Troubleshooting
Motor running over nameplate amps after soft starter bypass
Short answer
The motor averages 192.7 A against 164.0 A FLA with only about 0.90% voltage-magnitude unbalance, so the supply is not the proven cause. Treat excess shaft load or excess motor loss first, and confirm negative-sequence voltage with a three-phase power-quality analyzer reading V2/V1, total kW and speed before replacing the starter or the motor.
On this page
- Why is my motor drawing more amps than the nameplate?
- Can 0.9% voltage unbalance cause this much overcurrent?
- Can three RMS meter readings prove negative-sequence voltage?
- How much power is 192.7 A actually buying?
- What measurements settle it, and in what order?
- Could the soft starter or the padmount transformer be at fault?
- Should I raise the overload setting to stop the trips?
- What do I do with each measurement outcome?
- Questions asked
Average current is 192.7 A against a 164.0 A nameplate, and three line-to-line RMS readings — 476, 482 and 483 V — cannot prove negative-sequence voltage, because they carry no phase angles. With about 0.90% voltage-magnitude unbalance and 3.3% current unbalance, the supply stays on the checklist, but a 17.5% average overcurrent points first at excess shaft load or excess motor loss. Confirm with a three-phase power-quality analyzer reading V2/V1, total kilowatts and shaft speed, plus the pump's operating data. The soft starter is protecting the motor exactly as designed; the answer is not a new starter and not a higher overload setting.
Why is my motor drawing more amps than the nameplate? #
Current follows two things only — shaft load and losses — so at 192.7 A average the motor is demanding 17.5% more than its 164.0 A rating: more load or more loss, and every test below exists to tell those two apart.
The measured phases are 184 A, 195 A and 199 A — respectively 12.2%, 18.9% and 21.3% above the 164.0 A full-load amps on a 460 V, 3555 RPM nameplate carrying a 1.15 service factor, 94.5% efficiency and 90.8% power factor. A manufacturing error such as a short winding count is an unlikely explanation: nameplate full-load current is a tested figure, not a calculated one. This is heat, not logic. Stator copper loss rises roughly with current squared, so the 199 A phase dissipates about (199 ÷ 164)² = 1.47, or 47% more copper heating than at nameplate current, before any negative-sequence rotor heating.
Can 0.9% voltage unbalance cause this much overcurrent? #
No. NEMA MG1 Part 14.35 keeps voltage unbalance at 1% or less because current unbalance typically runs 6 to 10 times higher; your 3.3% current unbalance is consistent with 0.90% voltage unbalance, and neither number explains a 17.5% rise in the average.
Unbalance is calculated as 100 × (largest deviation from the three-reading mean) ÷ (that mean), the definition NEMA uses. Running voltages of 476, 482 and 483 V give a 480.3 V mean, a 4.3 V maximum deviation and 0.90% unbalance; the prestart set of 484, 488 and 488 V gives 0.55%. The currents give a 192.7 A mean, a 6.3 A deviation and 3.3% — below the 5.4% to 9.0% band that 0.90% voltage unbalance typically drives at six to ten times.
NEMA MG1 is the US motors-and-generators standard, and the guidance cited here traces to its Part 14.35 on unbalanced voltage; confirm the edition you hold and the motor documentation. MG1 also publishes a derating-factor curve for unbalanced operation — apply that curve rather than a guessed number.
The mechanism is real but it redistributes rather than inflates. The negative-sequence component produces a field rotating opposite the rotor, inducing rotor currents and heating out of proportion to the percentage, which is why 1% is the advisory ceiling. Unbalance can push the highest phase higher and cook it early; it cannot explain a 17.5% rise in the average of all three.
Can three RMS meter readings prove negative-sequence voltage? #
They cannot. Negative-sequence voltage is a symmetrical-component quantity set by both magnitudes and phase angles, and three line-to-line RMS magnitudes carry no angle information, so V2/V1 cannot be computed from your multimeter readings.
