CE vs FCC: EMC comparison for electronics
Guide · Cross-cutting comparison
Designing the same product for both the European Union and the United States is not just doubling the test campaign. The two EMC regimes share a common CISPR heritage, yet they diverge on points that turn expensive when discovered late: antenna distance, detector, mandatory immunity or not, class rules. This guide is written for the product engineer who has to build a single test plan calibrated to clear both CE and FCC, and for the project manager who has to budget it. You will find the real gaps, the limit cross-reference table, and the practical sequence that minimises retests.
Regulatory frames: two philosophies, one objective
Section titled “Regulatory frames: two philosophies, one objective”Before comparing numbers, the logic of each regime must be clear. The EU and the US share the same goal, prevent equipment from disrupting each other, but with opposite legal architectures.
EU side: framework directive plus harmonised standards
Section titled “EU side: framework directive plus harmonised standards”Directive 2014/30/EU sets generic essential requirements ("equipment shall not generate electromagnetic disturbances such that..."). It contains no numeric limits. Harmonised standards (published in the OJEU) translate those requirements into measurable thresholds. See CE harmonised standards for the full mechanism.
For a non-radio product, Directive 2014/30/EU applies directly. For a radio product (Wi-Fi, BLE, cellular), EMC is absorbed into Article 3.1(b) of RED Directive 2014/53/EU, which covers both generic EMC and radio-specific phenomena through the EN 301 489 series. See also RED standards.
US side: rules written directly into 47 CFR
Section titled “US side: rules written directly into 47 CFR”The FCC does not delegate to standards bodies: the technical requirements are written explicitly into 47 CFR Part 15 Subpart B for unintentional radiators (digital devices), and into sections 15.205, 15.247, 15.249, 15.407 for intentional radiators. Cellular bands fall under Parts 22, 24 and 27.
On the methodology side, the FCC references ANSI C63.4 for digital devices and ANSI C63.10 for unlicensed radios. Both are published by IEEE and stand as the US counterpart of CISPR/EN.
Practical consequence
Section titled “Practical consequence”| Aspect | EU | US |
|---|---|---|
| Source of limits | Harmonised standard (EN) | FCC rule (47 CFR) |
| Limit evolution | Standard update + OJEU publication | CFR amendment by the FCC (Notice of Proposed Rulemaking) |
| Presumption of conformity | Yes, through harmonised standards | No presumption; direct compliance with rules |
| Cited test method | EN 55032 / EN 55035 | ANSI C63.4 / ANSI C63.10 |
| Immunity required | Yes (EN 61000-4-x series, EN 55035) | No for Part 15 Subpart B |
Measurement methods: where divergence starts
Section titled “Measurement methods: where divergence starts”Both regimes measure the same phenomena, but with protocols different enough that a report does not transpose without correction.
Antenna distance: 10 m for CISPR, 3 m for the FCC
Section titled “Antenna distance: 10 m for CISPR, 3 m for the FCC”The historical CISPR reference for radiated emissions below 1 GHz is 10 m. The FCC references 3 m across nearly all of Part 15. This split is not cosmetic: converting between the two distances follows the inverse-distance law in free space, but that law only applies strictly in the far field and in a perfectly anechoic chamber.
Theoretical conversion: E(3 m) = E(10 m) × (10/3) = E(10 m) + 10.5 dB. In practice, inside a semi-anechoic chamber with a reflecting ground plane, the real factor varies with frequency and antenna height. A professional measurement report uses the chamber's site attenuation curves, never a flat multiplication.
CISPR also allows measurement at 3 m for small equipment (dimension < 1.2 m), with a limit correction. The FCC allows 10 m measurements under waivers (Section 15.31(f)(1)). But by default, each side expects its native distance, and the test plan must say so.
Detectors: peak, quasi-peak, average
Section titled “Detectors: peak, quasi-peak, average”Detectors are not equivalent either.
| Detector | CISPR / EN 55032 | FCC Part 15 |
|---|---|---|
| Peak | Used in pre-compliance, rarely as a final limit | Direct limit on conducted emissions above 5 MHz and at certain radiated points |
| Quasi-peak (QP) | Primary limit on conducted (150 kHz – 30 MHz) and radiated (30 MHz – 1 GHz) | Primary limit on conducted emissions (150 kHz – 30 MHz) |
| Average (AVG) | Complementary limit on conducted | Complementary limit on conducted |
Above 1 GHz, both the FCC and CISPR use peak and average detectors, but with different reference bandwidths (1 MHz on the FCC side for peak, 1 MHz on the CISPR side for average, the subtlety lies in the actual values applied). A pulsed signal that clears CISPR can fail under FCC simply because the peak detector captures the full crest energy, where the QP integrates over time.
