MOTOR CURRENT SIGNATURE ANALYSIS

(MCSA)
does not require any expensive tools and saves
Kavin time.
Department of EEE (Sandwich)
PSG College of Technology
Peelamedu, Coimbatore
[email protected]

Abstract- The common faults occurring in the

electrical motors like air gap eccentricity fault, rotor
fault, short-circuit fault, bearing fault etc. are
detected by conditional monitoring. But it requires
many tools which is quite expensive. The Motor Current Signature Analysis (MCSA)
advancement in digital signal processing (DSP) paves analyzes the frequency spectrum of stator current
a new solution to identify the faults in the motor. of healthy motor and how it varies from faulty one.
Motor Current Signature Analysis (MCSA) uses the By analyzing the frequency spectrum of fault
current signal to identify the faults. The frequency motors, different faults are located.
spectrum of the line current for the faulty motor
2. COMMON FAULTS
differs from the healthy one. Due to faults the
harmonic components induce in motor, changes the
a) Broken rotor fault: The broken rotor bars
frequency spectrum of current signal. The frequency strikes the stator core and windings of
spectrum is easily detected and analyzed by MCSA. It high voltage motor at very high speed.
depends on the locating specific harmonic component This may cause serious mechanical
in the line current. MCSA will be done through direct damage or insulation failure and
online monitoring by using LabVIEW software. The consequently that leads to winding
effect of faults in the machine and tracking the faults damage.
is briefly explained in this paper. The development of
frequency spectrum from the analog current uses
Fast Fourier Transforms (FFT) through data
acquisition. This paper explains the transformation
process and experimental procedure to detect the
common faults by means of MCSA.

1. INTRODUCTION

Induction motor called as asynchronous motors are

widely used in many industrial and engineering
applications. Generally induction motor is of low
performance and highly robust. The motors are
frequently exposed to worst environment. This
circumstances leads to overloading, inadequate
cooling, excess noise and vibrations etc. This Figure 1.Inducion motor with broken rotor
condition develops undesirable stress to the motor,
puts the motor in risk or faults. The most common Broken bars are due to brazing problem
faults occurs in the motor are bearing faults, broken (connecting bars part) or thermal stress or
rotor faults, winding faults, rotor eccentricity end-ring pressure.
faults, brushes etc. This requires a continuous and
conditional monitoring such as thermal monitoring, b) Air gap eccentricity: The rotor or stator
noise and vibration monitoring, torque monitoring. is not aligned centrally due to improper
But for such monitoring a special tools and skilled balancing while manufacturing is called
labors are required. It is quite expensive and time eccentricity. Air gap eccentricity causes an
constraints. Advancement in Digital Signal unbalanced magnetic pull (UMP). This
Processing (DSP) paves a solution through which damages the core and the windings.
the faults are detected by using software tools. This
 Vibrations and noise in the motor efficiency gets decreased. This further increases the
increases consequently. heat and noise. So it is very much important to
monitor for the faults in the motor. The cost of the
equipment is very high and so online monitoring is
the easiest way to detect the faults in the motor.

4. MOTOR CURRENT SIGNATURE

ANALYSIS (MCSA)

MCSA is the online analysis of current to

detect the faults in a three-phase induction
motor while it is in operational condition.
A basic MCSA instrumentation system
Static eccentricity will consist of the following:

 A current transformer (CT) to

sense the signal
 A shunt resistance across the CT
output
The air gap eccentricity is classified into three  MCSA instrument (Spectrum
depends on the direction of magnetic pull as analyzer) to produce current
signature
 Static eccentricity
 Dynamic eccentricity
 Mixed eccentricity

Static eccentricity: This cause the UMP on the

rotor in one direction that tries to pull the rotor
further from stator bore center in the direction of
minimum air gap as shown in Fig.2

Dynamic eccentricity: This cause the UMP on the

rotor that rotates at the rotor speed.

Mixed eccentricity: Combination of both static and

dynamic eccentricity is called mixed eccentricity.
Figure 3.Basic MCSA instrumentation system
Causes: If level of air gap eccentricity are not kept
within specified limits, then eccentricity causes
excessive stress and can increase bearing wear. 5. PRINCIPLE OF OPERATION OF
This act on the stator core and the rotor cage, MCSA
subjected to winding damages or insulation failure.

