Radiated emissions EMC test: CISPR 32 and FCC 15.109
A radiated emissions test measures the field strength that a product leaks into space, with an antenna at a set distance, a receiver with set detectors and bandwidths, and a test site whose behaviour is verified. On both sides of the Atlantic the texts split the job the same way: one document sets the limits (CISPR 32 in Europe, 47 CFR 15.109 in the US), others set the method and the instruments (the CISPR 16 series, ANSI C63.4). This page goes back to those texts: which edition of each CISPR 16 part is current in October 2026, the limit tables side by side at the same distance, how high the test has to go, what the uncertainty standard actually says about margin, and the common readings the texts contradict.
Which text sets what
Section titled “Which text sets what”Radiated emissions are the unintentional radio-frequency energy an equipment under test (EUT) releases into space, as opposed to conducted emissions, which travel along power and signal cables. The limit and the method come from different documents.
| Role | European side | US side (Part 15, unintentional radiators) |
|---|---|---|
| Limits, multimedia and IT equipment | CISPR 32 (EN 55032) Annex A, Tables A.2 to A.5 | 47 CFR 15.109(a) Class B, 15.109(b) Class A |
| Limits, ISM equipment | CISPR 11 (EN 55011) | 47 CFR Part 18 |
| Upper test frequency | CISPR 32 Table 1 | 47 CFR 15.33(b) |
| Measurement method | CISPR 16-2-3 | ANSI C63.4-2014, via 15.31(a)(4), excluding clauses 4.5.3, 4.6, 6.2.13, 8.2.2, 9 and 13 |
| Antennas and test sites | CISPR 16-1-4 | ANSI C63.4-2014 |
| Receiver and detectors | CISPR 16-1-1 | 47 CFR 15.35, with ANSI C63.4-2014 clause 4 for the quasi-peak instrumentation |
| Measurement uncertainty and decision rule | CISPR 16-4-2 | No equivalent rule in Part 15 |
CISPR 32 is a product family standard, not a generic one: it covers multimedia equipment (information technology, audio, video, broadcast receivers). Other families have their own emission standards, for example CISPR 14-1 (EN 55014-1) for household appliances and CISPR 15 (EN 55015) for lighting. The generic emission standards (EN IEC 61000-6-3 for residential, EN IEC 61000-6-4 for industrial environments) apply only where no product or product family standard covers the equipment. 47 CFR Part 15 Subpart B covers digital devices and other unintentional radiators, and the authorization route for Class A and Class B digital devices is Supplier's Declaration of Conformity or certification (15.101(a)).
Current editions, October 2026
Section titled “Current editions, October 2026”| Document | Current version | Date | Note |
|---|---|---|---|
| CISPR 16-2-3 | 2016 + AMD1:2019 + AMD2:2023, consolidated edition 4.2 | June 14, 2023 | Fourth edition September 15, 2016; covers 9 kHz to 18 GHz; basic EMC publication |
| CISPR 16-1-4 | Edition 5.0 | October 3, 2025 | Replaces the 2019 edition with AMD1:2020 and AMD2:2023; changes the CMAD definition and introduction (8.1.1) and adds the VHF-LISN as a cable termination device (8.2) |
| CISPR 16-1-1 | Edition 5.0 | May 22, 2019 | Receivers, 9 kHz to 18 GHz |
| CISPR 16-4-2 | 2011 + AMD1:2014 + AMD2:2018, consolidated edition 2.2 | August 15, 2018 | Values of U(CISPR) in Table 1 |
| CISPR 32 | Edition 2.0 (2015) with AMD1:2019 (edition 2.1) | March 31, 2015 / October 1, 2019 | Edition 3.0 at draft stage at IEC, not published |
| EN 55032 | EN 55032:2015 + A11:2020 | Cited under Directive 2014/30/EU | CISPR 32 AMD1:2019 is not part of the cited reference |
| EN 55022 (CISPR 22) | Presumption of conformity ended | March 5, 2017 | Replaced by EN 55032:2012 (OJ C 14, January 16, 2015) |
| ANSI C63.4 | C63.4-2014 | Referenced in 47 CFR 15.31(a)(4) | Unintentional radiators; intentional radiators use ANSI C63.10-2020 (15.31(a)(3)) |
How high the test goes
Section titled “How high the test goes”Neither regime fixes one frequency range for every product. Both tie the upper limit to the highest frequency generated or used inside the device, with the same breakpoints up to 1 GHz.
