Electrical Monitoring

Motor Current Signature Analysis (MCSA)

MCSA analyzes the current waveform drawn by an AC induction motor to detect rotor faults, eccentricity, stator winding issues, and load-side mechanical problems without interrupting production.

Key Principle

A broken rotor bar changes the motor's current waveform in a predictable way. MCSA finds that change at the frequency level, while the motor runs under load, before the fault progresses far enough to affect vibration or temperature.

What Is Motor Current Signature Analysis?

Motor Current Signature Analysis is a non-intrusive condition monitoring technique that captures the AC current waveform at the motor terminals and analyzes its frequency spectrum to detect developing faults. The name comes from the idea that each fault type produces a recognizable pattern, or signature, in the current signal.

Current clamps attach at the motor control panel or MCC. No physical contact with the rotating machine is required, and measurements are taken while the motor runs at normal load. This makes MCSA one of the safest and most practical monitoring methods for energized equipment in restricted or hazardous locations.

The technique applies primarily to three-phase AC induction motors, which account for the majority of industrial motor installations. The current drawn by a healthy motor, while not perfectly sinusoidal, has a predictable frequency content. Faults alter that content in ways the analysis can identify and quantify.

How MCSA Works

A healthy three-phase induction motor draws current at the supply frequency, 50 Hz or 60 Hz, plus harmonics introduced by magnetic saturation and winding distribution. The amplitude and phase of those harmonics remain relatively stable over time in a healthy machine.

Faults in the rotor, stator, or driven equipment create electromagnetic disturbances that modulate the current waveform. That modulation appears as sideband frequencies in the current spectrum, symmetric pairs of frequencies spaced above and below the supply frequency at mathematically predictable offsets.

The key analytical step is the Fast Fourier Transform (FFT). The FFT converts the time-domain current signal into a frequency spectrum where these sidebands appear as distinct peaks. Because the sideband frequencies are determined by the motor's operating parameters, specifically its slip and supply frequency, their locations can be calculated before measurement. The analyst looks for energy at those predicted locations and compares it against baseline readings.

Slip is central to the calculation. Slip is the difference between the synchronous speed of the rotating magnetic field and the actual rotor speed, expressed as a fraction of synchronous speed. A motor running at 1,746 RPM on a 60 Hz, 4-pole supply has a synchronous speed of 1,800 RPM and a slip of (1,800 - 1,746) / 1,800 = 0.030 (3.0%). A motor running at 1,752 RPM has a slip of 0.027 (2.7%).

For rotor bar faults, the sideband frequencies appear at:

f_sb = f₁ × (1 ± 2s) Where: f₁ = supply frequency (Hz) s = slip (dimensionless fraction) Example: f₁ = 60 Hz, s = 0.027 Lower sideband: 60 × (1 - 0.054) = 60 × 0.946 = 56.76 Hz Upper sideband: 60 × (1 + 0.054) = 60 × 1.054 = 63.24 Hz

The amplitude of these sidebands, expressed in decibels relative to the fundamental supply frequency component, indicates fault severity. A sideband that was not present in the baseline reading, or one whose amplitude has risen significantly, signals a developing fault.

Use the MCSA Fault Frequency Calculator to compute exact sideband frequencies for your motor.

Open Calculator

What MCSA Detects

Each fault type modulates the current waveform through a distinct electromagnetic mechanism. The table below maps fault types to their current spectrum signatures.

Fault Type Current Spectrum Signature Sideband Location
Broken or cracked rotor bars Asymmetric rotor resistance creates a backward-rotating flux component that induces current at the lower sideband; the interaction with the rotor produces the upper sideband f₁ ± 2sf₁
Air gap eccentricity (static or dynamic) Non-uniform air gap creates rotating flux density variation that modulates stator current at running speed intervals around the supply frequency f₁ ± (1 × running speed frequency)
Stator winding faults (turn-to-turn, phase-to-phase) Winding asymmetry increases the negative-sequence current component, elevating odd harmonic content relative to the fundamental Elevated 3rd, 5th, and 7th harmonics (180 Hz, 300 Hz, 420 Hz at 60 Hz supply)
Bearing faults Bearing defect impacts modulate rotor speed and therefore motor torque; the torque variation modulates stator current at bearing defect frequencies f₁ ± fbearing defect (calculated from bearing geometry and shaft speed)
Load-side mechanical faults (gears, pumps, fans) Periodic load torque variations caused by gear mesh, blade pass, or imbalance modulate the current at the corresponding mechanical frequency f₁ ± fmechanical (gear mesh frequency, blade pass frequency, etc.)

