Why a Smart Lock Will Not Sit on a Door Built in 1962

The thermostat learned a new schedule over the wire one night in February. The doorbell became a camera during an update nobody watched. The garage opener took a phone app, the lights answer a voice, and the water heater sends a message when it is unhappy. Not one of those needed a van at the curb or a person on the step.

The deadbolt is the exception, and it will stay the exception. It is a piece of machinery living in a hole cut through a piece of wood, and on a great many houses in this city that hole was cut between 1950 and 1975. Software cannot widen it. That is the whole reason a lock is still the one thing in the house that ends with somebody kneeling at the door with a jig and a chisel.

The instructions in the box assume a door that came off a line after about 1975. Plenty of Cupertino doors did not, and nothing on the packaging says so.

The first is the bore, the round hole through the face of the door. Today's standard is two and an eighth inches across, with a one-inch hole through the edge to meet it. Doors from the early tracts frequently carry an inch and a half, because that is what the knob set of the period wanted. A modern chassis will not go into that hole, and no amount of pressing the escutcheon against the wood will change it.

The second is the backset, the distance from the edge of the door to the center of the bore. Two and three-eighths and two and three-quarters are both normal, and most kits carry a latch that adjusts between them. What kits do not carry is two and a quarter or two and a half, which turns up on doors that were converted from an older mortise lock at some point in their life. On those the latch bottoms out before the bolt reaches the strike.

The third is thickness. Locks are rated for an inch and three-eighths through an inch and three-quarters, and a slab door on the older tracts is often a true inch and three-eighths. Some keypad units want the full inch and three-quarters for the interior housing to seat squarely, whatever the range on the box says. Stile width matters too: a deadbolt centered in a three-and-a-half-inch stile leaves under three-quarters of an inch of wood on either side of the hole.

Hinge screws pull out of a jamb that has been drying for sixty years. The leaf drops a few millimeters at the latch edge, and the bolt stops meeting the middle of the strike. Households adapt without noticing: they lift the handle, or lean a shoulder on the door, or pull it toward them and then turn the key. Ask anyone in a 1962 house how they lock their front door and you will get a small piece of choreography as the answer.

A person can do that. A motor cannot. A keypad deadbolt throws its bolt with a gearmotor sized for a bolt that moves with about the force of two fingers. Put it on a door that needs a shoulder and it stalls, reports a jam, retries, and works its way through a set of cells in a couple of months. Nothing is wrong with the lock. The door is wrong, and the lock is the only part of the arrangement that can complain.

The test costs nothing and takes ten seconds. Open the door wide. Throw the bolt with the thumb turn while the door hangs free. If it runs out smoothly in the air and fights you with the door shut, the fault is alignment. Fix that and half the problems people blame on the lock disappear.

The door comes first. Hinge screws are replaced with three-inch screws that reach the framing behind the jamb, which pulls the leaf back up where it was hung. The edge is planed only if it truly has to be. The strike pocket is re-cut and deepened so a one-inch bolt throws its full inch into solid wood instead of sitting halfway home against a lip of paint.

Then the hole. Where the bore is an inch and a half, it is opened to two and an eighth with a guide clamped across the door so the cut stays square to both faces. A hole started freehand wanders, and a wandering hole on a sixty-year-old slab is not something anyone can undo. Where the stile is too narrow to give up that much wood, we say so before the saw comes out and the conversation turns to a different lock rather than a damaged door.

Then the lock goes on. Codes are chosen by you rather than left on the factory setting, the bolt is thrown twenty times from both sides, and the mechanical key is cut and tested at the van before anybody drives away. All of it is one visit, because there is no sensible way to split it into two.

The spread is the door, not the brand of lock. A leaf that still hangs true with a two-and-an-eighth bore already in it sits at the bottom: the old lock comes off, the new one goes on, the visit is short. A leaf that has dropped, or a bore that has to be opened, or a strike that has to be moved and re-cut carries the middle. The top of the range is a door that needs all three, and it is still one visit and one figure said before anyone sets off.

The $89 service call is inside every one of those figures rather than added underneath. Between 8pm and midnight the figure carries $55, and between midnight and 7am it carries $95, both counted on the call rather than discovered on the receipt. The labor is warranted for ninety days. Say the age of the house when you ring and the range closes toward one number straight away, because the year is the single most useful thing you can tell a locksmith about a door.

Where a door has gone past correcting — a split stile, rot in the bottom rail, a leaf twisted enough to show daylight against a straightedge — that gets said plainly. The honest answer there is a new slab from a carpenter with our work after it, not a good lock screwed to a door that cannot hold it.

Rented houses are the clearest case, and there are a lot of them here: about 39 percent of the occupied homes in the city are rentals. A tenant should not be boring a new hole through a landlord's front door. Where the existing hole already takes a modern chassis, a keypad that drops into it is fine and reversible. Where it does not, the answer is a lock box or a conversation with the owner, not a saw.

The other case is simpler. Plenty of households have never once wanted to let a cleaner in from a phone, and for them a plain deadbolt at $129 to $339 with the same door work underneath it does everything the keypad would have done, with nothing to charge. The door work is the part that mattered either way. That is the only real argument this company makes, and it is the same argument whichever lock you end up choosing.