What is the scope of IEC 61000-4-6?

IEC 61000-4-6 defines the immunity test method for electrical and electronic equipment against conducted disturbances induced by radio-frequency fields, in the 150 kHz to 80 MHz band. The fourth edition (Ed 4.0:2013) consolidated with amendment A1:2017 is the international reference currently in force. The standard is published by IEC Technical Committee TC 77B. It is complementary to IEC 61000-4-3, which addresses radiated immunity above 80 MHz, and provides an indirect simulation of RF fields by injecting disturbances onto cables connected to EUT ports. The standard sets the test method; levels and acceptance criteria are selected by product standards or generic immunity standards that reference it.

Why use CDN networks rather than a radiated field?

Below 80 MHz, wavelengths exceed 3.75 m. The typical size of a semi-anechoic chamber (3 to 10 m EUT-antenna distance) becomes insufficient to generate a homogeneous and reproducible radiated field. IEC 61000-4-6 sidesteps this limitation by simulating the field effect: an RF field sweeping a cable produces in practice a common-mode current on that cable, and it is this current that the standard reproduces by injecting it directly through a coupling/decoupling network (CDN), an EM clamp or a current injection probe. The method is cheaper and more reproducible than a radiated test on this band, and reflects the real coupling mechanism of conducted disturbances.

What is the difference between CDN-M, CDN-S, CDN-T and CDN-A?

Coupling and decoupling networks are defined by port type. CDN-M (mains) addresses AC or DC power ports. CDN-S is intended for shielded signal ports, where the shield is returned to ground through the network. CDN-T addresses unshielded symmetrical signal ports (typically 2, 4 or 8 conductors, hence CDN-T2, CDN-T4, CDN-T8 variants). CDN-AF1 or CDN-AF2 is dedicated to antenna ports, asymmetrical by construction. Choosing an unsuitable CDN for the port type is one of the frequently observed errors; a CDN-M used on an unshielded signal port, for example, does not establish the 150 ohm common-mode impedance required by the standard, and the injected level becomes inaccurate.

What are the test levels and modulation depth?

The standard defines three severity classes: Class 1 at 1 V, Class 2 at 3 V, Class 3 at 10 V, expressed as the rms voltage of the unmodulated signal at the injection point. The carrier is amplitude-modulated at 80 % by a 1 kHz sine wave. The 80 % AM at 1 kHz modulation is common to IEC 61000-4-6 and IEC 61000-4-3, which allows severity comparison between conducted and radiated tests. The sweep covers 150 kHz to 80 MHz, sometimes extended to 230 MHz depending on the applicable product standard. Frequency is swept in logarithmic steps of at most 1 %, with a dwell time long enough for the EUT to respond, typically at least 0.5 s per point or one full cycle of the function under test, whichever is longer.

When should one use an EM clamp or the BCI method rather than a CDN?

The CDN is the preferred method whenever a CDN suitable for the port type exists. When no CDN fits (non-standard cable topology, multi-pair non-symmetrical port, EUT operating constraints incompatible with CDN insertion), the standard allows two substitution methods. The electromagnetic clamp (EM clamp) surrounds the cable without electrical contact and injects the disturbance through combined capacitive and inductive coupling. The current injection probe (current clamp) directly injects a common-mode current onto the harness. Both methods require a prior calibration step, verifying that the injected level matches what an equivalent CDN would have delivered at the reference measurement point. Bulk current injection in the automotive environment, separately standardised as ISO 11452-4 (BCI), follows the same philosophy on different bands and with different levels.

How does the standard relate to EMC product requirements (EN 55035, EN 60601-1-2, etc.)?

IEC 61000-4-6 is a horizontal standard: it sets the test method but neither the applicable levels nor the acceptance criteria. Product standards or generic immunity standards select the applicable class from the 1/3/10 V catalogue. EN 55035 (multimedia) and EN 55024 (ITE, normatively sunsetting) typically require 3 V (Class 2) on power ports and signal ports longer than 3 m. EN 60601-1-2 (medical, 4th edition) imposes 3 V outside the professional healthcare environment, and 10 V on ISM bands within that environment, with extended sweep and specific conditions for life-support equipment. EN 61000-6-2 (industrial) calls for 10 V. The A, B, C performance criteria are defined by the product standard, not by 4-6 itself.

