Custom enclosure design
The enclosure is part of the circuit. Most EMC failures and most field-thermal problems are mechanical decisions.
Most electronics projects treat the enclosure as packaging — something decided after the board works. That sequence is why so many of them fail their first EMC booking and run hotter in the field than they did on the bench.
The enclosure is part of the circuit. It is the return path for chassis-referenced signals, the shield around every emitter, the aperture that lets energy out, and the heat sink for everything that cannot cool itself.
Decided together, or decided badly
The board and the box constrain each other, and the constraints run in both directions:
| Enclosure decision | What it does to the board |
|---|---|
| Connector positions and orientation | Fixes edge assignments and forces routing before layout starts |
| Internal height budget | Caps component height, rules out through-hole, may force a second board |
| Mounting pattern and material | Sets keep-outs, and decides whether mounting holes are grounded |
| Cable entry points | Determines where filtering has to sit and where the shield lands |
| Aperture sizes | Sets the highest frequency the enclosure can contain |
| Wall thickness and material | Decides whether the enclosure is a heat path or an insulator |
Resolve these on drawings, together, before layout. Resolve them afterwards and one of the two has to be redone.
EMC-aware enclosures
An enclosure either contains energy or radiates it, and that is decided by three things.
Shield continuity. A shield with a gap is not a shield above the frequency whose half-wavelength matches the gap. Gasket paths have to be continuous, and the mating surfaces have to be conductive where the gasket lands — which rules out painted or anodised faces unless they are masked.
Aperture control. Every vent, display cut-out and connector opening is an antenna at some frequency. Slot length matters more than area: a long thin slot leaks far more than a round hole of the same opening.
Chassis grounding. Where cable shields terminate to chassis, and how — 360 degrees at the connector body, or through a pigtail that turns the shield into an inductor. This is the single most common cause of a cable failing a radiated emissions scan, and it is settled by connector and enclosure choice.
All three tie directly to our EMC design and pre-compliance work, which is where the shielding strategy usually originates.
Thermal, as a mechanical problem
A sealed enclosure has no convection path. A processor rated for 25 °C ambient inside a closed box in direct sun is not running at 25 °C, and the datasheet will not tell you what it is running at.
Conduction to the case, spreader design, the interface material and its long-term behavior, and where the hot package sits relative to the wall — these are the levers. They are mechanical, and they are cheaper than the alternative, which is finding out at qualification that the design needs a fan it has no room for.
What is in-house and what is not
Being precise about this, because it matters when you are choosing a supplier.
In-house: design for every process below, plus additive manufacturing in engineering polymers — functional parts, not visual mockups. Mechanical fits, brackets and enclosure iterations turned around in days on our own machines.
Through qualified manufacturing partners: CNC machining, sheet metal, injection molding. We produce the design, the DFM and the drawings; they produce the parts. We do not own a machine shop and will not imply otherwise.
Everything we produce transfers to you in native source, under EU jurisdiction — the mechanical files as completely as the electrical ones.
Common questions
Straight answers
Do you manufacture the enclosure?
We manufacture functional additive parts in engineering polymers in-house, which covers short-run enclosures, brackets and fit iterations. CNC machining, sheet metal and injection molding run through qualified manufacturing partners — we do the design and the DFM for those processes, and we say plainly which is which rather than implying a machine shop we do not have.
Why involve mechanical work in an electronics project at all?
Because the enclosure decides outcomes the schematic cannot. Where a cable enters, how a shield lands on chassis, how large an aperture is, whether the hot package has a conduction path to the case — these are mechanical decisions that show up as EMC failures and field thermal problems. Made deliberately and early they cost nothing. Discovered at a chamber booking they cost a respin and a redesign.
Can you work with our mechanical engineer rather than replace them?
That is often the better arrangement. We bring the electronics constraints to their CAD — connector positions, keep-outs, height limits, thermal interfaces, shielding requirements — early enough that they are inputs rather than late corrections. What we are protecting against is the sequence where the board is finished, the enclosure is finished, and neither fits the other.
What ingress ratings can you design to?
We design sealing concepts against the target rating — gasket selection and groove geometry, connector sealing, pressure equalisation where thermal cycling would otherwise pump moisture in. Formal ingress testing is done by an accredited house; our job is that it passes first time.
How early should the enclosure be decided?
Before layout. Connector positions, board outline, mounting hole pattern and maximum component height are enclosure decisions that become layout constraints. Settling them on drawings first is the difference between a routine layout and one fought against mechanics that were fixed while nobody was looking.
Evidence
Where this has been applied
Flight-controller carrier for a tube-launched UAV
Teardown, clean-room re-implementation, and a rigid-flex carrier that fits inside a launch tube.
mmWave presence detection for building automation
Radar instead of PIR, because PIR cannot see a person sitting still.