MCSA is most sensitive to rotor bar faults and eccentricity because those faults have a direct electromagnetic coupling to the stator current. Bearing and load-side fault detection via current is possible but less sensitive than direct vibration measurement at the fault location.

What MCSA Cannot Detect

Understanding the limits of MCSA is as important as knowing its strengths. Applying it outside its effective range produces unreliable results and wasted effort.

Mechanical looseness does not consistently appear in the current spectrum unless the looseness is severe enough to cause periodic load fluctuations that modulate motor torque. Slight foot looseness or soft foot conditions typically go undetected.

Seal leaks and cavitation are invisible to MCSA unless the fault is severe enough to alter the hydraulic load on the pump shaft, which then varies motor torque at a detectable frequency. Early-stage cavitation does not meet that threshold.

Misalignment produces an elevated 2x running speed component in vibration spectra, but only causes a detectable current modulation when the misalignment is severe enough to cause significant torque fluctuation at that frequency.

Low-load conditions limit sensitivity across all fault types. Below approximately 40% of rated load, slip decreases substantially. Rotor bar fault sidebands move closer to the supply frequency and their amplitude weakens. At low loads, sidebands may fall below the noise floor of the measurement and cannot be reliably distinguished from background noise.

VFD-driven motors require specialized approaches. A variable frequency drive introduces switching harmonics throughout the current spectrum, often at amplitudes that obscure the fault sidebands MCSA relies on. Standard MCSA procedures are not appropriate for VFD-driven motors without modification.

MCSA vs Vibration Analysis

MCSA and vibration analysis address overlapping but distinct fault domains. Treating them as alternatives misses the value each brings.

Criterion MCSA Vibration Analysis
Sensor placement Current clamps at motor control panel or MCC; no contact with the machine Accelerometers mounted directly on the motor bearing housings or driven equipment
Machine access required Electrical panel access only; motor can be in any physical location Physical access to the motor and driven equipment; restricted or guarded locations add complexity
Detects best Rotor bar faults, stator winding degradation, air gap eccentricity; electrical faults inside the motor Mechanical looseness, misalignment, bearing defects, imbalance; mechanical faults in the drivetrain
Load sensitivity Requires 40% or more of rated load; fault sidebands weaken at low load Less load-dependent; many faults detectable at low or no load
VFD compatibility Standard MCSA is unreliable on VFD-driven motors; requires specialized algorithms or alternative techniques VFDs can excite resonances at drive frequencies but do not fundamentally prevent vibration measurement

The practical conclusion: MCSA finds rotor and stator faults that vibration analysis misses. Vibration analysis finds mechanical looseness and misalignment that MCSA cannot reliably detect. Bearing faults are detectable by both, but vibration analysis provides higher sensitivity at earlier fault stages. Combined deployment gives substantially more complete motor health coverage than either method alone.

How to Take a Valid MCSA Measurement

A valid MCSA measurement requires attention to load conditions, measurement duration, and baseline availability. Skipping any of these produces data that cannot be reliably interpreted.

Load condition: The motor must be running at 40% of rated load or higher. Check nameplate full-load current and compare against the measured current to estimate load percentage before beginning. If the process cannot sustain sufficient load, document it and defer measurement.

Measurement duration and frequency resolution: Frequency resolution in an FFT is inversely proportional to the measurement duration. For a motor running at 60 Hz with 1% slip, rotor bar sidebands appear at approximately 58.8 Hz and 61.2 Hz, just 1.2 Hz away from the 60 Hz supply frequency. To resolve sidebands at that separation from the fundamental, you need a frequency resolution of 0.5 Hz or finer, which requires a measurement window of at least 2 seconds. In practice, 5 to 10 seconds of steady-state current data is standard to achieve reliable resolution at typical slip values.

