
A broken rotor bar is one of the most serious faults that can affect a squirrel cage induction motor. The damage may begin subtly — perhaps a slight increase in vibration or a minor fluctuation in motor current — but left unaddressed, it accelerates.
Intact bars are forced to carry higher-than-normal currents, which increases thermal and mechanical stress and dramatically shortens their service life. What starts as one broken bar can cascade into multiple failures, ultimately rendering the motor unsafe and unreliable.
The challenge is that broken rotor bars are notoriously difficult to detect. There is no single, foolproof test. Instead, technicians must rely on a combination of methods, some performed with the motor assembled and running, others requiring disassembly in a service center environment. Understanding how a squirrel cage rotor works — and what happens when it fails — is essential to choosing the right approach.
This guide covers the most important detection methods in depth, explains the physical principles behind each, and provides practical guidance on how and when to apply them.
How a Squirrel Cage Rotor Works — and How It Fails
The squirrel cage rotor is often described as the secondary winding of a transformer, with the motor stator acting as the primary. During operation, the rotating magnetic field of the stator induces voltages and currents in the rotor bars, which in turn create the forces that drive rotation. The rotor bars form parallel conductive paths, joined at each end by end rings that complete the electrical circuits.
Most modern rotors are constructed with bars and end rings made of aluminum alloy (die-cast in one operation) or copper/copper alloy (fabricated separately and then brazed or welded to the end rings). The two constructions fail in different ways: fabricated copper rotors most often develop fractures at the joint between a bar and an end ring, while die-cast aluminum rotors tend to develop porosity — small voids or blow holes — within the bars or end rings themselves.
Whatever the construction, the consequences of a broken or high-resistance bar are similar. That bar can no longer carry its share of current, so adjacent bars must compensate. Torque pulsations, speed fluctuations, vibration, and changes in the motor’s current signature all result. The motor may also run hotter than normal, draw higher current at rated load, and operate slightly below rated speed. In severe cases, arcing from the broken bar can scorch laminations and in fabricated rotors even cause a rotor bar-to-stator core rub.
An important and often misunderstood point: not every broken bar causes an immediate, obvious problem. The number of bars, the number of stator poles, and the specific bar/slot combination all determine how many broken bars a motor can tolerate before noticeable torque cusps appear. A motor may run for months or even years with broken bars — performing worse without an obvious cause — until enough bars fail to create an unmistakable fault signature. This is precisely why proactive testing matters.
For a closer look at why bars crack in the first place, see What Causes Broken Rotor Bars? (And How to Fix Them).
Detection Method #1: Visual Inspection
Visual inspection is always the first step after a rotor has been removed from the stator. It costs nothing and can immediately reveal obvious faults — or steer you toward the right follow-up tests.
Begin by examining the rotor before cleaning. Look for cracked or separated end rings, signs of arcing or burning along individual bars, and evidence of localized overheating. On painted rotors, overheated areas often appear as dark, burned patches where the paint finish has broken through. A completely broken bar sometimes causes current to arc through adjacent laminations, leaving scorched marks along the bar’s length.
Thoroughly clean the rotor (avoid grit blasting at this stage) and repeat the inspection. On die-cast rotors, examine the end ring surfaces and cooling fins carefully for porosity — small pits or voids — and for cracks. Inspect the inner diameter of the end ring where it meets the laminations, as this area is particularly prone to porosity. Also check for any skewed laminations that may have shifted during manufacture, partially or completely closing a rotor slot.
Pay attention to balance weights. A high concentration of added balance weight in one area of the end ring can indicate a void beneath the surface — the void reduces local mass, requiring correction at that point.
Interpret discoloration with care. A blue finish on rotor laminations may indicate overheating from open bars or unbalanced supply voltages, but some manufacturers apply a heat treatment to die-cast rotors after casting that produces a similar blue color as a normal process result. Look for other supporting evidence of a fault before drawing conclusions from color alone.
Detection Method #2: Single Phase Rotor Test
The single phase rotor test can be performed with the motor fully assembled but not running. It is one of the most accessible and reliable tests available — requiring only a single phase power source and an ammeter — and can be done both in the field and in the service center.
How to perform the test
Disconnect the motor from its normal power supply. Connect a single phase supply at reduced voltage — typically between 1/8 and 1/4 of rated voltage — to any two of the motor’s line leads. Use an ammeter to monitor current in the supply circuit, and adjust the voltage until the test current falls between 50% and 125% of the motor’s rated ampere draw.
