phase supply, at any instant, the potential of any
phase is exactly equal to and the opposite of the combination
(sum) of the other two phases. This means that - if the load on
the three phases is balanced - the return path for the current
in any phase conductor is the other two phase conductors.
Hence, the sum of the currents in the three conductors is
always zero and the current in each conductor is equal to and
in the opposite direction as the sum of the currents in the

SEMINARIO DE TESIS (MAYO-AGOSTO 2014)

Fig. 5. Three-phase signals overview.

Fig. 6. Three phase power (Wye configuration).

other two. Thus, each conductor acts as the return path for the
currents from the other two.
While a single phase AC power supply requires two conductors (Go and Return), a three phase supply can transmit
three times the power by using only one extra conductor. This
means that a 50% increase in transmission cost yields a 200%
increase in the power transmitted.
Three-phase systems may also utilise a fourth wire, particularly in low-voltage distribution. This is the neutral wire.
The neutral allows three separate single-phase supplies to be
provided at a constant voltage and is commonly used for
supplying groups of domestic properties which are each singlephase loads. The connections are arranged so that, as far as
possible in each group, equal power is drawn from each phase.
Further up the supply chain in high-voltage distribution the
currents are usually well balanced and it is therefore normal
to omit the neutral conductor.
Three-phase supplies have properties that make them very
desirable in electric power distribution systems:
The phase currents tend to cancel out one another,
summing to zero in the case of a linear balanced load.
This makes it possible to reduce the size of the neutral
conductor because it carries little to no current; all the
phase conductors carry the same current and so can be
the same size, for a balanced load.
Power transfer into a linear balanced load is constant,
which helps to reduce generator and motor vibrations.
Three-phase systems can produce a rotating magnetic
field with a specified direction and constant magnitude,
which simplifies the design of electric motors.
1) Three-wire and four-wire configuration: There are two
basic three-phase configurations: delta and wye (star). In
Figure 8 a delta configuration requires only 3 wires for
transmission but a wye (star) configuration may utilise a fourth
wire 6. The fourth wire, if present, is provided as a Neutral and
is normally Grounded. The 3-wire and 4-wire designations
do not count the ground wire used above many transmission
lines which is solely for fault protection and does not carry
current under non-fault conditions.
A four-wire system with symmetrical voltages between
phase and neutral is obtained when the neutral is connected

Fig. 7. Three phase power (Delta configuration).

to the common star point of all supply windings. In such

a system, all three phases will have the same magnitude of
voltage relative to the Neutral. Other non-symmetrical systems
have been used.
IV. FAST F OURIER TRANSFORM
The basic mathematical representation of periodic signals
is the Fourier series, which is a linear weighted sum of harmonically related sinusoids or complex exponentials [2].The
Fourier transform is a mathematical transformation employed
to transform signals between time (or spatial) domain and
frequency domain, which has many applications in physics
and engineering [3].
In the case of a periodic function over time, the Fourier
transform can be simplified to the calculation of a discrete
set of complex amplitudes, called Fourier series coefficients.
They represent the frequency spectrum of the original timedomain signal Fig. ??. Also, when a time-domain function is
sampled to facilitate storage or computer-processing, it is still
possible to recreate a version of the original Fourier transform
according to the Poisson summation formula, also known as
discrete-time Fourier transform.

SEMINARIO DE TESIS (MAYO-AGOSTO 2014)

Fig. 8. Fourier decomposition.

To implement this application in a digital platform [8], using

a discrete form of the signal is required, and this is named
Discrete Fourier Transform (DFT). To calculate the DFT the
equation (2) is used.
X(k) =

N
1
X

x[n]ej(2/N )kn

Fig. 10. Decimation-in-time algorithm for an 8 point FFT.

(2)

n=0

In this process, the periodicity and symmetry of the complex

exponential WNkn = ej(2/N )kn are used to accomplish the
DFT calculation by using smaller and successive calculations.
This scheme is based in the Decimation-in-time algorithm.
To get a clearer representation of the ej(2/N )kn factor is
substituted by WNkn , called Twiddle factor.
The implementation of the DFT involves a great set of sums
and multiplications, resulting into an N 2 computation factor.
This can be seen because for each N values of k, the expansion
of the sum needs N number of complex multiplications.
The algorithm for the decomposition is the Fast Fourier
Transform (3).
X(k) =

N 1
1 X
x[n]ej(2/N )kn
N n=0

(3)

The FFT let us to reduce the number of sums and multiplications of the computation factor to N log2 N . The reduction
of operations is significant for high values of N , like 1024.
Using N = 1024 the number of complex operations would
be:
N 2 = 1, 048, 576

(4)

N log2 N = 10, 240

(5)

with the FFT

With this structure, the algorithm that executes the operation

can be defined in the Figure 9. This module receives the data
from the input signal (real and imaginary part) previously
saved in a RAM to calculate the actual operations. This
operations include the lecture of the Twiddle factor.

Fig. 11. Digital blocks for the algorithm implementation.