| Highest internal frequency | CISPR 32 Table 1, upper test frequency | 47 CFR 15.33(b)(1), upper test frequency |
|---|---|---|
| Below 1.705 MHz | 1 GHz | 30 MHz |
| 1.705 MHz to 108 MHz | 1 GHz | 1000 MHz |
| 108 MHz to 500 MHz | 2 GHz | 2000 MHz |
| 500 MHz to 1 GHz | 5 GHz | 5000 MHz |
| Above 1 GHz | 5 times the highest frequency, or 6 GHz, whichever is lower | 5th harmonic of the highest frequency, or 40 GHz, whichever is lower |
So a product whose highest internal frequency is 25 MHz stops at 1 GHz in both regimes, while a product with a processor clocked above 1.2 GHz is measured up to 6 GHz in Europe and beyond 6 GHz for the FCC. For a device that is also an intentional radiator, 15.33(a)(5) says the higher of the transmitter range and the digital-device range applies.
Test sites and methods in CISPR 16-2-3
Section titled “Test sites and methods in CISPR 16-2-3”CISPR 16-2-3 describes several radiated methods, each with its own frequency range. The product standard decides which ones are allowed for compliance (CISPR 32 Table A.1 lists the basic standards and the limitations on the use of particular methods).
| CISPR 16-2-3 clause | Method | Frequency range |
|---|---|---|
| 7.2 | Large loop antenna system (LLAS) | 9 kHz to 30 MHz |
| 7.3 | Open-area test site (OATS) or semi-anechoic chamber (SAC) | 30 MHz to 1 GHz |
| 7.4 | Fully anechoic room (FAR) | 30 MHz to 1 GHz |
| 7.5 | Common emission and immunity set-up in a SAC | 30 MHz to 1 GHz (immunity from 80 MHz) |
| 7.6 | FAR and absorber-lined OATS or SAC | 1 GHz to 18 GHz |
| 7.7 | In situ measurements | 9 kHz to 18 GHz |
| 7.8 | Substitution method | 30 MHz to 18 GHz |
| 7.9 | Reverberation chamber | 80 MHz to 18 GHz |
| 7.10 | TEM waveguide | 30 MHz to 18 GHz |
| 7.11 | OATS or SAC | 9 kHz to 30 MHz |
Source: CISPR 16-2-3 edition 4.2, table of contents. Amendment 1 (2019) added EUT volume specifications that depend on the method and the measurement distance (clause 7.1.2 and Tables 11 and 12). The sites themselves, and the antennas, are specified and validated under CISPR 16-1-4. From 1 to 18 GHz, clause 7.6 treats the FAR and the OATS or SAC with absorber on the ground plane together. CISPR 32 Annex I describes GTEM and reverberation chamber procedures with "proposed limits"; that annex is informative.
On an OATS or in a SAC, clause 7.3.5 covers antenna height variation, because the ground reflection makes the received field depend on height. The standard also gives horizontal polarisation correction factors (Table 6) and recommended antenna heights for the pre-scan (Table 7).
Receiver, bandwidth and detectors
Section titled “Receiver, bandwidth and detectors”CISPR 16-1-1 specifies one receiver per detector, which is why "peak", "quasi-peak" and "average" are not just display settings.
| Detector | CISPR 16-1-1 clause | Frequency range of the specification | Where the limits use it |
|---|---|---|---|
| Quasi-peak | 5 | 9 kHz to 1000 MHz | CISPR 32 and 47 CFR 15.109, 30 MHz to 1 GHz |
| Peak | 6 | 9 kHz to 18 GHz | CISPR 32 and FCC limits above 1 GHz; pre-scans |
| Average | 7 | 9 kHz to 18 GHz | CISPR 32 and FCC limits above 1 GHz |
| RMS-average | 8 | 9 kHz to 18 GHz | Not used by the CISPR 32 radiated tables |
From 30 MHz to 1 GHz the quasi-peak receiver has a 120 kHz bandwidth with 1 ms charge and 550 ms discharge time constants; above 1 GHz the CISPR bandwidth is 1 MHz. For any signal, the peak reading is never lower than the quasi-peak reading, and the quasi-peak reading never lower than the average one. CISPR 16-2-3 Annex E sets how to check that a spectrum analyser is suitable for compliance tests, and its Table E.1 gives the maximum amplitude difference between peak and quasi-peak detected signals.
