Real-Time Motor Current

Signature Analysis Tool for

Undergraduate Laboratory
PETER IDOWU, CHRISTOPHER ROOT
Penn State University, Harrisburg, Pennsylvania

Received 6 July 2007; accepted 18 June 2008

ABSTRACT: The typical electrical engineering (EE) undergraduate curriculum is packed with foundational
materials and offers limited room for other desirable materials that could be readily applied in the power industry.
A Motor Current Signature Analysis (MCSA) tool is developed for effectively teaching the concepts of induction
motor fault detection within one lecture or laboratory period. ß 2009 Wiley Periodicals, Inc. Comput Appl Eng
Educ 18: 634
639, 2010; View this article online at wileyonlinelibrary.com; DOI 10.1002/cae.20263

Keywords: induction motor; laboratory; signature analysis; undergraduate education; power engineering

INTRODUCTION formers, DC machines, AC induction motors, synchronous

machines, and some non-conventional or renewable energy
The induction motor has enjoyed wide ranging applications in systems. It is logical to conclude that very little could be
industrial processes for decades because of its robust character- addressed in those topics beyond elementary models and analysis,
istics and lower cost when compared with the DC machine. The given that most junior or senior undergraduates come into the
profile of inductor motors is rising even higher as they are course with backgrounds that are limited to AC and DC circuit
becoming more competitive in adjustable speed applications. analysis. It is not very feasible to add more topics to cover in such
This success is owed in part to advances in power electronics and a course, especially as it is already overburdened, even when such
drive systems. The increasing critical role in industrial processes topics have wide application in the power-oriented industry.
and the need to closely monitor motor health [1
11] has spurned Is there a strategy or method for incorporating induction
lucrative professional services, as well as diagnostic and test motor fault diagnosis into the undergraduate curriculum without
systems for monitoring induction motor stator, rotor, bearings, significantly adding to the already congested course syllabus? Is
airgap, coupling, load, etc. The electric power industry appears to there an avenue to accomplish this with existing range of equip-
have ceded the chore of induction motor health monitoring and ment in a typical undergraduate EE energy conversion labo-
fault diagnosis to dedicated consultants who hold the expertise ratory? Is there an avenue to offer students the opportunity to
and requisite equipment for performing detailed studies. innovate and expand the learning system or tool? These are
Although these technologies and predictive assessment questions that are answered in this article.
strategies are gaining importance in the industrial environment, This article presents a Motor Current Signature Analysis
it is unlikely that a typical electrical engineering (EE) under- (MCSA) system developed for use in an EE undergraduate
graduate will have an exposure to fundamentals of induction course. The easy-to-use interactive MCSA system proves
motor health monitoring and diagnosis, or access to specialized adequate as a tool for introducing motor current-based incipient
equipment that professionals use in the trade. Acquisition costs fault detection, as well as for applying the concept in the labo-
and learning curve for the commercial tool would discourage ratory. Within a period of 1 h students without prior knowledge of
their use. motor current signature analysis techniques were able to demon-
In undergraduate electrical engineering programs where a strate some skills in diagnosing motor faults.
core course in energy systems and conversion is included in the Second section of this article discusses the design and
curriculum, the materials covered in the one semester course tend implementation of the MCSA system. Third section highlights
to be limited to traditional topics. These include: power trans- typical components of a 1-h laboratory introduction to signature
analysis and faults diagnosis. In fourth section students’ rating of
the MCSA system and laboratory experience is discussed, and
Correspondence to P. Idowu ([email protected]). concluding remarks and future plans for the project are included
ß 2009 Wiley Periodicals Inc. in fifth section.

