Before the work request
Three readings show a pattern. They do not name a cause.
A reading of 57 A beside 48 A and 49 A is hard to ignore. In a conveyor review, that high line current can quickly become the reason given for replacing the motor. Before accepting the work request, establish what the readings actually tell you.
The supply, a resistive connection and the motor itself all deserve a look. Changing conveyor duty raises a different question: were the three readings taken together, or one after another as the load changed? Sequential readings can make the lines look unequal. A difference that persists in a synchronised record still needs an explanation.
The calculation below uses assumed current and voltage readings to work through that decision. Its numbers are not operating limits.
Before the arithmetic
First make sure the three readings can be compared.
Start with the record collected by the qualified electrical team. Measurements and thermal surveys around energised equipment must follow the site's electrical safety rules; this review is not an instruction to open an enclosure, connect an instrument or tighten a terminal. Any isolation must follow the site's procedure. OSHA 29 CFR 1910.333 sets out relevant electrical work practices in the United States; local law and site rules apply elsewhere.
Four details make the record usable:
- Identity and locationThe motor, feeder and measurement point, whether the currents came from installed current transformers or a portable meter, and the instrument and range used.
- Supply arrangementLine-fed or drive-fed and the supply frequency. For a drive, whether each reading belongs to the drive input or the motor side, taken with an instrument suited to the drive waveform.
- Operating stateSpeed, process duty, any load indication, time since starting and anything that changed while the readings were taken.
- Comparison basisAll three line currents and all three line-to-line voltages, labelled and taken in a common time window, plus an earlier comparable record if one exists.
Even with steady throughput, avoid treating current as a direct load measurement. Tonnes moving along the belt do not give you shaft torque. Megger's article on instantaneous torque describes a motor whose load estimate from current alone was well below the load found by a fuller analysis.
Work through the record
Work through the conveyor record.
Use a line-fed induction motor on a 50 Hz supply for the worked calculation. Assume it is running at 1,480 rpm with steady conveyor throughput during the readings.
I-L1, I-L2 and I-L3, in that order. The average is 51.33 A, and the largest departure from it is 5.67 A, on I-L3.
L1-L2, L2-L3 and L3-L1. The average is 411 V, and the largest departure from it is 4 V, on L3-L1.
1,480 rpm with steady conveyor throughput during the readings. Throughput is context, not a torque measurement.
The example gives no motor rating, load percentage, protection setting or acceptance limit, so none can be inferred from it.
To summarise the spread, find the reading furthest from the three-value average. Divide that departure by the average, then multiply by 100. Fluke describes this maximum-deviation calculation for both voltage and current.
The tempting reading is that 0.97% looks small, so the supply must be fine and the motor must be at fault. That conclusion does not follow. In an induction motor, a small voltage imbalance can produce a much larger current imbalance. How much either figure matters depends on the motor, supply and protection requirements. These indices describe the spread; they are neither acceptance limits nor sequence-component measurements, and they do not establish whether operation may continue.
Keep the line labels straight. L1-L2 measures voltage between two lines; I-L1 measures current in one line. They do not describe the same winding. You cannot pair the largest voltage departure, on L3-L1, with a single line current to locate a fault. In a delta-connected motor, line current also differs from the current in an individual winding.
Each reading as a departure from its own three-value average. The voltage spread is small, but that alone does not clear the supply, and the voltage pairs do not line up with individual line currents.
Four places to look
Keep four explanations open until the evidence separates them.
Put each possible cause beside a check that could support or weaken it. A current alarm, a warm casing and a vibration change may be related, but counting all three as proof of a motor fault skips the diagnostic work.
- SupplyCompare the voltage record with a suitable upstream record or another motor on the same supply. That comparison helps distinguish a wider supply problem from one local to this feeder. Current readings alone cannot do that.
- Connections and switchingA resistive joint, worn contactor or poor termination can unbalance the voltage reaching the motor. Thermal imaging can help locate a connection for isolated inspection, although it cannot assess a contact hidden from the camera.
- The motorIf the pattern remains after the measurement, supply and connection checks, select a motor test for the suspected mechanism. A high line current alone does not identify broken rotor bars, a winding fault or a bad bearing.
- Driven equipment and dutyCheck the conveyor loading, transmission and mechanical condition. They affect current magnitudes. Establish whether the apparent inequality survives a simultaneous measurement at comparable duty before treating it as a persistent electrical pattern.
A second observation
Use temperature to choose where to look, not to deliver the verdict.
Now suppose the qualified electrical team takes a thermal record at the same steady duty. One feeder connection is warmer than the conductor around it, and the temperature falls away with distance from the joint. That finding gives the investigation somewhere specific to go.
Fluke's guidance on thermal imaging of motors and drives distinguishes this local pattern from even heating along an overloaded or undersized conductor. Read the image alongside current and duty, compare equivalent surfaces, and retain the location, visible image, thermal image and camera settings. A bright patch is not enough: emissivity and reflected temperatures can distort the result. Emissivity describes how effectively a surface emits thermal radiation; shiny metal can reflect heat from elsewhere.
