A headlamp that runs shorter than it used to is almost always a contact and storage problem, not a worn-out light. Corrosion on the battery contacts, a spring that has lost its grip, and cells that sat in the tube for a season all add resistance and burn runtime you thought you had. The fix is a short maintenance pass: clean the contacts, restore the spring tension, rinse the salt out of the strap, and store the light empty and dry.
I do that pass at the kitchen table the night before a hike, and it takes less time than filling a water bottle. Headlamp on the table with the battery door flipped open. Two cells rolled off to the side. A bottle of isopropyl alcohol, a handful of cotton swabs, a pencil with a clean eraser, and a scrap of fine abrasive cloth. That is the whole kit.
Look into the empty tube first. Two points carry the current: the flat pad on the inside of the door, and the spring at the far end. Both should read as bright metal. What I usually find is a dull gray haze, sometimes a ring of white crust around the base of the spring, and grit packed into the corner where the plastic meets the contact.
The contact is where the runtime goes
That gray haze is oxide, and oxide is a resistor. It sits between the cell and the metal, the light has to push through it, and the voltage at the emitter sags. You have not lost the cell's capacity, you have lost the ability to get it out. Contact resistance of a few tenths of an ohm looks like nothing on a meter and like everything on a beam that is supposed to run four hours. On a fresh set, a dirty contact might cost you a few minutes of top brightness. On a set that is already half down, the same film is the difference between usable and useless.
Cotton swab, isopropyl alcohol, scrub. Damp, not dripping. Alcohol lifts the residue and flashes off in seconds, which is why it beats water inside a battery compartment. Dry the tube with the other end of the swab and give it a minute before the cells go back in.
If the crust is white and crusty rather than hazy, you are looking at leaked electrolyte from an alkaline cell, which is potassium hydroxide. It is caustic and it eats plated contacts. Alcohol will carry it away and dry clean, but it will not neutralize it. A little white vinegar on a swab neutralizes what is left, then a distilled-water swab, then dry, then a final alcohol wipe. Keep all of that on the contact instead of running down into the housing. The nose of a pencil eraser lifts light corrosion without cutting into the plating. Fine abrasive cloth does the same job where the eraser cannot reach. What you do not want is a file or a knife. Once the nickel plating is gone, the base metal underneath corrodes faster than the contact ever did.
Grit in a slide switch causes the same flicker and can keep the light from shutting off cleanly, so a dry toothbrush across the switch slot is worth thirty seconds while the light is open.
The spring is a wear item, and nobody treats it like one
Battery springs take a set. Compress and release a coiled spring a few hundred times over a few seasons and it comes back shorter, and the pressure on the cells drops. Low pressure means a contact that is touching rather than pressed, and a contact that is touching is the one that flickers when you turn your head.
Here is how I test it without instruments. Cells in, door closed, hold the light a foot over the bed and give it a gentle shake. A steady beam is fine. A flicker, or a beam that cuts out until you press on the door, is a contact problem, not an emitter problem. Sometimes the light dims when you tilt it because the cell shifts a millimeter off the pad, which is the same diagnosis with a different symptom.
Bending a spring back is fine, in small steps. Coiled spring: bend at the middle of the coil, not where the wire is soldered or riveted to the board, and check the fit after each nudge. Flat leaf spring: bend from the middle of its length. You are looking for firm resistance when the door latches, not a fight to close it. If the spring will not come back without you leaning on the joint, stop. A cracked solder pad is a repair, not maintenance, and it is the kind of repair that ends with a light in a drawer.
While the door is off, look at the metal tab on the inside of it. That tab bends easily when cells get jammed in the wrong way, and a tab that has been pushed flat loses its pressure the same way a spring does.
Salt on the strap ends up inside the light
Sweat dries into salt on the strap, sunscreen goes on tacky and picks up trail dust, and the whole assembly gets stuffed in a pack where it stays damp for a day. Salt is hygroscopic. It pulls moisture out of the air, and a light sitting in a closed bag with a salty strap is sitting in a slightly wetter environment than the room around it. That is the mechanism connecting a grubby strap to a green contact six months later, and it is the part people skip because the strap does not look electrical.
Rinse the strap in cool water with a drop of mild soap, work it with your fingers, rinse it again, and hang it to air dry out of direct sun. Heat is what kills the elastic. A dryer or a hot dashboard will take the stretch out faster than any amount of trail use. Wipe the housing with a damp rag, and keep insect repellent and heavy sunscreen off the plastic and the gasket, because both will soften rubber over time.
Batteries out, dry, and not full
The cheapest thing you can do for a headlamp you are not using is take the cells out of it.
Alkaline cells left in a stored light can leak, and the leak is what kills the light, not the battery going flat. A headlamp that sits in a drawer for a year with cheap cells inside is a headlamp you may open to a corroded mess. Removal costs five seconds.
Rechargeable lights have a subtler version of the same problem. An electronic switch draws a small standby current even when the light is off, which is parasitic drain you cannot see, so a stored headlamp with the cell in it will be flat when you reach for it, and a cell left at zero for months will not come back all the way.
For lithium-ion, storage charge matters. Sitting at full voltage for months accelerates capacity loss, and sitting empty is not kind to it either. Partial charge in the middle of the range is the usual answer, and if the manual names a storage voltage, that number wins over anything you read online. Store the light cool and dry. A hot car in July is the worst place for both the cell and the gasket. A silica packet in the bin is cheap insurance if your storage spot is a garage or a basement.
One more thing about runtime numbers, because it explains a lot of false alarms. The figure on the box came from a bench test at room temperature, usually measured until output falls to a tenth of where it started. Alkaline cells sag hard in the cold. If your light quits early on a January hike but runs fine on the kitchen table, that may be chemistry, not maintenance.
What cleaning will not fix
If the contacts are clean, the spring is firm, the switch is dry, and fresh cells still fade in twenty minutes, you are past maintenance. That is an aged cell, a degraded emitter, or a driver issue, and no amount of scrubbing changes it. The cleaning pass is still worth doing first, because it tells you which one you have, and that is worth knowing either way.
Clean contacts, firm spring, rinsed strap, cells out and stored partial in a dry place. That order is the whole checklist, and it is the difference between a light that gets replaced every other season and one that keeps its runtime for years.
