Skip to content
Ice Maker Specialist
Anyone whose countertop, nugget, undercounter or refrigerator ice maker has stopped producing, or produces badly, and wants to know which part is at fault before spending money.10 min read · Updated July 2026

Ice Maker Not Making Ice: Symptom-to-Cause Diagnostic Chart

Almost every home ice maker failure resolves to one of five water-path faults or four thermal faults, and the symptom you can see narrows it to two candidates in about fifteen minutes. Find your symptom in the table, run the single test in its row, and compare the reading to the pass band. A reset clears the latched fault code but never changes the sensor reading that latched it, which is why it comes back on cycle three.

By the Ice Maker Specialist Editorial Team

Your machine is probably not dead. Home ice makers hardly ever fail outright — they degrade, and the degradation shows up as one of eleven observable symptoms that trace back to five water-path faults and four thermal faults. Find your symptom in the table below, run the one test named in its row, and compare the number you get against the pass band. Fifteen minutes, one multimeter, and you'll know whether you're buying a $14 pump or writing the unit off.

The brand badge is the least informative thing on the machine. A Euhomy bullet maker, a Silonn, a Frigidaire-branded countertop unit and a fifteen-year-old Whirlpool refrigerator module all use the same four building blocks: something that moves water, something that gets cold, something that releases the ice, and something that decides when the bin is full. Ninety near-identical pages exist for this category because everyone indexes by nameplate. The physics indexes by subsystem.

One warning before you start. Unplug the machine before you put a probe on anything, and never test a coil in circuit — you'll read the board, not the coil.

  1. Five subsystems own every fault on this page

    Before you look at the table, decide which of these is misbehaving. It cuts the candidate list by about eighty percent and it takes ten seconds of watching the machine run.

    • Water path — reservoir, float or level probe, pump (portables) or inlet solenoid valve (plumbed), supply line, filter. Owns: no fill, short fill, thin cubes, off-taste.
    • Thermal loop — compressor, condenser coil, condenser fan, evaporator prongs or cylinder, refrigerant charge. Owns: water that never freezes, halved output, long cycles.
    • Harvest mechanism — hot-gas valve or mould heater, ejector motor and cam, auger and die on nugget machines, the tray tip motor on bullet machines. Owns: ice that forms but won't drop.
    • Bin sensing — infrared emitter/receiver pair on most portables, mechanical bail arm on refrigerator modules, thermostat-in-bin on some undercounters. Owns: false full-bin, stops after one batch.
    • Control — main board, thermistor inputs, timers. Owns: cycles that never terminate, and everything that looks like two faults at once.

    If two subsystems look guilty, start with the water path. It's cheaper, it's more often the cause, and a starved water path fakes thermal symptoms convincingly.

