Garage Door Makes a Grinding Noise

The most common source, and the one people usually find first once they know to look.

A roller wheel spins on a bearing around its stem. On the plain steel rollers most builders install, that bearing is a ring of loose ball bearings in an open race — no seal, no cover, nothing keeping anything out. They work well for years and then they don't.

When the balls wear or escape, the wheel stops rotating cleanly. Sometimes it seizes altogether and skids along the channel, which produces a flat spot on the tread and a hard, rhythmic grind. Sometimes it keeps turning but on metal rather than on bearings, which produces a rougher, more continuous version of the same sound.

Both are consuming the stem. That matters more than the noise, because the stem is what holds the wheel perpendicular in the channel. A thinner stem means play, play means the wheel tips, and a tipped wheel is the first step toward one leaving the track.

How you can tell: The grinding travels with the door and has a rhythm to it that matches wheel rotation. Spinning the suspect wheel by hand shows it binding, notchy, or not turning at all — and there's often a visible flat spot.

The version worth checking before condemning any parts, because it's fixable with a brush and it's extremely common here.

Monsoon dust is fine silt — small enough to work past the shield on an unsealed roller bearing and into the race. Once it's in there, every rotation drags hard particles across the bearing surfaces, which is exactly the mechanism a lapping compound uses. It also collects in the bottom of the vertical channels, packs down after it gets damp and dries, and sets into something close to grout that the roller then has to grind its way through.

The sound is more of a gritty rasp than a hard grind, it's usually present on both sides rather than one, and it tends to appear in the days after a serious dust event rather than developing slowly.

The reason to act quickly: abrasive grinding starts as a cleaning job and becomes a parts job. Grit in a bearing destroys it over weeks, not years.

How you can tell: A rasping quality rather than a hard metallic grind, on both tracks, appearing after a dust storm. Visible silt in the bottom of the channels.

The version that stays put while the door moves past it.

The shaft above the opening turns in bearings — one in a plate at each end of the header, and on wider doors an additional center bearing. They carry the entire rotating counterbalance assembly, they turn under full spring load every cycle, and they're often the least-maintained parts on the door.

A bearing that has run dry squeals first. Left long enough, it wears through and starts to grind, and at that point the shaft is running in a race that has become oval. The shaft develops play.

That's the part that matters. The shaft's position determines where the cable drums sit. A shaft that can wander means drums that shift, and drums that shift don't spool cable evenly — which puts the door out of square, and the crooked-door sequence takes over from there. A bearing failure is one of the quieter routes to an off-track door.

How you can tell: Grinding fixed at one point on the header rather than traveling. Sometimes fine metallic dust on the floor directly below a bearing plate. Often it was squealing from the same spot months earlier.

Grinding that stays at the powerhead and doesn't move with the door.

Inside a chain or belt opener sits a nylon drive gear that transfers the motor's output to the drive shaft. It's deliberately the weakest link in the chain — it's designed to be the thing that fails when the opener is asked to do something it shouldn't, so the motor and the electronics survive. That's a sound engineering choice and it's also why grinding at the powerhead is so common.

What overloads it: lifting a door that has got heavy because the springs are tired, fighting a door that's binding, or trying repeatedly to move a door that's locked or jammed. Each of those puts the motor's full torque through nylon teeth.

The signature is unmistakable at the end: the motor runs its full travel, the rail sounds busy, and the door doesn't move at all. Before that, you get grinding on the heaviest part of the lift and a door that sometimes makes it and sometimes doesn't.

Two things to take from it. The gear is a repairable part rather than a dead opener. And a stripped gear almost always means something else made the door hard to move, so replacing it without finding that is a temporary fix.

How you can tell: Grinding located at the powerhead, worst at the start of the lift, sometimes with light-colored nylon shavings on the floor below the opener. Ultimately the motor runs and the door stays put.

The other opener-side grind, and it sounds different enough to separate.

A chain drive runs a roller chain over a sprocket at each end of the rail. Over years the chain stretches slightly and the sprocket teeth wear into a hooked profile. A stretched chain riding on hooked teeth doesn't mesh cleanly — it climbs and drops off each tooth, which produces a rasping, ratcheting grind that runs along the rail rather than sitting at the motor.

Slack makes it worse. A chain with visible sag slaps the rail and takes up its slack with a jerk at every direction change, which is both a noise source and a wear source on everything downstream.

This one is worth distinguishing from the drive gear because it doesn't have the same failure mode. A worn sprocket and chain get noisier and rougher; a stripped gear stops working outright.

How you can tell: A rasping or ratcheting sound that follows the trolley along the rail, on a chain-drive opener. Visible sag in the chain, or teeth on the sprocket that look hooked rather than symmetrical.

The version that happens at one specific height and nowhere else.

When a channel gets dented, creased or flared — usually from a vehicle contact, a bracket over-tightened onto the steel, or a previous off-track event — the internal width at that point stops matching the rest of the run. A roller arriving there is squeezed, or drops toward the lip, and either way it stops rolling cleanly and starts abrading.

The grind is short, hard, and utterly repeatable: same height, every cycle, in both directions. Over months it wears a bright polished patch into the channel and a matching flat into the wheel, and the two of them slowly consume each other.

The reason this one gets its own entry rather than being folded into the roller cause: replacing the roller doesn't fix it. A new wheel arriving at the same deformed spot does exactly what the old one did.

How you can tell: A hard grind confined to a narrow band of travel, identical every cycle. Marking the height on the jamb and looking into the channel there usually reveals the damage immediately.

Less common than a roller bearing, and it produces a grind that's easy to misattribute.

Each roller stem sits in a sleeve on the end hinge of its section. That sleeve is meant to hold the stem square while letting it turn. When it wears — and it wears faster on a door that has been running slightly out of alignment, because the stem is being pushed sideways rather than just rotating — the hole goes from round to oval.

An oval sleeve doesn't support the stem. The stem sits low in it, contacts on a narrow line rather than around its circumference, and grinds. It also lets the wheel tip, which means the wheel starts working against the channel wall at the same time. Two grinding contacts from one worn part.

The way to tell it from a bad roller: pull the wheel and the play stays. A worn bearing is play between wheel and stem. A worn sleeve is play between stem and hinge.

How you can tell: Grinding that travels with the door, plus a roller that visibly sits at an angle or can be moved up and down within its hinge. Often the hinge shows a bright wear ring around the sleeve.

The most urgent version on this page, and worth being direct about.

Each end of the shaft carries a grooved drum that the lift cable winds onto. If the cable has climbed out of its grooves and is riding over its own wraps, or has worked against a flange, it produces a grinding or grating noise localized right at that drum — and what's being abraded is the cable.

A lift cable carries the same tension the springs hold. Steel strands that have been crushed between wraps or ground against a flange are strands that have lost capacity, and cables don't degrade gracefully. They hold and then they don't.

There's a second version: a drum that has slipped on the shaft and is now turning against its set screws, which grinds against the shaft itself.

Either way, this is the one grinding noise where the correct response is to stop using the door immediately rather than to schedule something. And it usually comes with a visual: a door that no longer looks square as it rises.

How you can tell: Grinding localized at one end of the shaft, at the drum rather than the bearing plate. Often accompanied by a bottom edge that isn't staying parallel to the concrete as the door opens.