New Garage Door Springs, Won't Close Right: 4 Likely Causes
The spring job is finished. The door ran up and down while somebody watched, everything looked right, and the driveway is empty again. Then you press the button that evening and the door starts down, gets within a few inches of the concrete, and rolls straight back up. Or it drops the last two feet hard enough to rattle the shelves. Or it parks short and leaves a strip of daylight along the bottom seal. You already had this fixed, and the door still will not do the one thing it is supposed to do on its own.
New garage door springs that won't close right rarely mean the new spring is bad. They mean something that had to be reset alongside the spring was left where it was.
Why New Springs Change What the Opener Expects
An opener does not lift a garage door. A standard double door runs 150 to 250 pounds depending on material and insulation, and the motor overhead was never built to move that. The springs carry the weight. The opener supplies direction and a modest push, and it works only because the door has been made nearly weightless by the counterbalance.
The opener stores two numbers describing that arrangement. Travel limits mark where fully open and fully closed sit along the track. Force settings cap how much effort the motor may spend in each direction before deciding something is in the way. Both were set against the old spring.
A fresh spring delivers its rated torque. The one that came off had lost some of that over thousands of cycles, and the opener had quietly absorbed the difference for months. Change the spring, leave the opener on its old settings, and the two now disagree about how heavy the door is near the floor.
Matching the Misbehavior to Its Cause
The way the door misbehaves narrows the list fast, because each cause fails at a different point in the travel.
| What the Door Does | Likely Cause | What Is Actually Happening |
|---|---|---|
| Reverses a few inches above the floor, same height every time | Down force set too low for the new counterbalance | The opener meets more push-back near the floor than its setting allows and reads it as an obstruction |
| Touches down, then runs straight back up | Down travel limit set past the floor | The door lands, the motor keeps driving, and the current spike registers as an obstacle |
| Stops an inch or two short and stays there | Down travel limit set too high | The opener believes it has reached fully closed and shuts off normally |
| Falls fast through the last stretch and lands hard | Spring wound short of its turn count, or too light for the door | The counterbalance no longer offsets the weight near the floor, so gravity leads |
| Will not stay down, or creeps back up on its own | Spring wound past its turn count, or heavier than the door needs | The spring pushes up harder than the door weighs at the bottom of travel |
| Closes crooked, with a gap at one corner | Cables taken up unevenly on the drums when the shaft was reset | One drum pays out more cable than the other, so one corner lands first |
Two of these often arrive together. An over-sprung door reverses near the floor and also bangs into the open position on the way up.
Travel Limits and Force Settings After a Spring Change
The down travel limit tells the opener where the floor is. Set it too far and the trolley keeps pushing after the bottom seal has flattened, the motor loads up, and the opener reverses to protect itself. Set it too high and the door parks above the threshold, the seal never compresses, and the slide lock will not line up with its slot.
Force is the other half. Older openers set both with screws on the motor housing. Newer ones learn them electronically by running the door through a full cycle and recording what it took, so a unit that never got put through that routine is still holding numbers for the door it used to have.
The safety side matters more than the annoyance. Force set too high stops the door reversing when it should. The usual check is a flat board laid on the floor in the door's path: the door should strike it and reverse without hesitation. Both settings get corrected only after balance is verified, since tuning an opener around a badly balanced door buries the problem under a heavier motor load.
A reversal that happens the instant the door starts moving points to the photo eyes, separate from the new spring. A force-related reversal repeats at close to the same height every time, almost always in the last foot.
Too Many Turns or Too Few
A torsion spring does nothing until it is wound, and the installer sets that tension by turning the winding cone a specific number of times. The count is not a matter of feel. The working rule is roughly one turn per foot of door height plus a small allowance, so a 7-foot door lands near 7 1/2 turns and an 8-foot door near 8 1/2.
Miss low and the door is heavy. It feels fine at the top, gets heavier coming down, and drops through the final stretch faster than the opener intends. Miss high and the door is buoyant. It will not rest on the floor, the opener has to hold it down, and the down force limit trips before the seal touches concrete.
