Skip to content

Surge and Burst (IEC 61000-4-5 / 4-4): EMC transients

Guide · Conducted transient EMC immunity

Surge IEC 61000-4-5 and EFT/Burst IEC 61000-4-4 form the core of conducted transient immunity in the EMC reference framework. Mandatory for any product placed on the European market under EMC Directive 2014/30/EU, they are referenced by EN 55035 (multimedia equipment), EN 55024 (information technology equipment), EN 60601-1-2 (medical devices) and many vertical product standards. This page details waveforms, severity levels per class and port type, coupling and decoupling networks, the metrology of the capacitive clamp, performance criteria A, B and C, and the classic pitfalls observed in pre-compliance.

Surge and Burst represent two physical families of conducted transients that any electronic product connected to mains or to external cables must face. Separating them is a deliberate IEC committee choice, their energy, duration and coupling mechanism differ too much for a single standard to cover both.

  • Surge (IEC 61000-4-5). Single high-energy pulse, typically arising from induced lightning on the network or from distribution transformer switching. Slow waveform (microseconds), high energy, capable of destroying components if protection is insufficient.
  • EFT/Burst (IEC 61000-4-4). Repetitive burst of very fast pulses originating from inductive load switching (relays, contactors, motors, industrial lighting). Fast waveform (nanoseconds), low individual energy but wide spectrum extending to several hundred megahertz, capable of disturbing logic on digital links.

A product can be robust against one and vulnerable to the other. Protection mechanisms differ, varistors and gas tubes for Surge, decoupling capacitors and common-mode chokes for Burst. This is why both tests are systematically combined in harmonised EMC standards.

IEC 61000-4-5 applies to the following ports of an electrical or electronic equipment.

Port typeTypical couplingSource impedanceDefault waveform
AC powerCapacitive coupling through coupling/decoupling network (CDN)2 ohmCombination 1.2/50 us + 8/20 us
DC powerCapacitive coupling through CDN2 ohmCombination 1.2/50 us + 8/20 us
Short symmetrical signal port (under 30 m)CDN type S42 ohmCombination 1.2/50 us + 8/20 us
Outdoor telecom portDedicated CDN40 ohm10/700 us and 5/320 us
Ethernet link > 30 mEthernet CDN or clamp per product standard42 ohmCombination 1.2/50 us + 8/20 us or telecom variant

Ports strictly internal to an enclosure (internal bus between two adjacent boards) are not concerned. The practical rule, any port that may connect a cable leaving the enclosure must be evaluated.

The 1.2/50 us and 8/20 us combination wave

Section titled “The 1.2/50 us and 8/20 us combination wave”

The Surge generator, called Combination Wave Generator (CWG), simultaneously produces two profiles depending on the load.

  • Open circuit (high impedance). Pulsed voltage with 1.2 us rise time and 50 us duration at half amplitude. This is the nominal test voltage.
  • Short circuit (low impedance). Pulsed current with 8 us rise time and 20 us duration at half amplitude. This is the nominal test current.

The ratio of open-circuit voltage to short-circuit current defines the source impedance of the generator, fixed at 2 ohm for power ports. The same pulse thus becomes a voltage test for a product with high input impedance and a current test for a product with low input impedance. The CWG is calibrated individually and its calibration certificate is part of the test dossier.

For short signal ports, the CDN adds a series impedance bringing the effective source impedance to 12 ohm (high-impedance signal port) or 42 ohm (symmetrical signal port).

For outdoor telecom ports, the waveform switches to 10/700 us in voltage and 5/320 us in current, with a source impedance of 40 ohm, described in the corresponding annex of the standard and aligned with ITU-T Recommendations K.20 and K.21 for the resistibility of telecommunication equipment in a telecommunication centre.

The normative levels are organised in four classes corresponding to increasingly severe environments.

ClassTypical environmentLine-to-ground level (common mode)Line-to-line level (differential mode)
1Well-protected environment (lab, controlled room)0.5 kVnot applicable
2Partially protected environment (standard office)1 kV0.5 kV
3Moderate industrial, residential and commercial2 kV1 kV
4Severe industrial, long outdoor lines4 kV2 kV
XOpen class, value to be specified by product standardTo be definedTo be defined

For a consumer product or multimedia equipment under EN 55035, the typical AC-port target is Class 3 (2 kV common mode, 1 kV differential mode). For a medical device under EN 60601-1-2 edition 4, the AC port typically rises to 2 kV common mode and 1 kV differential mode, with mandatory criterion B and preferred criterion A.