634
 REAL-TIME MOTOR CURRENT SIGNATURE ANALYSIS 635

DESIGN AND IMPLEMENTATION OF THE believed to have insignificant wear and tear. The 5-HP adjustable
MCSA SYSTEM drive facilitates studies and diagnosis of motor conditions under
different operating frequencies. The real-time data acquisition
The MCSA system presented in this article is not conceived (DAQ) and signal conditioning devices can support line-to-line
to compete with the full-scale commercial signature analysis voltages exceeding 208 V, as well as currents exceeding 20 A.
systems. Rather, it is planned to be a learning system that is The signal acquisition system allows for real-time logging of all
adaptable to the changing needs of the academic environment. line voltages, line currents, speed and shaft torque. In addition, it
Since undergraduate curriculum needs are not the fundamental incorporates features for remote control of the inverter functions,
criteria in the development of commercial MCSA packages such as frequency adjustment and motor reversing. Sampling rate
they should not be expected to address conceptual difficulties of up to 50,000 samples per second is attainable for all the input
students may experience in the process of learning the basics of channels.
induction motor fault detection. The appropriate signature The user interface to the MCSA hardware and data analysis
analysis tool would facilitate delivery of core concepts of is developed using LabVIEW1 graphical programming tool. The
motor current signature analysis with the flexibility that is needed main MCSA control panel shown in Figure 2 has an intuitive
in the academic environment to advance learning goals design for ease of use. Users are presented with a wide range of
efficiently. It is fundamentally necessary to implement the system parameters to adjust and customize information display. The main
with as little additional cost as possible by integrating existing page displays all line voltages and currents, speed and torque. It
hardware and software components. The tool should offer the provides access to all input channel gain settings, input signal
environment for students to innovate and extend the learning filtering options, inverter control, sample rate control, data
system. logging to a spreadsheet and virtual instruments to register real
To accomplish these objectives, it was deemed most power, reactive power, power factor, etc.
appropriate to target an energy systems and conversion course The motor current signature diagnostic interface is shown in
as the right venue for exposing students to the MCSA system. Figure 3. Spectrum of line currents is displayed prominently
One laboratory session would be devoted to the subject, at along with user adjustable parameters. Users could control Fast
the conclusion of which it is expected that students demonstrate Fourier Transform (FFT) data averaging type (peak hold, RMS or
a measure of confidence in relating to the topic of signature vector), weighting, length of data, sampling rate, filter type, band
analysis. frequencies, etc., as well as a zooming feature on the frequency
The layout of the signature analysis system developed for scale.
the energy systems and conversion course at Penn State Uni- In order to monitor the condition of a given induction motor,
versity, Harrisburg is given in Figure 1. Eight three-phase induc- appropriate motor data such as number of poles, rotor slot
tion motors are available in the energy conversion laboratory, number, stator MMF harmonic order, etc., are entered directly
serving as specimen for demonstrating the MCSA system. Four of on the user interface, and may be adjusted in real-time. The
the motors are rated at 1-HP; they have been used heavily in the program computes and display frequency components that
energy conversion laboratory for over 15 years and have exhibited would indicate the presence of four types of rotor and stator
a number of stator as well as rotor related problems. The faults (Fig. 4). The MCSA analysis system is currently designed
remaining four are newer and used sparingly, rated 3-HP, and are to detect rotor eccentricity, broken rotor bars, stator winding
short-circuit, and mechanical influences on the rotor. More fault
types could be added as the MCSA system is expanded in the
future.
The LabVIEW1 interface and program provides an
environment where user programs developed in MATLAB1
could be easily and conveniently incorporated into the signature
detection section shown in Figure 4 (LabVIEW1 vi block
diagram). This offers students a creative avenue to test out more
advanced detection algorithms developed in MATLAB1, such as
techniques based on wavelets, fuzzy logic, neural network, etc.
This facility would encourage students to integrate material
learned in signal processing courses to problem solving in the
electric power engineering domain.
The assortment of three-phase motors and adjustable drive
system constitutes the most expensive pieces of hardware
equipment needed for the MCSA system. It should be noted that
theses were already in place at Penn State University, Harrisburg
at the time of development of the signature analysis system, and
would therefore not be factored into the development cost. The
older and heavily used three-phase motor units were in fact
already out of commission and on the way to salvage before being
drafted for this new mission. They are perfect candidates for fault
diagnosis. New data acquisition and signal conditioning devices
were purchased for the project, and LabVIEW1 interface
Figure 1 Layout of the Motor Current Signature Analysis (MCSA) program development was accomplished through the efforts of
system. a graduate student.
636 IDOWU AND ROOT

Figure 2 Motor Current Signature Analysis system interface and control panel.

ONE-HOUR LABORATORY INTRODUCTION TO machine faults within a 1-h laboratory period. Some of the
SIGNATURE ANALYSIS students in the first round of evaluation of this tool have
participated in a field trip to a local steel rolling plant where they
The purpose of the MCSA tool is to facilitate a meaningful observed an environment where large motors are used extensively
introduction of motor current signature analysis for diagnosing in difficult and hard-to-reach places. Undoubtedly, this exposure

Figure 3 Spectrum of motor currents and fault diagnosis.

