Hums briefly then falls silent, fan may still spin
Locked rotor, and the internal overload has opened to protect the windings. Cause is a failed start component far more often than a mechanically seized rotor.
Compressor diagnosis
AC compressor repair in Westchester, FL, tested electrically before it is condemned
The most expensive component in a cooling system is also the most frequently misdiagnosed. Winding resistance, insulation to ground and start component condition decide it, not a hum and an opinion.
A compressor that hums for three seconds and then goes quiet is the classic call. The homeowner has usually already been told the compressor is seized and that the system needs replacing. Perhaps a third of the time that is true. The rest of the time the compressor is mechanically sound and something feeding it has failed: a run capacitor that has drifted below its rated value, a contactor with burned contacts dropping voltage under load, or a start winding circuit that never energised.
Getting this wrong is expensive in both directions. Condemning a healthy compressor sells a replacement system that was not needed. Fitting a new compressor into a circuit contaminated by the failure of the old one produces a second failure inside two years. Neither mistake is visible on the day, which is why the test sequence below is not negotiable.
The tests that settle it
Winding resistance between all three terminals, insulation resistance from each winding to the shell measured with a megohmmeter, run capacitor microfarads against the nameplate, contactor voltage drop under load, and locked rotor amperage on attempted start. Five measurements, twenty minutes, and the answer is no longer a matter of opinion.
MegohmmeterInsulation to ground measured, not assumed
MicrofaradsCapacitor tested against its rated value
Under loadContactor voltage drop read while running
WrittenReadings left with you either way
Why so many working compressors get replaced
The failure modes that look identical from outside the cabinet, and the measurements that separate them.
A single phase compressor needs three things to start: adequate supply voltage at the terminals, a working start circuit to give the rotor its initial kick, and mechanical freedom to turn. Remove any one and the symptom is identical. The motor draws locked rotor current, the internal overload protector opens within a few seconds, and everything goes silent. From the driveway it sounds exactly like a seized machine.
The run capacitor is the most common culprit by a wide margin. It is an electrolytic component sitting inside a metal box in direct sun, and its capacitance falls steadily over years of thermal cycling. A capacitor rated at forty five microfarads that now measures thirty two will run a compressor that is already turning and will fail to start one that has stopped. It costs very little, it takes ten seconds to measure, and it is routinely never checked.
Contactors are second. The contact faces pit and carbonise from arcing every time they close, and the resistance across them climbs. The compressor then sees noticeably less than its rated voltage at the exact moment it needs the most, which drives current up and torque down. A voltage drop measurement across the closed contacts under load finds this immediately, while a visual inspection often does not.
Only when the supply, the start circuit and the contactor are all proven good does the compressor itself come under suspicion. Then the winding resistances are compared against each other and against the manufacturer values, and insulation resistance from each winding to the shell is measured. A reading in the tens of megohms is healthy. A reading in the low hundreds of kilohms means the windings are breaking down and the compressor is genuinely finished.
What the readings decide
Four outcomes, each with a different recommendation and a different price.
- Start components failed, compressor sound: an inexpensive repair with years of life left
- Contactor degraded: a cheap part that was quietly damaging the compressor every start
- Windings open or unbalanced: the compressor is mechanically or electrically finished
- Windings shorted to ground: a burnout, which contaminates the circuit and changes the whole job
What each compressor symptom actually indicates
Six presentations, and the fault family each one points at before any panel comes off.
The sound a failing compressor makes carries real diagnostic information, and describing it accurately on the phone shortens the visit. A hum is not a rattle, a rattle is not a screech, and a unit that is silent is telling you something different again.
What none of these symptoms can do is distinguish a dead compressor from a dead capacitor, because both produce the same behavior. That distinction always requires a meter.
Breaker trips immediately on a cooling call
Suggests windings shorted to ground or a hard short across terminals. Insulation resistance testing confirms it, and repeatedly resetting the breaker risks a fire.
Loud metallic rattle while running
Internal mounting springs have broken or a valve plate has come loose. The compressor is still turning but it is failing mechanically and will not recover.
Runs but the system barely cools
Possible internal valve failure, where the compressor turns without developing a proper pressure differential. Suction and discharge pressures converging is the signature.
Starts hard with a noticeable groan, then settles
A capacitor drifting out of tolerance or a compressor with rising internal friction. Measuring capacitance now is very much cheaper than the alternative.
Overheats and shuts down after twenty minutes
Thermal overload cycling, usually driven by a fouled condenser, an overcharge, low voltage supply or a failing condenser fan rather than by the compressor itself.
The test sequence, in order
Six steps that produce a defensible answer rather than an educated guess.
Order matters here more than in most diagnostics, because each step eliminates a cause that would otherwise produce a false reading in the next one. Measuring winding resistance before confirming the contactor has opened, for example, gives you the resistance of the whole circuit rather than of the motor.
Isolate and discharge
Disconnect pulled, capacitor safely discharged through a resistor. A charged run capacitor holds enough energy to injure and does not care that the power is off.