The application methodology is specified precisely by IEC 61000-4-5.

  • Five positive pulses and five negative pulses per level and per test condition, ten pulses minimum.
  • Minimum one-minute spacing between consecutive pulses to allow varistors, gas tubes and other protection elements to thermally reset.
  • Mains-phase synchronisation for AC ports, pulses applied at 0, 90, 180 and 270 degrees of the sine wave. The 270-degree phase is the worst case for a product whose rectifier diode conducts at the zero crossing.

For a single-phase AC port, the total typically reaches 40 pulses per level (5 positive polarity x 4 phases + 5 negative polarity x 4 phases). For a DC port, mains-phase synchronisation does not apply and the total stays at 10 pulses per level.

The CDN fulfils two critical simultaneous functions.

  1. Coupling. Inject the generator pulse onto the line of the equipment under test (EUT) at the desired point (line-ground, line-line).
  2. Decoupling. Prevent the pulse from feeding back into the upstream mains and into other lab equipment.

Decoupling is performed by common-mode and differential-mode chokes sized to block the high-frequency component of the pulse. An undersized CDN (impedance too low, parasitic capacitance too large) lets energy leak upstream, which falsifies the voltage actually applied to the EUT and invalidates the measurement. CDN selection depends on port type, nominal network voltage, maximum current and conductor count, and is documented in the report.

IEC 61000-4-4 applies to power, earth, signal and control ports of any electrical or electronic equipment. The functional earth port is tested only when the manufacturer explicitly declares it as a product function (chassis earth, not via the PE conductor).

The individual Burst pulse has the following characteristics.

ParameterNormative value
Rise time5 ns +/- 30 %
Duration at half amplitude50 ns +/- 30 %
Repetition frequency5 kHz or 100 kHz (per revision and level)
Burst duration15 ms
Period between bursts300 ms
Total application duration per polarity1 minute minimum
Generator source impedance50 ohm

A 15 ms burst at 100 kHz contains 1 500 pulses, and the test sequence applies this burst every 300 ms for at least one minute. The total reaches several thousand bursts per polarity, several million individual pulses end to end.

The 2012 edition of the standard introduced the choice between 5 kHz and 100 kHz as repetition frequency. The 100 kHz frequency is preferred for modern tests because it better reflects the fast switching observed on modern power networks (variable-frequency drives, switching power supplies). Some older product standards still reference the historical 5 kHz.

ClassTypical environmentPower port levelSignal port level
1Well-protected environment0.5 kV0.25 kV
2Protected environment1 kV0.5 kV
3Moderate industrial, residential2 kV1 kV
4Severe industrial4 kV2 kV
XOpen classTo be definedTo be defined

For a residential or multimedia product (EN 55035), the typical target is Class 3 on the power port (2 kV) and Class 2 on long signal ports (0.5 kV). A medical device under EN 60601-1-2 typically goes to 2 kV power and 1 kV on signal ports longer than 3 m.

The Burst pulse is coupled to the EUT through two methods depending on port type.

  • Direct coupling through CDN. Used for power ports. The CDN injects the pulse on each conductor in turn and on all conductors in common mode. Internal CDN coupling impedance is typically 33 nF.
  • Capacitive coupling through the clamp (capacitive coupling clamp). Used for signal and control ports whose connectors do not lend themselves to direct injection (ethernet, RS-485, audio, video). The clamp is a metallic structure in two half-shells closed around the cable, providing a distributed capacitance of 50 to 200 pF between the clamp body and the inner bundle. This capacitance distributed along the clamp length (typically 1 m) couples the pulse without galvanic contact.

The clamp position is standardised, 0.1 m above a ground reference plane, with the EUT cable leaving at 90 degrees to the main clamp axis and a termination on the EUT side at the product interface. An incorrect position introduces impedance variability that can exceed a factor of 2 on the level actually applied.

For EFT/Burst, criterion B is the default in current harmonised EMC standards (EN 55024, EN 55035, EN 60601-1-2). The product may show temporary disturbance during the burst but must return to nominal operation without intervention after application ends. Temporary data loss is generally accepted, configuration loss or a reset is rarely accepted.

To place these two tests in the IEC 61000-4 panorama of conducted and radiated immunity, the table below compares the three most frequent transient tests.

