Aircon thermal cutout: the part built to stop the unit
Most parts fail by not working. A thermal cutout does the opposite: it opens on purpose, the unit stops, and it closes again once things cool. Replacing it settles nothing, because the heat that moved it came from elsewhere.
By Team Snowflake | Updated 7 Aug 2026
What a thermal cutout is, and where these sit
A thermal cutout is a switch that answers heat by breaking the circuit it sits in. That is its entire vocabulary. It cannot report a value, slow anything down, or ask a board for permission. Current either passes through it or it does not.
Most versions are a small metal disc made of two bonded layers. The layers grow at different rates as they warm. Past a set point the disc snaps to the opposite shape and pulls the contacts apart. Copeland describes the device fitted to its scroll compressors as a temperature-sensitive snap disc, and the mechanism is the same wherever the part turns up.
These are scattered around a split system rather than concentrated in one spot. One rides inside the compressor shell among the motor windings. Another can be clamped to the outside of that same shell. Fan motors carry their own, packed in with the wire. Any heating element in the equipment will have one close by.
Where a motor is involved, the device is usually answering two things at once. Copeland's application data for its scroll range states that the overload protectors react to current and motor winding temperature. Current running through the disc warms it from within while the motor warms it from outside. So a machine pulling heavy current can open a heat-operated switch on a cool morning.
What a cutout is not
Several devices in the same machine end a run, and they are not variations of each other.
A breaker sits upstream in the supply and drops its whole circuit, while a cutout works inside the machine and no other appliance notices. Reading a trip by its timing is what aircon breaker-tripping patterns is for. A pressure switch answers pressure rather than heat, which is the territory of the high pressure switch. A thermistor measures and reports, leaving the decision to a board, and the thermostat vs thermistor split is worth reading if that boundary is new. A fuse on the indoor board is sized for current, not temperature. A PTC start device warms deliberately as part of how it does its job, so warmth there is normal rather than a warning.
Resetting, latching, and the one that never closes again
What the device does after it opens is what decides the household's experience of the fault. Three behaviours are in common use and they lead to three entirely different stories.
The self-resetting kind closes again on its own. It opens while hot, waits open while the metal around it comes down, and returns to a closed contact at its reset point. Nothing in the equipment writes any of that down.
The latching kind stays open. ATC Semitec, a supplier specialising in these components, describes thermal cut-outs as resettable devices available in auto-reset, manual-reset and self-hold forms. A manual-reset part waits for a person. A self-hold part waits for the supply to be interrupted. Either way the equipment stays down until somebody acts, which is far more visible than a stop that quietly repairs itself.
The third is not a switch at all. A thermal fuse is one-shot: it melts, opens the circuit permanently, and has to be renewed. The same supplier notes that these cannot be reset or repaired and must be replaced once the original fault has been rectified. That last clause carries the weight. Fit a fresh one over a heat source nobody found and the part becomes a consumable.
Which of the three ends up in a given machine is a safety decision taken at the design desk. Under the NEMA convention, a motor marked as thermally protected carries the automatic-reset type. The standing advice is to specify a manual-reset motor anywhere an unexpected start would be a hazard. Convenience and evidence pull against each other here, and household equipment is built for convenience.
| What the device does after opening | What the household sees | What is left for whoever has to find the fault |
|---|---|---|
| What the device does after openingCloses again once it has cooled | What the household seesA stop, then a restart with nobody touching anything | What is left for whoever has to find the faultNothing at all, unless someone noted the clock |
| What the device does after openingHolds open until a person resets it | What the household seesEquipment that is dead and refuses to come back | What is left for whoever has to find the faultThe opened device, still in the state that stopped it |
| What the device does after openingHolds open until the supply is cut and restored | What the household seesA dead unit that revives after the isolator is flicked | What is left for whoever has to find the faultThe same evidence, until somebody flicks that switch |
| What the device does after openingMelts, and never closes again | What the household seesA dead unit and a replacement part on the bill | What is left for whoever has to find the faultA renewed part, and the reason it went in the first place |
The device that resets is the device that hides
The self-resetting kind is the kindest to the household and the least useful to anybody investigating. A latched unit sitting dead in the corridor is an inconvenience that holds on to its evidence. A unit that stops and returns has already thrown its evidence away.
That is why a fault behind a resetting device can run for a long stretch before anyone treats it as real. The household adapts. Somebody notices the room went warm for a while, the cooling came back, and the story never gets told to anybody who could act on it.
Why a cutout fault refuses to appear on demand
An intermittent stop that recovers unaided is the signature of this part, and it is an outcome of the design rather than a defect in it. The device exists to interrupt and then restore. A fault that could be reproduced at will would be something else.
The gap between the stop and the restart is a cooling curve, not a countdown. The contact closes when the disc reaches its reset point. How long that takes depends on how hot the surrounding metal became and how fast the surroundings pull the temperature back. One fault therefore gives a long outage on a still afternoon and a short one after dark.
That is why the problem will not perform for a visitor. It needs the conditions that produced it, which usually means heat, load, or a run of a certain length. Somebody arriving on a cool morning is looking at equipment behaving exactly as designed.
A stop that comes back on its own is a different state from equipment that declines to try again. Where a controller has tallied repeated stops and stopped offering, compressor lockout runs on its own rules and is worth reading separately. Households usually describe this as an aircon that turns off on its own. Several unrelated causes sit behind that one description, and the page on it does the sorting.