On the FCC side, 15.35(a) bases the limits at or below 1000 MHz on the CISPR quasi-peak detector and lets the responsible party use a peak detector at the same bandwidth instead. 15.35(b) bases the limits above 1000 MHz on an average detector with a minimum 1 MHz resolution bandwidth, and caps the peak at 20 dB above the average limit. Above 18 GHz, the CISPR 16-1-1 receiver specification no longer applies; the FCC 1 MHz minimum bandwidth continues to the 40 GHz ceiling of 15.33(b)(1).
CISPR 32 limits (Annex A)
Section titled “CISPR 32 limits (Annex A)”CISPR 32 Class B is for equipment intended primarily for the residential environment; Class A covers the rest. Below 1 GHz the tables give OATS or SAC limits at 10 m and at 3 m (the 3 m values are 10 dB higher), plus separate FAR rows.
30 MHz to 1 GHz, OATS or SAC, quasi-peak
Section titled “30 MHz to 1 GHz, OATS or SAC, quasi-peak”| Band | Class A at 10 m | Class A at 3 m | Class B at 10 m | Class B at 3 m |
|---|---|---|---|---|
| 30 to 230 MHz | 40 dB(μV/m) | 50 dB(μV/m) | 30 dB(μV/m) | 40 dB(μV/m) |
| 230 to 1000 MHz | 47 dB(μV/m) | 57 dB(μV/m) | 37 dB(μV/m) | 47 dB(μV/m) |
Source: CISPR 32:2015 Tables A.2 (Class A) and A.4 (Class B). At the 230 MHz transition the lower limit applies.
1 GHz to 6 GHz, free-space conditions at 3 m
Section titled “1 GHz to 6 GHz, free-space conditions at 3 m”| Band | Class A average / peak | Class B average / peak |
|---|---|---|
| 1 to 3 GHz | 56 / 76 dB(μV/m) | 50 / 70 dB(μV/m) |
| 3 to 6 GHz | 60 / 80 dB(μV/m) | 54 / 74 dB(μV/m) |
Source: CISPR 32:2015 Tables A.3 (Class A) and A.5 (Class B). The product must meet both the average and the peak limit.
FCC limits (47 CFR 15.109)
Section titled “FCC limits (47 CFR 15.109)”The rule gives limits in microvolts per metre: Class B at 3 m in 15.109(a), Class A at 10 m in 15.109(b). The dB(μV/m) column is 20 log of the rule value. Under 15.109(c) the tighter limit applies at the band edges.
| Band | Class B at 3 m | Class A at 10 m |
|---|---|---|
| 30 to 88 MHz | 100 μV/m (40.0 dB(μV/m)) | 90 μV/m (39.1 dB(μV/m)) |
| 88 to 216 MHz | 150 μV/m (43.5 dB(μV/m)) | 150 μV/m (43.5 dB(μV/m)) |
| 216 to 960 MHz | 200 μV/m (46.0 dB(μV/m)) | 210 μV/m (46.4 dB(μV/m)) |
| Above 960 MHz | 500 μV/m (54.0 dB(μV/m)) | 300 μV/m (49.5 dB(μV/m)) |
Below 1000 MHz these are quasi-peak limits (15.35(a)). Above 1000 MHz the same "above 960 MHz" value is an average limit with a peak cap 20 dB higher (15.35(b)): 54 dB(μV/m) average and 74 dB(μV/m) peak at 3 m for Class B.
15.109(g) offers an alternative for digital devices: the limits of the third edition of CISPR 22. Two conditions come with it: the Part 15 test procedure still applies (Note 1 to 15.31(a)(4) points to ANSI C63.4-2014), and above 1000 MHz compliance is still shown against the 15.109(a) or (b) limits, which may be extrapolated to the CISPR 22 distance (the rule gives 150 μV/m at 10 m for Class B as the example).