Consequence on instrumentation
Section titled “Consequence on instrumentation”A modern EMI receiver (Rohde & Schwarz ESW, Keysight N9038A, etc.) handles both detector families simultaneously and can produce limits from both regimes off a single sweep. But the antenna, positioning and ground plane must match the cited method. A chamber certified for both FCC and CISPR (NSA validation, Normalized Site Attenuation, in both configurations) is the baseline of a dual test.
Emission limits side by side
Section titled “Emission limits side by side”To anchor orders of magnitude, here are Class B / residential limits compared.
Conducted emissions (150 kHz – 30 MHz)
Section titled “Conducted emissions (150 kHz – 30 MHz)”| Band | FCC Class B (15.107) | CISPR 32 Class B (EN 55032) |
|---|---|---|
| 150 kHz – 500 kHz | 66 → 56 dB(µV) QP, 56 → 46 dB(µV) AVG (log slope) | 66 → 56 dB(µV) QP, 56 → 46 dB(µV) AVG (log slope) |
| 500 kHz – 5 MHz | 56 dB(µV) QP, 46 dB(µV) AVG | 56 dB(µV) QP, 46 dB(µV) AVG |
| 5 MHz – 30 MHz | 60 dB(µV) QP, 50 dB(µV) AVG | 60 dB(µV) QP, 50 dB(µV) AVG |
This is where convergence is strongest: FCC Class B conducted limits are aligned with CISPR 22/32 Class B, by design. A product that clears one usually clears the other, provided measurement uses a LISN compliant with both (50 Ω / 50 µH, common-mode and differential-mode).
Radiated emissions (30 MHz – 1 GHz)
Section titled “Radiated emissions (30 MHz – 1 GHz)”| Band | FCC Class B at 3 m (15.109) | CISPR 32 Class B at 10 m | CISPR scaled to 3 m (calc.) |
|---|---|---|---|
| 30 – 88 MHz | 40 dB(µV/m) | 30 dB(µV/m) | 40.5 dB(µV/m) |
| 88 – 216 MHz | 43.5 dB(µV/m) | 30 dB(µV/m) (up to 230) | 40.5 dB(µV/m) |
| 216 – 230 MHz | 46 dB(µV/m) | 30 dB(µV/m) | 40.5 dB(µV/m) |
| 230 – 960 MHz | 46 dB(µV/m) | 37 dB(µV/m) | 47.5 dB(µV/m) |
| > 960 MHz | 54 dB(µV/m) | 37 dB(µV/m) (up to 1 GHz) | 47.5 dB(µV/m) |
Reading: between 30 and 230 MHz, the FCC is more permissive (40 to 46 dB(µV/m) against the CISPR-equivalent 40.5 dB(µV/m)). Above 230 MHz, CISPR becomes the tighter regime (37 dB(µV/m) at 10 m scales to 47.5 dB(µV/m) at 3 m, i.e. 1.5 dB above the FCC limit of 46 dB(µV/m)).
Design implication: for a product targeting both, treat CISPR as the lower envelope above 230 MHz and the FCC as the reference only between 30 and 230 MHz. A 6 dB design margin on the tighter envelope gives a strong first-pass probability.
Above 1 GHz
Section titled “Above 1 GHz”For equipment whose main clock exceeds 108 MHz, tests extend up to 5 or 6 GHz on the CISPR side, and up to 40 GHz on the FCC side in the worst case (very fast clocks). Above 1 GHz, limits are expressed in peak and average, with values close but not identical between Part 15.109 and EN 55032. See CE tests for the details.
Class A vs Class B: usage rules
Section titled “Class A vs Class B: usage rules”Both regimes define a "residential" class (B) that is stricter and a "commercial/industrial" class (A) that is more permissive, typically a 10 dB gap. But the rules for choosing one over the other diverge.
FCC side
Section titled “FCC side”The FCC allows Class A for equipment "marketed for use in a commercial, industrial or business environment" (Section 15.3(h)). The user must receive an explicit warning in the manual, with wording dictated by Section 15.105(a):
"This equipment has been tested and found to comply with the limits for a Class A digital device... operation of this equipment in a residential area is likely to cause harmful interference in which case the user will be required to correct the interference at his own expense."
The warning shifts responsibility to the end user. It is legally clear, and the FCC rarely pursues a manufacturer who has respected the formalism.