What are the most common pitfalls during the test?

Five errors recur. First, the wrong CDN type selected for the port, which falsifies the common-mode impedance and the actual injected level. Second, no conditioning of auxiliary equipment (AE): the EUT is not isolated from the laboratory and cables to AE must be decoupled by feedthrough CDNs to prevent a disturbance injected on the EUT port from being absorbed by AE. Third, a cable too short for the EM clamp or current probe: the standard requires a minimum length between the injection point and the EUT, typically 0.3 m. Fourth, CDN saturation by the operating current or voltage of the EUT, which can exceed CDN specifications on high-power supply ports. Fifth, neglecting dwell time: a sweep that is too fast passes through narrow EUT resonances without triggering a defect, and the report concludes a false conformity.

How does IEC 61000-4-6 relate to ISO 11452-4 (automotive BCI)?

The two standards are conceptually close: common-mode current injection on a harness to simulate the effect of an RF field. ISO 11452-4 is dedicated to the automotive environment and covers a wider band, typically 1 MHz to 400 MHz, with higher levels (up to 200 mA in BCI depending on OEM specifications such as GMW, Ford ESM or equivalent). CDNs are not used in automotive, which favours the current clamp because of harness topology and battery-voltage constraints. Calibration is performed by injected current, not by open-circuit voltage as in 4-6. A product targeting both markets (an embedded telematics module, for example) must be tested under both standards, with distinct test plans and benches.

IEC 61000-4-6: conducted RF immunity

Author Hugues Orgitello Last updated 27 May 2026 ~14 min read

Guide - IEC 61000-4-6

IEC 61000-4-6 is the horizontal immunity test standard for electrical and electronic equipment against conducted disturbances induced by radio-frequency fields. It covers the 150 kHz to 80 MHz band, where radiated simulation in a chamber becomes ineffective for lack of spatial homogeneity. The method injects disturbances onto cables connected to the equipment under test through coupling and decoupling networks (CDN), electromagnetic clamps or current injection probes. The consolidated fourth edition, IEC 61000-4-6:2013+AMD1:2017, is the reference currently in force, published by Technical Committee TC 77B of the International Electrotechnical Commission. The standard sets the method; test levels and acceptance criteria are selected by the product standards that reference it: EN 55035, EN 55024, EN 60601-1-2, EN 61000-6-1, EN 61000-6-2, and the entire body of vertical EMC standards. This page sets out the 4-6 / 4-3 split, the choice of injection device, the CDN catalogue by port type, the 1/3/10 V levels, the 80 % AM at 1 kHz modulation, sweep and dwell time, performance criteria, and the pitfalls observed in the laboratory.

Position in the IEC 61000-4 series

The IEC 61000-4 series groups electromagnetic immunity test methods. The 4-6 part addresses the low-frequency, conducted band; the 4-3 part addresses the high-frequency, radiated band. The pair reads as a spectrum split by dominant coupling mechanism.

Reference	Phenomenon	Coupling mechanism
IEC 61000-4-2	Electrostatic discharge (ESD)	Direct contact, air discharge
IEC 61000-4-3	Radiated RF field, typically 80 MHz to 6 GHz	Radiation in a semi-anechoic chamber
IEC 61000-4-4	Electrical fast transient bursts (EFT/Burst)	Capacitive coupling on cable
IEC 61000-4-5	Surge	Differential or common-mode impulse
IEC 61000-4-6	RF-induced conducted disturbances, 150 kHz to 80 MHz	Conducted injection via CDN, EM clamp, current probe
IEC 61000-4-8	Power-frequency magnetic field	50/60 Hz magnetic field
IEC 61000-4-11	AC voltage dips, interruptions, variations	Modulation of mains voltage

The 4-6 test is required by virtually every EMC product standard: EN 55035 (multimedia), EN 55024 (ITE, normatively sunsetting), EN 60601-1-2 (medical, 4th edition), EN 61000-6-1 and EN 61000-6-2 (residential and industrial environments). Combined with the IEC 61000-4-3 radiated test, it covers RF immunity over the entire 150 kHz to 6 GHz spectrum.

