Garage Door Won't Stay Up on Its Own

The answer in most cases, and it's a slow one rather than a dramatic one.

Springs don't hold their original lift indefinitely. As cycles accumulate and the steel fatigues, each one produces slightly less than it did. The door doesn't announce this. It gets marginally heavier, marginally slower under power, and at some point crosses the line where it can no longer sit still without help.

The place it usually shows up first is high in the travel, because the relationship between spring torque and door weight isn't constant through the door's arc. A door that holds at knee height but sinks from three-quarters open is a completely ordinary presentation of springs on their way down.

What makes this worth acting on rather than living with is what happens in the meantime. Every powered cycle, the opener is making up the shortfall — and an opener is a positioning device with a nylon drive gear deliberately chosen as the part that fails first. Waiting turns a spring job into a spring-and-opener job.

How you can tell: Sinks steadily and evenly, worse from higher up. The door has felt gradually heavier over a year or more. The opener sounds like it's working and travels more slowly than it used to.

The free answer, and it accounts for a real share of these calls.

The emergency release disconnects the door from the opener's carriage so the door can be moved by hand. Pull it during a power outage, work the door manually, close it, and it's easy to walk away without reconnecting. The opener then runs its motor and rail back and forth while the door does nothing, and the door itself is being held by the springs alone — which is precisely the condition that reveals a counterbalance that has drifted.

It also produces the other complaint we hear a lot: a door that “closes and then opens itself.” With nothing holding it at the bottom, a door sitting on marginal springs can follow them back up.

Two minutes to check, and it either solves the problem or eliminates it.

How you can tell: The opener runs but the door doesn't move, or the door moves independently of the opener. The release cord hangs slack and the carriage on the rail looks separated.

On doors with extension springs running back alongside the horizontal tracks, this symptom has its own particular cause and it confuses people because nothing looks broken.

Extension springs stretch to store energy. Over enough cycles, they stop returning fully to their original free length — they take a permanent set. The spring is intact, undamaged, and producing measurably less pull than it did. The door gets heavier and stops holding position, and anyone looking at the springs sees two perfectly normal-looking springs and rules them out.

The visual tell is length. A set spring hangs longer and lazier than it should when the door is closed, and if you have two you can sometimes see one sagging more than the other.

Worth checking at the same time whether a safety containment cable runs through the middle of each spring. It has nothing to do with holding position and everything to do with what happens when the spring eventually does part.

How you can tell: Springs alongside the tracks rather than a shaft above the door, sagging noticeably with the door closed, on a door that has been progressively losing its ability to stay open.

If the door sinks with one corner leading rather than coming down level, this isn't a balance fault and it belongs in a more urgent category.

Each cable drum is clamped to the shaft by set screws so both sides lift and hold by the same amount. If one loses grip, that drum stops turning with the shaft, that side stops being supported properly, and the door starts settling on that corner. Same result from a cable that has slackened or come off its groove.

The reason it's more urgent than a spring problem is what it does geometrically. A door that settles unevenly is a door being pushed into its track at an angle on every cycle, and that's the stage immediately before it comes off the rails.

One thing this is definitely not: a broken torsion spring. Both springs on a two-spring door sit on the same shaft and act as one counterbalance, so a spring failure makes the door uniformly heavy rather than one-sided.

How you can tell: The bottom edge tips as the door sinks. Sometimes a visibly slack cable, or a drum that looks out of position relative to its bearing plate.

If the door started failing to hold immediately after spring or cable work, this is usually the answer — and it's an adjustment rather than a fault.

Torsion springs are wound a specific number of turns to produce the lift a particular door needs. Slightly too few and the door is under-sprung from the moment the work finished: liftable, workable under power, but unwilling to hold position. It's an easy thing to end up with if the door's weight was estimated rather than measured, or if the winding was set by eye and pronounced close enough.

The distinguishing feature is timing. Springs that fatigue take years to produce this symptom. Springs wound short produce it on day one, and people often assume it's just how the new springs feel.

How you can tell: The problem appeared right after spring, cable or drum work, on hardware that's otherwise in good condition.

The other way to end up under-sprung without anything having worn out: the door got heavier while the springs stayed the same.

Insulation kits fitted into the sections. Replacement panels heavier than the originals. Window inserts added later. Years of repainting across a large steel surface, which adds more than people expect. Each of these moves the door's side of the balance, and unless the counterbalance was re-sized at the same time, the springs have been short ever since.

The tell is timing again, but on a different scale. If the door stopped holding position within a few months of anything being done to it, that's the link. It's covered more fully on the heavy-door page, since weight-gain shows up there as effort rather than as sinking — but it's the same underlying change.

How you can tell: Insulation, new sections, added glass, or repeated repainting somewhere in the door's history, followed by a door that has stopped staying where you put it.

Worth being precise about, because a fair number of people arrive here about a door that's behaving correctly.

The counterbalance is intentionally set very slightly under the door's weight. The reason is directional: if the door is going to move on its own, everyone would rather it moved down slowly than up unexpectedly. So a small amount of downward creep is designed in, not a symptom.

The rough line is this. A door that eases down an inch or two over a minute and then settles is fine. A door that keeps going, at any speed, until it reaches the floor is not. A door you have to catch is a stop-using-it result. DASMA's checklist puts the pass condition at a specific height — the door should stay in the partially open position 3-4 feet above the floor — which is a more useful thing to check than a general impression that it “doesn't stay up.”

The hand test that produces that reading belongs to another page, and we'd rather send you there than half-teach it here. Read How to Test Whether Your Garage Door Is Balanced and follow it as written, because two things in that method are not optional. First, the gate: look at the torsion spring for a gap in the coil and at both lift cables for slack before you touch the release cord, and if either one fails, don't run the test at all — on a door whose counterbalance has already gone, the opener's drive train is the only thing holding the weight and the release removes it. Second, where you hold the door: by the handles or gripping points, which the standard puts within 8 inches of the bottom edge of the bottom section on the second or third section, never by the seam between two panels and never with fingers inside a joint that is about to close.

How you can tell: You're after distance and time at a stated height, rather than an impression.