Conducted Emission Mechanisms

Conducted emissions originate from a variety of mechanisms within electronic equipment, each with distinct characteristics, frequency content, and propagation behavior. Understanding these generation mechanisms is fundamental to diagnosing emission problems and implementing effective mitigation strategies. Unlike radiated emissions that propagate through space, conducted emissions travel along physical conductors, making their behavior intimately tied to circuit topology, component parasitics, and interconnection impedances.

This section explores the primary mechanisms responsible for conducted interference generation, from the fundamental distinction between differential-mode and common-mode noise to specific sources such as power supply switching, ground loops, and parasitic oscillations. Mastering these concepts enables engineers to predict emission behavior during design, diagnose problems during testing, and implement targeted solutions that address root causes rather than symptoms.

Differential-Mode Conducted Noise

Nature of Differential-Mode Emissions

Differential-mode (DM) conducted noise, also known as normal-mode or symmetric-mode noise, flows in a loop between the power conductors, traveling out on the line conductor and returning on the neutral conductor (or vice versa). This noise appears as a voltage difference between the two conductors and represents unwanted high-frequency components superimposed on the intended power or signal waveform. The current loop is well-defined and follows the same path as the intentional power or signal current.

The primary sources of differential-mode emissions are circuits that draw pulsating or rapidly changing currents from the power supply. Switching power supplies are archetypal DM noise generators because their input current consists of high-frequency pulses synchronized with the switching frequency. The discontinuous conduction mode in power converters, where input current drops to zero during portions of each switching cycle, creates particularly rich harmonic content extending into the megahertz range.

Digital circuits also generate differential-mode noise through their power supply current demands. When large numbers of logic gates switch simultaneously, the aggregate current surge creates voltage drops across power distribution impedances. These transient current spikes contain frequency components extending well beyond the clock frequency, with spectral energy concentrated at harmonics of the switching rate and at frequencies corresponding to the rise and fall times of the current pulses.

Frequency Characteristics

The frequency spectrum of differential-mode emissions depends on the nature of the current waveform at the source. For switching power supplies operating in continuous conduction mode, the input current resembles a triangular waveform with a fundamental frequency at the switching frequency (typically 50 kHz to several MHz) and harmonics that decrease in amplitude at approximately 20 dB per decade. The switching transitions, with their fast edges, contribute higher-frequency components that extend the emission spectrum into tens of megahertz.

Discontinuous conduction mode, common in light-load conditions or certain converter topologies, produces current pulses with sharp leading and trailing edges. The spectral content of these pulses extends to frequencies determined by the edge rates, often reaching 30 MHz or higher. The envelope of the spectrum typically shows a first corner frequency related to the pulse width and a second corner frequency related to the edge rise time, with the amplitude rolling off at 40 dB per decade above the second corner.

Understanding these spectral characteristics guides filter design. Low-frequency DM noise requires inductors with high inductance values and capacitors with adequate voltage ratings. High-frequency DM noise demands attention to component parasitics, particularly the equivalent series resistance (ESR) and equivalent series inductance (ESL) of capacitors, and the self-resonant frequency and interwinding capacitance of inductors.

Mitigation Strategies for Differential-Mode Noise

Filtering differential-mode noise exploits the well-defined current loop. Series inductance inserted in the current path presents increasing impedance at higher frequencies, while shunt capacitance across the power lines provides a low-impedance bypass for high-frequency currents. The classic LC low-pass filter configuration is the foundation of DM filtering, with multiple stages providing steeper attenuation slopes when single-stage filters prove insufficient.

X-capacitors, connected between line and neutral, are primary DM filtering components. These capacitors must withstand the full line voltage plus transients and are typically metallized film types with excellent self-healing properties. Capacitor selection must consider not only capacitance value but also ESR, ESL, and current handling capability. At high frequencies, ESL causes the capacitor impedance to rise, so surface-mount multilayer ceramic capacitors (MLCCs) are often added in parallel with film capacitors to extend high-frequency performance.

Differential-mode chokes provide series inductance for DM filtering. These are wound on cores with air gaps or distributed gap materials to prevent saturation from DC or low-frequency currents. The inductance value must be sufficient to provide the required attenuation at the lowest frequency of concern, while the current rating must exceed the maximum operating current with adequate margin. Core losses at the switching frequency contribute to temperature rise and must be evaluated in the thermal design.

