Overload Trips on a 208 V Three-Phase Compressor

Average voltage

(210 + 205 + 196) / 3 = 203.67 V

Maximum deviation from the average

|203.67 − 210| = 6.33 |203.67 − 205| = 1.33 |203.67 − 196| = 7.67 V ← largest

Percent unbalance

% unbalance = (max deviation / average) × 100 = (7.67 / 203.67) × 100 = 3.8%

NEMA limits three-phase motors to about 2% voltage unbalance. At 3.8% the motor is roughly double the limit, and the effect on current is disproportionate - this is the classic 'the voltage looks okay on a digital meter' trap, because 196 V on one leg still reads as a plausible number to a technician who is not averaging the three legs.

The 196 V leg is also the low-current leg, which confirms the problem is upstream of the motor rather than a motor winding fault.

Average current

(16.0 + 15.2 + 10.5) / 3 = 13.9 A

Maximum deviation

|13.9 − 10.5| = 3.4 A ← largest

Percent current unbalance

(3.4 / 13.9) × 100 = 24.5%

What this means

NEMA's derating curve for polyphase induction motors shows that a small voltage unbalance produces a much larger current unbalance - roughly a 6–10× multiplier at these levels. The standard derating is approximately:

• 1% voltage unbalance → 98% of rating
• 2% → 95%
• 3% → 88%
• 4% → ~82%
• 5% → 75%

At 3.8% voltage unbalance, the motor should be treated as capable of only about 82% of its nameplate rating - an 18 A compressor is effectively derated to roughly 15 A. Two legs are already at 15–16 A. The negative-sequence current caused by the unbalance produces a reversing field in the rotor, which heats the rotor and one set of windings without producing useful torque. That is exactly the failure mode that opens a scroll's internal overload after 20 minutes and, left alone, eventually grounds a winding.

Recommendation

Do not condemn the compressor yet. Fix the voltage first: the single-phase lighting load on the same panel is the prime suspect, along with a loose or corroded connection on L3 at the disconnect, contactor line side, or panel lug. Re-measure at the compressor terminals under load after correction, and add a phase monitor if the building's loads cannot be redistributed.

The case against replacing the compressor now

• The compressor's own overload is opening, which is protective behavior, not a failed winding. It has been doing its job.
• Current unbalance of 24.5% is a supply problem, not a motor problem. A new compressor installed on the same 3.8% unbalance will fail the same way, and the customer will have paid for a compressor and a repeat trip.
• The failure is repeatable and conditional (20 minutes, 100°F day, loaded operation), which is the signature of a thermal/electrical supply issue rather than a mechanical failure.

The plan

1. With the unit off, check every connection from the panel through the disconnect, contactor line side, and compressor terminals for a high-resistance joint on the low leg. A voltage drop test under load on each leg is more telling than an open-circuit voltage reading.
2. Re-measure voltage and current with the lighting load on and off to see whether the unbalance tracks the single-phase load.
3. If the source is the utility or building distribution and cannot be corrected, install a phase monitor and derate/relocate load per NEMA guidance - and put that in writing to the customer.

Evidence that would change my mind and justify a compressor

• Normal voltage and current balance restored, or supply proven good, and the compressor still trips.
• Winding resistance to ground or an insulation test showing degradation.
• A winding resistance spread across the three phases beyond the manufacturer's tolerance.
• Low oil, acid, or a burned smell in the refrigerant circuit pointing to a mechanical/electrical failure inside the shell.

Only after the supply is proven good do we own the compressor. Until then we are replacing a part to fix a wire.