How Compressors Fail (Generic)
Why this matters
A compressor is usually the most expensive single component in a sealed or pressurized system, and replacing one is a major expense and a long job. That makes it the part you least want to misdiagnose. Most "bad compressor" calls are actually electrical starting problems, control problems, or a system condition that killed an innocent compressor. The compressor is also the part that pays for every upstream mistake: it inherits whatever the rest of the system does to it. Knowing how compressors really fail, and proving the failure before you condemn one, protects you from the costliest wrong call in the trade.
What a compressor does
A compressor raises the pressure of a gas (refrigerant, air, or another medium) by driving a piston, scroll, vane, or screw with a motor. It needs correct power to start and run, a load it can actually compress against, and protection from running in conditions that destroy it: liquid where it expects gas, no lubrication, or constant overheating. When you think "how could this fail," think in those three buckets: electrical, mechanical, and system-induced.
Electrical failures (test these first)
Most no-start compressors are an electrical problem, not a dead pump.
- Failed start components. Many compressors need a start capacitor, run capacitor, or start relay to get turning. A weak or dead capacitor leaves the compressor humming and straining but not starting, then tripping on overload. This is cheap to confirm and cheap to fix, and it is misdiagnosed as a dead compressor constantly.
- Open or shorted windings. The motor inside can fail open (no continuity through a winding) or shorted (winding to winding, or winding to the shell, which is a ground fault). Read winding resistance against spec and check for a short to the metal body, both with the unit isolated.
- Tripped internal overload. A compressor that ran hot may have an internal protector open. It will not start until it cools, then trips again if the cause remains. A compressor that "comes back after it sits" is overheating for a reason.
- No power or no call. Confirm correct voltage actually arrives and a control is telling it to run before you blame the compressor at all.
Mechanical failures
These are the genuine compressor deaths.
- Seized. It will not turn, draws locked-rotor current, and trips protection instantly. The mechanism is jammed.
- Worn or weak. It runs but no longer builds the pressure it should. Output is low across the board. This shows as a system that cannot reach its targets even though everything else checks out.
- Valve or internal damage. Internal valves leak so the compressor short-cycles pressure back on itself, running without producing.
System-induced failures (find the cause or you kill the new one)
This is the part techs skip and regret. A compressor often fails because of something else, and a replacement dies the same way.
- Liquid slugging. Liquid entering a compressor built to pump gas hammers the internals. Caused by upstream conditions that let liquid reach the inlet.
- Lost lubrication. Oil that left the compressor and got stuck elsewhere in the system, or a leak, runs the bearings dry.
- Chronic overheating. A dirty condenser, restriction, overload, or undersized cooling cooks the windings over time. The compressor dies of a problem somewhere else.
- Contamination. Moisture, debris, or acid from a prior burnout left in the system attacks the next compressor.
Reading the scene
| What you observe | Likely meaning |
|---|---|
| Hums, strains, will not start, trips | Start component (capacitor/relay) first |
| Dead silent, no draw | No power or no call |
| Tries, draws locked-rotor, trips instantly | Seized mechanically |
| Starts after cooling, repeats | Overheating from a system cause |
| Runs but system never reaches target | Worn compressor or internal valve leak |
| Winding shorted to shell | Grounded motor, compressor done |
Before you condemn one
Confirm correct voltage arrives, confirm the start components are good, confirm protection is not tripping on a cause you can fix, and confirm the system did not kill it. If a compressor truly will not turn, draws locked-rotor and trips, or reads a winding shorted to the shell, with good power and good start components, then it is genuinely failed. Even then, find why before you install the replacement: a system that murdered one compressor will murder the next on the same day.
References
- Manufacturer compressor documentation for winding resistance, start-component, and protection values
- Trade-standard sealed-system and compressor diagnosis (electrical-mechanical-system elimination)
- See related: How Pumps Fail (Generic); Confirming a Part Is Actually Bad