TestStandardWaveformPhysical mechanism simulatedTypical couplingTypical Class 3 level
ESD (electrostatic discharge)IEC 61000-4-20.7-1 ns / 60 nsDischarge from an operator or charged objectContact or air, by hand-held probe4 kV contact, 8 kV air
EFT/BurstIEC 61000-4-45/50 ns in burstsInductive load switching on the networkCDN or capacitive clamp2 kV power, 1 kV signal
SurgeIEC 61000-4-51.2/50 us voltage, 8/20 us currentInduced lightning, transformer switchingCDN, capacitive or direct2 kV common mode, 1 kV differential mode

Reading, ESD acts on very short time scales (sub-nanosecond) and high voltage levels but with low total energy. EFT/Burst acts on short time scales (nanosecond) repeated in bursts to excite the spectral band up to several hundred megahertz. Surge acts on long time scales (microsecond) with high unit energy capable of destroying an unprotected component. The three tests are systematically combined in a complete EMC campaign.

See RED tests for the panorama of EMC tests applicable under the RED directive, and CE tests for EMC directive 2014/30/EU.

The IEC 61000-4 family: placing Surge and Burst

Section titled “The IEC 61000-4 family: placing Surge and Burst”

Surge and Burst are two entries in the IEC 61000-4 chapter dedicated to immunity test techniques. The table below places the conducted tests and their direct neighbours.

StandardPhenomenonTest type
IEC 61000-4-2ESDConducted, point-like
IEC 61000-4-3Radiated RF fieldRadiated
IEC 61000-4-4EFT/BurstConducted transient
IEC 61000-4-5SurgeConducted transient
IEC 61000-4-6Conducted RFConducted, low frequency
IEC 61000-4-850/60 Hz magnetic fieldRadiated
IEC 61000-4-11Voltage dips and interruptionsConducted, mains
IEC 61000-4-12Damped oscillatory wavesConducted transient

IEC 61000-4-12 (Ring Wave) is less often required in European harmonised standards but remains applicable for some North American industrial environments and for power ports downstream of HV/LV transformers. The two variants (damped oscillatory wave at 100 kHz and 1 MHz) reproduce the so-called damped oscillatory wave transients characteristic of switching in electrical substations.

For a product intended for a road vehicle on-board network, IEC 61000-4-5 and IEC 61000-4-4 do not apply directly. The reference standard becomes ISO 7637-2, which defines its own pulses characteristic of automotive physics.

ISO 7637-2 pulseSimulated phenomenonPeak level
Pulse 1Switching of an inductive load in parallel with EUT-75 to -150 V
Pulse 2aInterruption of a current in nearby wiring+37 to +112 V
Pulse 2bContact interruption after engine stop+10 V (attenuated DC)
Pulse 3a / 3bFast bursts similar to EFT/Burst-150 to +100 V
Pulse 4Starter (slow voltage variation)-7 V
Pulse 5a / 5bLoad dump (battery disconnect with alternator loaded)up to +87 V peak

Pulse 5b in particular (load dump with alternator protection) remains the on-board network worst case and drives the input protection sizing of any automotive electronic module. A product certified to IEC 61000-4-5 at 2 kV is never automatically certified to ISO 7637-2, the physics of automotive transients does not share the same energy profile nor the same source impedance. A dedicated campaign is required.

Rejections observed in EMC labs come overwhelmingly from a few recurring causes.

  1. Wrong port classification. An ethernet port declared as a short signal port when it serves an outdoor or building cable should undergo the telecom 10/700 us test, not the classic 1.2/50 us combination. Classification appears in the test plan and drives the CDN, waveform and level. An error here invalidates the entire campaign on that port.

  2. CDN not matched to port impedance and current. The CDN must be selected for conductor count, nominal voltage, maximum current and frequency (50 Hz, 60 Hz, DC). An undersized CDN heats up and drifts in impedance, an oversized CDN has excessive parasitic capacitance that attenuates the pulse. The report must cite the model and calibration certificate of the CDN used.

  3. Non-compliant capacitive clamp position. Distance to the ground plane, coupling length, EUT cable termination, isolation of other cables, each parameter is specified by IEC 61000-4-4. A deviation introduces impedance variability that can reach a factor of 2 on the level actually applied to the cable. A photo of the clamp setup is required by some notified bodies.

  4. Source-impedance confusion. Surge uses 2 ohm on AC, 12 ohm on high-impedance signal, 42 ohm on symmetrical signal and 40 ohm on telecom. EFT/Burst uses 50 ohm. A test run with the wrong source impedance does not reflect the intended physical scenario and is non-compliant with the standard. The most frequent error is using the signal CDN for a test intended for power ports, which underestimates the energy actually delivered.