The complete algorithm to compute the FFT of 8 points is

shown in Figure 10. This arrange has the same geometry in
all the stages of the Decimation-in-time algorithm [7]. This
simplifies the sequential access and storage of the data, which
is better for a digital implementation.
This algorithm was designed in VHDL for a future implementation in an FPGA board. A general diagram can be shown
in Figure 11. In order to achieve an 8 point FFT calculation, a
technique named memory pagination is used, which consists
in using the output data of the RAM2, as the input to the
RAM1, as many time as needed, to reuse the same block of
the complex butterfly. This is useful and reduces the resources
that could be used during the hardware implementation.
Once the algorithm is done, the output coefficients of the
fast Fourier Transform are expressed as complex numbers.
As a design requirement, the output of the FFT computed
in VHDL must be shown in decibels (dB). The magnitude in
dB of the X(j) is described in equation 6.
y = 20 log10 (|X(j)|)
p
= 20 log10 ( Re2 + Im2 )

(6)

= 10 log10 (Re2 + Im2 )

The modules used to calculate the final result are shown in
Figure 12.
A. Test and results of the 8-point FFT in VHDL

Fig. 9. Basic Complex Butterfly algorithm.

The tests for the 8-point FFT in VHDL are shown in the
Table I. The input vector has 8 values input data for the FFT
computation. To compare the results of this design, the FFT
algorithm was developed in MATLAB step-by-step so both
VHDL and MATLAB results can be compared.

SEMINARIO DE TESIS (MAYO-AGOSTO 2014)

Sample
0
1
2
3
4
5
6
7

MATLAB
VHDL
-Inf
-96.3300
-Inf
-96.3300
-Inf
-96.3300
-Inf
-96.3300
0.0
0.0
-Inf
-96.3300
-Inf
-96.3300
-Inf
-96.3300
TABLE IV
R ESULTS OF T EST 3

Fig. 12. Digital block for the magnitude in dB.

Case
1
2
3
4
5

[1
[1
[1
[1
[1

0
1
-1
1
1

Input
0 0
1 1
1 -1
1 1
0 0

vector
0 0 0 0]
1 1 1 1]
1 -1 1 -1]
0 0 0 0]
0 0 0 0]
TABLE I

Description
Impulse
Unit step function
Nyquists limit
0.5Ts step function
0.25Ts step function

T EST PARAMETERS .

1) Test 1: The first test is the calculation of the impulse

response of the system. An 18 bit, 2.16 signed fixed point
format was used for the numerical representation. The VHDLs
module results are compared versus the MATLAB results, as
we can see in Table II.
Sample
0
1
2
3
4
5
6
7

MATLAB
VHDL
-18.0618
-18.0625
-18.0618
-18.0625
-18.0618
-18.0625
-18.0618
-18.0625
-18.0618
-18.0625
-18.0618
-18.0625
-18.0618
-18.0625
-18.0618
-18.0625
TABLE II
R ESULTS OF T EST 1

2) Test 2: The second test has as the input signal the unit
step function. The same codification was used for the input
data. The results are compared versus the FFT MATLAB
algorithm and are shown in the Table III.
Sample
0
1
2
3
4
5
6
7

MATLAB
VHDL
0.0
0.0
-Inf
-96.3300
-Inf
-96.3300
-Inf
-96.3300
-Inf
-96.3300
-Inf
-96.3300
-Inf
-96.3300
-Inf
-96.3300
TABLE III
R ESULTS OF T EST 2

3) Test 3: The third test uses an input signal as the limit

frequency for the Nyquist sample theorem, that is to say,
two samples per period of the input signal.The same data
codification as the previous examples was used. Again, the
results obtained by the VHDLs module is compared versus
the MATLAB algorithm. The results can be seen in Table IV
V. C ONCLUSION
The electric power quality is an actual issue that involves
every level of use of electrical energy. For this reason, it is

important to know and detail the features of the problems that

can be seen in the electric distribution line. The general idea
of the future work is to implement a measurement system
(placed in the instrumentation field of the electric power
quality stages). This instrumentation and measurement of the
power quality stage will be developed under FPGA hardware,
using the VHDL hardware description language. During the
research for this work, several documentation was revised to
explore the previous works and the state of the art in electric
power quality. In such documentation revision; it was found
that space transforms such as the Fast Fourier Transform and
the Wavelet transform are widely used to detect failures in
the electric power quality. For this reason, the core of an 8point FFT developed in VHDL is presented, leaving a 1024
point algorithm as future work. The purpose of this text is to
describe the core of the thesis work we will be working on
the next 8 months.
ACKNOWLEDGMENT
The authors would like to thank CONACYT, under the
scholarship number 541371 and DAIP for giving part of the
funds to make this project.
R EFERENCES
[1] Reid Edward. IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS.
[2] Proakis et al. Digital signal Processing, Year = 2000. Prentice-Hall
International, New Jersey.
[3] R.N. et al., Bracewell. The Fourier Transform and Its Applications (3rd
ed.), Year = 2000. McGraw-Hill, Boston.
[4] Santoso et al. Characterization of distribution power quality events with
fourier and wavelet transform.. IEEE Trans. Power Deliv., 15 (1):247
256, 2000.
[5] Surajit Chattopadhyay et al. Electric Power Quality. Springer, New York,
2011.
[6] Barry W. Kennedy. Power Quality Primer. McGraw-Hill, San Francisco,
2001.
[7] Romero-Troncoso Lara-Cardoso. IEEE International Instrumentation and
Measurement Technology Conference.
[8] Gregorio Romero Rey. POWER QUALITY HARMONICS ANALYSIS AND
REAL MEASUREMENTS DATA. inTech, Croatia, 2011.
[9] C. Sankaran. Power Quality Analysis. Springer, CRC Press, 2001.