Why 80 MHz as the 4-6 / 4-3 boundary

The 80 MHz boundary is not arbitrary. It corresponds to a wavelength of 3.75 m, on the order of typical EUT-antenna distances in a semi-anechoic chamber (3 m, 10 m). Below this frequency, the radiated wave is no longer fully established in the chamber: the field becomes inhomogeneous, the position of the EUT, supports and cables significantly alters the actual applied level. Above it, the wavelength is short enough that a standard chamber yields a homogeneous field measurable at the reference point.

IEC 61000-4-6 sidesteps this difficulty by exploiting an experimental observation: in this band, the dominant effect of an RF field on a product is through the cables leaving it, which behave as receiving antennas and reinject a common-mode current into the EUT. The 4-6 method reproduces this current directly by conducted injection, without going through the radiated field. The method is cheaper, more reproducible, and still representative of the real coupling mechanism.

Aspect	IEC 61000-4-6 (conducted)	IEC 61000-4-3 (radiated)
Frequency band	150 kHz to 80 MHz, optionally to 230 MHz	80 MHz to 1 GHz, typically extended to 6 GHz
Injection mechanism	CDN, EM clamp, current probe	Antenna in a semi-anechoic chamber
Controlled quantity	rms voltage at the injection point	Electric field at the measurement point
Main equipment	RF generator, amplifier, CDN or clamp, attenuator	RF generator, broadband amplifier, antenna, chamber
Reproducibility at low frequency	Good (controlled coupling)	Poor (inhomogeneous field)
Instrumentation cost	Moderate	High (chamber, broadband antennas, amplifiers)
Physical phenomenon simulated	Common-mode current induced on cables	Radiated E-field illuminating the product

Spectral continuity is typically ensured by an administrative overlap at 80 MHz: the product standard may call for a 4-6 sweep up to 230 MHz to partially cover the low end of 4-3, or conversely extend 4-3 down to 26 MHz in some configurations. The test plan must pin this articulation explicitly.

Coupling and decoupling networks (CDN)

The CDN is the preferred injection device in the standard. It performs two functions simultaneously: it couples the RF disturbance towards the EUT port at a controlled 150 ohm common-mode impedance, and it decouples the auxiliary equipment (AE) or the supply network from the disturbance, so that energy is not dissipated anywhere other than the EUT. Each CDN is specified for a given port type; an unsuitable CDN does not hold the impedance and falsifies the injection level.

Code	Port type	Characteristics
CDN-M1, CDN-M2, CDN-M3	AC or DC power port, single- or polyphase	Decoupling inductors rated to carry the EUT supply current; common-mode injection on all conductors; suffix = conductor count (M1 = neutral only, M2 = phase + neutral, M3 = phase + neutral + earth, with three-phase extensions)
CDN-S	Shielded signal port (coaxial, shielded pair)	Coupling via the cable shield; shield returned to ground through the network
CDN-Tx	Unshielded signal port, symmetrical pairs	x = conductor count: CDN-T2 (2 conductors), CDN-T4 (4 conductors, twisted pair + twisted pair or similar), CDN-T8 (8 conductors, Ethernet-style RJ45)
CDN-AF1, CDN-AF2	Antenna port (receiver, low-power transmitter)	Asymmetrical by construction; the useful RF signal passes through; the disturbance is injected in common mode on the outer conductor

CDN-M units are constrained by EUT supply current: a CDN-M2 rated for 16 A is unsuitable above that. CDN-Tx units require a precise match to the EUT cable: a CDN-T8 (Ethernet) does not properly apply to a four-pair RS-485 cable lacking pair symmetry. The CDN-S requires a continuous shield from EUT to network. When no CDN in the catalogue fits the port type, the standard allows substitution methods.

