Table of Contents
Optimizing switching frequencies in power electronic systems is a critical engineering challenge that directly impacts electromagnetic interference (EMI), system efficiency, component sizing, and regulatory compliance. Power electronic devices can lead to a higher rate of change in voltage and current during switching on and off processes, thus generating electromagnetic interference (EMI). Understanding the complex relationship between switching frequency selection and EMI generation enables engineers to design more efficient, compact, and compliant power conversion systems.
Understanding the Fundamentals of Switching Frequencies and EMI
Switching frequency refers to the rate at which power electronic devices—such as MOSFETs, IGBTs, and other semiconductor switches—transition between on and off states per second, typically measured in kilohertz (kHz) or megahertz (MHz). This fundamental parameter influences nearly every aspect of power converter design, from component selection to thermal management and electromagnetic compatibility.
The Relationship Between Switching Frequency and Component Size
Higher switching frequencies enable the use of smaller passive components, particularly inductors and capacitors, because the energy storage requirements decrease as frequency increases. This relationship allows for more compact power supply designs with higher power density—a critical advantage in applications where space is limited, such as mobile devices, automotive electronics, and aerospace systems. However, this benefit comes with trade-offs that must be carefully managed.
Conversely, lower switching frequencies require larger magnetic components and filter capacitors to maintain adequate energy storage and filtering performance. While this increases the physical footprint of the power supply, it can reduce switching losses and EMI generation, making lower frequencies attractive for high-power applications where efficiency and electromagnetic compatibility are paramount concerns.
EMI Generation Mechanisms in Switching Power Supplies
In power electronic equipment, since the semiconductor switching device will switch between the two states of turn-on and turn-off at high frequency with switching cycles, this is the main source of EMI in the system. The rapid voltage and current transitions during switching events create high-frequency harmonics that propagate through both conducted and radiated paths.
EMI interference in power electronic equipment is mainly determined by three factors: the interference source, conduction path, and the disturbed body. Therefore, EMI suppression techniques can be divided into three categories: (1) suppression of the interference source; (2) optimization of the impedance characteristics of conduction paths; and (3) shielding of the disturbed body.
Conducted vs. Radiated EMI
There are two major type of EMI: radiated and conducted. Understanding the distinction between these two forms of electromagnetic interference is essential for developing effective mitigation strategies.
Conducted EMI results from the rapid changes in a switching regulator’s conducted input current, including common-mode (CM) and differential-mode (DM) noise. Standard industry limits for conducted emissions usually cover a lower frequency range than radiated emissions, namely from 150 kHz to 30 MHz. This frequency range encompasses the fundamental switching frequency and its lower harmonics for most power electronic systems.
In a switch-mode supply, radiated EMI is usually generated by high dv/dt noise at the switching nodes. Industry standards for radiated emissions usually cover the frequency band from 30 MHz to 1 GHz. At these frequencies, radiated EMI from switching regulators is produced mainly by switching voltage ringing and spikes, and can depend heavily on PCB board layout.
Common-Mode and Differential-Mode Noise
CM conducted noise is measured between each input line and earth ground. CM noise is generated at high dv/dt switching nodes, couples through the device’s parasitic PCB capacitance, CP, to earth ground, then travels to the supply input LISN. Common-mode currents flow in the same direction through both power conductors and return through ground or parasitic capacitances.
DM noise is measured differentially between two input lines. Differential-mode currents flow in opposite directions through the power conductors, creating a current loop that generates both conducted and radiated emissions. The magnitude of differential-mode noise is directly related to the switching frequency, current ripple amplitude, and the di/dt during switching transitions.
CM currents are one of the main causes of EMI generation, and many articles have optimized active filters in terms of attenuating CM noise. Both common-mode and differential-mode noise must be addressed through appropriate filtering and design techniques to achieve regulatory compliance.
Strategic Approaches to Switching Frequency Optimization
Selecting the optimal switching frequency requires balancing multiple competing objectives: minimizing EMI, maximizing efficiency, reducing component size and cost, and meeting thermal constraints. This optimization process must consider the specific application requirements, operating environment, and regulatory standards that apply to the end product.
Frequency Selection Criteria
The choice of switching frequency significantly impacts both conducted and radiated EMI profiles. Lower frequencies (20-100 kHz) typically generate less high-frequency noise but require larger filter components. Mid-range frequencies (100-500 kHz) offer a balance between component size and EMI management. Higher frequencies (500 kHz to several MHz) enable very compact designs but demand advanced EMI mitigation techniques and careful PCB layout.