 REAL-TIME MOTOR CURRENT SIGNATURE ANALYSIS 637

Figure 4 LabVIEW1 implementation of signature detection algorithm.

served to underscore the critical role that motor diagnostic tools These results suggest that a 1-h exposure to the subject by
could play in early assessment of motor conditions before way of an interactive laboratory session proved somewhat
breakdown and interruptions of production in the typical steel- effective. Very few of the students had prior exposure to motor
processing environment. fault diagnosis. Consistently, ratings show averages of over 3.0 on
Prior to the laboratory session students were given a one- a scale of 5.0 indicating positive learning experience. Comments
page summary of test machine data and analytical expressions for from students reveal a desire to spend some more time of
frequency signatures of certain rotor and stator faults (Appendix). familiarization with the MCSA tool, a more detailed overview of
This pre-laboratory information provided critical insights on the hardware connections and software design, and extra data
how the MCSA system arrives at diagnosis displayed on the samples for homework projects.
LabVIEW1 interface.
The 1-h laboratory session started with PowerPoint
presentations on the basics of motor current signature analysis CONCLUSIONS
and the typical rotor and stator faults. Students were asked to
individually review the slides before running the motors and This article presents the development of a MCSA system for use
diagnosing faults. At the conclusion of the session students were by electrical engineering undergraduates. The tool offers an
given new sets of induction motor data to study and to analytically avenue for coverage of more topics efficiently without imposing
detect evidence of broken rotor bar, air gap eccentricity, stator undue burden on the already overcrowded curriculum. The easy-
winding short circuit and oscillation of rotor position. to-use interactive MCSA system proves adequate as a tool for
With the aid of the MCSA system the core concepts of introducing the concept in a 1-h laboratory session.
motor current-based fault diagnosis were successfully communi- The MCSA test platform would facilitate the development
cated to the Junior/Senior-level electrical engineering under- of a series of novel experiments on diagnostics of AC machines,
graduates within a 1-h laboratory session. The summary and study of load-induced harmonics, inverter-induced harmonics,
analysis of students’ evaluation of the learning experience are coupling condition, etc. The tool presents the right environ-
related in fourth section that follows. ment for creative opportunities to apply new signal analysis
techniques to fault signature detection by linking MATLAB1
and LabVIEW1 systems. It offers the flexibility that is needed
STUDENT EVALUATION OF THE MCSA in the academic environment to advance learning goals
LEARNING TOOL efficiently.
The MCSA system design started out as a graduate student
A group of 23 EE students in the junior or senior year participated project, it has however turned out to be an insightful illustration
in the inaugural 1-h laboratory exposure to the MCSA learning on incorporating graduate research into undergraduate education.
tool. Their assessment of the learning experience is summarized In the future, the MCSA system will be accessible to users
in Table 1. through the Internet. This would offer students some extra time
638 IDOWU AND ROOT

Table 1 Evaluation of Effectiveness of 1-h Exposure to MCSA

Evaluation of the MCSA laboratory experience

1. Rate your prior familiarity with induction motor faults diagnostic methods or tools
0 1 2 3 4 5 Average/(%)
Not familiar Very 1.04/(20.9)
familiar
2. The PowerPoint introduction of motor current signature analysis is very helpful
0 1 2 3 4 5 Average/(%)
Strongly Strongly 3.65/(73.0)
disagree agree
3. The MCSA tool is intuitive and easy to use
0 1 2 3 4 5 Average/(%)
Strongly Strongly 3.65/(73.0)
disagree agree
4. I am comfortable with applying this technique if necessary in some future projects
0 1 2 3 4 5 Average/(%)
Strongly Strongly 3.43/(68.7)
disagree agree
5. Given the opportunity I will like to explore motor diagnostic techniques further
0 1 2 3 4 5 Average/(%)
Strongly Strongly 4.04/(80.9)
disagree agree

for getting familiarized with the system prior to, as well as after n 1, 2, 3, . . . (integers)
the laboratory session. k 1, 3, 5, . . . (harmonics)
m harmonic number 3

APPENDIX Data for Machine Series Used in the Experiment:

Pre-Laboratory Summary for Nb 32 (3-hp series)

Frequency-Based Fault Detection Nb 28 (1-hp series)

