Direct Answer
**A motor's starting current surges to 5–7× rated for a few seconds — normal or faulty? The answer: for a direct-on-line start, 5–7× rated starting current lasting a few seconds is normal. The real abnormal signals are — starting current multiplier clearly too high, start time too long without settling, or obvious three-phase unbalance during start. Judge from the starting-current curve, not from "the current is huge" alone.**
At a campus, a 75 kW pump motor surged to 6.8× rated at every start and took 22 s to settle near rated — with an 18% three-phase deviation. The operator's first reaction was "starting current too high, the motor will burn." Reading the curve: the 6.8× peak was within the normal 5–7× band; the problem was the **settling time** — a healthy 75 kW motor settles in 3–8 s; 22 s was far too long, and the 18% deviation exceeded 10%. On-site investigation: bearing wear caused mechanical binding; the motor could not accelerate the load, dragging out the start. After bearing replacement, start time returned to 7 s with 3% deviation.
Because:
- **The large starting current is physics**: at the instant of starting, slip s=1 and rotor equivalent impedance is minimal, so current surges to 5–7× rated — normal; - **Abnormality is judged by settling time**: a normal start settles within seconds; stall or mechanical binding stays high — "not settling" is the danger signal; - **Abnormality also shows in phase balance**: a normal start is roughly symmetric (<10% deviation); broken rotor bars or winding faults amplify deviation; - **Start frequency affects life**: frequent starts accumulate rotor-bar thermal stress and winding temperature rise — every start consumes life.
Key insight: **5–7× starting current is not a fault; it is the "ID card" of a normal start. Excessive start time, phase unbalance, and high frequency are what to investigate.** Treating "big current" as a fault turns normal starts into false trip reasons; treating "normal peak" as safety misses stalls.
At a factory, a 30 kW fan was started 30+ times a day for the process. Six months later the insulation failed; teardown showed discolored overheated winding ends and multiple broken rotor bars. Review: start frequency far exceeded the motor's continuous-start allowance (roughly 3–4/hour); every start was a big-current + heat accumulation event, burning out six months of life — not "poor motor quality," but start-frequency overrun.
1. What a Normal Starting-Current Curve Looks Like
To judge an abnormal start, first know the shape of a normal one:
**Direct-on-line starting curve (three-phase induction motor)**: - **Peak segment**: at switch-on, current surges to 5–7× rated (locked-rotor multiplier) for a fraction of a second; - **Settling segment**: the motor accelerates, slip falls, current settles within 3–8 s to near rated (small motors 1–3 s, large motors 5–10 s); - **Steady segment**: at rated speed, current returns to load current (possibly 50–100% rated).
**Three features of a normal curve**: 1. Peak multiplier 5–7× (direct-on-line; exclude wound-rotor / soft-start / VFD); 2. Short settling time (seconds for small motors, within 10 s for large); 3. Roughly symmetric phases (<10% deviation).
**Mantra**: peak → multiplier (5–7 normal); settling → time (seconds normal); symmetry → deviation (<10% normal) — all three normal, the start is normal.
Key insight: **"big" is not the problem; "long" and "skewed" are.** The curve shape carries far more information than the peak number.
2. Where the 5–7× Comes From Physically
Why does starting current reach 5–7× rated? It follows from the motor's equivalent circuit:
**The instant of starting**: - Rotor not yet turning, slip s=1 (maximum relative speed to the rotating field); - Rotor equivalent impedance minimal (rotor current frequency = supply frequency, small reactance); - Starting current I_st = U / (Z₁ + Z₂′), roughly 5–7× rated (locked-rotor multiplier, on the nameplate or catalog).
**Key numbers**: - Direct-on-line locked-rotor multiplier: typically 5–7× (nameplate I_LRC/I_n); - Starting torque: typically 1.5–2× rated (torque ∝ U² — low voltage collapses torque fast); - Large motors (>100 kW) start longer with a slightly lower multiplier (4–6×) but far larger absolute current.
**Why "multiplier" matters more than "absolute current"**: - For the same motor, the multiplier is roughly constant (factory characteristic); - A clearly higher multiplier (7×+ sustained) hints at rotor or supply anomalies; - A normal multiplier with large absolute current is just a bigger motor behaving normally — judging by "huge current" unfairly penalizes large motors.
Key insight: **5–7× is a nameplate characteristic, not a fault value.** Judge against the motor's own characteristic (multiplier, settling time), not against expectation.
3. Why Start Time Becomes Too Long
Slow settling (long start time) has three main causes:
**1. Low voltage**: - Starting torque ∝ U²: at 90% voltage, torque drops to 81%, acceleration slows; - Start time stretches, locked-rotor current lasts longer, winding heat rises sharply; - Typical: long feeders with voltage drop, undersized transformer, multiple motors starting together.
**2. Mechanical binding**: - Bearing wear, pump impeller blockage, fan unbalance — load torque exceeds the starting-torque margin; - The motor "cannot carry it," accelerates slowly or stalls in the low-speed zone, current stays high; - Typical: start time stretches from 5 s to 20 s+, with noise and vibration.
