Short answer: the reed switch is not the part that fails outdoors — its soldered lead joints are. Glass capsules carry a -40 to +125 C operating range on typical datasheets, and the contacts sit sealed inside the glass, so rain, frost and sun never reach them. Water at the joints corrodes the copper and leaks enough current to mimic a closed switch. Clamp each lead before you bend or crop it, seal every joint with adhesive-lined heat shrink topped by two-part epoxy, and mount the assembly in a small stainless box rated NEMA 4X / IP66 by the manufacturer (not UL listed), entered from the bottom through an IP68 gland with a drip loop outside. Build it that way and a gate, mailbox or tank-level sensor runs for years untouched.
What the weather attacks, and what it leaves alone
Cold, heat and pressure all stay inside the capsule's comfort zone. Typical reed switch datasheets specify -40 to +125 C operation, and special high-temperature types reach a realistic 200 C. Solar gain sounds worse than it is: a dark housing in full sun runs 20-30 C above air temperature, so a 35 C day makes a 55-65 C interior — still 60 C below the capsule's limit. The glass is stable for the same reason: it was annealed from its forming temperature down to room temperature during manufacture, so decades of 10-20 C day-night swings add nothing new. Weather fronts move barometric pressure a few kilopascals, trivial next to those forming stresses.
Moisture is the attack that matters, and it lands outside the glass: on the leads, the solder and the flux residue. A thin water film starts corrosion within weeks and carries leakage current that a high-impedance input reads as a closed switch; thin electrolyte films are all atmospheric corrosion needs. The contacts themselves never see any of it, because they sit sealed inside the capsule.
Check the magnet as well. Ferrite magnets hold their field to about 250 C per magnet grade data, while standard neodymium grades begin to weaken near 80 C. A black neodymium magnet bolted to a sun-heated gate can be the weak link, not the switch.
What you need
You are sealing four things: two lead joints, one cable entry and one lid. The parts are ordinary; the details are what keep water out.
| Item | What to pick | Why this one |
|---|---|---|
| Reed switch | any 14 mm capsule rated -40 to +125 C | covers every outdoor climate on Earth |
| Sensing cable | 2-core outdoor cable, about 0.75 mm² [18 AWG] | stiff enough to strain-relieve the leads |
| Heat shrink | dual-wall adhesive-lined, 4:1 shrink ratio | the melted adhesive, not the plastic, makes the seal |
| Epoxy | two-part potting epoxy | grips the smooth glass where tubing creeps |
| Cable gland | M10 (10 mm entry), IP68, for 3-6.5 mm OD cable | typical M10 datasheet range; IP68 keeps the box's IP66 intact |
| Enclosure | small stainless box, NEMA 4X / IP66 manufacturer-rated | 304 stainless inland, 316L near salt |
| Tools | smooth-jaw pliers, heat gun, soldering iron | clamp, shrink, solder |
Pick the smallest box that takes the capsule, the gland and a short service loop; a 100 x 100 x 100 mm [about 4 x 4 x 4 in] interior is plenty for a bare 14 mm capsule. For what the Type 4X / IP66 pair covers in rain, hose spray and salt, the NEMA 4X vs IP66 guide maps the two ratings side by side. If a molded plastic box suits the site better, read whether a plastic junction box stays watertight outdoors before you commit. We build stainless to your own dimensions: set the size, 304 or 316L, and the cutouts in the 3D enclosure builder, or start from custom stainless enclosures and have every hole cut to the drawing.
Build the sealed assembly in six steps
-
Clamp, then form the leads. Grip each lead flat in smooth-jaw pliers, leaving at least 3 mm [1/8 in] of straight lead between the jaws and the glass, and make the bend beyond the jaws. Crop to length only where the pliers still grip. An unclamped bend or cut drives mechanical shock into the glass-to-metal seal, and the capsule can crack there or drift out of calibration.
-
Solder fast, with a heat sink. Isolate the sensing loop before the iron touches it. Work hot and quick — around 350 C for two to three seconds per joint — with pliers clamped on the lead behind the joint, so the heat soaks into the steel instead of the glass.
-
Shrink adhesive tubing over each joint. Slide a 15-20 mm length of 4:1 dual-wall tubing over each finished joint, one end overlapping the wire insulation, the other ending against the glass. Shrink it until melted adhesive beads out of both ends; the bead, not the plastic, is the seal. It is the same discipline that keeps 600 V conductor splices dry outdoors: seal to insulation at one end and to the component at the other.
-
Coat the joints with epoxy. Mix the two-part epoxy and cover each joint and both tubing ends, then let it cure hard before the lid goes on. Tubing grips insulation; epoxy grips glass. Together they leave no smooth surface for water to lift.
-
Enter through the bottom wall. Send the M10 gland hole with the drawing through the cutout designer so we cut it at order time, fit the gland, and form a drip loop in the cable below the gland so water runs off the loop instead of into the threads. Tighten the cap onto the cable jacket and blank any unused hole with a rated plug. The outdoor junction-box sealing guide covers gland choice, drip loops and drains in depth; on a shallow box where a gland stands too proud, a low-profile cable entry seal does the same job.
-
Mount and close. Fix the capsule with its leads strain-relieved, set the magnet gap to the operate distance on the datasheet, and pull the lid down evenly so the gasket compresses all the way round.
The condensation check: a 30 C afternoon, a 15 C night
Run this arithmetic once for your own climate and the sealing stops being a judgment call.
- Sealing conditions: the box closes on a warm, humid afternoon — assume 30 C and 80% relative humidity here; put in your own site values.