Common-Mode Conducted Currents

Origins of Common-Mode Emissions

Common-mode (CM) conducted noise flows in the same direction on all power conductors relative to the ground reference, returning through ground connections, earth, or parasitic capacitances. Unlike differential-mode noise, which follows the intended current path, common-mode noise arises from unintended coupling mechanisms that connect high-frequency voltage or current sources to ground through parasitic paths. These currents are particularly troublesome because they are difficult to filter and because conductors carrying CM currents act as efficient monopole antennas, converting conducted emissions into radiated emissions.

The most common source of CM emissions in switching power supplies is capacitive coupling from high dv/dt switching nodes to the chassis or ground. The drain or collector of a switching transistor experiences voltage transitions of hundreds of volts in nanoseconds, creating dv/dt rates exceeding 10 kV per microsecond. This rapidly changing voltage couples through parasitic capacitances to heatsinks, transformer windings, and chassis structures. The resulting displacement current must return to the source, typically flowing through the power cord and back through the mains to complete the loop.

Transformer interwinding capacitance is another significant CM coupling path in isolated power supplies. The primary-to-secondary capacitance, though typically only a few picofarads, provides a path for high-frequency currents to flow from the primary switching node to the secondary circuit and then to the chassis ground. Electrostatic shields between windings can reduce this coupling, but the shield must be properly connected, and the shield-to-winding capacitances still provide some coupling path.

Parasitic Capacitance Analysis

Understanding and quantifying parasitic capacitances is essential for predicting and controlling CM emissions. Every conductor in a system has capacitance to every other conductor and to ground. While most of these capacitances are negligible, those connected to high dv/dt nodes can drive significant CM currents even at picofarad levels. A 10 pF capacitance subjected to a 10 kV/microsecond voltage transition generates 100 mA of displacement current, more than enough to cause emission limit violations.

Key parasitic capacitances in switching power supplies include transistor case-to-heatsink capacitance (typically 10-100 pF depending on package and insulator), transformer interwinding capacitance (typically 5-50 pF), and PCB trace-to-chassis capacitance (highly variable depending on layout). Finite element analysis tools can estimate these capacitances during design, while impedance analyzers and network analyzers enable measurement in prototypes. Reducing parasitic capacitances or interposing shields and bypass paths are primary mitigation approaches.

The CM current driven by a parasitic capacitance equals the capacitance multiplied by the rate of voltage change (I = C dv/dt). This relationship reveals why slowing down switching transitions is an effective, though often costly, approach to reducing CM emissions. Doubling the transition time halves the CM current for a given parasitic capacitance. Gate resistors, snubbers, and soft-switching topologies all serve to reduce dv/dt and thus CM current generation.

Common-Mode Filtering Techniques

Common-mode chokes are the primary components for CM filtering. These consist of two or more windings on a common magnetic core, arranged so that DM currents (flowing in opposite directions through the windings) produce canceling magnetic fluxes, while CM currents (flowing in the same direction) produce additive fluxes. The result is high impedance to CM currents and very low impedance to DM currents, allowing the choke to selectively attenuate CM noise while presenting minimal impedance to the power or signal being filtered.

Y-capacitors, connected from each power conductor to ground, provide a low-impedance return path for high-frequency CM currents. By shunting CM currents to ground at the filter input, Y-capacitors prevent these currents from propagating along the power cord to the mains. However, Y-capacitor values are strictly limited by safety standards to control leakage current, typically to a few nanofarads total. This limits Y-capacitor effectiveness to higher frequencies, placing greater burden on CM chokes for low-frequency CM filtering.

The interplay between CM chokes and Y-capacitors creates a resonant filter structure. The resonant frequency and damping significantly affect filter performance. Underdamped filters can exhibit amplification at the resonant frequency, potentially worsening emissions in that frequency range. Proper filter design requires considering the complete impedance network, including source and load impedances, parasitic elements, and component tolerances. Simulation tools that model realistic component behavior are invaluable for optimizing CM filter performance.

Power Supply Switching Noise

Switching Transient Generation

Switch-mode power supplies generate conducted emissions through the fundamental mechanism of their operation: rapid on-off switching of transistors to control power flow. Each switching transition creates voltage and current transients with spectral content extending to frequencies determined by the switching speed. The faster the transitions, the higher the frequency content of the resulting emissions. Modern power semiconductors, particularly gallium nitride (GaN) and silicon carbide (SiC) devices, switch in nanoseconds, generating significant energy at frequencies extending to 100 MHz and beyond.