Both regimes at the same distance
Section titled “Both regimes at the same distance”To compare, bring the FCC Class B limits from 3 m to 10 m with the inverse linear distance rule that 15.31(f)(1) prescribes at or above 30 MHz (20 dB per decade, so 20 log(10/3) = 10.5 dB).
| Band | FCC Class B scaled to 10 m | CISPR 32 Class B at 10 m | Tighter limit |
|---|---|---|---|
| 30 to 88 MHz | 29.5 dB(μV/m) | 30 dB(μV/m) | FCC, by 0.5 dB |
| 88 to 216 MHz | 33.0 dB(μV/m) | 30 dB(μV/m) | CISPR, by 3.0 dB |
| 216 to 230 MHz | 35.5 dB(μV/m) | 30 dB(μV/m) | CISPR, by 5.5 dB |
| 230 to 960 MHz | 35.5 dB(μV/m) | 37 dB(μV/m) | FCC, by 1.5 dB |
| 960 to 1000 MHz | 43.5 dB(μV/m) | 37 dB(μV/m) | CISPR, by 6.5 dB |
| 1 to 3 GHz (3 m, average) | 54 dB(μV/m) | 50 dB(μV/m) | CISPR, by 4 dB |
| 3 to 6 GHz (3 m, average) | 54 dB(μV/m) | 54 dB(μV/m) | Equal |
The Class A comparison at 10 m follows the same pattern: FCC tighter from 30 to 88 MHz (39.1 against 40) and from 230 to 960 MHz (46.4 against 47), CISPR tighter from 88 to 230 MHz and from 960 MHz to 3 GHz, equal from 3 to 6 GHz. Neither regime is uniformly stricter, so a product sold in both markets has to clear the lower of the two limits band by band. Note that CISPR 32 steps by 10 dB between its own 3 m and 10 m rows, while the FCC rule extrapolates with 10.5 dB; the difference matters only when a result sits within a fraction of a decibel of the limit. See CE vs FCC, EMC comparison for the wider picture.
The measurement sequence
Section titled “The measurement sequence”A full quasi-peak sweep is slow: CISPR 16-2-3 clause 6.6 sets minimum measurement times and scan times per CISPR band for peak and quasi-peak detectors (Table 2), and the quasi-peak values are far longer than the peak ones. Clause 8, "Automated measurement of emissions", therefore describes a staged sequence.
| Step | CISPR 16-2-3 clause | What happens |
|---|---|---|
| Precautions | 8.1 | Limits of automation, operator checks |
| Generic procedure | 8.2 | Overall sequence |
| Pre-scan | 8.3 | Fast overview, usually with the peak detector; 8.3.2 covers the measurement time needed, 8.3.3 the requirements per measurement type |
| Data reduction | 8.4 | Selecting the frequencies that need a final measurement |
| Maximization and final measurement | 8.5 | Searching the worst case (antenna height, polarisation, EUT azimuth) and measuring with the detector the limit specifies |
| Post-processing and reporting | 8.6 | Corrections, comparison with the limit, report |
Strategies for obtaining a spectrum overview with the peak detector are in clause 6.6.5, and timing considerations for FFT-based instruments in 6.6.6 and 8.7. Because the peak reading is never lower than the quasi-peak reading, a frequency whose peak value is already below the quasi-peak limit cannot fail the quasi-peak limit; how far below the limit a frequency must sit to be dropped at the data reduction step is a choice the lab procedure has to state.
EUT arrangement
Section titled “EUT arrangement”CISPR 16-2-3 clause 6.4 covers the EUT arrangement and measurement conditions, including operation of the EUT, its mode of operation, multifunction equipment, and "determination of arrangement(s) causing maximum emissions" (6.4.8). For multimedia equipment, CISPR 32 Annex D gives the measurement arrangements (Table D.1) and the spacings, distances and tolerances (Table D.2), with example arrangements for table-top and floor-standing equipment; Annex B covers how to exercise displays, video and other ports. In the US, ANSI C63.4-2014 fixes the arrangement for Part 15 tests. Record the cable layout, the peripherals and the operating mode: a result that cannot be repeated cannot be defended.