With the supply connected, slowly rotate the motor shaft by hand — a full revolution — while watching the ammeter continuously. A healthy rotor will show a steady, consistent current reading throughout the rotation. A broken or high-resistance bar will cause the current to rise and fall as different sections of the rotor cage pass through the magnetic field.
Interpreting the results
If the ammeter reading varies by more than 3% between its minimum and maximum values during a single rotation, that is a strong indicator of an open or damaged rotor bar. The current variation will repeat as many times per revolution as there are winding poles — so a 4-pole motor will show the variation four times per revolution.
In addition to watching the ammeter, pay attention to physical feedback and sound. A cogging sensation as you rotate the shaft — where the rotor seems to catch or resist in a repeating pattern — is another sign of a torque disturbance. You may also hear an increase in electrical noise (buzzing) as the damaged section passes through the field.
This test is particularly valuable because it can be conducted on-site, without removing the motor from its installation, and without specialized equipment beyond a variable voltage single phase supply and an ammeter.
Detection Method #3: Vibration Spectrum Analysis
Vibration analysis is ideally performed with the motor operating under load — ideally at least 50% of rated load, since electrical vibration faults increase in magnitude as load increases. Modern Fast Fourier Transform (FFT) vibration analyzers can resolve differences of only a few cycles per minute even at frequencies in the thousands of cpm, making them capable of detecting the subtle signatures that broken bars produce.
Key frequencies to look for
Electrical faults in a motor — including open rotor bars — commonly produce vibration at twice the line frequency. At 60 Hz, this is 7,200 cpm. However, twice-line-frequency vibration is not exclusive to rotor bar faults; rotor eccentricity, stator/rotor misalignment, an elliptical stator core, and open or shorted stator windings can all produce the same frequency. To determine whether the source is electrical or mechanical, de-energize the motor while it is vibrating. If the vibration disappears the instant power is removed, the fault is almost certainly electrical in nature.
More specific to rotor bar faults is vibration at the pole pass frequency — the product of the number of stator poles and the slip speed. For example, a 2-pole motor running at 3,450 rpm (with a synchronous speed of 3,600 rpm) has a slip speed of 150 rpm. Its pole pass frequency is 150 x 2 = 300 cpm. A motor with open bars will show a noticeable vibration peak at this relatively low frequency. An audible beat — a cyclic rise and fall in noise — may also be detectable.
Rotor cage faults typically produce high 1X (once-per-revolution) vibration along with sidebands spaced at the pole pass frequency around the 1X peak. Cracks in rotor bars can generate pole pass sidebands at multiples of operating speed. Open or loose bars have also been observed to produce elevated vibration at the rotor bar pass frequency (number of rotor bars times operating speed) and at twice-line-frequency sidebands around that frequency.
For example, a rotor with 56 bars spinning at 1,750 rpm has a bar pass frequency of 98,000 cpm. The twice-line-frequency (60 Hz) sidebands would appear at 90,800 cpm and 105,200 cpm — well into the high-frequency range, but detectable by modern FFT analyzers.
A note of caution
Vibration at twice line frequency can indicate stator or rotor problems — or both simultaneously. Determining which component is the source requires additional analysis. Also be aware that motors with spider-arm shaft constructions may exhibit rotor fault frequencies even with perfectly sound bars, if the number of spider legs equals the number of winding poles.
Readings are typically most informative when taken in the horizontal direction. Load level matters significantly: always try to test at or above 50% load. A rotor fault that is invisible at light load may produce a clear signature under heavier loading.
Detection Method #4: High Current Excitation (Thermal) Test
The high current excitation test — sometimes called the high current rotor test — is performed with the rotor removed from the motor. It uses a high-current, low-voltage single phase supply (such as a core tester or high-current transformer) connected to both ends of the rotor shaft, passing current through the rotor assembly.
How to perform the test
Connect the current source to the shaft ends and apply enough current — potentially hundreds to thousands of amperes depending on rotor size — to excite the rotor magnetically. At lower current levels, iron filings can be sprinkled over the rotor surface to reveal the magnetic pattern of the bars, similar to the growler test described below. A bar that does not attract filings is likely open or severely defective.
If the high current is sustained for a longer period — typically 15 to 60 minutes depending on the size of the rotor — the thermal pattern of the rotor becomes diagnostic. A healthy rotor will warm slowly and evenly. An open or high-resistance bar will appear as a hot spot, as current finds alternate paths through adjacent bars and laminations, creating concentrated heat at the fault location.