EM clamp and current injection probe

When no CDN fits, the EM clamp and current probe provide alternatives. The three methods are not equivalent: the standard ranks them and requires prior calibration at the reference point.

Method	Principle	Conditions of use	Limitations
CDN	Network inserted between EUT and AE; coupling through internal capacitors or transformers; decoupling through inductors	Preferred method; 150 ohm common-mode impedance guaranteed	Requires a CDN matched to the port type; insertion modifies the EUT connectivity
EM clamp	Hinged clamp surrounding the cable; combined capacitive and inductive coupling without electrical contact	Substitution method when no CDN fits	No AE decoupling; needs a straight cable typically >= 0.3 m between clamp and EUT; frequency-dependent coupling
Current probe	Hinged ferromagnetic clamp; inductive common-mode injection	Substitution method, last resort	Less repeatable injection; probe calibration required; cable must be long and straight

Calibration at the reference point is the critical step of clamp-based methods. On the bench, a calibration jig defines the relation between amplifier output power and equivalent voltage at the injection point on the EUT. This calibration is specific to the cable, the clamp, and the frequency range; it must be verified periodically. A missing or stale calibration invalidates the test report.

Test levels and modulation

The standard defines three severity classes. The level is expressed as the rms voltage of the unmodulated RF signal at the injection point, that is, the voltage the signal would have in the absence of AM modulation.

Class	Level (V rms, unmodulated)	Typical environment	Usual product standards
Class 1	1 V	Lightly disturbed environment, low RF exposure	Sensitive equipment with low intrinsic immunity, applies by exception
Class 2	3 V	Residential, commercial, light industrial environments; mobile phones at moderate distance	EN 55035 (multimedia), EN 60601-1-2 (medical outside professional healthcare environment), EN 61000-6-1
Class 3	10 V	Severe industrial environment, proximity to high-power transmitters, ISM bands	EN 61000-6-2 (industrial), EN 60601-1-2 on ISM bands inside professional healthcare environment

The RF signal is amplitude-modulated at 80 % by a 1 kHz sine wave. The modulation simulates the envelope of a voice signal on an RF carrier, and provides realistic stimulation of an electronic product: it is the baseband audio component, recovered through unintentional demodulation in a PN junction, that triggers immunity failures, not the pure carrier. A 3 V unmodulated level becomes, at peak with 80 % AM modulation, a signal whose peak reaches 3 * (1 + 0.8) = 5.4 V rms-equivalent peak, but the standard expresses the level in unmodulated value for severity-calculation consistency with 4-3.

The sweep covers 150 kHz to 80 MHz, optionally extended to 230 MHz depending on the product standard. Frequency progression is logarithmic, with steps no greater than 1 % of the previous point. The dwell time, or hold duration at each point, must be long enough to let the EUT manifest a defect. The standard sets a minimum of 0.5 s per point or the duration of one full cycle of the function under test, whichever is longer. An EUT whose useful functional cycle (acquisition, transmission, processing, display) takes 2 s therefore requires a minimum dwell of 2 s per point. A sweep that is too fast misses narrow EUT resonances without exciting them, and conceals a real defect.

EUT setup and auxiliary equipment

The standard prescribes a horizontal ground reference plane (GRP), typically a copper or aluminium sheet of at least 1 m by 1 m, wider than the projection of EUT and CDNs. The EUT sits on an insulating support typically 10 cm above the GRP, or directly on the GRP if the product design requires it (metal enclosure with earth connection). CDNs sit on the GRP, in electrical contact with it, at least 30 cm from the EUT. Cables between EUT and CDNs are straight, free of parasitic loops, over a typical length of 30 cm.

Auxiliary equipment (AE) refers to any apparatus needed to operate the EUT in a representative mode: external power supply, control PC, input signal generator, loads, output signal receivers. Cables to AE must be decoupled by feedthrough CDNs: injection takes place on the EUT side of the CDN, the AE sits on the decoupled side. Without this decoupling, the disturbance injected on the EUT port partially dissipates into the AE, which reduces effective severity and falsifies the test. Verifying that all cables to AE pass through a CDN is one of the pre-test checks.