  2. The symptom table

    Symptom you can seeSubsystemLikeliest root causeTest to run (instrument)Pass bandPart class / cost bandDifficulty 1–5DIY or tech
    No water intake — reservoir full, nothing movesWater pathPump impeller seized with scale (portable) or inlet-valve solenoid gone open-circuit (plumbed)Resistance across the two coil terminals, unplugged and disconnected (multimeter, 2 kΩ range)Portable pump coil 40–120 Ω; plumbed inlet valve coil roughly 200–700 Ω. OL / open = dead coilPump or inlet valve, $12–$452DIY
    Water fills, still no ice after 25 minutesThermal loopCondenser fan stalled or condenser coil packed with dust; if the fan is fine, suspect charge lossFeel the condenser grille 10 min into a freeze cycle; then infrared thermometer on the evaporator prongsCondenser 15–35°F above room; prongs below 20°F within 6 min of cycle startFan motor $15–$40; sealed system, not a home repair2 (fan) / 5 (sealed)DIY / tech
    Cubes come out thin-walled and shatterWater pathReservoir below the minimum line, weak pump flow, or a freeze cycle terminating early on a drifted thermistorCheck level against the MAX line, then time three consecutive freeze cycles (phone timer)Bullet freeze cycle 8–13 min at 70°F ambient. Under 6 min means early terminationThermistor $8–$25; pump $12–$302DIY
    Makes one batch, then stops for goodBin sensingInfrared full-bin pair fogged, dusty, or reading the first batch sitting directly under the beamDry and wipe both lenses, empty the bin, restart; then deliberately block one lensFull-bin indicator responds within 10 s of blocking and clears within 10 s of unblockingIR emitter/receiver pair $10–$301DIY
    Ice forms but sticks in the tray or mouldHarvestScale film on the evaporator raising release temperature, or hot-gas valve not openingInfrared thermometer on the prongs during harvest, after a full descale cycleProng surface rises above 40°F within 60–90 s of harvest startDescaler $8–$20; hot-gas solenoid $35–$902 / 4DIY, then tech
    Full-bin light on with an empty binBin sensingCondensation film or dust on the IR lens, misaligned sensor pair, or a broken sensor harnessLens test as above; then DC volts on the receiver signal wire while blocking and clearing the beamSignal swings between roughly 0 V and board logic voltage (usually 5 V). A line that never moves is deadSensor pair or harness $10–$301DIY
    Freeze cycle never endsControl / thermal sensingThermistor drifted high, so the board never sees the harvest set point and holds the compressor onThermistor resistance in a stirred ice-water bath (multimeter, 200 kΩ range)10 kΩ NTC (B≈3435): 30–35 kΩ in the ice bath, 9.5–10.5 kΩ at 77°FThermistor $8–$253DIY
    Harvest motor hums, nothing dropsHarvestStripped gear in the ejector module, or cubes fused into a slab by an over-long freezePull the module and turn the cam by hand; on Whirlpool-type modular heads, jumper the T and H test points to force a harvestCam turns a full revolution under light finger pressure with no skip or clickIcemaker module / head $45–$1303DIY
    Water pooling in the bin instead of iceThermal loopBin drain plug unseated so melt passes straight through, or the harvest heater stuck energisedSeat the drain plug and re-run; then infrared thermometer on the evaporator 8 min into freezeEvaporator below 15°F at the 8-minute markDrain plug $5; control board $60–$1501 / 4DIY / tech
    Ice tastes of plastic or smells staleWater pathPlasticiser off-gassing on a new unit, or biofilm in the reservoir and pump line on an old oneNew machine: discard three batches. Old machine: wipe the reservoir wall with a white paper towelTowel comes back clean and dry, no slick film on the walls or under the lipDescaler and sanitiser $8–$201DIY
    Cycles look normal, output roughly halvedThermal loop / ambientAmbient derating, dust-blinded condenser, or inlet water above 70°F — often no fault at allWeigh 24 hours of output on a kitchen scale, log room temperature, compute Harvest Index0.90+ at 70°F ambient; the healthy floor drops roughly 10% per 10°F riseNothing, or a condenser brush and vacuum1DIY
    Find your symptom in column one, run only the test in that row, and act on the pass band. Do not run every test on the table — each row is a complete diagnosis on its own.
  3. Why the reset didn't fix it

    A reset is a latch clear. That's all it is. When a board sees a reading outside its window — thermistor open, level probe dry, harvest not confirmed within the timeout — it latches a fault state and stops, because continuing would run a compressor into a condition it can't survive. Holding the power button for eight seconds clears the latch and lets the machine start a fresh cycle with hope in its heart.

    It does not change the reading. So the machine runs one cycle, maybe two, sometimes three, and the same out-of-window value arrives at the same input, and the same latch trips again. Owners read that pattern as "it worked for a bit and then died again" and conclude the unit is finished.

    It isn't finished. It's telling you the same thing three times. The useful question is never "how do I reset it" — it's "which reading is out of window", and that's what the table above answers.

  4. Four instruments cover the entire table

    • A basic multimeter with resistance and DC volts. A $25 unit is fine — you are looking for open versus not-open and a resistance in a wide band, not laboratory precision.
    • An infrared thermometer. Anything that reads down to -20°F. Point it at the evaporator prongs, not at the ice, and not through the lid.
    • A kitchen scale that reads to at least 10 lb in ounces. This is the instrument that converts "not enough ice" into a number.
    • A phone timer. Cycle length is diagnostic on its own — an 18-minute bullet freeze cycle at 70°F is telling you something before you touch a probe.

    If you're plumbed in, add a $12 hose-thread pressure gauge. Short fills from low supply pressure masquerade as pump faults constantly, and the gauge settles it in thirty seconds.

