A milliwatt-class microcontroller board almost never overheats a small sealed metal enclosure on its own. The board that shut a project down last summer was probably cooked by the sun on the box, not by its own silicon. You can prove which one it is in an afternoon: measure the internal temperature rise on a mild day, add the solar gain, and extrapolate to your worst ambient. No simulation software, no heat wave required. This article walks through the measurement and the arithmetic for a small sealed box, and what to change if the numbers come back ugly. For the site-level decisions around an outdoor box, the outdoor enclosures guide is the starting point.
What actually heats the box
Three inputs set the internal air temperature:
- Internal dissipation. Everything that draws current and drops voltage turns into heat: the microcontroller, regulators, radios, LEDs, sensors.
- Solar gain. Sun on the walls pumps heat in whether the board is running or not. A dark box in full sun can sit 20 to 30 C (36 to 54 F) above ambient; a shiny metal box with shade sits far lower.
- Ambient air. The baseline everything else adds to. Design for the hottest afternoon the site sees, not the average.
Heat leaves only through the walls, because the box is sealed. The rise above ambient is roughly proportional to the total heat input, which is exactly what makes the bench test below work.
What you need
- The enclosure, closed as it will be mounted, with the real board or an equivalent resistive load inside
- Two temperature probes: one inside near the hottest component, one outside in the shade
- A multimeter or USB power meter to read the board's current draw
- A notebook, or better, a data logger
- For the sun test: a clear day, and the compass heading and tilt the box will really have
Any cheap probe works. A grill thermometer clipped through a grommet is ugly but honest; an indoor-outdoor unit with the sensor on a lead inside the box is cleaner.
Step 1: Add up the watts inside
Measure, do not guess. Power the board as it will run in the field, with radios transmitting and outputs driving real loads, and read voltage times current at the supply input. A low-power microcontroller board can idle at a few milliwatts and still be fine; the parts people forget are the regulators and anything that switches.
A linear regulator dropping 12 V to 3.3 V at 100 mA burns about 0.87 W as heat, more than the microcontroller it feeds. A radio that transmits in bursts may average low but peak high; log long enough to catch a representative average. Write down the total in watts. This is your internal heat input.
If the load turns out to be tens of watts rather than a fraction of one, stop here: the box needs a cooling strategy, not just a measurement.
Step 2: Measure the rise on a mild day
Thermal rise is linear over the range these boxes live in, so one mild-day measurement predicts the hot-day result. Pick a day with a shaded ambient near 20 to 25 C (68 to 77 F), in the shade and out of the wind.
- Mount or support the box the way it will be installed, in the shade.
- Power the board at its real duty cycle.
- Wait for equilibrium: the internal probe should stop climbing, typically after 1 to 3 hours for a small metal box.
- Record internal temperature and shaded ambient at the same moment.
- Subtract: that difference is your internal rise from electrical load alone.
Example: shaded ambient 22 C (72 F), internal 26 C (79 F). Rise from the board: 4 C (7 F). A 6 x 6 x 4 in stainless box with a fraction of a watt inside lands in this range, which is why the board is rarely the villain.
Step 3: Add the sun
Move the box into full sun, oriented as it will be mounted, with no board power or with the board powered, and repeat the equilibrium measurement. The difference between shaded and sunlit readings, at the same ambient, is the solar rise for that orientation.
Example: same day, ambient 22 C, box in sun, board off: internal 34 C (93 F). Solar rise: 8 C (14 F). With the board on: internal 38 C (100 F). Total rise above ambient at that orientation: 16 C (29 F). If the site's worst ambient is 43 C (110 F), the box interior will see roughly 43 + 16 = 59 C (138 F). A board rated to 85 C (185 F) has margin; a board in a black box with a 30 C solar rise does not.
Mounting height, orientation and the shading an eave or pole provides all change this number; the enclosure mounting guide covers the installation side.