Uncertainty and margin: what CISPR 16-4-2 says
Section titled “Uncertainty and margin: what CISPR 16-4-2 says”CISPR 16-4-2 gives a reference value U(CISPR) for each measurement type (Table 1 of the 2011 edition; Amendment 2 of 2018 added the LLAS row and did not change the other radiated rows).
| Radiated measurement | Frequency range | U(CISPR) |
|---|---|---|
| Disturbance current in an LLAS (added by AMD2:2018) | 9 kHz to 30 MHz | 3.3 dB |
| Electric field strength at an OATS or in a SAC | 30 MHz to 1000 MHz | 6.3 dB |
| Electric field strength in a FAR | 30 MHz to 1000 MHz | 5.3 dB |
| Electric field strength in a FAR | 1 GHz to 6 GHz | 5.2 dB |
| Electric field strength in a FAR | 6 GHz to 18 GHz | 5.5 dB |
The values are expanded uncertainties with a coverage factor of 2. Below 1 GHz they were calculated for the quasi-peak detector, above 1 GHz for the peak detector, and Amendment 2 adds that the OATS, SAC and FAR values assume a small EUT (fitting in a cylinder 1.5 m in diameter and 1.5 m high) at 3 m.
The decision rule (clause 4.2) is the part that matters for pass and fail. If the lab uncertainty U(lab) is at or below U(CISPR), compliance is decided by comparing the measured level directly with the limit. If U(lab) exceeds U(CISPR), the measured level is increased by the difference before the comparison, and the report must give U(lab). Nothing in CISPR 16-4-2, CISPR 32 or Part 15 requires a fixed margin such as 6 dB. A design team may set one to cover production spread, cable and peripheral changes, and the gap between a bench set-up and the lab arrangement; that target is an engineering choice, and it should be written as such in the test plan.
Pre-compliance and compliance
Section titled “Pre-compliance and compliance”A pre-compliance measurement runs on a development bench or in a non-validated chamber, often with a spectrum analyser in peak mode and a set-up without calibrated corrections. It finds problems and compares fixes. A compliance measurement uses a site validated under CISPR 16-1-4, a receiver meeting CISPR 16-1-1 and the arrangement of the product standard. Neither the EMC Directive nor the FCC SDoC procedure requires a particular kind of lab: under Article 14 of Directive 2014/30/EU the manufacturer may use internal production control (Annex II) without any notified body, and the EU-type examination of Annex III is optional. Many manufacturers still use an ISO/IEC 17025 accredited lab for traceability and credibility.
Common pitfalls
Section titled “Common pitfalls”- Cable layout changed between pre-scan and final measurement. The candidate list from the pre-scan no longer matches the set-up that is maximized. Photograph and record the arrangement before each step.
- Only one polarisation in the final measurement. The maximization in clause 8.5 covers polarisation as well as height and azimuth; a frequency measured in one polarisation only is not maximized.
- Antenna factor out of date or misapplied. Each antenna has its own calibrated factor, per polarisation where relevant, and it enters the result directly in dB.
- Operating modes missing. CISPR 16-2-3 clause 6.4.6 (mode of operation) and 6.4.8 (arrangement causing maximum emissions) mean every mode that changes the emission spectrum has to be considered, not only the idle state.
- A spectrum analyser at a default 100 kHz resolution bandwidth. The CISPR bandwidth from 30 MHz to 1 GHz is 120 kHz; whether an analyser is suitable for compliance is decided by CISPR 16-2-3 Annex E.
- Upper frequency taken from habit. A product with a clock above 108 MHz is not finished at 1 GHz (CISPR 32 Table 1, 15.33(b)(1)).
See also CE pitfalls and FCC pitfalls.
When a frequency fails
Section titled “When a frequency fails”The usual levers, from quickest to slowest to implement:
- Set-up and accessories: cable routing, separation of power and signal cables, clip-on common-mode ferrites selected for the failing frequency, conductive gaskets on enclosure seams. These need no board change, but any accessory that fixes the problem becomes part of the product and must ship with it.
- Shielding and filtering: a local shield over the emitting circuit (oscillator, switching converter), filters on cable entries, better cable screens, smaller enclosure apertures.
- PCB revision: shorter current loops for clocks and fast edges, a continuous reference plane under critical signals, local decoupling at each supply pin, series termination of clock outputs, spread-spectrum clocking where the interface allows it.
The PCB design for EMC guide covers the board-level levers.