During transistor turn-on, current rises rapidly while voltage across the device falls. During turn-off, voltage rises while current falls. The overlap of voltage and current during these transitions not only creates switching losses but also generates high-frequency current spikes that propagate through the power distribution network. The waveforms are complex, influenced by parasitic inductances, capacitances, and the reverse recovery characteristics of diodes and body diodes in the circuit.

Ringing and oscillation following switching transitions are common sources of EMI. Parasitic inductances in the power loop interact with junction capacitances and snubber capacitances to form resonant circuits. When excited by the rapid switching transitions, these circuits ring at frequencies typically ranging from 10 MHz to over 100 MHz. The damping of these oscillations depends on resistive losses in the circuit. Underdamped ringing produces sustained high-frequency emissions that can exceed regulatory limits even when the fundamental switching noise is well-controlled.

Input Current Harmonics

The input current drawn by switching power supplies contains rich harmonic content at multiples of the switching frequency. A flyback converter, for example, draws current in pulses during the on-time of the primary switch, with no input current flowing during the off-time. The Fourier spectrum of this pulsating current contains a fundamental component at the switching frequency and harmonics that extend to high orders. Active power factor correction (PFC) circuits, while improving the fundamental power factor, still generate high-frequency harmonics due to their internal switching.

The amplitude of switching frequency harmonics depends on the converter topology, operating mode, and load conditions. Converters operating in continuous conduction mode (CCM) typically have lower harmonic content than those in discontinuous conduction mode (DCM). Interleaved converter stages, with multiple phase-shifted switching channels, can provide harmonic cancellation at specific frequencies, reducing overall emission levels. Quasi-resonant and resonant converter topologies shape the switching waveforms to reduce harmonic content at the expense of more complex control.

Input filter design must address the full spectrum of switching harmonics. The fundamental frequency and low-order harmonics require substantial inductance and capacitance for attenuation. Higher harmonics benefit from the increasing impedance of inductors and the decreasing impedance of capacitors with frequency. However, component parasitics eventually limit high-frequency performance. Multi-stage filters with different component values optimized for different frequency ranges provide the best overall performance across the conducted emission spectrum.

Gate Drive and Control Circuit Contributions

Gate drive circuits, while operating at lower power levels than the main power stage, can contribute significantly to conducted emissions. The gate drive loop, from the driver output through the gate resistor and transistor gate to the driver ground and back to the driver, carries high-frequency currents during switching transitions. If this loop is poorly designed or if the driver ground couples to the main power ground, these currents can propagate to the input and output terminals, contributing to conducted emissions.

Control circuits, including pulse-width modulation (PWM) controllers, oscillators, and feedback networks, generate clock-related emissions at their operating frequencies. While these frequencies are typically lower than the main switching frequency, they can still fall within the conducted emission measurement band. Proper decoupling, layout, and shielding of control circuits prevents their emissions from coupling into power conductors. Spread-spectrum modulation, which intentionally varies the switching frequency, distributes emission energy across a wider bandwidth, reducing peak amplitudes at any single frequency.

Auxiliary power supplies within a main power supply unit are often overlooked sources of conducted emissions. These supplies, which provide bias power for control circuits and gate drives, typically operate at their own switching frequencies. Their emissions combine with those of the main power stage, potentially creating compliance issues at frequencies where the main stage alone would pass. Careful design and filtering of auxiliary supplies is essential for overall EMC performance.

Ground Loop Currents

Ground Loop Formation

Ground loops form when multiple ground connections create redundant paths for current flow. If two pieces of equipment are connected both through a signal cable and through the AC power system earth, a ground loop exists between them. Any voltage difference between the two ground points drives current through this loop. At low frequencies, earth resistance and building wiring impedance create ground voltage differences due to other equipment sharing the same grounding system. At high frequencies, electromagnetic induction couples energy into the loop, and the loop's resonant characteristics can amplify certain frequencies.

The impact of ground loop currents on conducted emissions depends on how these currents couple into the measurement path. Current flowing through a cable shield can induce voltage into the signal conductors through magnetic coupling. Current flowing through the ground reference of a circuit can modulate signal voltages relative to that reference. When ground loop currents contain high-frequency components, they directly contribute to conducted emissions measured at the power line input.

