Short answer: keep the whole charge path inside the box — charger, balancer and pack — and cross the wall once, with a low-voltage DC input through a single IP68 gland. For a 2S 1000 mAh pack charged at 0.5-1 A, that layout adds 0.6-1.1 W of heat and lifts a small stainless box about 2 K at the worst case. It leaves no hatch, no outside balance plug and no connector cap for salt water to creep into. Wireless through the wall is the second route, and it works only through plastic or fiberglass; stainless screens the field. The steps below size the gland, show the heat arithmetic and end with the rating and material to order for fresh water, salt water and true submersion.
Put the charger inside, cross the wall once
Flooded builds almost always share one design choice: a connector outside the wall. Salt spray sits in every crevice of a mated plug and creeps into the contacts, and a cap delays that by weeks, not forever. A hatch is worse. It adds a gasket and a ring of bolts that have to survive the same salt, and it opens the box every time you charge.
So invert the problem. A 2S charge controller with balancing is a standard part; set it to 8.4 V full charge, pick the current, and wire its output to the pack's own plug inside the box. The battery side of the circuit never crosses the wall again. The only penetration is one round DC input lead, and a gland seals a round lead well.
A gland keeps the box's rating when it is fitted the way its datasheet describes: the stated hole size, the O-ring against a flat wall, the cable inside the clamping range and the sealing nut at the datasheet's torque. Any wall accessory with a lower ingress figure drags the whole assembly down to its figure, so the gland is the only accessory this build takes. The fundamentals of cutting and sealing wall openings are in the cutouts and glands guide.
What you need
- A 2S charge controller with balancing, set to 8.4 V and 0.5-1 A for a 1000 mAh pack
- One PG7 gland with an IP68 figure: it seals a round lead of 4-5.5 mm OD and mounts in a 12.5 mm [0.49 in] hole
- A round 2-core input cable of 4-5.5 mm OD with 20 AWG [0.5 mm²] cores, about 5 mm OD in the common constructions
- A 2 A fuse, roughly twice the fast-charge current, for the input lead
- Four standoffs and 0.4 in [10 mm] of air gap around the board
- A pouch of silica gel for humid or salt-air sites
- For a plastic body only: a wireless charging pad and its receiver coil, feeding the same controller
Most small 2S packs carry a 2 mm-pitch JST-type plug, and that plug is happier indoors. Leads thinner than PG7's sealing range need a smaller gland, and the PG7 gland alternatives page shows how to match gland to lead before you order.
Run the charge current in through one gland
- Order the hole with the box. On a custom stainless build you set every cutout on the drawing. Put the 12.5 mm gland hole in flat wall, at least 1 in [25 mm] clear of corners and the gasket line; the factory cuts and deburrs it.
- Fit the gland dry: O-ring against the wall, sealing nut behind it, cable gripped between 4 and 5.5 mm OD. Tighten the nut to the datasheet's torque, then check that a hand-twist cannot turn the lead in the clamp.
- Make up the input lead: 2-core round cable through the gland, the 2 A fuse within 4 in [100 mm] of the entry inside, then the controller's input.
- Set the charge current. For a 1000 mAh pack, 1 A (1C) finishes in about 2 h including the balance taper, 0.5 A (0.5C) in about 3 h, and a 0.3 A trickle finishes overnight in 4-5 h.
- Keep the pack's own plug and the balance tap inside, connected for good. Service then means opening the box once, not mating a sealed connector in the rain.
- Stand the board off the wall on its standoffs, 0.4 in [10 mm] clear of the pack, so charger heat has air around it and a swelling pouch has room.
- Close the door and tighten the screws in a cross pattern, so the gasket compresses evenly along its whole length.
Check it before the first swim
- Pull test the gland: the lead must not slide, and a twist must not turn in the clamp.
- Hose or spray the closed box for the splash rating. Immersion is a different figure: IEC 60529's digit 7 means 1 m [3.3 ft] for 30 min, and only a box ordered as a submersion type should take that test.
- Measure at the end of a charge, when heat peaks. The box surface should sit within about 3 K of room temperature, and the pack below 40 °C (104 °F), leaving margin inside the 0-45 °C (32-113 °F) charge window printed on lithium cell datasheets.
- Compare cell voltages after the taper: within 0.05 V of each other means the balancer is working.
- After any water test, open the box, look for drops, dry the air and close it with the silica pouch inside.
How much heat a charge adds
Worked example, assumptions first: an 8 × 6 × 4 in [203 × 152 × 102 mm] 316L box, a switching 2S charger at 88% efficiency (a mid-range figure for small synchronous buck boards), a 12 V input, and heat leaving through the walls only, at k = 3.7 W/m²K, the stainless wall value this site's guides use.
At the end of charge, when the heat peaks:
- Charge power into the pack: 8.4 V × 1 A = 8.4 W
- Controller input: 8.4 W ÷ 0.88 = 9.5 W, so the loss inside the box is 1.1 W
- Box surface: 2 × (203 × 152 + 203 × 102 + 152 × 102) mm² = 134,000 mm² = 0.134 m²
- Rise over ambient: 1.1 W ÷ (3.7 W/m²K × 0.134 m²) = 2.2 K (about 4 °F)
The same charge through a linear charger from 12 V burns (12 − 8.4) V × 1 A = 3.6 W inside the box and lifts it 3.6 ÷ 0.496 = 7.3 K. That is the difference between a box that stays near room temperature and one that runs warm all night, so the switching board earns its place.
