Current IntelPower LibraryShop main store

Troubleshooting

Troubleshooting intermittent instantaneous motor start trips

Motors and motor starters Published 9 min read

Short answer

An intermittent instantaneous trip on motor start is usually the random closing angle of the contactor, not a failing motor. A first peak can reach about 2.8 times the symmetrical locked-rotor current, so only starts that close near a voltage zero hit magnetic pickup. Capture all three phase currents, rank contactor pole, winding, rotor and breaker causes, and let the trip follow the part you roll.

On this page
  1. Why does my motor trip instantaneously on start only sometimes?
  2. What should I measure before anyone swaps a component?
  3. Which causes do I check first, and in what order?
  4. Can one bad contactor pole trip a start that usually succeeds?
  5. What motor fault passes a megger test and still trips the breaker?
  6. Can rotor bar damage raise the first peak enough to trip magnetically?
  7. How do I calculate the worst-case first peak and check the breaker against it?
  8. When does this actually justify replacing the motor?
  9. Questions asked

If a motor trips the breaker instantaneously on start only sometimes, the motor itself is the least likely cause, and the random point on the voltage wave where the contactor closes is the most likely. Starting current lags the voltage by nearly ninety degrees, so each closure plants a decaying DC offset in the windings; close near a voltage zero and the first peak can reach roughly 2.8 times the symmetrical locked-rotor current, close near a peak and it barely rises above it. That is the intermittency. Before anyone orders a motor, capture all three phase currents at closure across a dozen or more starts and work the ranked causes in this order: contactor pole timing, breaker magnetic pickup, turn insulation, rotor cage. The trip follows the component that owns the abnormality.

Why does my motor trip instantaneously on start only sometimes? #

Because the first peak of starting current depends on where on the voltage wave the contactor closes: near a voltage zero, DC offset roughly doubles the peak, so only those unlucky starts reach a first peak near 2.8 times the symmetrical locked-rotor RMS current.

With the rotor stationary there is no rotational voltage in the winding; it behaves as resistance in series with inductance, so the current lags the applied voltage by close to 90 degrees. Inductive current cannot step instantaneously: at closure each phase carries a DC component that starts equal and opposite to its instantaneous AC value and decays exponentially with the loop's X/R ratio. The three poles close together, but each phase sits at a different point on its own voltage wave at that instant, so each phase gets its own offset — almost none if that phase closes at its voltage maximum, nearly a full peak if it closes at zero. The elevated peaks land on alternate half-cycles after closure, which is visible on any captured trace. This randomness is a property of switching, not of the machine: it looks exactly like an intermittent fault and is not one.

What should I measure before anyone swaps a component? #

Capture all three phase currents simultaneously from the instant of closure, on every start across at least a dozen attempts, and log the breaker's trip indication each time; single-phase or averaged readings miss the pole-to-pole differences that identify the cause.

  1. De-energize, lock out and verify absence of voltage with a tested meter before opening the starter or breaker enclosure: NFPA 70E, 2024 edition, covers safe work practices around electrical hazards, and your site's safe-work procedure and the local authority having jurisdiction govern the details.
  2. Fit three matched current channels rated for the circuit, triggered on the contactor coil command, capturing the first 20 cycles at high resolution and the whole acceleration at low resolution.
  3. Record the motor terminal voltage on the same trigger to expose supply-side dips or imbalance during the start.
  4. Repeat across a dozen or more starts and save every file; the comparison that matters is trip start versus clean start.
  5. Compare like with like: symmetrical RMS against an RMS reference, peak against a peak criterion. Never hold a scope peak against an RMS setting without the device's operating characteristic in between.

Which causes do I check first, and in what order? #

Check the contactor first, the breaker pickup second, the winding third and the rotor last; that order tracks both the probability of each cause and the cost of being wrong, and the first two explain most intermittent instantaneous trips on motor start.

  1. Contactor pole making late or with weak pressure. Signature: one phase's trace steps, lags or crosses the others in the first cycles after closure. Action: inspect springs and contact faces with the panel locked out, then roll the starter if a spare exists.
  2. Breaker magnetic pickup inside the real envelope. Signature: waveforms inside the expected starting envelope while the breaker still flags instantaneous. Action: verify the installed trip unit's labeled pickup against the worst-case calculation below.
  3. Turn-to-turn winding fault. Signature: an imbalance that repeats every start with the same phase worst, or a phase-to-phase resistance mismatch. Action: winding resistance comparison, then a motor-shop turn-insulation test.
  4. Rotor cage damage. Signature: normal first peak but slow acceleration with current high for seconds. Action: this explains thermal and overload-time trips, never a genuinely instantaneous one; pursue it only if the duration pattern is there.
  5. Supply-side problem. Signature: motor currents balanced while terminal voltage dips unevenly during the start. Action: chase upstream connections and impedance using the terminal-voltage capture.