Unused EUT cables (optional ports not connected in normal use) must be handled according to the product documentation. If normal use includes a cable, that cable must be present and terminated on a CDN or a representative load; an open port captures disturbances differently from a terminated port.

Performance criteria A, B, C

As with other tests in the 4-x series, acceptance criteria are not set by 4-6 itself. The classification structure is common to the generic immunity standards.

Criterion	Behaviour during test	Behaviour after test
A	Normal operation, performance maintained above the specified level	Normal operation without intervention
B	Temporary degradation allowed, performance may drop below the specified level	Automatic recovery without operator intervention
C	Loss of function allowed	Recovery acceptable after manual intervention (reset, restart)

For 4-6, the applicable criterion is typically A: the disturbance is continuous (unlike a 4-4 or 4-5 transient), and a product that fails during the sweep is likely to fail in actual service under influence of a nearby transmitter. Product standards reserve B or C for specific functions (acceptance of transient audible degradation on an audio headset, for example, outside the useful bands).

Medical case: EN 60601-1-2 and ISM bands

EN 60601-1-2 (4th edition) imposes a specific treatment of 4-6 on medical equipment. The base level is 3 V (Class 2), raised to 10 V on certain frequency ranges corresponding to ISM bands (industrial, scientific, medical), where intentional transmitters are likely in the professional healthcare environment. Equipment intended for hospital or clinical use in the presence of RF transmitters (electrosurgery, logistics RFID, contactless readers) must demonstrate immunity at these higher levels, on the bands where those transmitters concentrate their power.

For life-support equipment (critical systems for maintaining vital functions), EN 60601-1-2 imposes reinforced requirements: full sweep without acceptable degradation, criterion A on critical functions, extended dwell time. The life-support classification is carried by vertical medical standards and by the manufacturer's classification; it is not presumed by default.

Sub-GHz LPWA and coexistence

LPWA equipment (LoRaWAN, Sigfox, NB-IoT, LTE-M) transmits in sub-GHz bands (868 MHz in Europe, 915 MHz in the US). For them, IEC 61000-4-6 remains relevant in its own band (conducted disturbances on power and signal cables), but does not cover the transmit band itself, which falls under IEC 61000-4-3 and radio conformity tests under RED or FCC Part 15. An LPWA module whose power supply suffers from a current induced at 27 MHz (CB band) may experience clock drift or receiver-sensitivity loss; 4-6 detects these weaknesses, which a radiated test above 80 MHz would not see.

For cases where radio coexistence is the issue (sub-GHz modules sharing an enclosure with digital electronics having harmonised clocks), 4-6 is complemented by internal robustness tests (power-plane decoupling, RF-plane isolation, power-port filtering) that are outside its scope but rely on the same physical mechanism.

Automotive case: ISO 11452-4 BCI

The automotive environment has an equivalent standard, ISO 11452-4 (Bulk Current Injection, BCI), dedicated to embedded electronic components. The philosophy is the same (common-mode current injection on a harness), but parameters differ.

Aspect	IEC 61000-4-6	ISO 11452-4 (automotive BCI)
Typical band	150 kHz to 80 MHz, optionally 230 MHz	1 MHz to 400 MHz
Injection method	CDN preferred, EM clamp or current probe as substitution	Current probe only
Calibration	Equivalent voltage	Injected current
Levels	1, 3, 10 V rms	Defined by OEM specifications (GMW, Ford ESM, PSA, Renault, etc.), typically 60 to 200 mA depending on the level
Modulation	80 % AM at 1 kHz	80 % AM at 1 kHz and CW (continuous wave) depending on the level
Ground reference	Laboratory GRP	Simulated vehicle ground, harness length standardised at 1.5 m or 1.7 m

A product targeting the automotive market (ECU, smart sensor, embedded telematics, IVI) must be tested under ISO 11452-4, not under IEC 61000-4-6; the benches are distinct. An automotive aftermarket product bearing CE EMC marking may, depending on its classification, be tested under 4-6 and not under 11452-4. Dual marking is rare and requires both campaigns.