Detection and safety
Hot spots are best detected with an infrared thermometer or infrared camera. Do not use your hands to search for temperature differences — the rotor can become dangerously hot, and the magnetic field from the current source can affect some infrared instruments. Keep infrared devices at least 3 feet (1 meter) from the rotor core to avoid erroneous readings caused by the magnetic field.
This test is particularly effective for die-cast aluminum rotors, where porosity-related faults can be difficult to locate visually. It is generally performed in a service center environment, as it requires specialized equipment and safety precautions beyond what is practical in the field.
Note: Detection Method #6 (see below) with the growler also describes using iron filings and magnetic viewing paper in more detail, which can be done similarly with a core loss tester or a growler.
Detection Method #5: Current Spectrum Analysis (Motor Current Signature Analysis)
Motor Current Signature Analysis (MCSA) is one of the most precise and non-invasive methods available for detecting broken rotor bars in an operating motor. It works by analyzing the stator current to identify frequencies that are induced by rotor faults — frequencies that show up as sidebands around the fundamental supply frequency.
How it works
A current transformer is placed on one of the motor line leads to capture the current signal, which is then fed into an FFT spectrum analyzer. The fault frequencies of interest are the same pole pass sidebands described in the vibration section — specifically the sidebands appearing just above and below the line frequency at intervals equal to slip frequency times the number of poles.
The severity of a rotor bar fault is assessed by measuring the difference in amplitude (in decibels) between the line frequency component and the first lower sideband. A smaller dB difference indicates a more severe fault. As a general guideline: a difference greater than 54 dB typically indicates a healthy rotor, while differences in the range of 48 to 54 dB may suggest a developing fault. Differences below 48 dB are cause for concern and warrant more thorough investigation or repair.
These thresholds are guidelines, not absolute rules. MCSA is most powerful as a trend-monitoring tool: by taking baseline measurements when the motor is known to be healthy and comparing subsequent readings over time, gradual degradation can be detected before it causes failure.
Load requirements
MCSA requires the motor to be operating at load — at least 50% of rated load, and preferably closer to full load. At light loads, slip is low and the sideband frequencies move very close to the line frequency, making them difficult to resolve even with high-resolution analyzers. A motor that shows a borderline MCSA result at low load should be re-tested at higher load before conclusions are drawn.
MCSA is also sensitive to supply voltage imbalance, which can produce similar sidebands and lead to false positives. Confirm that supply voltages are balanced before interpreting MCSA results as definitive evidence of a rotor fault.
Detection Method #6: Growler Test
The growler test is one of the oldest and most direct methods for checking rotor bars. It requires the rotor to be removed from the motor but is straightforward to perform and immediately reveals missing magnetic responses from individual bars.
How to perform the test
A portable growler is placed against the side of the rotor, or the rotor is placed on a larger growler table. A sheet of paper is laid on top of the rotor surface, and iron filings are sprinkled onto the paper. When the growler is energized, the magnetic field it generates induces current in the rotor bars, which in turn creates localized magnetic fields that attract the iron filings. The filings will accumulate above each active bar.
A bar that is open — whether from a fracture, a void, or a bad joint — will not carry current and will not generate a magnetic field. The filings above that bar will not cluster, leaving a visible gap in the pattern. Work methodically around the entire rotor surface, re-sprinkling and checking until all bars have been assessed.
Using paper under the filings serves a practical purpose: it prevents the filings from sticking to the rotor after the growler is de-energized. An alternative is to place iron filings inside a sealed plastic bag and press the bag flat against the rotor surface — the filings rearrange inside the bag without touching the rotor directly.
Magnetic imaging materials
Some winding material suppliers offer magnetic imaging materials that produce a retained visual image of the bar pattern when the growler is energized. These materials have the advantage of holding the image even after the power is switched off, making it easier to document and evaluate results. The material can be reused, as each new energization overwrites the previous image.
Interpreting the results carefully
A gradual change in apparent magnetic field strength across the rotor surface does not always indicate a fault. If the rotor has been machined eccentrically, the varying distance between the rotor surface and the growler probe will cause differences in induced field strength that resemble — but are not — open bar signatures. Examine the rotor geometry before concluding that a gradual variation represents a defect.