Use a three-phase power-quality analyzer that reports sequence components directly: V2/V1, phase angles, per-phase and total kilowatts, power factor, harmonics and current unbalance. Measure at the padmount transformer secondary, the starter line side and the motor terminals; the point where V2 rises marks the section that creates it. A hand-held phase-sequence indicator confirms rotation order (A-B-C) only — it says nothing about sequence-component magnitudes. And the utility phases do not have to visibly shift from 120° to hurt the motor: unbalanced source impedance in a transformer, conductor or termination creates negative sequence on its own.
How much power is 192.7 A actually buying? #
On balanced assumptions it is about 137.6 kW at the shaft — roughly 184 hp — against a pump curve that promises 130 hp, and resolving that 54 hp gap with measured kilowatts is the whole diagnostic.
The screening calculation, with every input named:
- Apparent power: S = √3 × V × I = 1.732 × 480.3 V × 192.7 A = 160.3 kVA
- Input power: P = S × PF = 160.3 kVA × 0.908 = 145.6 kW
- Shaft power: 145.6 kW × 0.945 = 137.6 kW ÷ 0.746 kW/hp = 184 hp
It assumes balanced currents and that the 90.8% power factor and 94.5% efficiency hold at an operating point 17.5% above full load — both doubtful — so treat the result as an estimate to check against analyzer kilowatts and the equipment documentation. The pump's 130 hp is about 97 kW at the shaft.
Then the branch: if the analyzer reads total power near 145 kW, the current is buying real power and the question turns mechanical — measured speed against curve speed (pump power varies with the cube of speed; 5% fast is 1.05³ ≈ 1.16, or 16% more power), fluid density, rotation direction, rub. If it reads far below 145 kW while the amps stay high, the assumptions collapsed: look at power factor, the winding connection and the negative-sequence numbers.
What measurements settle it, and in what order? #
Trend one full cycle with one instrument — volts, amps, kilowatts, power factor and speed from the 60-second ramp through bypass into steady flow — then sweep from transformer secondary to motor terminals so each reading eliminates one mechanism.
De-energize, lock out and test for absence of voltage before opening any terminal box or making resistance checks; connecting an analyzer to energized terminals is energized work, so follow NFPA 70E (2024 edition, the US standard for electrical safe work practices), your site's program and the equipment documentation.
- Record the trip state: overload setpoint, trip class or curve, thermal state and time from start to trip, and verify the settings were reviewed against this motor's 164.0 A nameplate rather than the old motor's.
- Trend the entire cycle: three voltages, three currents, total kilowatts, power factor, flow, suction and discharge pressure, from ramp through bypass to steady state — this separates a pipe-filling transient from a steady overload.
- Measure simultaneously at the starter input and the motor terminals under load; a phase-specific difference locates resistance in the disconnect, bypass pole, cable or termination.
- Read V2/V1 and phase angles at the transformer secondary, starter line side and motor terminals.
- Measure shaft speed with a strobe or from the vibration spectrum and compare against the 3555 RPM nameplate; slip scales roughly with load.
- Put flow, pressures and fluid density against the pump curve at the measured speed, not the catalog speed.
- Record the transformer nameplate — kVA, winding configuration, tap range, present tap — and compare loaded with unloaded secondary voltages.
- Run the motor uncoupled: no-load current and no-load kilowatts separate motor loss from pump demand.
Could the soft starter or the padmount transformer be at fault? #
Not the SCRs — your readings were taken in bypass, where the contactor shorts them out of the circuit — but a high-resistance pole, termination or cable shows up as loaded voltage drop, and the transformer stays on the list until its secondary is measured.
Test the bypass path by measuring each phase under load at both ends and subtracting, the same method used to fix voltage drop on a long 277 V lighting feeder. A pole with excess resistance drops voltage on its own phase and appears as unbalance at the motor, not as a uniform 17.5% rise, so it is a check to close out rather than the leading theory. On the padmount: a defective supply transformer can push a motor above nameplate and eventually burn the winding, which is why the secondary measurement comes before any motor replacement. If it has a tap changer, change taps only with the transformer de-energized unless the nameplate explicitly identifies an on-load tap changer, and arrange the switching with qualified utility or facility personnel.
