Short answer: write IP68 on the order, with depth and duration in the wording, because IP67 is tested at 1 m for 30 min and a floating tracker capsizes and stays down longer than that. A 15 x 10 x 5 cm [5.9 x 3.9 x 2 in] enclosure displaces about 0.75 L, so it floats roughly 750 g [1.7 lb] of total weight, and your assembled tracker has to come in under that — under 500 g [1.1 lb] if you want a third of the box riding clear. If it must float by itself, use a polymer case with a screwed, gasketed lid. Save stainless for a box tethered to a buoy, because steel this size weighs about what the water it pushes aside weighs.
Two metres of water is a pressure test, then a cold soak
Water adds about 10 kPa [1.5 psi] per metre of depth. At your 2 m operating depth the box sees about 20 kPa [0.2 bar, 3 psi] above the air trapped inside, which sounds small until you spread it over the lid: on a 15 x 10 cm face that is about 300 N [67 lbf] pressing inward. External pressure seats a gasketed lid, so depth helps a good seal and exposes a poor one — the load bows a thin lid and opens a gap at the corner farthest from a compression point.
Above the waterline the same box takes spray, wave slap and sun. A black polymer case in direct sun runs 20-30 °C above air temperature, then falls to water temperature within minutes of a capsize; that swing works on the gasket and on your battery voltage. Fresh water at 1000 kg/m³ gives you 750 g of support per 0.75 L displaced; seawater at 1025 kg/m³ gives 770 g, and its chlorides pick the material: 316L stainless or a coated metal, not bare aluminum.
IP68 or Type 6P: the rating to write
IEC 60529 tests the second digit 7 by immersion at 1 m for 30 min in still water, once. For anything that floats, capsizes or sits tethered at depth, write IP68 and make the declared depth and duration part of the spec — “IP68, 2 m, continuous” is common wording for small screwed-lid cases. The article on why IP68 ratings carry depth and time limits shows what that wording buys you, IP67 vs IP68 for battery projects compares the two across a build, and the IP rating chart covers what each digit tests.
On the NEMA side, Type 4X covers hose-directed water and corrosion: splash, rain, washdown, never immersion. The submersion type is 6P — prolonged submersion at a limited depth — and NEMA's own conversion guidance pairs it with IP68. Our custom line offers Type 6P as a manufacturer rating for the design you approve; those enclosures are not UL listed. On a US spec, NEC 110.28 states an IP code cannot replace an enclosure type rating, so write both: IP68 with depth and duration, NEMA Type 6P.
| Where the tracker lives | Rating to write |
|---|---|
| Rain, spray, splashes | IP66 (or NEMA 4X) |
| Dropped in, out within 30 min, no deeper than 1 m | IP67 |
| Floating, capsizing, sitting submerged | IP68 rated for continuous immersion at your depth, or Type 6P |
| Tethered under the surface for weeks | IP68 with depth and time in writing; see continuous immersion enclosures |
Will it float: run the mass budget first
Buoyancy is displaced volume times water density, and on a box with straight sides the submerged fraction of volume equals the submerged fraction of height. Your 15 x 10 x 5 cm envelope is 750 cm³: 750 g of support in fresh water, 770 g in seawater, at full submersion. A box riding that low is one wave from swamping, so design to two-thirds of displacement, about 500 g all-in, and a third of the box stays dry.
Worked example, polymer case:
| Item | Mass |
|---|---|
| Polycarbonate case, 15 x 10 x 5 cm, 2.5 mm wall | 165 g |
| 4 AA alkaline cells at 23 g each | 92 g |
| Tracker board, GPS and LoRa module | 60 g |
| Antenna, standoffs, 10 g desiccant sachet | 25 g |
| Total | 342 g |
342 g against 750 g of buoyancy floats the box 46% submerged: 2.3 cm of draft, 2.7 cm of freeboard, antenna clear. The same load in seawater rides at 44%.
Now run the same envelope in stainless. A 15 x 10 x 5 cm box carries about 550 cm² of surface; at 1.2 mm [18 ga] wall that is 66 cm³ of steel, and 304 at 7.9 g/cm³ puts 520 g in the shell before the lid frame, gasket and screws — call it 600 g of enclosure. Add the 177 g of tracker and you are at 777 g against 750 g of support: it sinks in fresh water, and only a foam collar saves it.
One size up, the smallest box in the stainless size finder at 6 x 4 x 4 in [152 x 102 x 102 mm] displaces about 1.6 L, but at the same 1.2 mm wall it carries roughly 0.8 kg of steel, so with your tracker aboard it rides about two-thirds submerged — call it 35 mm [1.4 in] of freeboard, and every chop ships water over the lid. That paragraph is the material decision: self-floating means polymer, metal means a buoy or added flotation.
Polymer, aluminum or stainless: pick by the antenna and the mooring
UV-stabilized polycarbonate or FRP is the default for a self-floating tracker. A 2-3 mm polymer wall passes an 868 or 915 MHz signal with under 1 dB of loss, so the whip lives inside the box, dry, with no penetrations at all — the cleanest seal is the one you never cut. Cases this size weigh 100-200 g, which is what makes the mass budget close.
Aluminum splits the difference. The same 550 cm² shell in 1.5 mm aluminum weighs about 220 g, a real floater with a metal body. The RF then has to leave through an antenna on the top face, and salt water pits bare aluminum, so specify it anodized or painted and touch up every screw hole.