The warm connection moves the feeder path up the list, but it need not explain the whole pattern or sit on the line carrying 57 A. A resistive joint changes the voltages reaching the motor; the highest current does not necessarily identify that joint. The motor may still have a fault, and electrical and mechanical problems can coexist.
Mechanical evidence
Ask vibration and ultrasound a mechanical question.
Vibration and ultrasound become useful when you ask them something specific. Is the change concentrated at one motor bearing? Does a repeatable vibration feature follow the transmission? Does a comparable ultrasonic recording support a local friction concern?
Keep the marked measurement point, sensor setup, speed and load with each record. The UE Systems Ultraprobe 9000 manual calls for the same bearing type, similar load and the same rotational speed in ultrasonic comparisons. A stronger bearing finding may justify a parallel mechanical check. It does not finish the electrical investigation, just as an unchanged overall vibration value cannot clear every motor fault.
ISO 13373-9:2026, published in August 2026, covers vibration-based diagnostics for electric motors. Its public scope excludes using voltage and current to identify vibration-related motor faults. Both approaches can inform the review, but they are not interchangeable. Oil analysis belongs here only if an oil-lubricated compartment, such as the conveyor gearbox, has its own question; it cannot assess the motor's electrical circuit.
Electrical signature analysis
Three current readings are not a current signature analysis.
Motor current analysis includes a more detailed online technique: motor current signature analysis, or MCSA. It examines the operating current signal, including its frequency content, in the context of the motor's design and duty. Three RMS values are a useful starting observation, but they do not provide that signature.
ISO 20958:2013 covers online electrical signature analysis of three-phase induction motors. Its public catalogue page describes that scope, and this article does not reproduce the standard's procedures.
Our field note on five monitoring methods introduces MCSA alongside the other disciplines. To take this record further, a specialist needs recorded current and voltage signals, pole count, supply or drive details, speed, load and acquisition settings.
Speed matters because of slip: the difference between rotor speed and the rotating magnetic field's speed. If the nameplate confirms four poles, the field at 50 Hz turns at 1,500 rpm. A measured 1,480 rpm would then give 20 rpm of slip, about 1.3%. Slip changes with load and shifts some signature features, so a spectral peak needs operating context and a comparable record before it earns a fault label.
Online signature analysis and offline circuit testing answer different questions, and neither replaces the other. Choose the test for the question the evidence has raised.
Offline tests examine an isolated, verified circuit. Depending on the selected method, they measure properties such as winding resistance balance or insulation condition. They cannot reproduce the operating signature, while an online record cannot replace every offline test. Megger's article on combined static and dynamic testing uses current imbalance as an example of an online finding that offline tests can follow up.
One abbreviation needs care in a work request. In motor-testing literature, MCA can also mean motor circuit analysis, an offline test, rather than motor current analysis. Write the method out in full so the right test is booked.
Technical sources
Primary references behind this field note.
Standards links point to public catalogue pages; the full standards were not reviewed for this article. They do not imply ConditionCore certification or conformity with a particular standard.
- ISO 20958:2013Condition monitoring and diagnostics of machine systems, electrical signature analysis of three-phase induction motors. Public scope for online analysis. Published August 2013, confirmed in 2023.
- ISO 13373-9:2026Vibration condition monitoring, diagnostic techniques for electric motors. Edition 2, published August 2026. Its public scope does not cover using voltage or current to identify vibration-related motor faults.
- Fluke: Transients and voltage unbalanceThe maximum-deviation calculation for voltage and current, and supply and connection causes of imbalance.
- Fluke: Using thermal imaging to troubleshoot motors and drivesLocalised connection heating compared with even heating along a conductor, with load context.
- Fluke: 10 things you need to know about thermal imagersEmissivity and reflected temperatures, item 8.
- UE Systems: Ultraprobe 9000 instruction manualVersion 2, comparative bearing test guidance on printed page 28.
- Megger: The Case for Combined Static and Dynamic TestingOffline and online motor testing. Published 4 March 2026.
- Megger: Instantaneous torque as a predictive maintenance toolThe limits of estimating load from current alone. Published 4 December 2020.
- OSHA 29 CFR 1910.333Selection and use of electrical safety-related work practices. A US regulation, not a substitute for local requirements.
Back at the work request
Replace the motor when the evidence points to the motor.
For this worked record, investigate the feeder path and connection first. Keep the motor test ready if those checks leave the current pattern unexplained.
After authorised corrective work, repeat the observations at comparable duty and record what was found and changed. Closer current readings support the repair, but also revisit the original hot connection and any unresolved motor or load findings. If the difference persists, investigate the next explanation instead of repeating the same repair. Our field note on post-repair verification covers the recheck in more detail.
For a second view of a current, temperature or vibration record, bring the raw readings, their locations and times, the operating state and the explanations still open to a ConditionCore machine review.
Review a machine with ConditionCore