Claims the texts contradict
Section titled “Claims the texts contradict”| Claim | What the texts say |
|---|---|
| "CISPR 16-2-3 does not cover below 30 MHz" (this page said so before October 2026) | CISPR 16-2-3 covers 9 kHz to 18 GHz, with LLAS (7.2) and OATS or SAC (7.11) methods from 9 kHz to 30 MHz. The 30 MHz start comes from the CISPR 32 and 15.109 limit tables. |
| "The pre-scan and final scan strategy is in clause 6 of CISPR 16-2-3" (this page) | Clause 6 holds the general requirements, including measurement times (6.6). The automated sequence (pre-scan, data reduction, maximization, final measurement) is clause 8. |
| "CISPR 16-4-2 quotes up to 5.2 dB for radiated emissions from 30 MHz to 1 GHz" (this page) | U(CISPR) is 6.3 dB at an OATS or in a SAC and 5.3 dB in a FAR from 30 to 1000 MHz. 5.2 dB is the FAR value from 1 to 6 GHz. |
| "A 6 dB margin is the standard target" | No CISPR or FCC text requires a margin. CISPR 16-4-2 clause 4.2 compares the measured value directly with the limit when U(lab) is at or below U(CISPR). |
| "Full compliance needs a lab recognised by a notified body for CE or a TCB for FCC, and its report is legally binding" (this page) | The EMC Directive allows internal production control (Annex II) with no notified body. Part 15 Class A and Class B digital devices may use SDoC (15.101(a)), which involves no TCB. A test report is evidence in the technical file, not a legal act. |
| "CISPR 32 and CISPR 11 are the generic standards" | They are product family standards. The generic emission standards are IEC 61000-6-3 and IEC 61000-6-4, used only where no product standard applies. |
| "CISPR RBW is 1 MHz up to 40 GHz" (this page) | CISPR 16-1-1 specifies receivers up to 18 GHz. Above that, the FCC 1 MHz minimum of 15.35(b) applies to Part 15 measurements. |
| "15.109(g) is the rule for measurements above 1 GHz" | 15.109(g) is the option to use the CISPR 22 limits; 15.109(g)(2) sends measurements above 1000 MHz back to the (a) and (b) limits. The detector rule above 1 GHz is 15.35(b). |
| "CISPR 32 replaced CISPR 22 in 2017" | In the EU, EN 55032:2012 replaced EN 55022 as the cited standard on March 5, 2017 (OJ C 14, January 16, 2015). CISPR 32 itself was first published in January 2012. |
See also
Section titled “See also”- Conducted emissions: LISN, CISPR 32 and FCC 15.107
- FCC standards and measurement procedures
- What is EMC? Emissions, immunity and coupling
- Radio: RX blocking, selectivity and intermodulation tests
- HEMP and IEMI: IEC 61000-4-25 and hardened electronics
- SAR procedures: absorption rate (IEC 62209, EN 50360)
- IEC 61000-4-3: radiated RF field immunity
- IEC 61000-4-6: conducted RF immunity
- Pre-compliance EMC: TEM cell, near-field probes, LISN
- EMC chamber types: SAC, FAR, OATS, GTEM, reverberation
- Calibration and measurement uncertainty (GUM, CISPR)
- PCB design for EMC: return paths, decoupling, stackup
Where the test sits in an EMC campaign
Section titled “Where the test sits in an EMC campaign”| Family | Test | Standard | Reference |
|---|---|---|---|
| Emissions | Radiated emissions | CISPR 32 / 47 CFR 15.109 | This guide |
| Emissions | Conducted emissions on a LISN | CISPR 32 / 47 CFR 15.107 | Conducted emissions guide |
| Immunity | Electrostatic discharge (ESD) | EN 61000-4-2 | See CE tests |
| Immunity | Radiated RF field | EN 61000-4-3 | See CE tests |
| Immunity | Fast transients (EFT) | EN 61000-4-4 | See CE tests |
| Immunity | Surge | EN 61000-4-5 | See CE tests |
| Immunity | Conducted RF | EN 61000-4-6 | See CE tests |
See RED tests for radio products and FCC tests for the US sequence. The glossary covers the acronyms (ANSI, QP, SAC, EUT, dwell time).
Key takeaways
Section titled “Key takeaways”- Limits and method live in different texts: CISPR 32 or 15.109 for the numbers, CISPR 16-2-3, 16-1-4 and 16-1-1 or ANSI C63.4-2014 for how to measure.
- The upper frequency follows the product: 1 GHz up to a 108 MHz internal frequency, then up to 6 GHz under CISPR 32 and up to 40 GHz under 15.33(b)(1).