An intact spring in sound condition can be adjusted rather than replaced, which is the good news. A technician adds or removes turns at the cone to match the door's actual weight. Noting the turn count and watching how the door behaves is something you can do from the ground. Turning the cone is not.
Winding is where a torsion spring's stored energy lives, and a slipped bar can break a hand or a jaw. Turn adjustment is professional-only work, done with a matched pair of solid steel bars sized to the cone holes.
When the New Spring Is the Wrong Size
Three dimensions define a torsion spring: wire diameter, inside diameter, and wound length. Together they set how much torque it produces per turn and how far it can be wound safely. Wind direction counts too, since left-hand and right-hand springs are not interchangeable. A spring that is close but not correct can be wound until the door feels acceptable at chest height while still being wrong at the floor, which is exactly where closing trouble shows up.
Doors also drift from their original specification. Insulation added to the panels, a replacement section in a different material, or a wood door that has taken on moisture all add weight the factory chart never counted.
On a two-spring door, pairing one new spring with the surviving old one produces the same symptoms. The new spring makes more torque per turn than its worn partner, so the sides pull unevenly and one corner reaches the floor first.
Testing Balance Before Blaming the Opener
Every diagnosis here starts the same way: the door gets disconnected from the opener and moved by hand. A properly balanced door lifts with roughly 10 pounds of effort and holds where you leave it around waist height. A door that sinks from there is under-sprung. A door that climbs on its own is over-sprung. Neither is an opener problem, and no amount of limit adjustment fixes either one.
That test is worth leaving to the visit. Pulling the release cord on a door whose balance you do not yet know means letting go of a panel that might drop or might rise, and you find out which only after it moves.
What a Follow-Up Visit Covers
Order matters more than the individual steps. The spring gets compared against the door's measured weight, the turn count gets verified at the cone, and balance gets tested by hand. Cables and drums get checked so both sides take up evenly. Only then do the travel limits get reset for the new closed position, followed by force in both directions.
Until then, holding the button down to force the door closed is worth avoiding. Driving a motor against a door that keeps fighting back strips drive gears and works cables loose on their drums.
Frequently Asked Questions
Yes, mostly below eye level. Repeated hard landings elongate the hinge holes in the bottom section, split the rubber bottom seal, and knock a roller out of line near the floor. On a chain drive, the plastic drive gear takes the abuse from the other direction, and glazed panels feel the shock through the insert frame.
It could, though not from an ordinary outage. Openers normally hold learned travel limits in memory that survives a power loss, so a routine flicker rarely explains this by itself. A logic board that is failing, or a full factory reset performed during the spring visit, is the more likely source if the timing lines up with the service call.
New springs take a small initial set as the coils seat under load, shifting the balance slightly during the first days of use. Cold weather adds to it by thickening the grease in the rollers and bearings. A down force setting sitting at the edge of acceptable passes on day one and trips a week later.
It spreads. An unbalanced door loads the cables and drums unevenly, wears the end bearing plates faster, and puts steady strain on the top section where the opener arm bolts on. That bracket area is a common place to find elongated holes after a long run out of balance.
It does, through a different adjustment. Extension springs are tensioned by where the hook sits in the hanger holes and by pulley position rather than by winding turns, so correcting the counterbalance means moving the attachment point. The safety containment cable running through each spring has to go back in after the swap.
Some settling sound is normal, but it should not be coil noise. A new spring's factory coating is intact, so the coils themselves should stay quiet the way a healthy spring does. What you may hear for the first few cycles instead is the cone seating fully against the shaft, or a faint metallic settle as the mounting hardware finds its final position. A rhythmic pop or clack as the door moves is a different signal, usually coils binding or a dry torsion shaft turning in its bearings.
Have a freshly re-sprung door rechecked — a technician verifies spring size and turn count against the door's measured weight, tests balance by hand, and resets opener travel limits and force in both directions. Arnold's Garage Door & Gates serves Fort Worth, Alvarado, and the DFW metroplex. Call Arnold: 682-337-7220.
