Sealed enclosure SOM cooling without a fan

A sealed enclosure can cool a system-on-module (SOM) without ventilation when a conductive path carries its heat outside. The design must control every interface between the module, heat spreader, enclosure wall and surrounding air. An internal heatsink alone does not establish that path.

Start with the module's documented temperature limits, then calculate a thermal budget and verify it in the finished assembly. Include carrier-board losses, mounting orientation and the highest expected ambient temperature. A successful room-temperature bench run does not demonstrate enough cooling margin for the installed system.

What you need

Collect the module documentation before choosing a heatsink, thermal pad or enclosure depth. You need the manufacturer's temperature measurement locations, operating limits and approved mechanical attachment method.

Prepare these inputs and tools:

• Module heat dissipation under sustained processing, communications and peripheral activity.
• Carrier-board, power-converter and other internal heat loads.
• Maximum ambient temperature, installation orientation and any solar exposure.
• Module mounting dimensions, component heights and permitted mounting forces.
• Thermal-interface conductivity, compressed thickness and required contact pressure.
• Small temperature sensors and a logger suitable for the required measurement accuracy.
• Access to module temperature telemetry, workload monitoring and thermal-throttling status, where available.
• A representative enclosure, mounting arrangement and external heatsink, if the design uses one.

Use documentation for the actual module revision and operating configuration. A processor power estimate does not automatically include memory, regulators or carrier-board losses.

Install sensors with power isolated, then run tests with the enclosure closed and suitable guarded connections. Keep sensor wiring clear of power terminals and moving door hardware.

Calculate the thermal budget

1. Choose the temperature limit and measurement point. Use the module manufacturer's defined case, board or junction limit. Set a lower design ceiling that allows for measurement uncertainty, operating variation and aging. Do not treat case temperature and junction temperature as interchangeable.

2. Define the worst sustained heat load. Record module dissipation separately from the remaining enclosure heat. Include startup or processing bursts when their duration could produce the highest temperature. Electrical input power is a useful starting point, but subtract power delivered outside the enclosure when estimating internal heat.

3. Calculate the available temperature rise. Subtract maximum ambient temperature from the selected design ceiling. Divide that difference by module heat only as an initial, single-source screening calculation. Then account for other heat entering the shared enclosure wall.

For an illustrative design, assume a module dissipates 12 W at 113°F [45°C] ambient. Assume its documented case measurement point has a chosen design ceiling of 185°F [85°C]. The available rise is 40 K, giving an initial case-to-ambient budget of approximately 3.3 K/W.

That number is not a heatsink specification. It must cover the complete path, including interfaces, wall spreading and heat rejection outside. Other internal heat sources consume part of the same allowance.

The heat sink thermal resistance explanation covers the resistance terms and their limitations. Here, the essential task is separating the module's dedicated path from the wall path shared with other equipment.

Build a defined conductive path

4. Identify each physical contact in the heat path. Sketch the module contact surface, interface material, spreader, wall and external cooling surface. Mark which parts touch and how each joint receives controlled pressure. If a heatsink only warms internal air, do not count it as a direct module-to-wall connection.

A simplified path can be written as:

module case → interface → spreader → interface → wall → outside air

The enclosure wall may spread heat sideways before rejecting it through convection and radiation. An external heatsink adds surface area, but its mounting interface also adds resistance. Use the conductive cooling through the wall article for that wider enclosure arrangement.

5. Choose the interface from the mechanical stack-up. Calculate the largest and smallest gaps using board, standoff, spreader and enclosure tolerances. Select a thermal pad within its documented compression range. Do not tighten a spreader until the board bends simply to close an uncertain gap.

For a uniform interface, conduction resistance is approximately R = t / (k × A). Here, thickness t is in metres, conductivity k is in W/(m·K), and contact area A is in square metres.

For example, a 0.039 in [1 mm] pad with conductivity of 3 W/(m·K) over 1.55 in² [1,000 mm²] gives approximately 0.33 K/W. This calculation excludes contact resistance and spreading resistance. Use compressed thickness and relevant supplier data, then confirm performance by measurement.

A thick pad can accommodate tolerances but still consume considerable thermal margin. Conductivity alone is not enough to compare interfaces.

6. Support the thermal hardware independently where necessary. A heavy spreader or external heatsink must not load small board connectors or unsupported PCB areas. Follow the module's mounting-force limits and accommodate differential thermal expansion. Keep electrical isolation requirements separate from thermal performance requirements.

Stainless steel can carry heat through its thickness, but lateral spreading may become a significant restriction. Do not assume a small contact patch makes the entire wall an equally effective cooling surface. Define the spreader footprint and wall contact location before finalizing the enclosure.