Stainless earns its keep when salt water and decades matter more than floating. 316L shrugs off chlorides, and the box sinks or rides low, so you moor it beneath a buoy or bolt it to a rail and let the buoy carry the foam. The antenna goes out through an IP68 bulkhead connector on the lid, a few grams of seal you accept because metal blocks the signal. The GRP vs stainless junction box comparison walks the material trade in more depth, and every route starts at custom enclosures, where you set material, width, height, depth and every cutout.
Lid, gasket and screws: what holds 0.2 bar
Between 1 m and full depth the failure mode shifts from leakage to flex, and the closure carries it. A screwed lid with a continuous gasket is the build for a box that rides under: six M4 stainless screws on a lid this size, spaced no more than about 100 mm [4 in] apart so the face cannot bow between them. Compress a shore-A 50-60 EPDM or silicone gasket 20-30%; that band closes a sponge seal's voids without taking a set. The screw cover vs latch comparison covers when a latch is the better daily answer, and cover flatness and gasket compression shows why an even, flat squeeze beats a thick gasket crushed at the corners.
Molded clip-top cases earn IP67 at 1 m in the maker's test, with clips snapping a tongue-and-groove O-ring shut. Give one a dunk to depth and a cold soak and two things relax: the polymer stiffens as it cools, and 300 N [67 lbf] of external pressure bows the lid away from the clips. A clip lid is right for a case you open weekly that takes splashes; for a box that lives on the water, screws win.
Condensation: a sealed box still breathes
The 0.6 L of air sealed inside your case holds about 8 mg of water vapour at 25 °C and 60% humidity — a film, not a flood. The danger is breathing: if the smallest leak lets air move, each thermal cycle draws humid air in as the box cools on the water and leaves the moisture behind when it warms. Grams add up over a season, and they condense on the coldest surface, which is your PCB at night. Fix it with a 10 g desiccant sachet changed at every battery swap, a hydrophobic membrane breather on the top face, or both. The same mechanism drives condensation in a sealed IP67 box, where the cable entry was the breathing path, and condensation in an outdoor electronics case, where the day-night cycle does the pumping.
What to put on the order
| Line of the spec | What to write |
|---|---|
| Rating | IP68, 2 m, continuous — a manufacturer rating, not UL listed; add NEMA Type 6P on US specs |
| Material | Self-floating: UV-stabilized polycarbonate or FRP. Buoy-mounted: 316L stainless. Metal and floating: anodized aluminum |
| Envelope | 150 x 100 x 50 mm [5.9 x 3.9 x 2 in], loaded weight under two-thirds of displacement, about 500 g here |
| Closure | Screwed lid, 6 x M4 A2 stainless screws, EPDM gasket at 20-30% compression |
| Penetrations | None below the waterline; antenna connector or gland on the top face only |
| Vent and drying | Hydrophobic membrane breather on top; 10 g desiccant clipped inside |
| Internals | Boss for the tracker board; battery holder that grips cells when the box inverts |
Custom work runs from one piece: you set the material, the three dimensions and every cutout, and the enclosure is delivered in 10 business days anywhere in the world, counted from your approval of the drawing. If your next step is narrowing sizes and ratings first, the enclosure selector filters by environment and inside dimensions.
Prove the seal before you commit the electronics: ballast the empty box to the rated depth for 30 min, open it over a dry cloth, and look for beads at the gasket and the gland. Then decide by the mooring. The box floats alone with its antenna inside: polymer, IP68 continuous, screwed lid. It rides a buoy in salt water: 316L, Type 6P, antenna through the lid, foam in the buoy. You want metal and it must float: 1.5 mm anodized aluminum, coated, under 500 g all-in.
Enclosures for this job
Catalog sizes that suit the job this article answers, picked from its subject; open one for its full specifications.
| Enclosure | Material | Outside H x W x D / display | Rating |
|---|---|---|---|
| 6 × 4 × 4 in | 304 stainless | 6 x 4 x 4 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 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 |
Questions
Is an IP67 case good enough for a tracker that floats?
For the floating itself, yes — spray and wave slap are well inside IP67. The risk is the unattended dunk: IPX7 is tested at 1 m for 30 min, so a capsize that leaves the box deeper, or down longer, sits outside the rating. If you can guarantee recovery inside 30 min and 1 m, IP67 is enough; if the box spends nights on the water, order IP68 rated 2 m continuous.
How much weight can a 15 x 10 x 5 cm enclosure float?
750 g in fresh water and 770 g in seawater at full submersion, but design to two-thirds of that, about 500 g, so a third of the box rides dry. The sensitivity is worth knowing: on this 150 cm² footprint every extra 100 g of payload costs about 7 mm of freeboard, so a heavier battery choice shows up fast in how the box sits.
Do you need a breather vent in a sealed floating box?
Fit a membrane breather on the top face and you solve two problems at once. When the box hits water 15 °C colder than the air inside, that gas contracts by roughly 5% of its volume and would otherwise pull replacement air in through the weakest seal. A hydrophobic membrane lets pressure equalise while blocking liquid water, and most breathers carry an IP68 rating of their own.
Which batteries last longest in a box floating on cold water?
Water at 10 °C pulls an alkaline cell well down its curve: at 0 °C and moderate drain you get roughly half of the 20 °C capacity, and the transmit pulses sag first. Lithium iron disulfide AA cells hold capacity to about -20 °C and weigh about 15 g each against 23 g for alkaline, so four cells also save you 32 g of mass budget.
Where should cable entries go on a floating enclosure?
On the top face, and only if you truly need them. A gland rated IP68 with a clamping range matched to your cable — a 3 to 6.5 mm range covers most small sensor and antenna cables — holds 2 m; a gland on the side or bottom sits at or below the waterline and wets the seal the whole time afloat. An internal antenna with zero penetrations is the cleanest build.