The cable adds no heat to the box, but it drops voltage. Ten feet [3 m] of 20 AWG is 0.033 Ω/m × 6 m round trip = 0.2 Ω; at the 0.8 A a 12 V input draws, that costs 0.16 V. The same charge from a 5 V USB brick pulls 1.9 A and loses 0.38 V, a real slice of a 5 V budget, so 12 V in keeps the arithmetic comfortable.
The sealed enclosure conductive cooling article shows how wall conduction scales when you change the box size, and where you mount the board moves that heat around inside the box, as the component layout article shows.
Wireless through the wall: plastic only
An inductive pad couples coil to coil through a few millimetres of plastic. Typical Qi-class links deliver 50-80% of the transmitter's power with the coils aligned, and the figure falls fast as they offset. Through stainless the route dies: the alternating field drives eddy currents in the wall, which screen the link and turn charge power into wall heat. Fiberglass and polycarbonate pass the field, so the wireless build pairs with a non-metallic body — the polycarbonate vs fiberglass guide covers that choice.
Budget the heat before you pick the wireless route. At a 60% link, delivering the example's 8.4 W takes 14 W from the pad, and 5.6 W is lost; assume the receiver coil and rectifier take half of it, and 2-3 W of heat lands inside the box — two to three times the wired route's 1.1 W. Give the coil its own corner, away from the pack. Why stainless fights inductive links, and what works instead, is in the wireless power inside a stainless steel enclosure article.
Mistakes that flood the box
- A cap over an outside plug: it is a splash delay, not a seal, and salt sits in the crevice it leaves.
- A gland gripping a lead below its clamping range, padded with filler: filler is not a rated seal, so pick the gland size instead.
- A rough field-drilled hole: burrs cut the gland's O-ring and the seal never recovers. Order the cutout, or deburr and fit exactly to the gland datasheet.
- A breather or drain on this box: the weakest part sets the assembly rating, so one low-IP vent hands the whole box its figure. Keep the box sealed and dry the air instead.
- Charging a cold pack: below 0 °C (32 °F) lithium plates onto the anode, so bring the pack inside the 0-45 °C window before the charger switches on.
- A second wall crossing for data: each opening is another seal. If the build needs USB through the wall, the rated ways to do it are in the USB pass-through article.
What to order
| Where the box lives | Body | Rating to order | The one wall crossing |
|---|---|---|---|
| Rain, splash, hose-down, always above water | 304 stainless (fresh water) or 316L (salt splash) | Type 4X | one PG7 gland cutout |
| Salt water contact: docks, brine, chloride wash-down | 316L stainless | Type 4X | one PG7 cutout with an IP68 gland |
| It can go under: a bilge, a bucket, flood water | 316L stainless | Type 6P | IP68 gland picked for the depth |
| Wireless build on a plastic wall | polycarbonate or fiberglass | Type 4X | none; the pad stays outside |
304 pits under chloride films even at room temperature, so salt work goes 316L. Type 4 covers hose-directed water and stops there. Type 6P is the prolonged-submersion type, the nearest neighbour of IEC 60529's digit 8, which means continuous immersion at the depth the maker states. What each type and code is tested against is in the NEMA 4 vs 4X vs IP66 guide. For a build that swims, pick the gland for depth as well as diameter: a walkthrough on selecting an IP68 M20 stainless gland for a 40 m [131 ft] depth rating shows the checks that matter.
Send the gland hole on the drawing with the rest of your cutouts. The box is delivered in 10 business days from your approval of that drawing, anywhere in the world, and the minimum order is one piece. The types and IP codes here are manufacturer ratings: our boxes are rated NEMA 4X / IP66 by the manufacturer; not UL listed.
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 |
|---|---|---|---|
| 8 × 6 × 4 in | 304 stainless | 8 x 6 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
Can the balance lead go out through a second gland?
It can, but every extra wall crossing is another seal to keep dry, and small multi-pin glands that seal a fine balance harness rarely carry an IP68 figure at hobby sizes. Keeping the balance inside needs nothing on the wall: a 2S board with balancing uses the pack's own plug indoors, so one PG7 gland in a 12.5 mm hole carries the entire build.
What gland fits a thin charge lead from a small pack?
A 2S pack's charge pair is often only 2-3 mm OD, below the 4-5.5 mm range a PG7 gland seals well, and padding the gap with filler is not a rated seal. Order the input as a round 2-core cable of 4-5.5 mm OD instead, and keep the pack's own plug inside the box where no seal has to work.
Will an overnight charge overheat a sealed box?
At 0.3-1 A the charger wastes 0.3-1.1 W, and the example 8 × 6 × 4 in stainless box rises 2.2 K at the 1 A worst case, so heat is not the constraint. The pack's own 0-45 °C charge window is: a box left in the sun, or below freezing, fails the window long before the charger warms it.
Can a wireless pad charge through a stainless steel wall?
Stainless screens the field: eddy currents in the wall absorb the link and turn it into wall heat. Wireless works through a plastic or fiberglass wall a few millimetres thick, at a typical 50-80% link efficiency with aligned coils, and the receiver's share of that loss adds 2-3 W of heat at a 1 A charge. Keep the coil away from the pack.