One more rule: a component that merely prolongs the current without raising its peak cannot trip a magnetic element on its own, so a longer start changes what a thermal element sees and nothing else.

Can one bad contactor pole trip a start that usually succeeds? #

Yes: a pole that closes late or with weak contact pressure lets two phases establish their current division first, then drops the third in milliseconds later, and that abrupt re-distribution can push the transient past anything a balanced closure produces.

On a wye-connected winding the two leading phases set a temporary voltage and phase relationship that the late pole then destroys. The motor runs normally afterward, which is why the fault hides: the evidence lives in the first three or four cycles, not in the running current. With the panel locked out as in the measurement section, inspect for loose, weak, broken or burned contact-pressure springs, burned or contaminated faces, tracking on the pole structure and mechanical binding, and check the line- and load-side connections to the manufacturer's requirements rather than an assumed torque.

If a similar starter is available, roll it — starter to another motor, or motor to another starter — changing one variable at a time; the trip follows the component that owns the abnormality. Judge the contactor on what each pole does in the first cycles after closure, never on the fact that the motor runs fine after the event.

What motor fault passes a megger test and still trips the breaker? #

Turn-to-turn insulation failure inside a winding: a megger applies test voltage from the winding to the grounded frame, so a short between adjacent turns can pass that test while still shifting the phase's impedance and unbalancing the starting current.

The consequential evidence is collectable: a starting-current imbalance that repeats on every start with the same phase worst (random closing angle moves the peak, but not a persistent imbalance), a phase-to-phase resistance mismatch, localized heating, or recurring trips that insulation testing never explains. Measure winding resistance phase-to-phase with a low-resistance ohmmeter or bridge, correct for temperature where the method requires it, and compare the three readings against each other and the manufacturer's comparison limit.

While the terminal box is open, check the connection configuration against the nameplate diagram, because a deteriorated or altered connection mimics a winding fault. If resistance or the captured waveforms point at the winding, the verdict on turn insulation belongs to a motor shop equipped to stress insulation between turns: a ground-insulation pass is not a verdict on turn insulation, and that gap is what keeps a failing motor looking healthy across repeated megger checks.

Can rotor bar damage raise the first peak enough to trip magnetically? #

No: broken bars, cracked end rings or high-resistance cage joints cut starting torque and stretch acceleration, but the first-cycle peak is set by locked-rotor impedance and DC offset, not by rotor condition.

The rotor signature is duration: current above the normal acceleration band for seconds, slow speed build-up and torque pulsation as damaged bars pass the stator field. That pattern loads the thermal and overload-time elements, so it can produce nuisance trips — just not instantaneous ones. Tell them apart on the captured trace: read the peak from the first few cycles and the duration from the whole acceleration.

Comparing measured current with the nameplate only works when you compare the same kind of quantity, the distinction we drew in a motor running over nameplate amps after a soft starter bypass.

How do I calculate the worst-case first peak and check the breaker against it? #

Read the nameplate code letter to get locked-rotor kilovolt-amperes per horsepower, compute the symmetrical locked-rotor RMS current, multiply by √2 for the peak and allow up to 2 again for maximum DC offset: a screening bound of roughly 2.8 times the symmetrical RMS value.

Worked example, illustrative — substitute your own nameplate before using it: a 460 V, three-phase, 50 hp motor carrying NEMA code letter G, which NFPA 70, the National Electrical Code, 2023 edition, Table 430.7(B) defines as 5.6–6.3 locked-rotor kVA/hp.

  • Locked-rotor kVA at the top of the band: 6.3 kVA/hp × 50 hp = 315 kVA
  • Symmetrical locked-rotor RMS current: 315,000 VA ÷ (1.732 × 460 V) = 315,000 VA ÷ 796.7 V ≈ 395 A
  • Symmetrical peak: 395 A × 1.414 ≈ 559 A
  • Screening bound for the first peak, fully offset: 2 × 559 A ≈ 1120 A, about 2.8 × 395 A

The real peak lands between 559 A and 1120 A depending on the loop's X/R ratio, the residual voltage at closure and the decay constant. Treat 1120 A as the screening bound to compare against the breaker curve, not as a measurement, and verify the inputs against the governing code and the equipment documentation.

Then read the installed trip unit, not the historical one: a replacement breaker can carry a different pickup strategy or tolerance than the device it replaced, and many molded-case breakers express magnetic pickup as a multiple of the trip unit's rating with a published tolerance band per model. The decisions fall out directly:

  • Labeled pickup clear of 1120 A with margin: closing-angle randomness cannot trip this breaker, so an instantaneous trip means a real transient (contactor, winding) or a defective device. Find the transient.
  • Labeled pickup inside or below the 559–1120 A envelope: the intermittency is fully explained without any motor fault. Set the pickup above the envelope within the breaker's listed range while it still protects the conductors as NFPA 70, 2023 edition, Article 430 Part IV requires — the same branch-circuit logic as in a #14 AWG motor branch circuit on a 20 A breaker — and confirm the change with your authority having jurisdiction.