Pitfalls observed in the laboratory

Several recurring issues appear in conducted-RF test reports.

Wrong CDN class selected. An Ethernet port tested with a CDN-S instead of a CDN-T8 fails to hold the 150 ohm common-mode impedance, and the actually injected level is imprecise. CDN choice follows from port type, not from laboratory inventory convenience.
AE not decoupled. A disturbance injected on the EUT port must stay within the EUT for it to act. If the cable to the control PC or the measurement probe is not decoupled by a feedthrough CDN, part of the disturbance escapes. The test appears more lenient than it actually is.
EUT-AE cable too short for the clamp. With an EM clamp or current probe, the standard requires a straight cable of sufficient length (typically >= 0.3 m) between the clamp and the EUT, and between the clamp and the AE. A product with a short harness (a 0.2 m USB cable, for example) does not lend itself to this method and requires an adapted setup or a CDN.
CDN saturation. On a high-power supply port, the EUT operating current may exceed CDN specifications (a CDN-M2 rated 16 A does not sustain 20 A in continuous regime). The CDN saturates; decoupling becomes imprecise; the inductor changes behaviour; the injection level becomes non-reproducible.
Insufficient dwell time. A sweep that moves too quickly across each point lets immunity defects with narrow resonance escape. An EUT with a long functional cycle (slow acquisition, periodic transmission) requires a dwell time aligned with the useful cycle duration.
Modulation omitted. Testing with a pure carrier (CW) instead of 80 % AM at 1 kHz reduces actual severity; the demodulated baseband component does not appear, and several real defects are not triggered. The report must explicitly establish that the prescribed modulation was applied.
Confusion between 4-6 and 4-3. The 4-6 scope ends at 80 MHz, where 4-3 begins. Testing a product under 4-6 up to 230 MHz by option does not waive 4-3 on its own band. The two tests are independent and required in parallel in nearly every product standard. See the IEC 61000-4-3 radiated RF immunity guide for the radiated counterpart.
Confusion between 4-6 and conducted emissions. The 4-6 test measures immunity (the product suffers), not emission (the product emits). Conducted emission is measured by CISPR 16 methods through a LISN chain. See the conducted emissions and LISN test guide for emission measurement.

Connection with CE marking and the EMC Directive

The 4-6 test is required by virtually every EMC product standard harmonised under the EMC Directive 2014/30/EU and the RED 2014/53/EU for radio equipment. For the broader CE EMC test context, see the CE tests page and the RED tests page which applies to radio equipment.

For the definitions of the terms used on this page (CDN, EM clamp, current probe, common mode, GRP, AE, dwell time, criterion A/B/C, EUT, port), see the glossary.

Going further

Several related topics deserve a dedicated page:

4-3 (radiated RF immunity, 80 MHz to 6 GHz), the high-frequency counterpart of 4-6,
calibration of EM clamps and current probes at the reference point,
the articulation between 4-6 and specific medical requirements on ISM bands,
convergences and divergences between 4-6 and ISO 11452-4 for dual-market products.

Sources & references

IEC 61000-4-6 Electromagnetic compatibility (EMC), Part 4-6, Immunity to conducted disturbances induced by radio-frequency fields , IEC webstore.iec.ch/publication/4197
IEC TC 77B Electromagnetic compatibility, High frequency phenomena , IEC www.iec.ch/dyn/www/f?p=103:7:::::FSP_ORG_ID:1304
EN 55035 Electromagnetic compatibility of multimedia equipment, Immunity requirements , CENELEC www.cenelec.eu/
EN 60601-1-2 Medical electrical equipment, Part 1-2, Electromagnetic disturbances, Requirements and tests , IEC www.iso.org/standard/59789.html
ISO 11452-4 Road vehicles, Component test methods for electrical disturbances from narrowband radiated electromagnetic energy, Part 4, Harness excitation methods , ISO www.iso.org/standard/77983.html
EN 55024 Information technology equipment, Immunity characteristics, Limits and methods of measurement , CENELEC www.cenelec.eu/