Step 4: Sanity-check the hot spot
The air temperature is not the whole story. Components in still air inside a sealed box run above the internal air temperature, and a part mounted flat against a sunlit wall can be hotter than the probe reads. Check the component on the board that ran hottest in the mild-day test; a probe or an infrared thermometer on that package, extrapolated the same way, tells you the real margin. Keep the hottest parts away from the sunlit face, and stand the board off the wall so air can move.
If the numbers come back too high
Work the list in order of payoff:
- Shade or shield the box. A sun shield, a light finish, or simply mounting under an eave removes most of the solar rise.
- Mount the box out of direct sun if the site allows it. Orientation and shading are free; everything else costs parts.
- Improve conduction inside. Stand the board off the wall on standoffs, put the hot package near a metal wall, and let air move inside by convection.
- Open the box, carefully. A filtered vent trades some sealing for convective cooling, and changes the ingress rating; the guide to NEMA 4, 4X and IP66 explains how the ratings behave.
- Move the heat out. If the load is real watts rather than milliwatts, plan active cooling: fans with filters, heat exchangers or air conditioners for sealed boxes.
Mistakes that void the test
- Measuring with the board idle. An idle board that sleeps between transmissions understates the duty cycle in the field.
- Testing in wind or afternoon shade. Wind cools the walls; shade during the sun test hides the solar gain.
- Ignoring the regulator. The drop from 12 V or 24 V to 3.3 V can dwarf the board's own draw.
- Extrapolating the wrong baseline. Add the total rise to the shaded ambient at the site, not to the sunlit wall temperature.
- Forgetting the rating. If you later add vents or cutouts, the box no longer seals the way it did during the test. Any NEMA type or IP code on such a box is a manufacturer rating, not a UL listing, and modifying the enclosure voids it.
What to specify once the numbers are in
If the extrapolated internal temperature stays comfortably below the board's rating, order the sealed box as planned: a stainless wall-mount enclosure, manufacturer-rated NEMA 4X / IP66 and not UL listed, in the size you tested, delivered in 10 business days. If the margin is thin, spec the same box with a light finish and plan the shade before it ships, or send us the layout for a custom enclosure with the board mounted to conduct heat into the wall. Pick the size with the enclosure sizing guide so the next revision of the board still fits, and check the rating wording against the guide to NEMA 4, 4X and IP66 before it goes on a drawing. For a door mounted where the sun hits it, the outdoor enclosures guide rounds out the rest of the site-level decisions.
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 × 6 × 4 in | 304 stainless | 6 x 6 x 4 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 |
| 24 × 20 × 8 in | 304 stainless | 24 x 20 x 8 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
| 30 × 24 × 10 in | 304 stainless | 30 x 24 x 10 in | Rated NEMA 4X / IP66 by the manufacturer; not UL listed |
Questions
How many watts can a board draw before a small sealed box overheats?
There is no single number, because it depends on box size, material and sun. As a rough check, a 6 x 6 x 4 in (150 x 150 x 100 mm) stainless box in shade sits about 2 to 5 C (4 to 9 F) above ambient per watt of internal load. One watt from a microcontroller board is rarely a problem; twenty watts from a power supply is a different conversation and needs a real cooling strategy.
My board is rated to 85 C (185 F) ambient. Is that the same as inside the box?
No. The 85 C figure is the air temperature around the board in free air with its own heat dissipated. Inside a sealed box in the sun, the air can be 20 to 30 C (36 to 54 F) above outdoor ambient. Subtract the internal rise and the solar rise from 85 C to find the real outdoor temperature your design tolerates.
Does a stainless box run cooler than a plastic one?
Usually yes, in sun. Metal conducts heat from the sunlit wall to the shaded walls and sheds it to the air, so the whole shell works as a fin. A plastic box insulates, so the sunlit face gets hot and stays hot. In shade the difference is small, because both boxes just track ambient plus the internal load.
Can I test the whole thing in an oven instead of waiting for summer?
You can, if the oven holds a low setting reliably and you keep the powered board under observation. A bench test at a mild ambient plus the extrapolation method in this article gives the same answer with less risk, because the rise is linear over the range these small boxes work in. Use the oven only to confirm a borderline result.
