LG Washer OE Drain Error

Symptom

An LG WM3470HWA front-load washer stopped mid-cycle with water still in the tub and an OE error on the display.

Diagnosis

The whole drain path checked out: hose, filter, and the pump itself, whose winding measured a healthy 10.9 Ω. The fault was in the air path the washer uses to sense its own water level: restricted, so the level value in service mode sat around 250 while the pump emptied the drum, never climbed to the 260 that means empty, and the control timed out the drain step.

Fix

Cleared the restriction in the pressure-sensing air path and refitted the cleaned original pump. The washer drained and advanced normally through a full cycle. No parts replaced.

What an OE Code Actually Reports

On an LG front-loader, OE is a drain error. It's tempting to read that as "the drain pump is dead," but that isn't what the machine is telling you, and on this washer the difference was worth the entire cost of a pump.

There's no sensor on the pump. The control board can't tell whether the pump is spinning, whether the impeller is turning water, or whether anything is coming out of the drain hose. What it can do is run the drain step and watch the water level. If the level hasn't reached the point the machine considers empty by the time the drain step is over, it stops and posts the code.

Starting With the Checks That Cost Nothing

Before anything came apart, the drain path got the basic inspection it always should:

• The external drain hose, checked for kinks, crushing, and a blockage at the standpipe
• The pump filter, opened, inspected, and cleaned (the usual haul of lint, hair, coins and buttons collects here)
• The hose running from the tub sump to the pump, checked for a restriction upstream of the pump itself

These go first for a practical reason, not just a frugal one: a clogged filter and a failed pump produce an identical code, and one of them is free to fix. Plenty of OE calls end right here.

This one didn't. Every part of the drain path was clear, and the washer still wasn't completing a cycle. The pump itself was the next thing to look at.

Testing the Pump: What the Meter Could and Couldn't Say

With the machine unplugged, I pulled the drain pump, inspected it, and measured across its motor winding:

The impeller and housing were clear and undamaged, so the pump got cleaned while it was out and went back into the machine.

Now, an honest word about that reading, because it would be easy to present it as more conclusive than it is. A resistance measurement looks at the winding sitting still, cold, unpowered, and carrying no load. Several things that stop a pump from moving water leave that measurement looking perfectly normal:

• Mechanical failure in the pump end. A worn, cracked, or obstructed impeller has no electrical signature at all. The motor can run exactly as designed and still move nothing.
• Bearing drag. A rotor that spins freely by hand can still fail to reach speed against a tub full of water.
• A shorted turn. Shorting a handful of turns out of several hundred barely moves the reading on a meter, but it costs the motor real torque and makes it run hot.
• Anything that only appears under load or once warm. A bench reading captures none of it.

So the meter didn't clear this pump on its own. What it did was rule out the one failure mode it's genuinely good at, an open winding, and leave the pump as an unconfirmed suspect rather than a condemned part. That distinction is what the next decision turned on.

How a Washer Knows How Much Water Is In It

No float, and nothing electrical down in the water. The machine measures air.

At the bottom of the tub there's an air chamber, a small trap that holds a pocket of air. A thin hose runs from that pocket up to a pressure sensor mounted high and dry near the control board. As water rises in the tub it compresses the trapped air, and the pressure in that hose rises with it. The sensor converts that pressure into an electrical signal, and the control board reads the signal as a water level.

It's an elegant design: the sensor never touches water, so it never corrodes and never gets fouled with detergent. The cost of that design is that the machine's only knowledge of its own water level travels through a narrow tube full of air. Interfere with the tube and the washer goes blind, while every part that looks important is still working perfectly.

Which raises the obvious question. If the machine's entire understanding of its own water level is a column of air in a tube, how do you find out what it's actually reading? You ask it.

Asking the Washer What It Sees

Like most modern appliances, this LG has a service test mode, a diagnostic mode entered at the control panel, separate from the normal wash programs. It exists so a machine can be exercised and interrogated one function at a time, rather than judged from the outside by whether a cycle happens to finish.

The function that mattered on this job displays the water level the control is currently reading. That takes the invisible middle of this whole problem, air pressure in a hidden tube converted to a signal on a wire, and puts it on the front of the machine as a number.