- Trapped air: a 100 x 100 x 100 mm interior is 1.0 L.
- Vapour load: saturated air at 30 C holds 30.4 g of water per m³ in standard psychrometric tables, so at 80% RH the trapped litre holds 30.4 x 0.8 = 24.3 mg.
- Dew point: air at 30 C and 80% RH condenses at about 26 C on the psychrometric chart.
- Overnight wall temperature: assume 15 C on a clear night; an exposed metal wall also radiates to the sky and sits 2-3 C below night air.
Rule. Vapour condenses on any interior surface colder than the dew point of the trapped air.
Arithmetic. The wall at 15 C sits 26 - 15 = 11 C below the dew point. Air cooled to 15 C saturates at 12.8 g/m³, so it keeps 12.8 mg as vapour and releases 24.3 - 12.8 = 11.5 mg of water onto the walls. Spread that over the box's roughly 600 cm² of interior surface: 0.0115 cm³ / 600 cm² ≈ 0.00002 cm — a film about 0.2 µm thick.
Result. A fifth-of-a-micron film of water across the glass and every bare joint. Invisible, and precisely the thin electrolyte film that corrodes copper and carries leakage current.
Decision. Seal the joints wherever the overnight low can fall below the afternoon's dew point, which on clear nights is most climates; with walls running 2-3 C below night air, even a night within 3 C of the dew point produces the film. Sealed joints make that film harmless — it lands on the outside of the seal and dries away. Cut the cable to length rather than coiling spare metres, because every extra litre of trapped air brings another 24 mg. Leave the vent off: it trades this film for blown rain and dust, and the vent-or-sealed comparison shows the rare cases where airflow genuinely wins.
Checks when the box is closed
- Actuation. On a continuity meter, bring the magnet to its working gap: the switch pulls in and releases cleanly. Wiggle the cable at the gland; a flickering reading means a joint is moving under the seal.
- Isolation. Move the magnet away and read resistance across the leads. Capsule datasheets list isolation of 100 MΩ or more, so a reading below a few megohms says moisture or flux residue is bridging the joints.
- Water. Hose the closed box for two minutes from all sides — a Type 4 rating is defined against hose-directed water, so this is the field version of that test. Open the box and look for droplets at the gland and the lid gasket.
- First freeze. After the first hard freeze-thaw cycle, repeat the isolation reading. It should sit where it was.
Mistakes that cost the rating
- Bending or cropping an unclamped lead. The shock travels into the seal and the capsule cracks or drifts out of spec — sometimes sticking closed, sometimes never pulling in again. Clamp first, every time.
- Plain heat shrink instead of adhesive-lined. Standard tubing seals nothing; water wicks along the strands under the wall and reaches the joint within a season. The 4:1 dual-wall adhesive type is the one that seals.
- A gland rated below the box. The weakest part sets the assembly: an IP54 gland on an IP66 box is an IP54 installation. Match the gland at IP66 or better — IP68 is the usual datasheet figure for M10 glands.
- Entry through the top or an upward-facing gland. Water pools on an upward cap and runs into the threads; enter through the bottom, or sideways under cover. The article on cable glands facing up on an outdoor junction box walks through that placement.
- A spare-cable coil inside the box. Extra trapped air means extra condensate on the first cold night; cut the cable to length or leave the coil outside.
- A field-drilled hole left bare. Drill, deburr, and fit a gland or a rated blanking plug; a bare hole is an open hole whatever the rating says. The cutouts and glands guide covers hole sizing and spacing.
Catalog dimensions to compare
Compare these catalog dimensions with your component layout. Open each enclosure for its rating, material and usable space, then check those specifications against your application.
| Enclosure | Material | Outside H x W x D / display | Rating |
|---|---|---|---|
| 12 × 10 × 6 in | 304 stainless | 12 x 10 x 6 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 16 × 14 × 8 in | 304 stainless | 16 x 14 x 8 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 20 × 16 × 8 in | 304 stainless | 20 x 16 x 8 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 16 × 12 × 8 in | 304 stainless | 16 x 12 x 8 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
Questions
Can a reed switch crack in freezing weather?
No. Glass capsules are rated down to -40 C operating, and some datasheets list storage to -65 C; the glass was also annealed from its forming temperature during manufacture, so winter sits well inside its history. Cracks come from bending or cropping an unclamped lead, not from cold. Keep water out of the joints and there is nothing left to freeze and expand.
Is adhesive-lined heat shrink enough, or do I need epoxy too?
For rain and sprinkler splash, 4:1 dual-wall tubing with adhesive beads at both ends seals the joint on its own. Add the epoxy coat where the joint can sit wet for days: salt air, washdown, ground-level splash. Tubing grips the wire insulation while epoxy grips the smooth glass capsule, and together they leave no surface water can lift.
What enclosure rating does an outdoor reed sensor need?
A NEMA 4X / IP66 stainless box covers rain, hose spray and corrosion for a gate, mailbox or tank sensor. The rating holds only if the entry matches: an IP68 M10 gland keeps the assembly at IP66, while a lower-rated gland pulls the whole box down to its own figure. Use 304 stainless inland and 316L within sight of salt water.
How close does the magnet need to be?
Work to the numbers rather than a guess: read the capsule's sensitivity in ampere-turns and the magnet's pull-in curve, then set the gap so the switch still releases with the door at rest. Off-the-shelf magnetic door contacts of this kind commonly actuate between 10 and 25 mm with their matched magnets; your datasheet pair decides the exact figure.





