Two blades in glass

Magnets made by a magnet

Inside the tube are two flat blades of a magnetic alloy, one sealed into each end, with their tips overlapping in the middle and a small gap between them. The contact faces are plated with ruthenium, and the glass keeps dust and damp away from them. Nothing else is in there.

Slide the magnet in. The blades are not magnets on their own, but in a magnet's field each one becomes one, and the two overlapping tips come out as opposite poles. Opposite poles attract. Once that pull beats the blades' own springiness, the tips touch and the switch is closed. Take the magnet away, the blades lose their magnetism and spring apart.

Turn the magnet over and every pole swaps, the tips are still opposite, and it closes the same way. Some Hall sensors answer to one pole only; a reed switch has no wrong way round.

Ampere-turns, not millimetres

A switch's sheet cannot say "closes at 10 mm", because that depends on the magnet. It says how strong a field it needs instead, in ampere-turns (AT): the current in a standard test coil times its turns. The MKA-14103's coil has 5000 turns, so 15 AT is 3 mA through it.

The type ranges from 8 to 35 AT. This board's part is ordered as "MKA14103 15-18", sorted into the 15 to 18 AT band, so one board closes at nearly the same field as the next: the two ends of the band are only about 6 % apart in distance. Your magnet decides the rest.

The figure's magnet is a model, a field that falls as the cube of the distance, set to close the switch at 15 mm. A real magnet can do better or much worse. Hold yours to the glass, watch the LED, and move it away until the LED goes out: that is your distance.

Where to hold it

Along the tube works best: a bar magnet lying parallel to the glass, near its middle. A magnet pointed straight at the middle of the tube, one pole facing it, often does nothing, because then both blades can get the same pole at their tips. If one position fails, slide the magnet toward an end before deciding the switch is dead.