Multi-equipment systems are particularly susceptible to ground loop problems. Audio-visual systems, computer networks, industrial control systems, and laboratory instrumentation often connect multiple devices that are also connected to the AC power system at different outlets. The potential for ground loops is inherent in such configurations. Understanding the complete ground current flow, from noise sources through the various ground paths to the measurement point, is essential for diagnosing and resolving ground-loop-related emission problems.

Ground Impedance Effects

Real ground connections have impedance that varies with frequency. At DC and low frequencies, resistance dominates, typically milliohms for good bonding connections. At higher frequencies, inductance becomes the controlling factor, with even short ground leads presenting significant impedance. A 10 cm wire has approximately 100 nH of inductance, which represents over 6 ohms of impedance at 10 MHz. This frequency-dependent impedance causes the voltage drop across ground connections to increase dramatically with frequency for a given current.

When ground currents from multiple sources share a common ground impedance, they interact through that impedance. Current from one source creates a voltage drop that appears as noise to other circuits using the same ground reference. This common-impedance coupling is a fundamental mechanism for conducted noise propagation within systems. High-frequency ground currents from switching circuits can corrupt sensitive analog circuits, digital communication signals, or even cause logic errors in digital systems if the ground voltage transients exceed input noise margins.

Minimizing ground impedance is a primary defense against ground loop problems. Wide, short ground connections reduce both resistance and inductance. Ground planes in PCB design provide very low impedance paths for return currents. Multiple parallel ground connections distribute current and reduce total impedance. However, even with best practices, some ground impedance is unavoidable, and design must ensure that the resulting voltage drops are tolerable for all circuits sharing that ground.

Breaking Ground Loops

Several approaches exist for eliminating or mitigating ground loop problems. Galvanic isolation, using transformers, optocouplers, or capacitive isolators, breaks the conductive path and eliminates the ground loop entirely. Isolation is often the most effective solution but requires isolated power for the isolated circuits and introduces propagation delay and potential signal integrity challenges. The isolation barrier must provide adequate voltage rating, low coupling capacitance, and sufficient bandwidth for the signals being transmitted.

Balanced signaling reduces susceptibility to ground loop currents by rejecting common-mode interference. Differential transmission, where the signal is represented by the voltage difference between two conductors rather than the voltage on one conductor relative to ground, is inherently immune to ground potential variations that affect both conductors equally. The common-mode rejection ratio (CMRR) of the receiver determines how effectively ground loop interference is rejected. Professional audio, Ethernet, and many industrial protocols use balanced signaling specifically for ground loop immunity.

Frequency-selective grounding provides multi-point grounding at high frequencies for effective shielding while preventing low-frequency ground loops. Capacitors in the ground path pass high-frequency currents to the ground reference while blocking DC and low-frequency currents. The capacitor value must be chosen to provide low impedance at the frequencies of concern while maintaining adequate impedance at lower frequencies. This approach is commonly used for cable shield grounding and for equipment enclosure bonding in systems where true multi-point grounding would create unacceptable ground loops.

Power Line Harmonics

Harmonic Generation Mechanisms

Power line harmonics are integer multiples of the fundamental power frequency (50 or 60 Hz) that appear in the current drawn by electronic equipment. These harmonics arise from the nonlinear current-voltage characteristics of the load. Linear loads draw sinusoidal current in phase with the applied voltage, but nonlinear loads draw distorted current waveforms that contain harmonic components. Virtually all electronic equipment with power supply front ends behaves as a nonlinear load to some degree.

The classic example is the simple rectifier-capacitor input stage found in many power supplies. Current flows only during the portion of each AC cycle when the instantaneous line voltage exceeds the capacitor voltage, resulting in narrow current pulses near the peaks of the voltage waveform. These pulses are rich in odd harmonics, particularly the 3rd, 5th, 7th, and 9th. The harmonic content depends on the relative values of the source impedance, rectifier characteristics, and capacitor size, with higher source impedance generally reducing harmonic distortion.

Modern power factor correction (PFC) circuits are specifically designed to reduce harmonic current by shaping the input current to approximate a sinusoid in phase with the voltage. Active PFC using boost converters is highly effective, achieving power factors above 0.99 and total harmonic distortion below 5%. However, the high-frequency switching of the PFC converter introduces its own emissions at the switching frequency and its harmonics, representing a trade-off between low-frequency harmonic reduction and high-frequency switching noise generation.