Instrument and test the assembled path

7. Measure temperatures at identifiable nodes. Place sensors at the module's prescribed measurement point, the spreader and the wall near the contact footprint. Add an outside-air sensor positioned away from the enclosure's warm exhaust plume or heated surfaces. Record internal air temperature to help identify heat bypassing the intended path.

Attach sensors without lifting the thermal interface or significantly changing contact pressure. A sensor trapped inside a joint can create the very temperature rise you are trying to measure. Use adjacent accessible locations and account for the resulting uncertainty.

Do not estimate junction temperature from case temperature using an unrelated thermal-resistance figure. Follow the module manufacturer's documented method. An infrared reading from reflective stainless steel is also unreliable without appropriate emissivity control and measurement technique.

8. Run the complete assembly at representative worst conditions. Use the intended mounting orientation, wall clearance, workload and power-converter loading. Close the door and fit the actual seals, cable entries and external cooling hardware. Test relevant operating modes rather than assuming maximum processor activity always produces the hottest condition.

The component layout article explains how nearby heat sources affect local temperatures. Keep those sources operating during the validation run.

Log temperature, ambient conditions, module power and workload together. Continue until temperatures stabilize under a predefined criterion appropriate to the enclosure's thermal mass. For example, a project might require less than 1.8°F [1°C] change over the final hour, with stable ambient conditions. That is a project criterion, not a universal test requirement.

9. Compare the measured drops with the budget. Divide each measured temperature difference by the heat actually crossing that segment. Use module heat for its dedicated conductive path. Use total relevant heat for a shared wall-to-ambient path only when the simplified model reasonably represents the assembly.

Consider this illustrative lumped model:

If all 20 W reaches the modeled wall node, its temperature is 45 + (20 × 1.2) = 69°C. The module case adds 12 × 0.75 = 9 K, reaching 172°F [78°C]. That leaves 7 K below the assumed design ceiling.

These are hypothetical inputs, not performance claims for an enclosure or module. Real wall temperatures are uneven, and parallel heat paths can invalidate this simple model. Use the measurements to refine the model rather than forcing the measurements to match it.

Checks before releasing the design

Confirm that temperatures meet their individual limits, not just the processor limit. Carrier-board regulators, storage devices, connectors and power supplies may establish a lower allowable internal temperature.

Check the following before approving the thermal design:

• The module remains below its design ceiling without unintended thermal throttling.
• Every monitored component remains within its documented operating limits.
• Contact pressure and interface compression remain acceptable across manufacturing tolerances.
• External surfaces meet the project's touch-temperature requirements.
• The installed mounting clearance matches the tested arrangement.
• Cable routing and service access do not disturb the spreader or its supports.
• Reassembly produces repeatable temperatures without damaging the thermal interface.

Use the sealed-enclosure thermal soak test to develop the broader acceptance record. Include ambient temperature, workload, sensor locations, stabilization criterion and temperature margins.

For outdoor service, add solar loading rather than treating shaded ambient temperature as the complete boundary condition. The outdoor enclosure temperature in sun article addresses that additional load.

Mistakes that cost cooling or sealing

A loose internal heatsink may reduce a local hotspot while leaving enclosure heat rejection unchanged. It is not equivalent to a measured conductive connection through the wall.

An internal fan does not inherently breach an ingress rating, but it cannot eliminate the wall-to-ambient heat rejection requirement. For a fanless design, validation must use natural convection and conduction throughout the intended assembly.

Through-wall heatsink fasteners, sensor leads and cooling assemblies can create leakage paths. Specify their sealing arrangement rather than adding unsealed holes after fabrication. For our enclosures, ratings are manufacturer ratings for the specified design; the enclosures are not UL listed.

Avoid using thermal compound as an enclosure seal unless the complete joint is specifically designed for that purpose. Thermal contact, environmental sealing and electrical bonding are separate requirements, even when one assembly must satisfy all three.

What to put on the enclosure drawing

Send the module mounting pattern, heat-transfer footprint and complete tolerance stack with the enclosure requirements. Identify the required flat contact area, spreader support points and any external heatsink attachment locations. State whether cooling hardware needs removal during normal servicing.

Include maximum ambient temperature, module heat, other internal heat and the selected temperature measurement limit. Mark installation orientation, required external clearance, sealing details and the applicable enclosure type or ingress protection requirement.

We build custom enclosures with customer-defined material, width, height, depth and cutouts, from one piece. Define the thermal contact features before approving the drawing so fabrication and thermal design use the same geometry.

Once usable dimensions and environmental requirements are established, the enclosure selector can help narrow the enclosure choice. A dimensional match does not establish thermal suitability; release the design only after its measured heat path meets the temperature budget.

Enclosures for this job

Catalog sizes that suit the job this article answers, picked from its subject; open one for its full specifications.

Questions