Why Your Compressor Keeps Tripping the Overload: Six Causes and How to Tell Them Apart

A compressor that trips its overload again and again is usually not lying to you, and raising the dial is usually the wrong response. Six distinct faults produce the same symptom — the thermal element opens, the contactor drops out — and each leaves a different signature in a voltmeter and clamp meter reading. By the end of this article you’ll run through all six in order, know exactly which measurement separates each from the rest, and understand why treating the trip as the problem instead of the message destroys compressors.

Why the overload is usually telling the truth

A thermal overload relay is a thermal model of the winding it protects: load current passes through a heating element wound around a bimetallic strip in each phase, the strip bends as it warms, and past a set deflection it opens a contact in the contactor coil circuit. Under IEC 60947-4-1, an overload set to nameplate FLA must not trip at 1.05 times setting within two hours, and must trip at 1.20 times within two hours. For an 18 A compressor that means it tolerates 18.9 A indefinitely and acts above 21.6 A — not oversensitive, doing exactly what it’s built to do.

Raising the setting to stop the tripping is the single most damaging habit in this trade. Insulation life halves for roughly every 10°C above rated temperature, and winding heating rises with the square of current. Raise an 18 A overload to 24 A and you allow (24 ÷ 18)² = 1.78 times the heat input the winding was designed for. The compressor fails in eight or eighteen months, by which point nobody connects the two events. Treat every trip as a measurement result, and work through the six causes below before touching the dial.

Safety note: every measurement below is taken with the circuit live — voltage at terminals under load, current on a clamp meter, voltage drop across a closed contactor pole. Use a CAT-rated meter appropriate to the panel, keep one hand clear where practical, and treat every live panel per your site’s OHS Act 85 of 1993 procedure.

The six causes, and the measurement that separates each

Cause 1: low supply voltage

Starting torque in an induction motor varies with the square of applied voltage. Drop the supply from 230 V to 200 V (87% of nominal) and torque falls to 0.87² = 0.756 — roughly 76% of rated. A compressor with margin to spare shrugs that off; one already marginal may simply not break away, humming and tripping on overload instead. On three-phase, expect 400 V ±10% (360-440 V) line-to-line and 230 V ±10% (207-253 V) line-to-neutral — outside that band, don’t expect normal starting or running behaviour.

This is exactly the pattern seen around load shedding: voltage sags under heavy demand, and the surge of simultaneous reconnections after a stage lifts puts supply voltage below nominal at the moment compressors across a site try to restart together. A compressor that starts reliably alone and fails during a recovery window is often a voltage problem, not a compressor problem — confirm it with a meter at the terminals during the actual start attempt before condemning anything.

Cause 2: voltage imbalance or phase loss

Measure all three line-to-line voltages, average them, and express the largest single deviation from that average as a percentage of the average — the NEMA method. The limit is 2%. Above that, current imbalance in the motor runs at roughly 6 to 10 times the voltage imbalance percentage, because the negative-sequence component sees a very low rotor impedance. A 3.9% voltage imbalance predicts current imbalance somewhere between 23% and 39% — enough on its own to trip an overload with no fault in the compressor at all. Full detail on running this calculation is in our guide to measuring voltage imbalance on three-phase refrigeration.

Phase loss is the acute version of the same fault: lose one line while the compressor runs and current in the remaining two phases rises to roughly 1.73 times normal, overheating windings within minutes. The signature is unmistakable — one line at or near 0 A with the other two high and roughly equal. Check upstream for a blown fuse, one contactor pole not closing, an open overload element, or a lost phase at the supply.

Cause 3: a pitted contactor pole single-phasing the motor

A contactor pole that has pitted or partially failed can either add resistance without fully opening, or fail to close at all — both produce a trip that looks electrically like a motor fault. Measure voltage drop across each closed pole with the load running at normal current: below 0.1 V per pole is healthy, 0.1-0.3 V means pitting has begun, 0.3-0.5 V is deteriorating fast, and above 0.5 V means replace now. A pole that has failed to make contact at all produces the same one-line-low signature as phase loss above, traced to the contactor rather than the supply. Cold resistance across a closed contact is not the test — a healthy contact is 1-5 mΩ, well below what a multimeter’s ohms range can resolve, so only a voltage-drop reading under load catches it. The full procedure is in our guide to testing a contactor properly.

Cause 4: mechanical load or a seized bearing

A mechanically seized compressor — a stuck scroll, a seized bearing, hydraulic lock from migrated oil or liquid refrigerant — draws current at or near locked rotor current continuously, because the rotor cannot turn, tripping the overload within seconds every time it’s reset. Isolated and cooled, winding resistance and insulation resistance both test normal, because the problem is not electrical. That distinction matters: a compressor condemned “electrically faulty” on this symptom alone, without a resistance and insulation check, is sometimes a sound motor bolted to a seized mechanism.

A lesser version of the same cause is a running compressor with current sitting persistently 15-30% above nameplate FLA, with supply voltage and superheat both confirmed normal — pointing to a mechanical drag or a partially shorted winding rather than an outright seizure, and worth investigating rather than dismissing as a nuisance trip.

Cause 5: incorrect overload setting

Set the dial to nameplate FLA, not to whatever number stops the tripping. A dial left at a previous compressor’s rating, or nudged down during a service, trips a perfectly healthy motor at normal running current. Trip class matters too: Class 10 is the default for hermetic and semi-hermetic compressors, which accelerate to speed in under two seconds and hold at 7.2 times set current for only 2-10 seconds from cold. Class 20 or 30 belong on high-inertia loads — open-drive and screw compressors, large flywheels. A Class 10 relay tripping on every start is often correctly protecting a start that’s taking too long, not an incorrectly rated device.

Overloads on a compressor should always be manual reset, never automatic. An automatic-reset overload recloses once the bimetal cools, typically after 2-5 minutes, restarting a stalled or faulted compressor into the same condition every few minutes at locked-rotor current — 5 to 8 times FLA — until the windings carbonise. It also destroys the evidence that it tripped, which is exactly the evidence needed to work through the other five causes.

Cause 6: genuine winding failure

This is the cause to confirm last, not first, and only with two independent tests. Measure resistance across all three winding pairs (three-phase) and require them within 5% of each other, or measure C-R and C-S on a single-phase compressor and check the arithmetic identity C-S + C-R = S-R. Then test insulation resistance winding-to-earth at 500 V DC (230 V equipment) or 1000 V DC (400 V equipment): reject anything below the absolute floor of 1 MΩ, and treat a healthy reading as well above 100 MΩ. A shorted turn is genuinely hard to catch on resistance alone — a couple of percent of turns shorted barely moves the reading — so a compressor with balanced, in-tolerance resistance and clean insulation resistance is very likely sound even if it has been tripping repeatedly. The fault, in that case, is one of the other five causes, not the winding.

Final Thoughts

In 28 years of chasing overload trips, the pattern repeats: a technician replaces the compressor, the new one trips within the hour, and only then does someone reach for a voltmeter. Work the six causes in the order given here — supply voltage, imbalance and phase loss, the contactor, the mechanical load, the setting itself, and only then the winding — and a six-minute test sequence with a clamp meter and a voltmeter separates a lug, a pole, or a bad phase from an actual dead compressor almost every time. The full method behind every one of these tests is in the complete guide to electrical troubleshooting and fault-finding.