Regulatory Limits on Harmonics

International standards, particularly IEC 61000-3-2, establish limits on the harmonic current that equipment can draw from public power networks. Equipment is classified into categories based on type and power level, with different limits for each class. Class A (balanced three-phase equipment and most other equipment), Class B (portable tools), Class C (lighting), and Class D (equipment with specific input current shapes, primarily televisions and computers) each have specific harmonic current limits expressed in absolute amperes per harmonic order.

Compliance with harmonic standards is mandatory for products sold in many markets and is enforced through product certification. Testing measures the RMS current at each harmonic frequency from the 2nd through the 40th using specified source impedance and voltage conditions. The equipment must operate at its rated load during testing. Results are compared to the applicable limits, and products exceeding any limit cannot be legally marketed in jurisdictions that enforce these standards.

The rationale for harmonic limits stems from the adverse effects of harmonic currents on power distribution systems. Harmonic currents cause additional heating in transformers and conductors, reduce the power-handling capacity of the infrastructure, can cause resonances in power factor correction capacitor banks, and create neutral conductor overloading in three-phase systems when triplen harmonics (3rd, 9th, 15th, etc.) add rather than cancel. By limiting individual equipment harmonics, standards prevent cumulative degradation of power quality.

Harmonic Mitigation Approaches

Passive harmonic filtering uses LC networks tuned to attenuate specific harmonic frequencies. Series inductors increase the source impedance seen by the rectifier, reducing harmonic current by extending the conduction angle. Shunt filters tuned to the 3rd, 5th, or higher harmonics provide low-impedance paths that divert harmonic currents from the mains. Passive filters are simple and reliable but bulky, and their effectiveness depends on the source impedance and the presence of other nonlinear loads on the same network.

Active power factor correction dynamically shapes the input current waveform to track the input voltage. A boost converter between the rectifier and bulk capacitor draws continuous current that follows the rectified sinusoidal voltage envelope. Control loops adjust the switching to maintain the desired current shape and regulate the output voltage. Active PFC achieves excellent power factor and very low harmonic distortion but adds complexity, cost, and its own high-frequency switching emissions that require filtering.

Valley-fill and other passive PFC circuits provide a compromise between simple rectifier-capacitor inputs and full active PFC. These circuits use diodes and capacitors to extend the conduction angle, reducing harmonic content without active switching. Performance is not as good as active PFC, but for lower-power applications, passive PFC may meet regulatory requirements with lower cost and complexity. The choice of harmonic mitigation approach depends on power level, regulatory requirements, cost constraints, and acceptable complexity.

Voltage Fluctuations and Flicker

Causes of Voltage Fluctuations

Voltage fluctuations occur when equipment with varying power demands causes variations in the voltage at the point of common coupling with the power distribution network. Large motors, arc furnaces, welding equipment, and some electronic loads can draw currents that change rapidly, creating voltage drops across the source impedance that modulate the supply voltage. Even relatively small loads can cause noticeable fluctuations if the source impedance is high or if the load changes are particularly rapid or repetitive.

Switching power supplies with dynamic load characteristics can generate voltage fluctuations if the load changes rapidly between high and low power states. Computer power supplies during processor load transitions, laser printers during fusing cycles, and electronic equipment with standby modes that wake periodically all represent fluctuating loads. The magnitude of the fluctuation depends on the load current change, the repetition rate, and the impedance of the power distribution from the equipment back to the public network.

The frequency of voltage fluctuations significantly affects their perception and impact. Fluctuations at frequencies between about 0.5 Hz and 25 Hz are most noticeable as flicker in incandescent and some LED lighting. The human visual system is most sensitive to brightness variations at approximately 8 Hz, making fluctuations in this range particularly objectionable. Standards define flicker severity metrics that weight the fluctuation amplitude by frequency to correlate with human perception.

Flicker Severity Assessment

IEC 61000-3-3 and related standards establish limits on voltage fluctuations and flicker for equipment connected to public low-voltage networks. The standard defines a flicker meter algorithm that simulates the response of the lamp-eye-brain chain to voltage fluctuations, producing short-term flicker severity (Pst) measured over 10 minutes and long-term flicker severity (Plt) calculated from multiple Pst values over 2 hours. Equipment must not cause Pst values exceeding 1.0 or Plt values exceeding 0.65 under normal operating conditions.

Testing for flicker compliance uses specialized flickermeters that implement the standardized algorithm. The equipment under test operates through its normal operating cycles while the flickermeter monitors the voltage at the input terminals. Various operating modes, including start-up, maximum power, standby transitions, and any repetitive operating cycles, must be evaluated. The test results depend on the reference impedance, which is defined in the standard to represent typical public network conditions.

