Intermittent Breaker Trip on a 3-Ton Split

Conductor and breaker sizing are correct; the voltage drop is not.

• Conductor: MCA 19.5 A means the minimum circuit ampacity required is 19.5 A. 10 AWG copper is well above that at any normal termination rating, and the 30 A breaker matches the MOP exactly. You may not exceed the MOP - installing a 40 A breaker to stop tripping would be a code violation and would remove the protection the listing assumes.
• Expected voltage drop: for 85 ft one-way (170 ft of conductor) at ~18.5 A, 10 AWG copper (~1.24 Ω/1000 ft) gives about 0.21 Ω and roughly 4 V. You should see ~236–237 V at the unit, not 219 V.
• Measured drop is ~22 V with only a 1 V drop at the panel. That is about 1.2 Ω of resistance in the branch circuit - five to six times what the wire can account for. Something in the circuit is a high-resistance point: a back-stabbed or corroded splice, a burned breaker terminal or bus connection, a marginal contactor line-side termination, a damaged conductor in conduit, or a loose disconnect.

So: do not re-wire or re-break the circuit on a guess. The evidence points to a connection defect, and heat at that defect plus the low running voltage (which raises compressor current) explains a breaker that trips only after the load runs and the day gets hot. Low voltage also makes the compressor run hotter and draws more current at the same load - 18.5 A against an RLA of 16.9 A is already over nameplate.

Next measurement: voltage across each termination (breaker line/load, disconnect, contactor) while running, plus a thermal scan. Any termination dropping more than a fraction of a volt is the culprit.

Sequence (each step before the next):

1. Isolate the circuit. With the unit running, measure voltage at the panel and at the disconnect simultaneously; then measure across each termination (breaker, disconnect, contactor, capacitor terminals) - a termination with more than a few tenths of a volt across it is defective. Thermal-scan the panel and disconnect. Repair/replace the bad termination; re-measure until the running drop is ~4 V.
2. Clean the condenser coil. Lint-blocked coils raise condensing pressure, which raises compressor amps directly. All charge and amp readings taken on a dirty coil are meaningless - do not adjust charge yet.
3. Verify the capacitor properly. 41 µF against a 45 µF rating is outside a typical ±6% tolerance band; check it under load and replace with the exact rated part. A weak run capacitor raises running amps and is a classic intermittent-trip cause.
4. Re-check amps and pressures after cleaning. Compare to the data plate RLA and to the manufacturer's charging chart. Only then evaluate subcooling: 3°F on a TXV system suggests undercharge, but it was read on a dirty coil, so it proves nothing yet.
5. Check the compressor itself: winding resistance balance, insulation resistance to ground, start components, and whether it is running on the correct phase/rotation.
6. Check the breaker for a hot bus connection, trip curve age, and whether it has been repeatedly tripped under load. Replace like-for-like at 30 A if it is damaged.
7. Record ambient, indoor wet bulb, suction/liquid pressures, superheat, subcooling, and amp draw at completion.

Note that a low-voltage condition at the compressor and an over-amp condition reinforce each other - fix the termination first, or every later reading is misleading.

Explanation for the homeowner, in plain terms:

• The 30 A breaker is not too small - the unit's listing says 30 A is the maximum protection allowed. A larger breaker does not make the circuit stronger; it removes the protection and moves the failure into the compressor, the wiring in the wall, or the panel. A tripping breaker is doing its job: it is reporting a fault.
• The real fault is measurable: 22 V lost in 85 ft of wire is a bad connection, not wire size, and low voltage makes the compressor pull more current and run hotter. Replacing the breaker would hide the symptom while the bad connection keeps cooking.
• The compressor is already drawing 18.5 A against a 16.9 A RLA on a dirty coil with a weak capacitor. Left alone, the likely outcome is a seized compressor rather than a tripped breaker.

In writing, on the invoice/ticket:

1. Measured voltages (panel vs. unit, running), amp draw, capacitor readings, static/subcooling readings, and the nameplate MCA/MOP/RLA.
2. The statement that the branch circuit conductors and 30 A breaker are correct per nameplate and that no breaker larger than 30 A may be installed.
3. The identified defects and the recommended repairs, itemized, with the consequence of declining each (compressor failure, repeat trips).
4. A clear "we are not responsible for equipment failure if the electrical defect is not corrected" note - and, if the homeowner refuses, a documented refusal.

Do not restore the system to service with a known high-resistance connection energized.