Compressor Trips on Overload with 'Normal' Voltage on All Three Phases
A 7.5-ton, 460-volt, three-phase rooftop unit cools a church fellowship hall. The compressor starts, runs about six minutes, then the internal overload opens and the unit locks out; it restarts after roughly 25 minutes and repeats the cycle. Line voltage at the disconnect is 472, 470 and 471 volts phase to phase. With the contactor held closed, a meter placed on the compressor terminals reads 471, 470 and 470 volts. A clamp meter shows 22.0 amps on L1, 21.8 amps on L2 and 0.0 amp on L3. Compressor nameplate RLA is 14.5 amps. The contactor was replaced last summer. The condenser coil is clean and the outdoor fan is running. Refrigerant readings while running are 118 psig suction and 355 psig discharge on R-410A with an 82°F outdoor ambient. The building owner has already approved a compressor replacement because 'the windings must be burned.'
1. Explain how three apparently normal voltage readings can coexist with zero current on one leg, and state the most probable fault.
Read the worked answer
How the readings coexist: The compressor motor is a three-phase winding set tied together at a common neutral point. If L3 is not being delivered to the compressor, terminal T3 is electrically floating - but it is still connected to T1 and T2 through the motor windings. A high-impedance voltmeter placed between L1 and T3 completes a circuit through two windings, so it displays essentially full line voltage. The same happens L2 to T3. T1 to T2 reads normal because that path is intact. The voltage test therefore looks perfect while the compressor is actually running on single phase.
Fault: The open circuit is between the point of measurement and the compressor - most likely the L3 pole of the contactor not making (pitted or burned contacts, a broken shunt, a loose line-side lug), or an open fuse/disconnect blade on that phase upstream. The reading at the compressor terminals proves nothing because back-feed masks it.
Supporting evidence: 22 amps on two legs and zero on the third is textbook single-phasing. Nameplate RLA is 14.5; running single phase, the current in the two energized legs rises roughly 1.5 to 1.7 times normal - right where these readings sit.
2. Describe the test that confirms the open pole without disassembling the compressor, and describe the winding checks that would prove the compressor is not the problem.
Read the worked answer
Confirming the open pole: With the contactor coil energized and the compressor attempting to run, put the voltmeter across each pole - line terminal to load terminal on the same phase. A closed, healthy contact shows nearly 0 volts (a few tenths). An open contact shows full line voltage across it. Repeat for all three poles, then de-energize and lock out and inspect/pull the contactor to see the burned L3 contacts. A low-impedance (loaded) meter or a voltage drop test under load removes any doubt that a digital meter is reading a ghost voltage.
Winding checks (power off, locked out, conductors disconnected and tagged):
- Resistance T1-T2, T2-T3, T1-T3 should be essentially equal (within a few percent) and match the manufacturer's values.
- Resistance from each terminal to the shell/ground should read megohms, not near zero.
- If the internal overload is accessible, confirm it is closed when cold.
Balanced windings plus good insulation to ground on a compressor that was single-phasing points squarely at the contactor. Also check phase balance: average is 471 V, maximum deviation is 1 V, so unbalance is about 0.21% - well inside the 2% limit, so supply quality is not the culprit.
3. Explain why replacing the compressor would be the wrong decision, and list the repair and verification steps including the likely root cause of the contactor failure.
Read the worked answer
Why replacement is wrong: The evidence describes a compressor being electrically abused, not a failed compressor. It starts, runs, pumps refrigerant, and trips on overload after six minutes - that is exactly what single-phase running does: the current in the energized windings rises, the motor heats, and the internal protector opens. Replacing it without fixing the open phase would destroy the new compressor the same way and leave the owner with a second bill.
Root cause to investigate: Contactors do not usually burn a single pole by chance. Check for a loose line-side lug (heat and arcing at that termination), control voltage that is marginal or chattering the coil (long thermostat runs, a failing transformer, a dirty control circuit), short-cycling from a bad low-pressure or freeze control, and undersized or corroded wiring at that pole.
Repair and verify:
- Lock out and replace the contactor (match coil voltage and pole rating); torque all lugs to spec and inspect wire ends for heat damage.
- Verify voltage on all three phases at the load side under load and confirm current balance within ~10% across legs.
- Confirm amps settle near RLA as head pressure normalizes, discharge temperature is reasonable for 82°F ambient, and the overload no longer trips through a full cycle.
- Record the readings and the phase unbalance so the next visit has a baseline.