How to Test Whether Your Garage Door Is Balanced

Run this with the door closed and the opener disconnected, in that order. Five minutes, no tools.

Three things people get wrong. Pulling the release with the door open, which changes what's holding the door up rather than just how it's controlled. Testing at only one height — a door can be balanced near the floor and out of balance near the top, because the relationship between spring torque and door weight shifts through the travel. And grabbing the door wherever it's convenient, which usually means the seam between two sections: the one place DASMA names outright as unsuitable, because that joint closes on itself as the door moves.

Stand to one side rather than directly underneath while you're working, and keep the opening clear the whole time.

How you can tell: You're looking for two things: whether the door moves freely and no faster than the force you apply, and whether it holds position when you let go.

The one you want. The door moved freely, it held at 3-4 feet and at every other height you tried, and lowering it took no more effort than lifting it.

A small amount of downward drift is normal and intended. The counterbalance is deliberately set very slightly under the door's weight, so that if the door does move on its own, it moves down rather than up. An inch or two of slow creep over a minute is fine. Steady sinking isn't.

The useful thing to do with a good result is record it. Balance changes gradually, and the value of this test compounds — a good result this month plus a marginal one in the spring tells you something neither reading gives you alone.

How you can tell: Moves freely under your hands, holds at 3-4 feet, half open and three-quarters, and comes down under easy control.

The most common failing result, and the direction doors nearly always drift.

It means the springs are producing less lift than the door weighs. Under power nothing looks wrong — the opener has ample authority to move a moderately under-sprung door — which is exactly why this test exists.

How far and how fast matters. A door that creeps down slowly is mildly under-sprung and worth booking. A door that sinks steadily is substantially under-sprung. A door that drops fast enough that you have to catch it should end the test on the spot, because that's a door whose counterbalance is close to not working at all.

The cause is usually springs that have fatigued and lost lift with accumulated cycles, though weight added to the door and springs that were never correctly sized produce the same reading. Telling those apart is a measurement rather than a guess.

How you can tell: The door moves downward on its own from wherever you stop it, at any height or at the higher ones first.

Less common and less discussed, but it's a genuine fault rather than a bonus.

An over-sprung door has more lift than it weighs, so it wants to rise. In practice it means the door pulls up out of your hands, slams into the fully open position, and puts real load into the opener every time it closes — because now the motor is fighting the springs on the way down rather than being helped by them on the way up.

It also makes the door harder to hold closed, and on a door with a manual lock it can put strain on hardware that was never intended to hold a door down.

Usual causes are springs wound too far during a previous adjustment, springs specified for a heavier door than the one fitted, or weight removed from a door — a heavy wooden door replaced with a lighter steel one on the original springs, for instance.

How you can tell: The door rises when you let go, opens faster than it closes, and bangs at the top of its travel.

This isn't a balance result, and the right response is to stop rather than to keep testing.

A door that racks as it rises has lost support on one side. The candidates are a lift cable that has come off its drum, a drum that has slipped on the shaft, or a cable that has parted. All of them mean the two sides are no longer taking up equal cable, and continuing to work the door drives the rollers into the track at an angle.

Worth being explicit because it comes up so often: this is not what a broken torsion spring does. Both springs on a two-spring door sit on the same shaft and act as a single counterbalance, so a spring failure makes the door uniformly heavy, not crooked. Crooked means cables.

How you can tell: One corner leads the other as the door moves. The door looks twisted in the opening. Often a visibly slack cable on the trailing side.

Friction and weight feel identical in your hands, and they have completely different answers, so it's worth separating them before drawing conclusions.

The signature of friction is a door that's hard to move in both directions — as much effort to lower as to lift — and that feels notchy or catches at particular points. The signature of an under-sprung door is one that's hard to lift and easy to lower. A door that's heavy but holds position is usually both: friction is holding it where the springs no longer can, which is a genuinely deceptive result.

The way through it is to remove the friction and test again. That's an hour's work with a can of the right lubricant, and it makes the second measurement honest.

How you can tell: Similar effort in both directions. Grinding, notchy or catching movement. Hardware that looks dry, dusty and gray, or rollers that skid rather than turn.

A different test that belongs in the same five minutes, because it answers a question balance doesn't: will the door stop if it meets something on the way down?

UL 325 is the operator safety standard behind the requirement, and residential openers have had to reverse on contact for a long time. What the standard can't do is confirm that yours still does — that's what the 2x4 test is for, and it's the only way to know.

On how often, the sources are unusually consistent, and they say monthly. CPSC's announcement of the final opener rules states the reversing feature should be tested monthly. LiftMaster's manual for the 8500WSM says the safety reversal system must be tested every month, and must be re-tested after any adjustment. DASMA asks homeowners to make monthly inspection and testing part of the routine and to add an annual visit from a trained technician on top of it. So: monthly for the contact-reversal and photo-eye tests, yearly for a professional inspection. The twice-a-year deep clean and lubrication we recommend further down the page is our own addition for Arizona conditions, not an industry standard — don't substitute it for the monthly test.

Reversal matters more after a balance change than at any other time, because force settings interact with door weight. A heavier door means an opener working closer to its limits, and any adjustment made to compensate changes how the door behaves on contact.

How you can tell: You're testing whether the door reverses promptly when its bottom edge meets a solid object.

Two loose ends worth closing properly.

Reconnection first. The trolley relatches either by pulling the release cord back toward the opener or by running the opener and letting the carriage catch as it passes, depending on the model. What matters more than the method is the position: with the door closed, every time. Re-engaging with the door raised means that if the springs are marginal, the door can drop the instant the trolley catches. A disconnected trolley left forgotten is also the single most common reason we get called to a door that “closes and then opens itself.”

Frequency second, and it's worth separating three different intervals that often get blurred into one. The monthly pass is the balance check plus the contact-reversal and photo-eye tests — that's the cadence CPSC, LiftMaster and DASMA all state for the reversal test, and the balance check rides along with it for free. Once a year, a trained technician looks at the whole system, which is a different kind of look than you can give it. And twice a year we suggest a deeper clean and lubrication pass, which is our own recommendation for Mesa and Gilbert garages rather than anybody's standard — it exists because of what dust and heat do here, and it is not a replacement for the monthly test.

How you can tell: You'll know the trolley has relatched when the opener moves the door normally through a full cycle.