The evaporator fan: the cheapest part that fails most

What the fan does, and what a Sub-Zero does differently

Every refrigerator makes cold in one place, the evaporator coil, and then has to move that cold somewhere useful. The evaporator fan is what moves it. A Sub-Zero built-in has two of almost everything: a sealed system, an evaporator and an evaporator fan for the refrigerator, and a second, completely separate set for the freezer. The two compartments never trade air, which is the entire point of dual refrigeration and the reason each side holds its own humidity and its own odors. For diagnosis it is a gift. A warm refrigerator sitting above a rock-hard freezer has already cleared half the machine: the side that still works has settled the power, settled the controls, and demonstrated one complete refrigeration circuit moving heat. Whatever is wrong lives on the refrigerator circuit, and nothing done to the freezer will change it. On a conventional single-system refrigerator, where one evaporator sits in the freezer and feeds the fresh food compartment through a duct and a damper, a warm refrigerator over a cold freezer looks the same from the outside but narrows things far less: the fan, the damper and the defrost heater all sit on that one shared coil, and a freezer that is still cold rules none of them out.

• Refrigerator warm, freezer hard frozen: the fault is on the refrigerator circuit, and its own evaporator fan is the first candidate
• Freezer soft while the refrigerator holds: same logic, other half of the cabinet
• Both compartments warm at once: two independent systems rarely fail on the same day, so look at what they share, which is the condenser, the condenser fan and the power supply
• Cold within a few inches of the vent and several degrees warmer at the far end of the same shelf: air is being made cold but not moved

The symptom pattern that points at the fan, and the one that does not

A stopped evaporator fan has a shape to it. The temperature climbs gradually over a day or two rather than falling off a cliff, because the compartment is still losing heat to a cold coil, just very slowly and only near the coil. Inside the affected compartment there is a large gradient: the milk on the shelf nearest the vent is still cold while the far end of that same shelf has come up several degrees. And the compartment is quiet in a way it was not before, although silence is the least reliable clue on this list, for reasons in the next section. A fan on the way out rather than fully stopped is louder, not quieter. A dry bearing chirps, warbles or whines, and the sound rises and falls with the compressor cycle. A blade catching the edge of a frost buildup ticks or clatters intermittently and then stops for a day. Both of those are worth acting on the week you hear them.

• Gradual warming over a day or two, not a sudden failure
• A wide temperature spread from one end of a shelf to the other
• A chirp, warble or whine from inside the cabinet that stops the moment the door opens
• Ticking or clattering that comes and goes, which is usually a blade touching frost
• Not the fan: a unit that is not running at all, a compartment that went warm within an hour of a power interruption, or a rattle at the lower grille, which is the condenser fan and a different part

What you can check yourself, safely

Start with a thermometer rather than the display, because the display reports what a sensor near the coil sees and that sensor is not sitting where your food is. Put a probe or a dial thermometer on the middle shelf, close the door and leave it four hours. Then try to establish whether air is moving. Holding a strip of tissue at the outlet vent with the door open sounds like the obvious test, but on most built-ins the door switch cuts the evaporator fan the instant the door opens, so an unmoving tissue proves nothing at all. Some generations use a small plunger switch in the door frame that can be pressed and held to simulate a closed door; hold it ten or fifteen seconds and listen for the fan to come up to speed. Other generations use a magnetically operated switch with nothing to press, and the test is simply not available to you. Either way, do not stand with the door open for minutes running experiments in a humid kitchen, because you are handing the coil a great deal of water to deal with.

• A thermometer on the middle shelf for four hours, doors closed, instead of the panel reading
• Return and outlet vents clear of a box, a bag or a platter pushed to the back wall
• Whether the other compartment is unaffected, which is the single most useful thing you can tell us on the phone
• When it started, and whether anything changed: a power interruption, a door left open, a very full weekend
• No chipping at any ice you can see: the refrigerant tubing behind the rear panel is thin aluminum and punctures easily

The tests that belong to a technician

A fan that will not run is not the same thing as a fan that has failed, and this is where money gets wasted. On newer generations the evaporator fan is a brushless DC motor with a control line, and the board commands its speed rather than simply switching it on. That motor can be perfectly good and stationary because the control never asked it to run, because a door switch is stuck reporting an open door, or because the unit is sitting in a defrost cycle at that moment. So the order is voltage at the connector first, then current draw against the motor's rating, then a check that the rotor turns freely by hand with no drag or grinding. Only then is the motor condemned. The second reason for care is that the fan is often the victim rather than the culprit. When a defrost circuit stops finishing, the evaporator grows a jacket of ice, that ice grows into the blade path, and the motor either stalls against it or wears its bearing out fighting it. Fit a new motor alone and it runs beautifully for a few weeks until the ice comes back, and you have paid for the wrong part.

• Supply voltage measured at the fan connector, not assumed from the fact that the light works
• Current draw compared against the motor's rating, which catches a bearing that is dragging
• The control line checked on variable-speed motors before the motor is condemned
• Door switch operation, because a switch stuck open holds the fan off permanently
• Defrost heater continuity, thermistor accuracy and a drain proved clear at the same visit, so the fan does not have to be replaced twice

Why it fails, and the part of it that is local

Two things end evaporator fans: bearings and water. The motor runs for long stretches whenever the compartment is calling for cooling, so run hours accumulate quickly in a house where the doors get opened often. The moisture half is worth being precise about, because the local version of it is not the one people expect. On this side of the county salt air is the story behind most of what wears out on a built-in, but the evaporator fan never meets it: it lives inside a sealed food compartment and breathes air that has already been through the coil. What it sees is humidity, and a kitchen that stays open to onshore air most of the year puts moisture into the cabinet on every door opening. That moisture ends up as frost on the coil, and a coil that frosts heavily is one whose fan spends its life in wet, near-freezing air. The salt gets the condenser fan behind the lower grille. Humidity gets the evaporator fan behind the rear panel. Two different motors, two different failure stories, and worth being clear about on the phone, because they are not in the same place and they are not the same part.

Where it sits on the cost scale, and why waiting moves it

It helps to think of the parts in a built-in as three tiers. At the bottom sit the evaporator fan, the door gasket, a thermistor and a door switch: small, commonly stocked, fitted in one visit. In the middle sits a control board. At the top sits anything that requires opening the sealed system, and that is the tier where owners start weighing repair against replacement at all. The evaporator fan is firmly at the bottom, which is why it is worth dealing with the week it starts making noise rather than the month it stops. A noisy fan is still moving air and the food is still cold. A seized fan gives you a warm compartment within hours, and then the consequences: food to throw out, and a coil that has been running with nothing blowing across it. Nothing about the part changes. What changes is how much else has to be put right at the same time.