Equipment that causes significant voltage fluctuations, particularly for larger loads such as motors and heating elements, may require specific mitigation measures. Soft starters reduce motor inrush current by gradually applying voltage during acceleration. Electronic phase control can modulate heating element power without the abrupt switching that causes step voltage changes. For very large loads, dedicated transformers or power factor correction equipment may be necessary to meet flicker requirements.

Design Considerations for Flicker Compliance

Designing equipment for flicker compliance requires attention to how power demand varies during operation. Sudden load changes should be minimized or ramped gradually when possible. Repetitive load cycles should avoid frequencies in the 0.5 to 25 Hz range where human sensitivity to flicker is highest. If substantial load changes are unavoidable, the duty cycle and repetition pattern should be chosen to minimize the weighted flicker severity as calculated by the standard algorithm.

Input capacitance provides some buffering against rapid load changes. Larger bulk capacitors can absorb brief load increases without drawing proportionally increased current from the mains. However, very large capacitors increase inrush current and can create their own power quality issues. Active power management, where the equipment's power consumption is controlled to limit the rate of change, may be necessary for demanding applications.

Coordination with facility power distribution is important for larger equipment. Understanding the actual source impedance at the installation point, which may be much lower than the reference impedance used in standard testing, helps predict real-world flicker impact. For industrial installations, power quality analysis may be required to ensure that new equipment will not cause unacceptable disturbances for other users sharing the same distribution network.

Transient Conducted Emissions

Sources of Transient Emissions

Transient conducted emissions are brief, non-repetitive or irregularly occurring disturbances on power lines and signal cables. Unlike continuous switching noise, transients are often associated with specific events such as equipment startup, shutdown, mode changes, or fault conditions. The brief duration of transients makes them challenging to characterize and often requires time-domain analysis rather than the frequency-domain measurements used for continuous emissions.

Power supply inrush current at turn-on creates significant transients. The charging of input filter capacitors from zero to operating voltage draws peak currents that can be many times the steady-state current. This inrush produces voltage disturbances on the power line and can trip protective devices or cause disturbances in nearby equipment. Inrush limiting circuits, including NTC thermistors and active soft-start circuits, reduce these transients but add complexity and power loss.

Switching of reactive loads, particularly inductive loads like motors and relays, generates transients when the current path is interrupted. The energy stored in the magnetic field must be dissipated, and without proper suppression, the resulting voltage spike can reach hundreds or thousands of volts. These transients propagate through the power distribution and can appear as conducted emissions on equipment power inputs. Snubber networks and transient voltage suppressors (TVS) across inductive loads control these emissions.

Electrostatic Discharge Effects

Electrostatic discharge (ESD) is a specific transient mechanism that occurs when accumulated static charge discharges through an arc. While ESD is primarily an immunity concern, the discharge event also creates emissions. The ESD current waveform has sub-nanosecond rise time, containing significant spectral energy into the gigahertz range. The discharge current flowing through equipment and grounding conductors creates transient conducted emissions that can propagate to other equipment through shared power and ground connections.

The magnitude and spectrum of ESD-induced conducted emissions depend on the discharge path and the impedances along that path. Direct discharge to equipment chassis typically produces the most severe transients. Indirect discharge to nearby objects couples through air and capacitance, with reduced amplitude but still significant spectral content. The transient conducted emissions from ESD events can cause errors or malfunctions in sensitive electronic equipment sharing the same power or signal connections as the discharge target.

Controlling ESD-related conducted emissions requires attention to grounding, shielding, and layout. Low-impedance ground connections ensure that ESD currents flow through intended paths rather than coupling into sensitive circuits. Adequate spacing and shielding between ESD entry points and sensitive traces prevents direct coupling. EMI filters at cable entry points attenuate the high-frequency components of ESD transients before they can propagate to other equipment.

Transient Characterization and Control

Characterizing transient emissions requires time-domain measurement techniques. Oscilloscopes with high bandwidth and sampling rate capture the transient waveform, revealing peak amplitude, rise time, duration, and ringing characteristics. Spectral analysis of the captured waveform using FFT shows the frequency content of the transient. Repetitive transients can be characterized using triggered averaging, while single-event transients require single-shot capture and real-time spectral analysis.