  5. Reference readings

    What you're measuringInstrument and rangeWhere to put the probeHealthy bandWhat a failed reading means
    Thermistor at 32°FMultimeter, 200 kΩAcross both thermistor leads, disconnected, sensor bulb in a stirred ice-water bath for 3 min30–35 kΩ for a 10 kΩ NTC (B≈3435)Reading high means the board thinks it's warmer than it is — the freeze cycle never terminates
    Thermistor at 77°FMultimeter, 20 kΩSame leads, sensor sitting in still room air for 10 min9.5–10.5 kΩOpen circuit (OL) is a broken sensor or a chafed lead at the grommet
    Portable pump coilMultimeter, 200 Ω or 2 kΩAcross the pump's two spade terminals, connector unplugged40–120 Ω depending on wattageOL means a burnt winding. A near-zero reading means a shorted winding — replace either way
    Plumbed inlet valve coilMultimeter, 2 kΩAcross each solenoid's two terminals, harness unpluggedRoughly 200–700 Ω per coil, and both coils within about 15% of each otherOne coil far off the other is a failing valve even if it still opens sometimes
    Supply water pressureHose-thread gaugeAt the shutoff feeding the fridge or undercounter unit, with a second fixture running20–120 psi. 40 psi minimum if the feed comes off a reverse-osmosis tankBelow 20 psi gives short fills — small, hollow, or half-formed cubes that look like a freezing problem
    Evaporator mid-freezeInfrared thermometerDirectly on the prongs or cylinder wall, 8 minutes into a freeze cycleBelow 15°FAbove 25°F at 8 minutes points at a blocked condenser, a stalled fan, or lost charge
    Condenser outletInfrared thermometerOn the discharge side of the condenser coil, 10 minutes in15–35°F above room temperatureBarely warm means the compressor isn't pumping. Scorching means airflow is blocked
    Prong temperature at harvestInfrared thermometerSame spot, from the moment the harvest phase startsAbove 40°F within 60–90 sSlow rise is scale insulating the prong, or a hot-gas valve that isn't shifting
    Use this only to interpret the test named in your symptom row. Record the number you actually got — if the fix doesn't hold, that number is the first thing a technician will ask for.
  6. Refrigerator modules branch at the bail arm

    Built-in freezer ice makers share a design across Whirlpool, KitchenAid, Maytag and Amana, and it behaves nothing like a countertop machine. There's no pump and no bin sensor — there's a mould with a heater under it, a rake driven by a cam, and either a wire bail arm or an optical beam that stops production when ice piles up.

    Check these four in order before you condemn the module.

    • Is the bail arm in the down position? Half of all "my fridge stopped making ice" calls are an arm knocked up by a bag of frozen vegetables.
    • Is the freezer at 0–5°F? Above about 10°F the mould thermostat never reaches its harvest set point and the module simply waits. Put a thermometer in there overnight rather than trusting the display.
    • Does the fill tube have an ice plug? A slow-closing inlet valve dribbles after the fill and freezes a plug in the tube. Thaw it with warm water, then replace the valve — the plug is a symptom, not the fault.
    • Does a forced harvest work? On the modular head, jumpering the T and H test points drives one full cycle. If the rake sweeps and the valve fills, the module is fine and your problem is upstream in water or freezer temperature.

    A refrigerator module that makes three to four pounds a day is working correctly, and that surprises people. It is not a substitute for a dedicated machine, and no repair will make it one.

  7. Where to stop

    Sealed-system work is where a home diagnosis ends. If the condenser is clean, the fan spins, the compressor is warm and running, and the evaporator still won't get below 25°F eight minutes into a cycle, you're looking at lost refrigerant charge or a dead compressor. Recovering and recharging needs certification and equipment, and on a countertop machine the labour alone usually exceeds what the unit is worth.

    Say it plainly: on a bullet maker in the $90–$220 class, a sealed-system fault is a write-off. On a 15-inch undercounter machine it's worth a service call. The dividing line is roughly whether the repair quote lands under a third of replacement.

    Everything above the sealed system is fair game. Pumps, fans, valves, thermistors, sensor pairs and ejector modules are all bolt-in parts in the $8–$130 range, and none of them need refrigerant handling.

  8. Prove the fix with a number

    Don't declare victory because ice came out. Weigh a full 24 hours of production on a kitchen scale, note the room temperature while it ran, and divide the measured pounds by the nameplate rating on the sticker. That ratio is the Harvest Index, and it's the only honest answer to "is it fixed".

    Anything at 0.90 or above at 70°F is a healthy machine. Between 0.75 and 0.89 in a warm kitchen is normal derating and not a fault. Below 0.60 means something on the table above is still wrong, and you fixed the wrong row.

More guides