- Neither regime is uniformly tighter: at 10 m, the FCC Class B limit is lower from 30 to 88 MHz and from 230 to 960 MHz, CISPR 32 Class B elsewhere up to 3 GHz.
- CISPR 16-4-2 sets a decision rule, not a margin: U(CISPR) is 6.3 dB at an OATS or SAC from 30 to 1000 MHz; a lab above that adds the excess to its readings.
- The staged sequence is CISPR 16-2-3 clause 8: pre-scan, data reduction, maximization and final measurement, reporting.
- Current editions: CISPR 16-1-4 edition 5.0 (October 3, 2025), CISPR 16-2-3 edition 4.2 (June 14, 2023), EN 55032:2015 + A11:2020 cited in the EU.
For the EU side, see CE tests; for the US side, FCC tests; for radio products, RED tests.
Sources & references
- CISPR 32:2015 (edition 2.0), Electromagnetic compatibility of multimedia equipment, emission requirements, IEC webstore.iec.ch/en/publication/22046
- CISPR 32:2015 RLV preview (table of contents, Table 1 and Annex A table list), via iTeh, IEC (preview hosted by iTeh Standards) cdn.standards.iteh.ai/samples/20486/c899a27bb5574bdaa724e48fe2f56fd5/CISPR-32-2015.pdf
- CISPR 16-2-3:2016+AMD1:2019+AMD2:2023 CSV (edition 4.2), Radiated disturbance measurements, IEC webstore.iec.ch/en/publication/86318
- CISPR 16-2-3 edition 4.2 preview (foreword, table of contents), via iTeh, IEC (preview hosted by iTeh Standards) cdn.standards.iteh.ai/samples/21346/57cf4ccc48d24d06886c88f790ca0184/CISPR-16-2-3-2016.pdf
- CISPR 16-1-4:2025 (edition 5.0), Antennas and test sites for radiated disturbance measurements, IEC webstore.iec.ch/en/publication/68250
- CISPR 16-1-1:2019 (edition 5.0), Measuring apparatus, IEC webstore.iec.ch/en/publication/60774
- CISPR 16-4-2:2011+AMD1:2014+AMD2:2018 CSV (edition 2.2), Measurement instrumentation uncertainty, IEC webstore.iec.ch/en/publication/63750
- CISPR 16-4-2:2011 preview (clause 4, Table 1 reference uncertainty values), via iTeh, IEC (preview hosted by iTeh Standards) cdn.standards.iteh.ai/samples/16836/3d649487728c4d77ae1b73d819da7aa5/CISPR-16-4-2-2011.pdf
- CISPR 16-4-2:2011/AMD2:2018 preview (changes to Table 1), via iTeh, IEC (preview hosted by iTeh Standards) cdn.standards.iteh.ai/samples/22906/35527691ab084b33a990c564c15dca0f/CISPR-16-4-2-2011-AMD2-2018.pdf
- CISPR 11:2024 (edition 7.0), Industrial, scientific and medical equipment, RF disturbance characteristics, IEC webstore.iec.ch/en/publication/66118
- 47 CFR 15.109, Radiated emission limits (unintentional radiators), eCFR www.ecfr.gov/current/title-47/chapter-I/subchapter-A/part-15/subpart-B/section-15.109
- 47 CFR 15.33, Frequency range of radiated measurements, eCFR www.ecfr.gov/current/title-47/chapter-I/subchapter-A/part-15/subpart-A/section-15.33
- 47 CFR 15.35, Measurement detector functions and bandwidths, eCFR www.ecfr.gov/current/title-47/chapter-I/subchapter-A/part-15/subpart-A/section-15.35
- 47 CFR 15.31, Measurement standards (ANSI C63.4-2014, measurement distance), eCFR www.ecfr.gov/current/title-47/chapter-I/subchapter-A/part-15/subpart-A/section-15.31
- 47 CFR 15.101, Equipment authorization of unintentional radiators, eCFR www.ecfr.gov/current/title-47/chapter-I/subchapter-A/part-15/subpart-B/section-15.101
- IEEE/ANSI C63.4-2014, Methods of measurement of radio-noise emissions, 9 kHz to 40 GHz, IEEE standards.ieee.org/ieee/C63.4/5841/
- Commission communication, EMC Directive harmonised standards (OJ C 14, January 16, 2015): EN 55032:2012 replaces EN 55022 from March 5, 2017, EUR-Lex eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52015XC0116(03)
- Commission Implementing Decision (EU) 2019/1326, EMC harmonised standards, consolidated text of June 10, 2022, EUR-Lex eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:02019D1326-20220610
- Directive 2014/30/EU on electromagnetic compatibility, EUR-Lex eur-lex.europa.eu/eli/dir/2014/30/oj
Frequently asked questions
- Why does a radiated emissions test usually start at 30 MHz?