Setting an instantaneous pickup above a calculated first peak is the same exercise as configuring a relay's instantaneous overcurrent element pickup. The same first-peak physics, with no fault at all, is why transferring onto a motor with residual voltage can trip a magnetic element instantly, as covered in ATS transfer surges that exceed breaker instantaneous pickup.

When does this actually justify replacing the motor? #

Replace or rewind only when the abnormality follows the motor: a controlled swap where the trips move with the machine, a persistent same-phase imbalance or resistance mismatch across repeated starts, or a failed turn-insulation test at the motor shop.

  • Trip follows the starter in the roll → rebuild or replace the contactor, and recheck pole timing and spring condition after the swap.
  • Waveforms clean but the pickup sits inside the envelope → set pickup margin or fit a trip unit whose band clears the calculated bound; the motor is innocent.
  • Resistance mismatch or every-start imbalance follows the motor → motor shop for turn-insulation testing before spending on a replacement.
  • Slow acceleration with a normal first peak → rotor condition and the driven load; that was never an instantaneous problem.
  • A dozen or more clean captures with the pickup verified clear → the remaining trips are the closing-angle lottery, and the fix is pickup margin, not a new motor.

Twenty years of quiet service makes a motor neither guilty nor exempt. The fault follows whoever owns the abnormal waveform, and for most intermittent instantaneous trips on motor start that is the contactor, the pickup setting or plain closing-angle statistics.

Questions asked

We already replaced the breaker after the first trip, so can the breaker still be the cause?

Yes. A like-for-like replacement with the same pickup band reproduces the same marginal setting, and a substitute from a different manufacturer can change the magnetic pickup or its tolerance. Compare the installed trip unit's labeled pickup against your calculated worst-case first peak of about 2.8 times symmetrical locked-rotor current before treating the breaker as proven.

Which trip element matters here, the magnetic one or the thermal one?

A trip at or just after closure points to the magnetic element. If the breaker holds through the first half-second and trips seconds into the start, you are dealing with a thermal or overload-time problem, and slow acceleration from rotor or load trouble becomes credible. Duration alone never explains a genuinely instantaneous trip.

How many starts should I record before I trust the pattern?

The closing angle is effectively random, so treat each start as a sample. Record at least a dozen starts with all three phases captured, including any start the breaker flags. If the worst first peak stays below your calculated bound and no phase shows a step at closure, the evidence points away from the motor.

Can I simply raise the instantaneous pickup to stop the nuisance trips?

Only within the breaker's listed adjustment range, and only while it still protects the conductors: NFPA 70, the National Electrical Code, 2023 edition, Article 430 Part IV requires branch-circuit short-circuit protection adequate for the circuit. Check the manufacturer's published curve, keep your peak calculation on file, and confirm the change with the authority having jurisdiction.

Keep reading

  • Selection guide

    Selecting overload and jam protection for a reversible motor

    Set a bimetallic or electronic overload relay from the motor's nameplate full-load current — never from wattage — and give the electronic-clutch trip to a separate current-sensing jam or shock relay set above the highest normal process current. Its adjustable reset delay must outlast the cycle timer's worst-case command drop-out, and auto reset needs a deliberate-restart decision because a live run command restarts the motor.

    Motors and motor starters

  • How-to

    Wiring a three-pole contactor for a single-phase motor

    Run the single-phase current through two poles in series per the overload maker's diagram — supply to terminals 1 and 5, a link between 3 and 6, motor leads at 2 and 4 — and never parallel all three poles. Set the overload at 125% of the 6 A nameplate full-load current (7.5 A) when the service factor is 1.15 or higher or the temperature rise is 40 °C or less, otherwise 115% (6.9 A), per NFPA 70 (NEC) 430.32.

    Motors and motor starters

  • How-to

    Motor starting voltage dip calculation on an 11 kV bus

    A starting medium-voltage motor dips the bus in proportion to its locked-rotor kVA against the source's short-circuit strength. Screen it as dip ≈ 100 × S_start/S_sc at the point of common coupling, then refine with ΔV = I_start(R cosφ + X sinφ) per phase. In the worked 11 kV, 1.5 MVA chiller example a direct-on-line start gives 3.6% at the PCC and 4.0% at the plant bus, failing a 3% limit; a 65% tap autotransformer or a soft starter limited to 3× rated current holds both points under 2%.

    Motors and motor starters