  5. Omission of mains-phase synchronisation. For Surge tests on AC ports, pulses must be injected at 0, 90, 180 and 270 degrees of the mains sine wave. A test run in asynchronous mode (random pulse phases) does not cover the worst case and is non-compliant with the normative procedure. The Surge generator must therefore feature a mains synchronisation input or an internal phase sensor.

Articulation with the EMC and RED directives

Section titled “Articulation with the EMC and RED directives”

Within the European Union, Surge and Burst are referenced in all harmonised EMC immunity standards listed in the OJEU for directives 2014/30/EU (EMC) and 2014/53/EU (RED).

  • EN 55035 (residential and commercial multimedia equipment), Surge 2 kV / 1 kV on AC, EFT/Burst 1 kV on AC, 0.5 kV on signals.
  • EN 55024 (information technology equipment, transitioning to EN 55035), identical levels.
  • EN 60601-1-2 edition 4 (medical devices), Surge 2 kV / 1 kV, EFT/Burst 2 kV / 1 kV, typically stricter criteria (A preferred).
  • EN 61000-6-1 (generic immunity residential/commercial/light industrial), Surge 1 kV / 0.5 kV, EFT/Burst 1 kV / 0.5 kV.
  • EN 61000-6-2 (generic immunity industrial), Surge 2 kV / 1 kV, EFT/Burst 2 kV / 1 kV.

Selection of the applicable product standard depends on product classification. The generic EN 61000-6-x standard applies in the absence of a dedicated vertical standard. See the Glossary for the definitions of the terms used here (CDN, EFT, Surge, criteria A/B/C).

To plan a Surge and Burst campaign integrated into an EMC dossier, here are the orders of magnitude observed in a European ISO/IEC 17025 accredited laboratory.

Configuration4-4 + 4-5 campaign duration
Single-port AC product, no long signals0.5 to 1 day
AC + 2 to 4 short signal ports1 to 2 days
AC + signals + 10/700 us telecom port2 to 4 days
EN 60601-1-2 medical device, multi-mode3 to 5 days
Class 4 industrial equipment2 to 4 days

Duration depends on the number of ports to test, the target level (Class 4 requires more powerful generators that are scarcer in the lab fleet) and the number of operating modes to cover for criteria A and B.

  • Surge IEC 61000-4-5 and EFT/Burst IEC 61000-4-4 are the two conducted transient immunity tests required by every European product EMC standard.
  • Surge. Single pulse 1.2/50 us (voltage) and 8/20 us (current), impedance 2 ohm AC, 12-42 ohm signal, 40 ohm telecom (10/700 us). Levels 0.5 to 4 kV per class.
  • EFT/Burst. Repetitive bursts 5/50 ns at 5 kHz or 100 kHz, 15 ms per burst every 300 ms. Levels 0.25 to 4 kV per class and port. Source impedance 50 ohm.
  • Coupling. Dedicated CDN for power and some signals, capacitive clamp for signals not suited to direct injection.
  • Performance criteria A (no degradation), B (temporary degradation with automatic recovery), C (loss accepted with operator intervention). Criterion B is the common default.
  • ISO 7637-2 replaces Surge and Burst for automotive products. Pulses specific to the 12 V / 24 V on-board network, levels reaching hundreds of volts in transient, lower source impedance.
  • Classic pitfalls. Port classification, CDN mismatch, capacitive clamp position, source-impedance confusion, omission of mains-phase synchronisation.

For the EMC directive context, see CE tests. For RED health and EMC tests, see RED tests. For the definitions of terms used here, see the Glossary.

Sources & references

  1. IEC 61000-4-5:2014+AMD1:2017, Testing and measurement techniques, surge immunity test , IEC webstore.iec.ch/publication/4079
  2. IEC 61000-4-4:2012, Testing and measurement techniques, electrical fast transient burst immunity test , IEC webstore.iec.ch/publication/4193
  3. ISO 7637-2:2011, Road vehicles, electrical disturbances from conduction and coupling, transients on supply lines , ISO www.iso.org/standard/50925.html
  4. EN 55035:2017+A11:2020, Electromagnetic compatibility of multimedia equipment, immunity requirements , CENELEC www.cenelec.eu/dyn/www/f?p=104:110:::::FSP_PROJECT,FSP_LANG_ID:64483,25
  5. ITU-T Recommendation K.20, resistibility of telecommunication equipment installed in a telecommunication centre , ITU-T www.itu.int/rec/T-REC-K.20
  6. IEC 61000-4-12:2017, Testing and measurement techniques, damped oscillatory wave immunity test , IEC webstore.iec.ch/publication/28729