EU side
Section titled “EU side”The EU applies the opposite presumption: a product likely to be used in a residential setting must meet Class B. Class A is allowed for equipment clearly aimed at industry, documented as such in the manual, and explicitly excluded from consumer placement on the market. The distinction rests on the analysis of the use environment (CISPR 32 clause 4), not on a formal user warning.
In practice, declaring Class A on an IoT product that everyone knows could end up in an office or home is a flag for market surveillance recalls. Default to Class B for the EU, unless backed by a solid technical justification. See CE pitfalls for real-world cases.
Immunity: the big asymmetry
Section titled “Immunity: the big asymmetry”This is the point that most surprises teams coming from a US-only history.
What the FCC does not require
Section titled “What the FCC does not require”For Part 15 Subpart B digital devices, the FCC requires no immunity testing. No ESD, no surge, no radiated RF, no voltage dips. The American logic is that the equipment being disturbed must self-protect, and that responsibility for coexistence rests mainly on emission control.
What the EU requires
Section titled “What the EU requires”The EU mandates a full immunity battery, codified by EN 55035 for multimedia equipment and by the generic standards EN 61000-6-1 and EN 61000-6-2, declined through the EN 61000-4-x series:
| Phenomenon | Standard | Typical level (residential) |
|---|---|---|
| Electrostatic discharge | EN 61000-4-2 | ± 8 kV contact / ± 15 kV air |
| Radiated RF field | EN 61000-4-3 | 3 V/m from 80 MHz to 6 GHz |
| Fast transients (EFT) | EN 61000-4-4 | ± 2 kV mains, ± 1 kV I/O |
| Surge | EN 61000-4-5 | ± 2 kV differential, ± 4 kV common |
| Conducted RF | EN 61000-4-6 | 3 V (150 kHz – 80 MHz) |
| Voltage dips and interruptions | EN 61000-4-11 | 0 %, 40 %, 70 % at various durations |
Design consequence
Section titled “Design consequence”A product designed for the FCC only has no calibrated ESD protection, no varistors on the mains input, possibly no common-mode filtering on I/O. Bringing it to the EU exposes these weaknesses at the lab, often via a microcontroller reset at 4 kV ESD, or a comms loss at 3 V/m near Wi-Fi bands. Retro-fitting downstream costs more than building in protection on the first board spin.
Radio equipment: a special case
Section titled “Radio equipment: a special case”For products with intentional emitters (Wi-Fi, BLE, cellular), EMC is treated differently on each side.
EU: RED Article 3.1(b)
Section titled “EU: RED Article 3.1(b)”The EN 301 489 series covers radio EMC under Article 3.1(b) of the RED Directive. It builds on EN 55032 + EN 55035 testing, but adds:
- Emission tests with modulated carrier (the radio transmits during EMC tests),
- Immunity tests with performance criteria specific to radio (packet loss, BER, receive sensitivity),
- Out-of-band tests specific to the technology (e.g. EN 301 489-17 for Wi-Fi/Bluetooth, EN 301 489-52 for cellular).
US: Part 15 Subpart B + KDB
Section titled “US: Part 15 Subpart B + KDB”The FCC treats them separately:
- The unintentional emitter (internal clocks, power supply, digital logic) → Part 15 Subpart B + ANSI C63.4,
- The intentional radiator (the radio itself) → Part 15 Subpart C + ANSI C63.10, with conducted power, EIRP, occupied bandwidth, spectral mask, and spurious emissions measurements.
KDB publications clarify the application for non-standard cases (certified modules, multiple antennas, MIMO, beamforming). See FCC Parts.
Consequence for the test plan
Section titled “Consequence for the test plan”For a dual-market Wi-Fi/BLE product, the test plan combines:
- Unintentional EMC: EN 55032 + ANSI C63.4 in the same lab session,
- Intentional EMC: EN 301 489-17 + Part 15.247 tests (ANSI C63.10),
- Spectrum: EN 300 328 on the EU side + Part 15.247 on the US side.
An experienced team structures the report so every section serves both dossiers, with no duplicated tests or repeat lab visits.
Report reuse through MRA agreements
Section titled “Report reuse through MRA agreements”The EU–US Mutual Recognition Agreement (MRA) signed in 1998 lets designated US labs test for the EU, and conversely lets EU labs accredited to ISO/IEC 17025 and designated under the FCC-MRA produce reports accepted by the FCC.
For the FCC
Section titled “For the FCC”An EU lab must be accredited by an ILAC signatory body (in France, COFRAC) and designated by the FCC for the relevant Parts. The list of designated labs is public. A report from such a lab is usable directly by a US TCB for the FCC grant.
