Aircon solenoid valve: a coil opening a refrigerant path

A solenoid valve is a short length of pipe with an electromagnet built around it. Wind wire into a coil, pass current through that coil, and it pulls a small steel plunger along a sealed tube running through its middle. The plunger rests on a seat inside the refrigerant path. Lift it and refrigerant travels through. Let it fall and the path is shut.

The coil never touches refrigerant, and that fact carries the rest of this page. Danfoss describes the arrangement plainly: the valve is a coil and a valve body, with the coil fitted onto an armature tube. Everything electrical sits on the outside of that tube. Everything that moves sits inside it, in the same pressurised space as the refrigerant. One half of this component can be unclipped in the open air. The other half is behind the seal that keeps the whole system alive.

There is no middle setting. Current on means the path is open, current off means the path is shut, and the valve holds one of those two positions at any moment. Which position it takes when nothing is powering it is a design choice. Refrigerant work mostly uses the version that closes on losing power, so a system that shuts down for the night seals that path by itself rather than letting refrigerant wander. The opposite arrangement exists and holds open until told otherwise.

That resting position decides what a power failure looks like. A valve that closes when de-energised turns a lost signal into a blocked path. A valve that opens when de-energised turns the same lost signal into a path nobody is watching. The fault is identical on the wiring side and opposite in the refrigerant.

A plain wall-mounted split has nothing for a solenoid valve to do. One outdoor unit feeds one indoor coil through one pair of pipes, and a metering device already sits where the pressure has to drop. There is a single route to a single destination. Nothing has to be chosen, so nothing has to be commanded, and most household equipment in Singapore is built exactly this way.

These valves earn their place once the circuit has more than one route through it. Equipment serving several indoor units on shared pipework has to settle which branch gets fed and which sits isolated. Equipment that takes heat out of one room and puts it into another has to route each branch between three pipes rather than two. Each of those decisions is a path opened or closed, which is the only thing this valve does.

Larger outdoor units carry a set of them for internal housekeeping nobody sees. The Daikin VRV service guide lists solenoid valves for hot gas bypass, for oil return, for the liquid pipe stop, and for charging and discharging the refrigerant receiver. None of those sit near a room. They are named here because a fault on one of them is a fault on your system, and the description of it will never mention a bedroom.

Two further places are worth knowing about. Refrigeration equipment in a shop or a food business commonly carries a valve on the liquid line. Its job is to shut that line before the compressor stops, so the cold side clears instead of holding refrigerant that will drift back overnight. That is an automatic pump down, run by the controls, and what a manual pump down protects during a relocation is a separate subject. Reversible equipment that heats as well as cools uses a small pilot solenoid to shift the large valve that swaps flow direction, which is a part almost no cooling-only flat carries.

This component has two halves that fail independently, and the repair for each sits in a different category of job. The coil is an electrical part living outside the sealed circuit. The body is a mechanical part living inside it. A report saying the solenoid valve has failed has named a component without naming a side, and the missing half is the part that decides the bill.

Coils fail the way small windings fail. The winding goes open and produces no magnetic pull at all. Insulation breaks down and the coil cooks itself. Water finds a way in, which Danfoss names as the usual reason behind a burst coil, with the plastic shell bulging and rust visible inside where the sealing was inadequate. Every one of those is confirmed and corrected without the refrigerant being disturbed. A coil comes off its stem and a matched one goes on.

Bodies fail from what travels through them. The plunger and its seat stand in the refrigerant stream, sharing it with oil, moisture and any debris the pipework has collected since installation. Debris caught on the seat holds the valve a fraction open when it should be shut. Varnish or corrosion holds the plunger where it stands, and a magnetic field that would normally lift it achieves nothing. Reaching that part means recovering the charge and cutting into the pipework, which is a different order of intervention.

Two more situations produce the same complaint with both halves in perfect health. The first is an instruction that never arrived, whether from a board output that has stopped switching, a broken wire, a blown fuse on the control side, or a control transformer no longer supplying it. The second is more particular to this part. A coil can only pull its plunger open against so much pressure difference, and manufacturers publish that limit for each valve and coil pairing. Danfoss notes that a valve which closes hard enough to stop the system on low pressure then faces the full gap between the high and low sides. The published limit has to be large enough to open against that gap. Past that point a sound coil on a sound body simply stays shut.

A commanded valve rarely announces itself as a valve. It announces itself as whatever a wrongly closed or wrongly open path does to everything downstream, and everything downstream is where attention goes first.

A liquid line solenoid held shut starves the circuit past it. The compressor keeps running and draws the low side down with nothing arriving to replace what it removes. Pressure falls until protection acts and the machine stops. What reaches the owner is equipment that runs briefly and cuts out, or a low pressure fault on the display, or a technician saying the gas is low. Nothing in that picture points anybody at a wire, and a compressor lockout that follows repeated attempts only buries the cause further.

The same valve held open causes trouble at the other end of the cycle. With the system off, refrigerant drifts toward whichever part is coldest and settles into the oil sitting in the compressor. The next start then pulls liquid into a machine built to handle vapour. What gets reported afterwards is noise at start-up, or a compressor that fails outright, and the valve that permitted it is never part of the conversation.

On shared pipework the fault isolates itself, and that isolation misdirects the search. Cooling disappears in one room, the rest of the flat sits at its usual setting, and suspicion falls on that indoor unit and on the metering device serving it. The branch valve upstream of both sits outside the discussion entirely. Room selectivity is genuine evidence, and it narrows the fault to one branch, but it does not say which part on that branch is responsible.

Ask for the finding on each half separately. The solenoid valve is faulty is a conclusion with the working left out, and it can be reached from a coil result, from a symptom, or from a hunch. Which of those produced it decides whether the quote in front of you is for a part that clips on in the open air or for an operation on the sealed circuit.

Ask next about the instruction. A coil that was never energised is not a faulty coil, and a valve never asked to move is not a stuck valve. Establishing that the command was present while the fault was happening is a separate finding from establishing that the coil itself was sound, and both need to exist before the body is condemned. On equipment whose controller reports valve state, that finding is available without anything being dismantled.

Ask what travels with the repair if the body is coming out. Opening the circuit means the refrigerant is recovered, the pipework stands open to room air, and a filter drier belongs in the same job. A vacuum and pressure test before the charge goes back is the standard close, and the charge itself goes in by weight. A quote listing a valve and nothing else has priced a part rather than a repair.

Ask what the conclusion stood against. This fault shares its symptom with several others, so a report naming it should also say what was cleared along the way. Whether the charge was confirmed by weight, whether the metering device on the affected branch was checked, and whether the search covered the places refrigerant actually escapes from are the usual companions. Where gas keeps going missing rather than merely failing to move, the schrader valve core is a different suspect with its own page. A conclusion reached by elimination can name what it eliminated.