EMC Design and Certification for Embedded Products: CE, FCC, CCC Essentials and Fix-It Strategies (2026)
Key takeaway: EMC is not a step before the test house — it is a design input at kickoff. Most projects that end up with expensive rework were not badly designed; they simply did not consider EMC at all. Getting return paths, filtering and partitioning right on the schematic and PCB is far cheaper than re-spinning boards in the lab.
1. Two dimensions of EMC, clearly separated
- EMI (emissions): how loudly your device "shouts" — conducted emissions (escaping along power/signal cables) and radiated emissions (escaping into space). Over-limit means you disturb others;
- EMS (immunity): how well your device "tolerates noise" — ESD, EFT bursts, surge, radiated RF fields, conducted immunity, voltage dips. Failing means the environment can make your device hang or reboot;
- They are two sides of one coin: a board with good filtering and clean return paths usually emits less and tolerates more. Conversely, a board that just bolts on shielding to pass emission tests often has poor immunity too.
2. Certification requirements by market (overview)
| Market | Typical requirements | Notes |
|---|---|---|
| Chinese Mainland | CCC (mandatory), SRRC (radio type approval), CTA (network access) | In-scope products must have CCC; Wi-Fi/4G adds radio approvals |
| EU | CE (EMC Directive + LVD; RED when wireless is present) | Self-declaration plus technical file; some categories need a notified body |
| US | FCC Part 15 (unintentional / intentional radiators) | Verification, DoC or certification depending on product class |
| Others | Country-specific (e.g. VCCI in Japan, KC in Korea, RCM in Australia) | Confirm the target-country list before shipping to avoid duplicate testing |
Safety (e.g. IEC/UL 62368-1), energy efficiency and environmental requirements (RoHS/REACH) usually run alongside EMC — planning the test schedule together saves time.
3. Six things to get right at design time
- Power entry filtering: common-mode choke + X/Y capacitors + TVS, placed as close to the connector as possible, with clean/dirty ground partitioning so filtered power does not re-couple to noise;
- Ground planes and partitioning: a solid ground plane, digital/analog/power partitioning and a defined single- or multi-point grounding strategy. Ground is EMC's first principle — slots and detours become antennas;
- High-speed return paths: impedance-controlled, length-matched differential pairs (MIPI/HDMI/PCIe/USB), continuous return across splits, and stitching vias at layer transitions. Most radiation issues trace back to a broken return path;
- Clock and crystal placement: minimise loop area for crystals, DDR clocks and switching regulators, keep them away from connectors and board edges, and use local ground pours or shielding when needed;
- Interface protection: TVS/ESD arrays and common-mode suppression on all external interfaces (Ethernet, USB, RS485/CAN, power, buttons, LEDs) — protection works both ways, limiting ingress and egress;
- Enclosure and shielding: bond impedance of metal enclosures, slot length (slots leak more than holes), aperture arrays for cooling, and reserved room for conductive coating inside plastic housings.
4. Leave headroom for the fix-it stage
Even a good design can miss a limit in the lab. The proven approach is to reserve room for fixes at the prototype stage:
- Reserve filter component footprints on the PCB (common-mode chokes, ferrites, Y capacitors) so parts can be added without a re-spin;
- Reserve shielding/conductive-foam space and mounting points in the mechanical design before tooling is cut;
- Reserve ferrite bead locations and cable slack for clamp-on cores;
- Run a pre-scan on prototypes to see which band and which cable is over-limit — far more efficient than going straight to formal testing;
- Keep software levers such as spread-spectrum or clock trimming available (where the platform supports it) as debugging variables.
5. A smoother certification flow
- Step 1 — Confirm the list: target markets + product category + wireless or not determines which tests apply;
- Step 2 — Pre-test: run a pre-scan in your own or a third-party chamber and fix obvious issues first;
- Step 3 — Formal test: execute EMI and EMS in one pass to avoid repeated teardowns;
- Step 4 — Fix and retest: act on the pre-scan findings and retest only the failing items;
- Step 5 — Reports and certificates: archive the technical file and control consistency during mass production.
Wireless products also need RF and network-access testing — schedule it together with EMC.
6. Common pitfalls
- Myth 1: think about certification after the prototype is done. By then the enclosure, PCB and filtering are frozen — fixing them means redoing the project;
- Myth 2: aim only at "passing", ignoring long-term stability. Products that barely scrape through can still misbehave across batches and conditions; design margin matters more than the certificate;
- Myth 3: shielding solves everything. Shielding changes thermals and assembly, and can trap noise inside the cavity — evaluate holistically;
- Myth 4: change parts after mass production without retesting. Swapping a power module, connector or crystal can change EMC behaviour; key changes need assessment;
- Myth 5: treat EMC as "the test department's problem". It emerges from hardware, mechanical and software together — someone must own the system-level target from kickoff.
7. FAQ
- Q1: When should certification work start? Ideally at kickoff: fix the target markets and test list, write EMC requirements into the hardware and mechanical specs, and pre-scan at the prototype stage.
- Q2: Our SoM already has module-level certification — does the system need retesting? Yes. Cables, power, enclosure and connectors all change electromagnetic behaviour. Module-level certification is a starting point, not a substitute.
- Q3: Do we have to use a metal enclosure? Not necessarily — a plastic housing with sound board-level design and selective coating can pass too. The choice depends on frequency bands and mechanical constraints; decide after the pre-scan.
- Q4: How do we debug ESD failures? First locate the injection path (discharge point, reset line, interface ground), then work three tracks: interface protection, ground impedance, and reset/power integrity. It is usually a grounding impedance and return-path imbalance issue.
- Q5: Can certification be handled as part of a customization project? Hongyin Technology applies the target market's EMC requirements as design constraints from the schematic and carrier design stage, supports pre-scan and remediation together with the customer, and provides design files and test support. Send your target markets and product form factor for advice at review.
8. Related solutions & contact
Related solutions: Industrial Vision Inspection、Security NVR Recording、Industrial HMI & Panel PC. Further reading: RK3588 Carrier Hardware Design Review Checklist、RK3588 Power Budget & Thermal Design、Evaluating SoM Industrial Reliability.
Shenzhen Hongyin Technology Co., Ltd. (HONGYIN TECH) provides embedded project customization on RK3588/RK3576/RK3568 platforms — free requirement review, engineer-direct communication and low-volume ordering. Send your requirement list and receive selection advice and quotes within 48 hours.
phone +86 13728726926 | email sales@hoyin-tech.com | website www.hoyin-tech.com | address: Room 301, Building A, Fanrong Science Park, Sanwei Community, Hangcheng Street, Bao'an District, Shenzhen, China.
Hongyin Tech provides full-stack SoM/SBC customization on Rockchip & HiSilicon platforms with Android/Linux. Engineers respond within 1 business day.
Get a Free Quote →