Supply voltage at the terminals
Measured at the compressor itself, not at the panel. Voltage lost along an undersized or corroded run is a real and frequently missed cause of hard starting.
Capacitor microfarads
Read on a capacitance meter and compared with the nameplate tolerance. Anything more than ten percent below rating is replaced regardless of what else is found.
Contactor under load
Voltage drop across the closed contacts measured while the unit attempts to start, which is the only condition that exposes a degraded contact face.
Winding resistance
Common to start, common to run and start to run compared with each other and with the manufacturer table. An open or badly unbalanced set ends the assessment.
Insulation to ground
Each winding to the compressor shell with a megohmmeter. This is the test that distinguishes a tired compressor from a burnout, and the two are very different jobs.
When a burnout changes the entire job
A compressor that has shorted to ground has put acid and debris through the whole refrigerant circuit, and swapping the compressor alone guarantees a repeat.
There is an important distinction between a compressor that has failed mechanically and one that has suffered a motor burnout. In a burnout the winding insulation breaks down and the motor arcs internally, and because the motor sits inside the refrigerant circuit, the products of that arcing go everywhere. Acid, carbon and varnish travel through the line set into the evaporator, the metering device and the accumulator.
Fitting a new compressor into that circuit without cleaning it up is the most expensive shortcut in this trade. The contamination attacks the new windings from the first hour of operation, and the replacement typically fails within a season or two. The correct scope is a suction line filter drier, a liquid line drier, a full flush or line set replacement depending on severity, an acid test, and a follow up oil check.
Once that scope is priced honestly against the age and condition of the rest of the system, replacement is often the better decision, and the numbers get laid out rather than asserted. On a unit past a decade old running a refrigerant no longer manufactured, a proper burnout cleanup approaches the cost of new equipment that carries a fresh warranty. On a five year old system still in warranty, the cleanup is clearly worth doing.
Scope after a motor burnout
- Acid test on the oil to establish how far contamination has traveled
- Suction line filter drier fitted and rechecked after a running period
- Liquid line drier replaced, never reused from the previous circuit
- Line set flushed or replaced depending on the severity of the burn
- Evacuation to a micron target and charge weighed in, not gauged
Compressor questions
Cost, warranty, second opinions and whether a replacement compressor is ever the right call.
Yes, and it is one of the most worthwhile second opinions in this trade, because the difference between a capacitor and a compressor is the difference between a small repair and a system replacement. The test sequence takes about twenty minutes and produces numbers rather than an opinion.
Bring whatever paperwork you were given. If the previous visit recorded winding resistance and insulation readings, that is useful evidence. If it recorded nothing, that is informative too.
Sometimes. On a system still inside its manufacturer parts warranty, where the failure was mechanical rather than a burnout, and where the rest of the equipment is in good condition, replacing the compressor is clearly sensible because the part itself is covered and you are paying labor and materials.
Outside warranty the arithmetic shifts hard. The labor, the recovery, the driers, the evacuation and the refrigerant charge add up, and you end up with one new component inside an aging system. At that point the comparison against replacement deserves an honest look.
It is a start capacitor and relay that gives the compressor extra starting torque. Fitted appropriately it helps a compressor with rising internal friction, or one on a long supply run with voltage sag, get turning reliably.
Fitted as a substitute for diagnosis it hides a problem. A hard start kit on a compressor that is failing internally buys a few months and masks the deterioration, and it should never be the answer to a unit that used to start fine.
The most likely explanation is that the original failure was a motor burnout and the circuit was never properly cleaned. Acid left in the system attacks the new windings from the first hour, and a second failure within a season or two is the predictable result.
The other common cause is that the underlying stressor was never addressed. A fouled condenser, a restricted duct system or a persistent overcharge will destroy a second compressor exactly as reliably as it destroyed the first.
In a system that is correctly charged, has adequate airflow and gets its condenser cleaned, a well past a decade of service is unremarkable. In a system running short of air across the evaporator or with a condenser packed with debris, the same compressor can fail in half that.
Age alone is a poor predictor. What matters is how many hours it ran under stress, which is why the annual readings are worth having on record.
Not reliably, and anyone who says otherwise is guessing. What the phone call can do is narrow it: whether the fan is turning, whether the breaker trips, whether there is a hum, and how old the system is all shape what parts travel on the van.
The one thing worth saying on the phone is whether the breaker has tripped. That moves the call up in priority and it means nobody should be resetting it in the meantime.
Get the compressor tested before anyone quotes a new system
Twenty minutes of electrical testing in Westchester, FL decides whether this is a small repair or a replacement conversation.
Call with the age of the system, what the outdoor unit does when cooling is called, and whether the breaker has tripped. That is enough to load the right parts and to tell you whether this needs attention today or can take a scheduled slot.
If you are already holding a quote for a new system, say so. Comparing it against the actual test readings is a more useful conversation than comparing it against another quote.
What the assessment includes
- Supply voltage measured at the compressor terminals, not at the panel
- Run capacitor capacitance read against its nameplate rating
- Contactor voltage drop measured under load
- Winding resistance and insulation to ground recorded in writing
- A flat rate price for whichever repair the readings point to