For the EU
Section titled “For the EU”A US lab accredited by A2LA and designated by the Designating Authority (NIST) can produce reports acceptable under Directive 2014/30/EU and the RED. The EU technical file then cites that report without retesting.
Typical savings
Section titled “Typical savings”Running the campaign in a dual ILAC + FCC-MRA lab avoids:
- Physical shipment of the product between two labs,
- Duplicated test setups (same cables, same peripherals, same software configurations),
- Repeated common tests (LISN conducted, certain radiated bands).
Observed savings on full Wi-Fi/BLE campaigns are around 25 to 30 % on the lab invoice, and more on the schedule. Asking the lab upfront for its FCC designations at quotation time is the first check to make.
Design strategy: where to start?
Section titled “Design strategy: where to start?”Three campaign orders are possible. Experience shows they are not equivalent.
Option A: EU first (recommended for most IoT products)
Section titled “Option A: EU first (recommended for most IoT products)”- Design to Class B + full EU immunity,
- Run the complete CE campaign (EMC + RED if radio),
- Top up with FCC-specific tests (Part 15.247 or 15.407 for radio, ANSI C63.4 for anything not yet covered),
- Submit to the US TCB.
Typical total cost on a Wi-Fi/BLE module: ~30–40 k€. Lead time: 8 to 10 weeks.
Option B: US first
Section titled “Option B: US first”- Design to FCC Class B + radio tests,
- Get the FCC grant,
- Re-engage for the EU: add immunity + RED 3.1(b) tests,
- Often: re-test emissions because the ESD protection changes (varistors, TVS, ferrite beads) shift the EMC footprint.
Typical total cost: ~45–55 k€. Lead time: 12 to 16 weeks.
Option C: Unified campaign in a dual lab
Section titled “Option C: Unified campaign in a dual lab”- Single test plan defined during specification,
- Robust design on the tighter envelope per phenomenon,
- Single campaign in an ILAC + FCC-MRA designated lab,
- Parallel submission of EU (DoC) and US (TCB grant) dossiers.
Typical total cost: ~25–32 k€. Lead time: 6 to 8 weeks. This is the most efficient option as soon as both markets are planned together.
Common pitfalls in dual EMC certification
Section titled “Common pitfalls in dual EMC certification”Five recurring mistakes show up in EU + US dossiers:
- Converting units by hand-wave. dB(µV/m) at 3 m ≠ dB(µV/m) at 10 m without the site-specific conversion. A homemade spreadsheet is not enough; the report must be recomputed by the EMI receiver or the chamber.
- Forgetting immunity when coming from the US market. Engineering hands off a "FCC-compliant" prototype, and the first week of EU testing surfaces ESD upsets at 4 kV. Board redesign, 6 to 10 weeks lost.
- Using the wrong detector in pre-compliance. Measuring in peak gives a false comfort margin. The QP run at the final lab brings out brutal overshoots. Pre-test in QP as early as possible.
- Picking Class A to clear more easily, then selling consumer. Market surveillance and the FCC do prosecute. A multi-SKU recall costs far more than a properly done Class B campaign.
- Underestimating radio tests. For Wi-Fi/BLE products, EN 301 489-17 + EN 300 328 + Part 15.247 are each as heavy as the generic EMC campaign. Budget at least three lab weeks for a complete product.
See also FCC pitfalls and CE pitfalls for the per-regime detail.
Key takeaways
Section titled “Key takeaways”- CE and FCC are not symmetric. The EU mandates immunity; the US does not. That is the most structural gap.
- Emission limits are close but not identical. FCC more permissive in HF, CISPR stricter above 230 MHz.
- Methods differ. 10 m vs 3 m, QP vs peak, a report does not transpose by simple arithmetic.
- A dual ILAC + FCC-MRA lab enables a unified campaign with 25–30 % savings.
- "EU-first" design is the lowest-cost path when both markets are in scope.
For implementation on the EU side, see CE tests. On the US side, see FCC tests.
Sources & references
- CISPR 32:2015+A1:2019: Multimedia equipment, emission requirements , IEC webstore.iec.ch/publication/26241
- ANSI C63.4-2014: American National Standard for methods of measurement of radio-noise emissions , IEEE / ANSI standards.ieee.org/ieee/C63.4/5536/
- 47 CFR Part 15: Radio frequency devices , FCC www.ecfr.gov/current/title-47/chapter-I/subchapter-A/part-15
- Directive 2014/30/EU on electromagnetic compatibility , EUR-Lex eur-lex.europa.eu/eli/dir/2014/30/oj