Hard water and the ice maker: the failure that arrives one month at a time

What hardness is doing while nothing looks wrong

Hardness is dissolved rock, mostly calcium and magnesium that the water picked up on its way to the tap. Moving and cold, it stays in solution and does no harm at all. It comes out of solution in three situations, and an ice maker manages all three several times a day. Water standing still in a warm place evaporates and leaves its mineral behind. Water that gets heated deposits it faster, which is why the harvest heater under a mold builds a film on the mold face over the years. And water that freezes rejects it outright: as the crystal forms it pushes dissolved solids into whatever water is still liquid, which is why the last part of a cube to freeze is the cloudy part, and why a machine that freezes ice slowly over a plate with water circulating ends up with a sump far harder than the water feeding it. None of this is a failure. It is a deposit, and it lands in the same handful of places every single cycle.

• The screen inside the inlet valve, which is the finest opening in the entire water path
• The fill tube, where a small amount of water stands between cycles and warms through every defrost
• The face of the mold, worked over by the harvest heater against wet metal several times a day
• The sump, pump and evaporator plate of an undercounter clear-ice machine, which recirculate the same water for a whole freeze
• The filter housing and the fittings either side of it, where flow slows and turbulence drops mineral out
• The saddle or stop valve at the back of the cabinet, which is often the oldest fitting on the line

Why it is always the fill side that suffers

The fill on a refrigerator ice maker is timed, not measured. The control energizes the inlet valve for a fixed number of seconds and trusts that the usual quantity of water arrives. Nothing in the appliance counts ounces, watches a level or compensates for a slow supply. So every restriction upstream converts directly into a smaller cube: a screen half blinded, a fill tube narrowed by a third, an inline filter three years past its date. The timer does not care. That is the single most important thing to understand about an ice maker on hard supply, because it explains why the machine keeps behaving perfectly normally while delivering less and less. There is a second effect that is easier to see. A narrowed tube holds water it cannot drain, that water freezes across the bore, and the following fill backs up and runs down the outside of the tube instead of into the mold, leaving a lump of ice at the back of the compartment and cubes fused into each other.

The order the symptoms arrive in

Scale is a slope rather than a cliff, and it produces its symptoms in a fairly reliable sequence. Recognizing where you are on that slope is worth more than any single observation, because the cycle count never changes. The machine makes the same number of attempts it always made. What changed is what each attempt delivers.

• Cubes slightly smaller than they used to be, which nobody notices and nobody should be expected to
• Hollow cubes, or shells with a dip in the top that crush between two fingers
• Trays filled on one side only, or a mold that has plainly filled part way
• A lump of ice at the back of the mold, or cubes fused into a slab, meaning the fill tube is icing
• Production roughly halved, the bin never full, and somebody in the household starting to buy bags
• Nothing at all, at which point the module usually gets blamed for a fault that lives upstream of it

What helps, and what people buy that does not

The most common purchase after a run of bad ice is a new water filter, which is the right thing to change and does not do what most owners think it does. A carbon filter is for chlorine, taste and particulate: the water leaving it is exactly as hard as the water going in. What a filter change genuinely fixes is a restriction in the filter itself, which is real and worth having, and it is why the ice can pick up for a short while after a change and then slide back. If the improvement does not last, the restriction is somewhere past the filter. Softening is the only one of these that addresses hardness, by exchanging calcium and magnesium for sodium, and it protects everything downstream at the cost of a small sodium addition and a change in the taste of the ice. Reverse osmosis makes beautiful ice and starves more ice makers than anything else on this list, because a tank part way through refilling delivers water at well under line pressure and the timed fill simply ends before enough has arrived.

• A carbon filter changed on schedule: fixes taste and its own restriction, does nothing whatsoever to hardness
• A softener: stops scale forming downstream, adds sodium, and changes how the ice tastes in a drink
• Reverse osmosis: the clearest ice available and the most commonly starved supply, unless pressure at the valve is checked
• An inline filter fitted under the sink at installation and never changed since, quietly the tightest point on the line
• A pressure reading at the inlet valve, static and running, which settles in a minute what a year of guessing has not

Descaling: what an owner can do and where it stops

There are two different appliances hiding under the words ice maker, and they want opposite treatment. A refrigerator or freezer mold-type maker is not a descaling job for an owner. The mold is a coated casting with a heater and a thermostat bonded underneath it, and pouring acid descaler in, or levering the assembly out to soak it, is how a working ice maker becomes a replacement ice maker. An undercounter clear-ice machine is the reverse: it is designed to be cleaned, it has a cleaning cycle in its own control, and the manufacturer names the cleaner. Use that one. Most of these evaporators are plated, general household descalers attack the plating, and a nickel-safe scale remover exists precisely because of that. On hard supply, the published cleaning interval is a floor rather than a target, and a machine in a covered patio bar working through a hot valley afternoon will want it more often than the manual assumes.

• Bin: empty it, wash with mild soapy water, dry it, and accept that old ice picks up odor no matter how good the water is
• Filter: change on the published schedule and write the date on the cartridge, because the housing hides its age completely
• Clear-ice machine: run the machine's own cleaning cycle with the cleaner the manufacturer specifies, then sanitize
• Valve screen and fill tube: water off, pressure measured, screen out and inspected, which is technician work rather than a Saturday job
• Never: acid descaler down the fill tube of a refrigerator ice maker, or a mold assembly taken out to soak

Module or water: the repair that gets bought twice

The two repairs sit at opposite ends of the appliance and at opposite ends of the invoice. The water side is a filter, a valve screen, a fill tube or an inlet valve, and it is the cheap end. The harvest side is the module, the mold thermostat, the ejector motor and, on some designs, a mold and heater assembly that is not sold as a screen and a spring. Cubes plus one pressure reading decide which of those you are buying, and the ordering matters because a new module fitted to a starved line produces exactly the hollow cubes the old module produced. That is the fault this trade fixes twice more often than any other, and on hard inland supply the odds sit firmly with the water side.

• Hollow, small or partly formed cubes with the cycle still running on time: water side
• No ice at all from a freezer that is genuinely cold, with a dry mold: harvest side
• A lump of ice at the back of the mold or cubes fused together: fill tube, narrowed or icing
• Output that halved across a year rather than stopping in a day: scale, almost every time
• Two good weeks after a filter change and then a return to hollow cubes: the restriction is past the filter