Nothing about the recovery is a sign of health. The cooling came back because the disc cooled, not because anything was resolved. The condition that pushed it there is still sitting in the machine, waiting for the next warm afternoon.
The recovery gap is the closest thing to a reading
The figure worth collecting is how long the equipment stays down, not how often it goes down. A disc closes once it has cooled to its own reset point. The length of the outage therefore tracks how much heat was stored in the metal around it.
Outages growing longer describe a machine arriving hotter each time. Outages of roughly constant length describe a condition that is stable, which is a different kind of news and a less urgent one. That figure costs nothing to collect, needs no tools, and no part of the system stores it for you.
The heat came from somewhere else
A thermal cutout hardly ever breaks, and saying so plainly matters, because it is the part that collects the blame. The device is simple, cheap, and doing precisely the job it was fitted for at the moment it opens. Calling the stop a fault in the switch skips the only question worth asking, which is what got hot.
Heat with no way out is the first family. The outdoor coil is where the system dumps everything it collected indoors, and anything narrowing that path lifts the temperature the compressor works against. Copeland names condenser or evaporator fan blockage among the failures that leave a compressor cycling on its internal overload protection. Locally the two usual culprits are coil fouling on the surface and lost condenser clearance around the outdoor unit.
Heat with nothing to carry it away is the second. Gas returning from the room passes over the compressor motor and takes heat with it, so a circuit short of gas runs that motor hotter than intended. Copeland's guidance states that loss of system charge results in overheating and recycling of the motor overload protector, and that letting it carry on can finish in bearing failure. A refrigerant restriction arrives at the identical condition by starving the line that feeds it, and the refrigerant charge page explains what the right amount is measured against.
Heat made in the electrics is the third, and this family catches people out. A motor struggling to get itself turning pulls heavy current for as long as the struggle lasts, and the protector answers current as readily as temperature. Capacitor weakening puts a machine in exactly that state. So does a loose or corroded connection, which manufactures its own heat at the very spot the device is watching.
Room conditions belong on the list and they are not faults. A condenser standing in full afternoon sun inside an enclosed service yard can walk a healthy machine up to the boundary. So can a room carrying more people and equipment than it was ever sized for. So can a long spell of still, heavy air. Sorting the machine from the conditions is what a visit is for.
| What got hot | What put the heat there | What the household could have noticed |
|---|---|---|
| What got hotThe gas leaving the compressor | What put the heat thereHeat failing to clear the outdoor coil | What the household could have noticedStops that gather in the fiercest afternoon heat |
| What got hotThe motor inside the compressor | What put the heat thereToo little gas returning to carry its heat off | What the household could have noticedRooms cooling slower than they used to, well beforehand |
| What got hotThe wire inside a fan motor | What put the heat thereA seized bearing, a fouling blade, or a choked air path | What the household could have noticedA scraping or rising noise before the fan went quiet |
| What got hotThe disc inside the protector | What put the heat thereHeavy current drawn while a motor fights to start | What the household could have noticedA hum or a click as the equipment tries to start |
| What got hotMetal around a joint or a terminal | What put the heat thereA connection that has worked loose or corroded | What the household could have noticedHot plastic on the nose, or browning near a terminal |
Replacing the messenger costs twice
A cutout is among the cheaper items in the machine, and a quote to swap one is easy to approve without thinking. Ask what got hot before agreeing to it. A device renewed with no heat source named will open again, because the condition that moved it was never touched, and the second visit costs what the first one did.
The risk of leaving a machine to keep cycling is the sharper half of this. Copeland's position is that a compressor left to cycle after a fault has been detected stands a high probability of damage. Debris and decomposed oil then get carried around the system. Where that contamination ends up, and what clearing it costs, is the subject of compressor burnout. It is why a stop that keeps returning is worth chasing early.
What to write down while the pattern is still running
This part keeps no history, so whatever gets noted at the time is the entire case file. Four things are worth capturing and none of them call for tools.
- The clock time of each stop, and how long the machine had been running by then
- How long the equipment stayed down before it came back, since that gap is the nearest thing to a temperature anyone will get
- Whether it returned unaided or somebody had to switch something to bring it back
- Outside conditions and room use at that moment: sun on the outdoor unit, how many people were in, which other rooms were on
Before anyone changes the part
The general method for pinning down a fault that vanishes is already written up, and the intermittent aircon fault log covers the rest of what is worth keeping. The list above is only the part specific to a heat-operated device.
One short question separates a finished job from a parts swap. What got hot, and which reading settled that? Somebody who found the heat source answers without hesitating and names a condition. Somebody who changed the device and left will describe the part instead.
Some of what sits beside these devices is live, and some of it is hot enough to burn. None of it is worth opening a panel to look at. Note the pattern, keep the equipment off if there is any smell of burning or sign of scorching, and hand it to somebody who can measure it.
Where the trail continues
The protector buried inside the compressor is a special case, because it can read as a broken circuit while it is simply open and hot. What a reading across compressor windings settles, and what it leaves unanswered, has a page to itself.
If the outdoor unit has also grown quieter, slower to start, or weaker at holding a room, then compressor failing shows itself in ways that are earlier and cheaper to catch. A heat-operated stop is often the second warning rather than the first.
Ready to get started?
Tell us what’s going on. Symptoms, setup, photos, anything we should know. We’ll assess and come back with the right next step.