c) Rotor faults: Insulation failure is the In the induction motor, fundamental wave of
main drawback in the armature of the frequency (f) is always accompanied with a
motor. This is mainly because of abrupt harmonic component of frequency (mf); where m
temperature rise in the windings. Wedges is the harmonic number. This presence of harmonic
inserted in the slots may damage the component varies the amplitude in a small
insulation. Loose bracing for end windings magnitude. This harmonics induces the flux that
or contamination due to oil, moisture and cuts the rotor in associate with flux due to
dirt causes insulation wear. Short circuit fundamental supply. This induces the rotor current
may leads to insulation failure. that in turn produces the flux that links with the air
gap at the specific harmonic frequency. If the
fault occurs in the rotor, the amplitude of the stator
current varies at the specified harmonic frequency.
3. ILL EFFECTS DUE TO FAULTS By comparing the amplitude of stator current for
healthy motor and the faulty motor, the fault is
The faults in the machine may cause the detected. The different fault produces a harmonics
unbalanced voltage and current in the motor, this of different frequency. This frequency is
further changes the torque. Because of high voltage associated with the fundamental frequency called
and current, loss in the motor increases and as sideband frequency. In general, stator current
contains unique frequency for different faults. Here through data acquisition by using DATA
it is very important to transform the time domain ACQUISITON CARD (DAC). DAC consist of
current into frequency domain. This is done by digital to analog converter and vice versa. DAC
application of Digital Signal Processing (DSP). also forms an interface between the computer and
Various analyses such as Fast Fourier Transforms the motor. Then the acquired signal is converted
(FFT), Wavelet analysis, etc. are used to determine into frequency domain by means of FFT. This is
frequency content in the time domain function. easily done by using software such as MATLAB or
Laboratory Virtual Instrumentation Engineering
Workspace (LabVIEW). Calculation of sidebands
for each fault is the crucial process which will be
explained clearly in the next sections.

7. BROKEN ROTOR ANALYSIS

Usually, a decibel (dB) versus frequency spectrum

is used to detect unique current signature patterns
of different faults. The rotating magnetic field
(RMF) induces the rotor voltage and current at slip
frequency, and this simultaneously produces a
3s1111phase RMF at slip frequency with respect to
the rotor. If rotor fault occurs there will be a
Figure 4.Illustration of current signal in frequency resultant backward RMF at slip frequency with
domain showing the fundamental frequency f1 and regards to the forward RMF. This backward RMF
two sidebands. induces a voltage in the stator at the corresponding
frequency. This changes the spectrum of current.
6. STEPWISE PROCEDURE
The forward RMF produced in the induction motor
Stator current of the rotates at synchronous speed.
motor
N1 =120f1/p
Where f1 is supply frequency and p is
the poles
Signal condition and Data
Slip (s) = N1– N
acquisition N1
Where N is the rotor speed

Slip speed (N2) =N1 – N

Fast Fourier Transform
(FFT) Slip (s) = N2/N1

N2= s.N1

Calculate the sideband We get N = (1-s) N1

frequencies of each fault
The backward rotating magnetic field produced by
the rotor due to broken bar and w.r.t the rotor is

Observation and analysis Backward RMF speed Nb = N – N2

at different load
Nb= [(1-s) N1] – s.N1
conditions
Nb = (1-2s) N1
Location of various faults In terms of frequency

fb = (1-2s) f1
Block 1.Steps in MCSA
Due to slip frequency two sidebands occurs and
The analog stator current is converted into discrete the which equals
time signal (DTS) to detect the faults. This is done
 fb = (1±2s) f1
8. AIR GAP ECCENTRICITY
Many broken rotor bar occurs then it forms a
sequence as
fb = (1±2ks) f1 The sideband frequencies due to air gap
eccentricity are found by means of magneto motive
where k = 1, 2, 3 etc. force (MMF) – Permeance Wave approach. The
frequencies are identified (Cameron, 1986 and
sobezya, 1988) as

fecc = f1 (R±nd) [(1-2s)/p]±nws

where fecc = frequency component of air gap

eccentricity

R = Number of Rotor slots

nd = 0 for static eccentricity ; 1, 2, 3, etc. for

dynamic eccntricity.

Figure 5.Frequency spectrum of broken rotor bars P = pole pairs

Due to rotor broken bar, two sidebands due to slip nws = 1, 3, 5.etc
frequencies are obtained. If the amplitude of the
sideband exceeds certain limit there is a broken 9. SIGNAL PROCESSING
rotor bar, generally amplitude limit is considered to TECHNIQUE
be 40dB. The experimental observation is done
which is shown in Fig.6 as Signal processing technique is used to convert the
current signal into useful information for diagnose
the fault. Some of the methods of transformation is

 Fast Fourier Transform (FFT)

 Short Time Fourier Transform (STFT)
 Gabor Transform
 Wavelet Transform

For analysis FFT is useful and easy . Discrete

Fourier Transform (DFT) is the most straight
mathematical procedure to determine the frequency
component of the time domain sequence. FFT is
the simplest and efficient way to calculate DFT.
This transformation can be done through LabVIEW
Figure 6a.Power spectrum of healthy motor software.

Block 2.Obtaining power spectrum using LabVIEW

Figure 6b.Power spectrum of motor with broken bar under full
load
10. LIMITATIONS
This method is limited to light load conditions. The
power spectrum obtained while the machine under
light load condition is quite small amplitude. It is
very difficult to measure the amplitude of the
sidebands under such condition. REFERENCES:

1) Thomson.W.T,1988,”conditional monitoring on induction motor”,IEEE.

2) Gilmore.J,”a paper on MCSA”,IEEE.
3) Mehala Neelam,2010,NIT,”fault diagnosis of induction machine by MCSA”,thesis for Degree
of Doctor of Philosophy.