**3. Motor body degradation**: - Broken rotor bars / cracked end rings — lower starting torque, slower acceleration; - Inter-turn winding short — phase unbalance, slower start with local overheating; - Bearing jamming (electrical + mechanical combined).
**Engineering judgment**: - Start time above 2× the normal value for the same motor model → investigate; - Low voltage: judge by the voltage dip at start (bus dropping below 85% rated at switch-on is a warning); - Mechanical binding: combine post-start noise + vibration + slow settling.
Key insight: **start time is the most-overlooked abnormal-start signal.** It reveals "can the motor actually carry the load" better than the peak multiplier.
4. Distinguishing Start Data from Stall
The worst misjudgment is confusing an abnormal start with a stall (or vice versa). Three differences:
**1. Settling time**: - Normal start: current settles within seconds, even a long start is accelerating; - Stall: current stays at 5–7× without settling; rotor locked; - Criterion: start time over 2–3× normal with no settling trend = near-stall.
**2. Phase balance**: - Normal start: <10% deviation; - Broken rotor bars: unbalance grows during start (higher current in the broken phase), shrinking at steady state; - Winding fault: one phase clearly abnormal, deviation persists.
**3. Curve shape**: - Normal start: smooth settle after the peak; - Stall/binding: slow settle or a "plateau" (current parked above 4×); - Low voltage: slightly lower peak (not enough torque) but slower settle and more heating.
**Protection coordination**: the thermal relay's trip time must clear the normal start time (set by start duration); stall protection (e.g., 7× for 10 s) coordinates with the starting curve — no false trips on normal starts, fast trip on real stalls.
Key insight: **stall is not "big current"; it is "big current that does not settle."** With a curve, read the curve; without one, use time + balance.
5. Start Frequency and Life
"Starting a few times is fine" is a misconception — every start consumes motor life:
**Thermal effect of starting**: - Each start drives winding and rotor bars through a big-current thermal shock (I²t accumulation); - At 5–7×, heat scales with current squared — one start equals minutes to tens of minutes of rated operation; - Bars endure electric + thermal stress at start; repeated starting causes fatigue fracture.
**Frequency limits (engineering experience)**: - Typical motors: no more than 3–6 continuous starts per hour (fewer for large motors, load-dependent); - Leave a cooling interval between starts (typically 10–30 min, rise-dependent); - Cold state allows 2–3 consecutive starts; hot state suggests 1 start then a cooling interval; - Frequent-start duty (hoists, presses) needs "frequent-start" motors (thicker bars, better cooling).
**How to read the data**: - Log daily start count and intervals; - Count over the limit plus high temperature rise = heat accumulating, life burning faster; - Persistently abnormal starting curve + high frequency = double damage, address first.
Key insight: **start frequency is an invisible life tax.** One start looks harmless; 30 starts a day, 10,000 a year, bars and winding age fast — continuous start-counting puts the life ledger in clear view.
6. Product Landing: Continuous Monitoring of the Starting Curve
FEXLINK's **ESB Three-Phase Unbalance Monitor** turns "starting-curve judgment" into automatic recognition:
- Samples three-phase current continuously, auto-detects every start event (surge + settling features); - Auto-logs three metrics per start: peak multiplier, settling time, three-phase deviation; - Threshold alerts: start time too long (mechanical binding / low voltage), phase deviation >10% (broken bars / winding fault), start frequency overrun (heat-accumulation risk); - Starting curves are replayable: each start's waveform stored, trend comparison shows "the start is getting worse" at a glance; - Connected to the **Digital Electrical Intelligent Supervision System**, start events, protection actions, and temperature share one dashboard for clear motor-health status.
Design principle in one sentence: **turn "startle at the starting surge" into "automatic starting-curve reading."** 5–7× is normal; a bad curve is the signal — the monitor tells them apart.
7. Common On-Site Mistakes
1. **Treating 5–7× as a fault**: reporting "motor abnormal" at a normal peak, even tampering with protection settings. Consequence: false trips on normal starts, protection neutered. Correct: check multiplier + settling time + phase deviation first. 2. **Watching only the peak**: reassured by 6×, missing the 22 s without settling. Consequence: stall/binding missed. Correct: settling time is the key stall indicator. 3. **Ignoring phase unbalance**: 18% deviation at start brushed off. Consequence: broken bars discovered only after full fracture. Correct: >10% deviation at start → inspect rotor and winding. 4. **Not chasing long start time**: 20 s+ every start "at least it starts." Consequence: low voltage / binding keep worsening. Correct: 2× the normal value demands investigation. 5. **Ignoring start frequency**: 30 starts a day "always done this way." Consequence: broken bars, insulation failure, premature scrapping. Correct: count starts; over the limit, change the process or choose a frequent-start motor.
Scope and Limitations
This article discusses starting-current curve reading for three-phase induction motors in direct-on-line, power-frequency starts (wound-rotor, soft-start, and VFD characteristics differ; thresholds need adjustment). Locked-rotor multiplier and start time follow the nameplate and manufacturer data. Protection settings (thermal relay, stall protection) must be coordinated with the starting curve; the values here are engineering references. All cases are anonymized (a certain campus / factory) and do not refer to any specific customer.