Getting Into It on This Machine

On this WM3470HWA, with the drum empty and the machine powered off:

1. Press and hold Spin Speed and Soil Level together.
2. With both still held, press Power. The panel comes up in test mode rather than in a normal wash program.
3. Press Start/Pause repeatedly to step through the component tests one at a time. Each press advances to the next stage.
4. On this control platform the drain sequence lands around the eleventh press. That's the stage where the pump runs and the display shows the water-level value the control is reading.

The value on screen is a frequency the pressure sensor returns, not gallons or inches, and it runs backwards from what you'd expect: the sensor's output frequency falls as pressure rises, so a low number means a full tub. With water in the drum it sits somewhere in the 230 to 250 range on this machine, and as the pump clears the water it should climb steadily to 260, the value this control reads as a completely empty drum.

So rather than inferring what the washer believed about its own tub, I could read it directly. Stepped to the drain test and watching the display while the pump ran, the value sat at around 250 and stayed there. The pump ran. Water came out of the drain hose. The number on the panel never climbed.

That's the observation the rest of the job turned on. It split the two halves of the OE definition cleanly apart: the water, and the report of the water. One of them was still working, and it wasn't the one the error code points at.

The Same Readout Verified the Fix

This is worth pulling out on its own, because it's the part that turns a plausible repair into a confirmed one. The level display found the fault, and then it proved the fix: once the air path was cleared, I went back into the same drain test, ran it again, and watched the value climb up to 260 as the water left, the way it should have all along.

Without that, the strongest available claim is "it seems to work now," which is a statement about one cycle on one afternoon. Watching the level signal follow the water is a statement about the specific thing that was broken. Those are very different standards of proof, and only one of them is worth putting a guarantee behind.

Finding the Restriction in the Air Path

I disconnected the pressure hose and tested whether air moved through it freely. It did pass air, but with noticeably more resistance than it should have. It would barely bubble unless I blew through it hard.

Restricted rather than blocked is exactly the right fault for this symptom. A restricted air path still equalizes, just slowly, and "slowly" against a drain step that lasts a few minutes is indistinguishable from not at all, which is why the readout looked frozen. So the tub empties, the pressure behind the restriction lags far behind reality, the reported level never climbs to 260, and the drain step runs out of time waiting for a number that is going to arrive long after the machine has given up on it. The washer posts OE over a tub it has already emptied.

The Test That Settled It

With the pressure hose disconnected from the sensor, the washer advanced through its cycle normally.

That result is worth unpacking, because it's easy to over-read. An open port leaves the sensor breathing plain atmospheric pressure, which is exactly what it would see on a genuinely empty tub, and it reports the value to match: 260. The machine read "empty," was satisfied, and carried on. What that proves is that the sensor and the control board are both doing their jobs, and that the thing holding the cycle hostage was the level signal rather than anything in the drain system.

On its own it doesn't convict the hose. It's the combination that does: a sensor and board demonstrably working, plus a directly measured restriction in the air path between them and the tub.

I cleared the restriction by blowing the air path through with it disconnected at both ends, checked the air chamber at the tub, and reconnected everything. One detail matters here: clear the hose with it off the sensor. Pressurizing a level sensor by blowing into a hose that's still attached to it risks pushing debris and moisture straight into the part you're trying to prove is healthy.

Why These Restrict in the First Place

The tube sits above a tub full of warm, soapy water for years. Detergent residue, softener, lint and sludge gradually work their way into the air chamber and the first stretch of hose, and condensation collects in any low spot in the run. Front-loaders are more prone to it than top-loaders because they use so little water that the wash is more concentrated. If you want to make it less likely on your own machine, the two things that help most are dosing detergent for the actual load rather than by habit, and running an occasional hot cycle to keep residue from setting up, the same maintenance that keeps the door gasket from turning black.

The Diagnosis In Order

Read down that list and the shape of the job is clear enough. One fault, in the system nobody thinks about, wearing the error code of the system everybody does.

What This Job Shows About How We Work

A Note on This Write-Up

This describes one specific repair on one specific machine, not a universal procedure. Washer designs vary considerably between manufacturers and between models from the same manufacturer, and an OE error on a different LG can easily have a different cause than the one found here, including, quite often, a genuinely failed pump. If you're working on your own machine: unplug it and shut off the water first, and expect the drain filter to hold more water than you think it does.