Faults in induction motors are usually classified as stator related,

rotor related or bearing related. Studies show that 30
40% of
induction motor failure is stator related, directly linked to some REFERENCES
form of winding breakdown [2] and 10% are related to rotor
problems [7]. The frequency characterizations of various fault [1] M. Benbouzid, A review of induction motors signature analysis as a
medium for faults detection, IEEE Trans Ind Electr 47 (2000),
conditions have been established from previous research on
984
993.
current signature detection. Table A.1 shows typical fault con- [2] A. Siddique, G. Yadava, and B. Singh, A review of stator fault
ditions and the frequency component indicators of abnormality. monitoring techniques of induction motors, IEEE Trans Energy
Conver 20 (2005), 106
114.
Table A.1 Typical Faults and Their Associated [3] T. Chow and H. Shi, Induction machine fault diagnostic analysis
Frequency Components with wavelet technique, IEEE Trans Ind Electron 51 (2004),
1 Broken rotor bar [7] fsb ¼ f1 ð1  2sÞ
Hz 558
565.
2 Air gap eccentricity [10] fec ¼ f1 fðNb  nd Þ 1
sp  nw g Hz [4] W. G. Zanardelli, E. G. Strangas, H. K. Khalil, and J. M. Miller,
3 Stator winding short circuit [7] fst ¼ f1 fn=pð1
sÞ  kg Hz Wavelet-based methods for the prognosis of mechanical and
4 Oscillation of rotor position [7] frot ¼ f1  mf2 Hz electrical failures in electric motors, Mech Syst Signal Process 19
(2005), 411
426.
[5] G. K. Singh, S. Kazzaz, and S. Ahmed, Vibration signal analysis
using wavelet transform for isolation and identification of electrical
fsb frequency sidebands due to broken rotor bars (Hz) faults in induction machine, Electr Power Syst Res 68 (2004),
fec frequency components linked to airgap eccentricity (Hz) 119
136.
fst frequency components linked to shorted turns (Hz) [6] E. A. Ebrahim and N. Hammad, Fault analysis of current-controlled
frot frequency component linked to oscillation of rotor position PWM-inverter fed induction-motor drives, Proc IEEE Int Conf
(Hz) Propert Appl Dielectr Mater 3 (2003), 1065
1070.
f1 supply frequency (Hz) [7] W. T. Thomson and M. Fenger, Current signature analysis to detect
induction motor faults, IEEE Ind Appl Mag 7 (2001), 26
34.
f2 rotor frequency—sf1 (Hz)
[8] Z. Zhengping, R. Zhen, and H. Wenying, A novel detection method
Nb number of rotor bars (or slots) of motor broken rotor bars based on wavelet ridge, IEEE Trans
nd dynamic eccentricity index (approximately 1) Energy Conver 18 (2003), 417
423.
s per unit slip [9] M. Benbouzid and G. Kliman, What stator current processing-based
p pole-pairs technique to use for induction motor rotor faults diagnosis?, IEEE
nw MMF time harmonics (1, 3, 5, 7, . . .) Trans Energy Conver 18 (2003), 238
244.
 REAL-TIME MOTOR CURRENT SIGNATURE ANALYSIS 639

[10] W. T. Thomson, D. Rankin, and D. G. Dorrell, On-line current [11] A. Bellini, G. Franceschini, and C. Tassoni, Monitoring of induction
monitoring to diagnose airgap eccentricity in large three-phase Machines by maximum covariance method for frequency tracking,
induction motors—industrial case histories verify the predictions, IEEE Trans Ind Appl 42 (2006), 69
78.
IEEE Trans Energy Conver 14 (1999), 1372
1378.

BIOGRAPHIES

Peter Idowu (S’84-M’86-SM’03) received the Christopher Root received the BSET in
PhD degree from the University of Toledo, Electrical Engineering degree from Blooms-
Toledo, OH, in 1989. He is currently an burg University, Pennsylvania in May 2006.
associate professor of Electrical Engineering He is currently enrolled in the MEng Electrical
at Pennsylvania State University, Harrisburg, Engineering program at Penn State University,
Middletown, PA. His research interests include Harrisburg. His research interest includes
modeling, simulation and control of electric power systems analysis, signal processing
machines, power systems, and remote labora- and remote laboratory systems.
tory systems. Dr. Idowu is a registered
Professional Engineer in the State of Ohio.