Transient suppression techniques depend on the nature of the transient source. Surge protective devices (SPDs) and transient voltage suppressors (TVS) clamp voltage spikes by providing low-impedance paths when threshold voltages are exceeded. Metal oxide varistors (MOVs) are commonly used for AC mains transient protection, while TVS diodes provide faster response for signal lines. Filter networks attenuate high-frequency transient components, and proper grounding ensures that transient currents flow through intended paths.

Testing for transient emissions compliance, when required, uses specialized equipment to capture and analyze the transient events. Some standards specify limits on transient emissions, particularly for automotive and aerospace applications where susceptibility of nearby equipment is a concern. Design validation should include testing under various operating conditions, including startup, shutdown, load changes, and abnormal operating modes, to identify and address all significant transient emission sources.

Parasitic Oscillations

Origins of Parasitic Oscillation

Parasitic oscillations are unintended self-sustaining oscillations that arise when feedback paths exist with sufficient gain and phase shift to satisfy the Barkhausen criterion. In power electronics, parasitic oscillations often occur in the gate drive circuit, where the Miller capacitance of the power transistor provides feedback from the drain/collector to the gate. If the gate drive circuit presents sufficient impedance and the layout creates inductance in the gate loop, oscillation at tens or hundreds of megahertz can occur during switching transitions.

The oscillation may be continuous, occurring throughout the operating cycle, or it may be triggered by specific events such as switching transitions or load changes. Transition-triggered oscillations are particularly common, as the power stage passes through states of high transconductance where gain is maximized. These oscillations may not be visible on low-bandwidth test equipment and may cause EMI problems that are difficult to diagnose without high-frequency measurement capability.

Parasitic oscillations can also occur in filter networks and feedback loops. LC filter stages can ring or oscillate if not properly damped. Feedback loops with inadequate phase margin can become unstable under certain load or line conditions. While these oscillations may occur at lower frequencies than gate-drive parasitics, they can still contribute to conducted emissions, particularly if the oscillation frequency or its harmonics fall within the measurement band.

Detection and Diagnosis

Detecting parasitic oscillations requires appropriate measurement bandwidth and technique. High-frequency oscillations may appear as noise or thickening of switching waveforms on inadequate oscilloscopes. A true representation requires bandwidth well above the oscillation frequency and appropriate probe techniques. Current probes in the gate and drain circuits can reveal high-frequency current oscillations even when voltage probes with bandwidth limitations show little indication of the problem.

Spectrum analyzer or EMI receiver measurements of conducted emissions may show unexplained peaks at frequencies not related to the switching frequency or its harmonics. These anomalous peaks warrant investigation for parasitic oscillation. Time-domain correlation, using triggered oscilloscope captures synchronized to the emission events, can identify whether the emissions correspond to switching transitions or other operating events, providing clues to the oscillation mechanism.

Temperature and operating condition sensitivity often characterizes parasitic oscillations. The gain and phase relationships that sustain oscillation depend on component parameters that vary with temperature, voltage, and current. An oscillation that appears only at cold start, at high temperature, at light load, or at high line voltage suggests marginally stable conditions that cross into oscillation only under specific circumstances. Thorough testing across the full operating envelope is essential to identify all parasitic oscillation conditions.

Suppression Techniques

Gate resistors are the primary defense against gate-drive oscillations in power transistors. The resistor damps the resonant circuit formed by gate capacitance, lead inductance, and Miller capacitance. Values from a few ohms to tens of ohms are typical, with the optimum value balancing oscillation suppression against increased switching losses and slower switching speed. Ferrite beads in series with the gate provide frequency-selective damping, offering high impedance at oscillation frequencies while maintaining low impedance at switching frequencies.

Layout optimization reduces the parasitic inductances and capacitances that enable oscillation. Short, wide gate drive traces minimize loop inductance. Kelvin-source connections, which use separate sense and power connections to the source/emitter, prevent voltage drops in the power current path from affecting gate drive. Proper bypass capacitor placement and ground plane design ensure low-impedance return paths for high-frequency currents.

Filter network damping prevents LC oscillations in EMI filters. Resistors in series with filter capacitors (damped capacitors) or in parallel with filter inductors provide the losses needed to prevent resonance. The damping must be sufficient to avoid oscillation while not excessively degrading filter performance or causing power dissipation problems. Simulation with realistic component models helps optimize damping network values for the specific filter configuration.