- Because that is where the limit tables start. The CISPR 32 radiated limits (Tables A.2 to A.5) and the 47 CFR 15.109(a) and (b) tables begin at 30 MHz, and 15.33(b)(1) says the spectrum is investigated without going below the lowest frequency for which a radiated limit is specified. The measurement standard itself goes lower: CISPR 16-2-3 covers 9 kHz to 18 GHz, with loop-antenna methods for 9 kHz to 30 MHz used by product standards that set limits there.
- What is the difference between peak, quasi-peak and average detectors?
- CISPR 16-1-1 specifies one receiver per detector: quasi-peak from 9 kHz to 1000 MHz (clause 5), peak, average and RMS-average from 9 kHz to 18 GHz (clauses 6 to 8). From 30 MHz to 1 GHz the quasi-peak receiver uses a 120 kHz bandwidth with 1 ms charge and 550 ms discharge time constants. CISPR 32 limits are quasi-peak up to 1 GHz and peak plus average above. 47 CFR 15.35 sets the same split for the FCC, with a 1 MHz minimum resolution bandwidth above 1000 MHz and a peak limit 20 dB above the average limit.
- Do the standards require a 6 dB margin below the limit?
- No. CISPR 16-4-2 sets the decision rule: if the lab uncertainty is at or below the CISPR value (6.3 dB for field strength at an OATS or in a SAC from 30 to 1000 MHz, Table 1), the measured level is compared directly with the limit. If the lab uncertainty is larger, the excess is added to the measured level first. Any margin above that is an internal design target, not a requirement of CISPR 32 or Part 15.
- Where is the pre-scan and final measurement sequence described?
- In clause 8 of CISPR 16-2-3, "Automated measurement of emissions": pre-scan measurements (8.3), data reduction (8.4), emission maximization and final measurement (8.5), post-processing and reporting (8.6). Minimum measurement times and scan times for peak and quasi-peak detectors are in clause 6.6, and Table E.1 gives the maximum amplitude difference between peak and quasi-peak detected signals.
- Why is the antenna raised and lowered and turned through two polarisations?
- On an open-area test site or in a semi-anechoic chamber, the ground plane reflection makes the received field depend on antenna height. CISPR 16-2-3 clause 7.3.5 covers antenna height variation for 30 MHz to 1 GHz, and the standard gives correction factors for horizontal polarisation (Table 6) and recommended pre-scan antenna heights (Table 7). The maximum over heights, polarisations and turntable positions is the value compared with the limit.
- What separates FCC Class A from Class B?
- Distance and level. 47 CFR 15.109(a) gives Class B limits at 3 m (100, 150, 200 and 500 microvolts per metre for 30-88, 88-216, 216-960 and above 960 MHz). 15.109(b) gives Class A limits at 10 m (90, 150, 210 and 300 microvolts per metre). Brought to the same distance with the 20 dB per decade rule of 15.31(f)(1), Class A is about 10 dB looser below 960 MHz and 6 dB looser above.
- Is testing above 1 GHz always required?
- No. It depends on the highest frequency generated or used in the product. In CISPR 32 Table 1, up to 108 MHz the test stops at 1 GHz; then 2 GHz up to 500 MHz, 5 GHz up to 1 GHz, and above that five times the highest frequency or 6 GHz, whichever is lower. 47 CFR 15.33(b)(1) uses the same steps but stops at the fifth harmonic or 40 GHz, and it adds a 30 MHz upper limit when the highest frequency is below 1.705 MHz.
- Which EN 55032 version gives presumption of conformity in the EU?
- EN 55032:2015 with A11:2020, cited under Directive 2014/30/EU (Implementing Decision (EU) 2019/1326 as amended by Decision (EU) 2020/1630). EN 55022:2010 stopped giving presumption on March 5, 2017, when EN 55032:2012 took over. CISPR 32 edition 3 is still a draft at IEC in October 2026.