Steady-state operation: The motor must be running at stable speed and load during the measurement window. Transient events, motor starting, load steps, or process disturbances invalidate the FFT by introducing non-stationary frequency content that obscures the fault signatures.

Baseline: A baseline measurement taken when the motor is known to be in good condition is essential. All subsequent measurements are compared against it. A single snapshot without context cannot distinguish a manufacturing-related asymmetry from a developing fault. When no baseline exists for a specific motor, fleet comparison against similar motors under similar load conditions can serve as a reference, with the caveat that motor-to-motor variation can complicate interpretation.

Interpreting Results

MCSA results are expressed as the amplitude of fault-frequency sidebands relative to the supply frequency component, in decibels (dB). A lower dB value (more negative) means the sideband is weaker relative to the fundamental.

For rotor bar faults in most squirrel-cage induction motors, the following amplitude thresholds serve as general guidelines:

These thresholds are reference points, not universal cutoffs. Motor design, number of rotor bars, rated load, and operating history all affect baseline sideband amplitudes. A motor with inherent manufacturing asymmetry may show a sideband of -50 dB in a healthy state; the same amplitude in another motor might indicate damage.

The most reliable indicator is always a rising amplitude trend across multiple measurements over time, not a single snapshot value. A sideband at -48 dB that has been stable for 18 months is less concerning than one that has risen from -60 dB to -48 dB over three months.

When interpreting results, also consider whether the sideband exists on both sides of the fundamental. A true rotor bar fault produces symmetric sidebands. Asymmetric sidebands, or a sideband present on only one side, may indicate a different fault mechanism or an artefact of the measurement.

MCSA on VFD-Driven Motors

Variable frequency drives are increasingly common across industrial motor installations, and they present a specific challenge for MCSA. A VFD controls motor speed by generating a pulse-width-modulated (PWM) current waveform at a programmable carrier frequency, typically 2 kHz to 16 kHz. The harmonics of that carrier frequency, and the sidebands around them, occupy much of the current spectrum and can mask the low-amplitude fault sidebands MCSA depends on.

Three approaches are available for VFD-driven motors:

Offline motor circuit analysis (MCA): With the VFD bypassed and the motor de-energized, a low-voltage AC test signal is injected across each phase pair. The test measures resistance, inductance, capacitance, and impedance balance without the VFD's interference. MCA identifies rotor bar faults, stator winding degradation, and turn-to-turn shorts in a controlled measurement environment. The limitation is that the motor must be taken offline.

VFD-aware MCSA algorithms: Specialized software applies adaptive filtering to remove the known VFD carrier harmonics from the current spectrum before performing fault frequency analysis. The effectiveness depends on how stable the VFD carrier frequency is and how well the algorithm models the drive's harmonic signature. This approach allows online measurement but requires software that understands the specific drive topology.

Stray flux monitoring: A search coil placed near the motor frame picks up the stray magnetic flux radiating from the motor. This flux signal contains fault information similar to the current signal, but is less contaminated by VFD switching harmonics because the flux integrates across the motor's magnetic circuit. Stray flux monitoring requires positioning the coil consistently across measurements and is more sensitive to environmental electromagnetic interference than current-based methods.

For any VFD-driven motor, document the drive's carrier frequency and PWM switching pattern before designing the monitoring approach. Different drive topologies produce different harmonic structures, and the monitoring method should be selected with that in mind.

Continuous MCSA, Not Periodic Snapshots

Tractian's platform runs MCSA continuously on connected motors, trends sideband amplitude over time, and generates fault alerts before severity reaches a critical threshold.

See Electrical Monitoring
Content on electricalmonitoring.com is written for electrical and reliability engineers. Threshold values and diagnostic guidelines are general references; consult motor manufacturer specifications and site-specific baselines before making maintenance decisions. Electrical work on energized equipment must comply with applicable safety standards including NFPA 70E and local regulations.