Additional Detection Methods
The six methods above are the most commonly applied, but several other techniques are available for specific situations or to confirm suspected faults:
Tap test: On fabricated rotors, broken bars can sometimes be detected simply by tapping along each bar from end ring to end ring with a hammer and screwdriver. A loose or broken bar responds with a distinctly different sound and feel compared to a tight, intact bar. This method works best with two technicians — one tapping and one sensing movement at a point 3 to 4 inches away by resting a screwdriver on the bar.
Dye penetrant inspection: When visual inspection doesn’t reveal a fault but an open bar is suspected, dye penetrant testing can reveal surface cracks. Cleaner is applied, followed by a penetrating dye that flows into any cracks, and then a developer that draws the dye back out as a visible colored line. This method is especially useful for checking brazed joints on fabricated rotors and end ring surfaces on die-cast rotors. If the rotor has been painted, the paint must be removed first.
Digital low resistance ohmmeter (DLRO): A DLRO or resistance bridge measures micro-ohm resistance across individual rotor bars. A void, crack, or blow hole within a bar increases its resistance relative to adjacent bars. A variation of more than 5% from the average bar resistance is cause for suspicion. The same instrument can check end ring integrity by measuring resistance from bar to bar around each end ring. Tests can be repeated with the rotor heated in a bake oven to detect faults that only open up under thermal expansion.
Two-ammeter comparison: This assembled-motor test uses two identical analog ammeters to distinguish electrical faults (such as an open rotor) from mechanical causes of current variation. Both ammeters are first connected to the same line lead to confirm they track together. One is then moved to a different line lead. If the needle swings are simultaneous, the cause is mechanical. If one needle leads the other, the cause is electrical — an open rotor creates a phase lag that produces this sequential needle movement.
Stroboscope: A stroboscope synchronized to motor speed can reveal cyclic speed variations caused by rotor bar faults. A motor with broken bars produces torque pulsations that cause the rotor to repeatedly slow down and speed up. Under strobe illumination, a reference mark on the shaft or rotor will appear to shift back and forth rather than holding still. Note that a fluctuating mechanical load can produce similar effects.
Stray flux analysis: A sensing coil mounted concentrically on the outboard end of the motor captures stray magnetic flux leaking from the motor exterior. This flux contains information about internal electrical conditions. The coil output is analyzed with an FFT analyzer, using the same pole-pass sideband methodology as MCSA. The technique requires repeatable coil placement across tests; small position changes will affect results. There are no universally accepted pass/fail thresholds — trending over time is the primary value.
Ultrasonic testing: The end rings of die-cast rotors can be checked for internal voids using ultrasonic non-destructive testing equipment. The probe is placed on the outer periphery of the end ring and moved around its circumference. Internal voids reflect the ultrasonic signal differently from solid metal, identifying problem areas that are invisible on the surface.
Choosing the Right Approach
No single test is definitive in all circumstances. The best strategy depends on where the motor is, whether it can be removed and disassembled, and what equipment is available. The following practical principles apply:
- If the motor is installed and running: start with vibration spectrum analysis or MCSA. Both are non-invasive and can be performed without any downtime.
- If the motor is offline but still assembled: the single phase rotor test is the fastest and most accessible option.
- If the rotor has been removed: combine visual inspection, growler testing, and DLRO resistance measurements. For die-cast rotors, add high current excitation with infrared imaging. For fabricated rotors, add tap testing and dye penetrant on end ring joints.
- When a result is ambiguous: use a second method to confirm. A positive result from two independent methods is much stronger evidence than a single test.
Early detection is always preferable to waiting for a fault to become obvious. A motor running with broken bars draws higher current, runs hotter, and operates less efficiently — all conditions that shorten the life of other components. With the testing methods described here, broken rotor bars can be identified long before they cause catastrophic failure, allowing planned repair rather than emergency replacement.
When to Call a Professional Repair Shop
Several of the tests described above — including high current excitation with infrared imaging, dye penetrant inspection, ultrasonic end ring testing, and DLRO bar resistance measurement — require specialized equipment and expertise that are typically found only in a qualified motor repair shop.
If your field inspection or on-site testing indicates a probable rotor bar fault, the safest and most cost-effective next step is to send the rotor or the complete motor to a reputable service center for comprehensive evaluation.
At Electrical Apparatus, we have been diagnosing rotors and providing squirrel cage rotor repair since 1974. Our team will recommend the most appropriate repair strategy, whether that means rebarring, replacing laminations, or building a new rotor entirely. Contact us to discuss your motor’s symptoms and get an expert assessment.