Impedance Stabilization Networks

Purpose and Function of LISNs

The Line Impedance Stabilization Network (LISN), also known as Artificial Mains Network (AMN), is a fundamental measurement accessory for conducted emission testing. Its purpose is threefold: to present a defined, standardized impedance to the equipment under test (EUT) regardless of the actual mains impedance; to isolate the measurement from noise present on the mains supply; and to provide a measurement port where conducted emissions can be extracted and measured with a spectrum analyzer or EMI receiver.

Without a LISN, conducted emission measurements would vary depending on the impedance of the local power network, which changes with wiring, transformer characteristics, and other connected loads. The LISN ensures that measurements made at different times and locations are comparable, which is essential for compliance testing and correlation between pre-compliance and compliance measurements. Standards specify the LISN impedance characteristics, typically 50 ohms at measurement frequencies for power line conducted emissions.

The LISN also performs critical filtering functions. Inductors in series with the mains input block high-frequency noise from the mains from contaminating the measurement. Capacitors from the power conductors to ground provide the defined impedance at high frequencies while allowing the power frequency current to flow to the EUT. The measurement port, typically a BNC connector, provides a 50-ohm output for connection to measurement equipment through a coaxial cable.

LISN Specifications and Variations

Different conducted emission standards specify different LISN characteristics. The most common is the 50 microhenry LISN specified by CISPR 16 for measurements from 9 kHz to 30 MHz. This LISN presents 50 ohms impedance in parallel with 50 microhenrys from each power conductor to ground. The impedance varies with frequency due to the inductor, being lower at low frequencies and asymptotically approaching 50 ohms at high frequencies. The phase angle of the impedance is also specified and must meet tolerances for accurate measurements.

Automotive EMC testing uses different LISN configurations. CISPR 25 specifies 5 microhenry LISNs for measurements extending to 108 MHz, reflecting the higher frequency range of concern in automotive applications. The lower inductance maintains the 50-ohm impedance to higher frequencies without the resonances that would occur with larger inductors. Different LISN configurations exist for testing DC power ports, signal lines, and antenna ports.

LISN calibration is essential for accurate measurements. The impedance and measurement port response must be verified periodically using network analyzers and standard test equipment. Phase angle measurements are particularly important because phase errors can cause significant measurement uncertainty. Many standards specify calibration requirements and acceptable tolerances for LISN performance.

Proper LISN Usage

Correct LISN installation and usage significantly affects measurement accuracy. The LISN must be well-bonded to the ground reference plane with short, wide connections. Any impedance in this connection affects the measurement by modifying the CM path impedance. The LISN housing is typically the ground reference, and the EUT should be positioned on a non-conductive surface at a specified height above this reference, with cables arranged according to standard procedures.

Multiple LISNs are required when testing equipment with more than two power conductors, or when auxiliary equipment (such as loads or monitoring equipment needed to operate the EUT) must be isolated from the measurement. Each power conductor requires its own LISN channel, and all LISN ground connections must be bonded together at the common ground reference. When testing three-phase equipment, three single-phase LISNs or dedicated three-phase LISNs are used.

The LISN measurement port must be properly terminated when not connected to measurement equipment. An unterminated port presents a high impedance that disrupts the defined network impedance and invalidates measurements on other ports. Typically, the line and neutral ports are measured sequentially, with the unused port terminated in 50 ohms. Some modern test systems measure both ports simultaneously using switching networks or dual-channel receivers.

Summary and Best Practices

Conducted emission mechanisms span a wide range of phenomena, from fundamental current and voltage transients in switching circuits to complex interactions involving parasitic elements, ground systems, and measurement networks. Differential-mode and common-mode noise, though arising from different mechanisms, often coexist and require coordinated mitigation strategies. Power supply switching generates the primary emissions in most electronic equipment, with the spectral characteristics determined by switching frequency, waveform shape, and transition speeds.

Ground loops and power line phenomena including harmonics, flicker, and transients represent additional emission mechanisms that extend beyond the equipment boundary to involve the power distribution infrastructure. Understanding these mechanisms enables design choices that minimize interference generation and enables diagnosis when problems arise. Parasitic oscillations represent a particularly challenging category because they may not be apparent until specific operating conditions reveal the instability.

Effective conducted emission control requires a systematic approach that considers all generation mechanisms, propagation paths, and measurement conditions. Designing with EMC in mind from project inception is far more effective than attempting to fix problems after compliance testing reveals failures. The impedance stabilization network, as the standardized interface for conducted emission measurement, defines the conditions under which compliance is evaluated and must be understood to interpret measurements and correlate test results.