Should I raise the overload setting to stop the trips? #
No. The relay is responding to real current, so set it from the 164.0 A nameplate and the starter manual — not from the 192.7 A you are measuring — because the overload is the messenger, and raising it only moves the failure into the winding.
The service-factor arithmetic — 164.0 A × 1.15 = 188.6 A — is a comparison point that your 199 A phase already exceeds; it is not a continuous-current allowance and not an overload setting. The setting logic in selecting overload and jam protection for a reversible motor applies the same way here: the relay protects the winding, and the winding has not changed its rating because the current went up.
What do I do with each measurement outcome? #
Compare measured total kilowatts against the 145.6 kW estimate and the pump's 97 kW demand: high kilowatts means the load is real, low kilowatts means the estimate's assumptions collapsed, and both branches have a defined next test below.
| Measurement outcome | Likely mechanism | Deciding next step |
|---|---|---|
| Total kW near the 145.6 kW estimate, speed at or below 3555 RPM | Real shaft demand: operating point off the curve, speed, density, rotation, rub | Flow, pressures and density against the pump curve at measured speed; uncoupled run |
| Total kW well below the estimate, current still high | Power-factor collapse, winding or connection fault, current distortion | Connection diagram check, phase-resistance comparison, surge test |
| Voltage balanced at the transformer, dropped or unbalanced at the motor | High-resistance termination, bypass pole or cable | Per-phase loaded voltage drop, subtract both ends |
| Current peaks during pipe filling, then settles | Transient operating point crossing the trip curve | Time trend of current, flow and thermal state from start to trip |
| Uncoupled no-load kW high | Motor internal loss or friction | Vibration, winding tests, current signature, shop test |
Commit on this evidence: with the pump and impeller verified and flow tests pointing to 130 hp, negative sequence is not the overcurrent — it cannot explain a 17.5% rise in a three-phase average, and the 0.90% magnitude unbalance sits inside the 1% band NEMA advises living with. The measured kilowatts then decide the branch: near 145 kW and the problem is mechanical load; far below it and the problem is in the electrical path, starting with the winding connection and the transformer secondary. Replacing the five-year-old soft starter is not supported by any reading taken so far.
Questions asked
A megohmmeter passed the windings. Is the motor cleared?
No. An insulation-resistance test to ground finds ground faults only. Turn-to-turn shorts, a wrong internal connection and rotor bar defects can all pass a megohmmeter while pushing running current up. Compare phase-to-phase resistances, surge-test the winding and read the running current signature before you either condemn or clear the motor.
How do I estimate the real load from shaft speed?
With a strobe, or from a vibration spectrum, measure shaft speed and compare it with the 3555 RPM nameplate: 3600 RPM synchronous minus 3555 RPM is 45 RPM of slip at 150 hp full load. Slip scales roughly with load, so 130 hp implies about 39 RPM of slip and roughly 3561 RPM. A reading at or below 3555 RPM says the load is at or above full load. It is an estimate; confirm with measured kilowatts.
Could the winding connection be wrong?
Check the connection diagram on the nameplate against the actual lead hookup. A dual-voltage winding connected in wye instead of delta at the same voltage draws about 1.73 times nameplate full-load current. Your 192.7 A average is 17.5% over the 164.0 A rating, not 73%, so a full misconnection is unlikely, but the check costs minutes with the circuit locked out.
Is the 480 V at the motor a problem for a 460 V winding?
The 480.3 V running average is 4.4% above the 460 V rating (480.3 divided by 460 equals 1.044). On a loaded motor, slightly high terminal voltage tends to reduce load current while raising magnetizing current, so it does not explain a 17.5% overcurrent. If anything, a true 460 V at the terminals would read slightly higher current for the same shaft load.