As a rule of thumb, suppressing EMI becomes harder as the operating frequency of a switch-mode converter rises. This fundamental trade-off drives many design decisions in power electronics, particularly in applications with stringent EMI requirements or limited space for filtering components.
Switched-mode power supplies (SMPS) are the most efficient way to supply regulated electrical power, but they generate a significant amount of EMI. Increased switching speeds and switching frequencies result in higher power density but also tend to make EMI worse. Engineers must carefully evaluate whether the benefits of higher frequency operation justify the additional complexity and cost of EMI mitigation measures.
Application-Specific Frequency Considerations
Different applications have unique requirements that influence optimal frequency selection. Consumer electronics often prioritize compact size and low cost, favoring higher switching frequencies despite increased EMI challenges. Industrial equipment may emphasize reliability and ease of EMI compliance, leading to more conservative frequency choices. Automotive applications must navigate strict EMI standards while operating in electrically noisy environments, requiring careful frequency selection and robust filtering.
Some applications, such as radio detection and ranging or ultrasound, use particular bandwidths in operation that must be avoided by the surrounding electronics at all costs. In these applications, the SMPS must have a fixed switch-node frequency. This constraint eliminates certain EMI mitigation techniques like spread spectrum modulation, requiring alternative approaches to achieve compliance.
Wide Bandgap Semiconductors and Frequency Optimization
The third-generation semiconductor devices composed of silicon carbide (SiC) and gallium nitride (GaN) have higher blocking voltages, higher operating temperatures, and higher switching speeds than Si-based power electronic devices. These advanced materials enable operation at significantly higher switching frequencies while maintaining high efficiency.
The third generation of semiconductors can greatly improve operating frequency, reduce the volumes of devices and radiators, and increase power density in converters. However, it must be noted that the high-frequency switching device is the main high-frequency noise source in a converter. The switching device generates electromagnetic interference (EMI) through the capacitive loop formed between the insulating sheet, the radiator, and the ground, bringing threats to the stable operation of the new energy system.
The faster switching transitions enabled by SiC and GaN devices create steeper voltage and current edges (higher dv/dt and di/dt), which can exacerbate EMI issues even as they improve efficiency. Designers working with these advanced semiconductors must employ sophisticated EMI mitigation techniques to fully realize their benefits.
Advanced Frequency Modulation Techniques for EMI Reduction
Beyond simply selecting an appropriate fixed switching frequency, engineers can employ various frequency modulation techniques to spread the electromagnetic energy across a wider spectrum, reducing peak emissions at any single frequency. These techniques have proven highly effective in achieving EMI compliance without significantly increasing filter complexity or cost.
Spread Spectrum Frequency Modulation (SSFM)
Dithering switching-voltage-regulator switching frequency—through the use of spread-spectrum frequency-modulation (SSFM) techniques—will reduce EMI and help designers pass compliance testing when filtering and optimized layout fails to meet EMI standards limits. Dithering slightly modulates a fundamental switching frequency in the switching regulator. Slightly varying the fundamental frequency (using a typical modulation of 3% or so) would modulate the peak switching noise energy, which would be shared among multiple fundamental frequencies.
Frequency Spread Spectrum (FSS) with a switching frequency jitter of ±6%, reduces EMI by not allowing emitted energy to stay in any one frequency for any significant amount of time. This technique effectively converts the discrete spectral lines characteristic of fixed-frequency operation into a continuous spectrum with lower peak amplitudes.
Spread spectrum modulation is one of the best EMI mitigating techniques for SMPS design. The technique is particularly effective for reducing conducted emissions in the frequency range where regulatory limits are most stringent, typically between 150 kHz and 30 MHz.
Frequency Jittering and Randomization
Frequency jitter is one of ways for power converters to reduce conducted EMI without extra hardware cost. Unlike spread spectrum modulation, which uses a deterministic modulation pattern, frequency jittering introduces pseudo-random variations in the switching frequency. This randomization prevents energy from accumulating at specific harmonic frequencies, distributing it more evenly across the spectrum.
Another positive side effect of quasi-resonant switching is that because switching is triggered when a valley is detected, rather than at a fixed frequency, a degree of frequency jitter is introduced, spreading the RF emission spectrum and further reducing EMI. This inherent jitter in quasi-resonant converters provides EMI benefits without requiring additional control circuitry.
The QuarEgg EPIC controller also includes integrated cable drop compensation and frequency jittering that reduces the EMI signature of the solution. Modern integrated controllers increasingly incorporate frequency modulation features, making these techniques more accessible to designers.
Chaotic Pulse Width Modulation
The method of combining chaotic pulse width modulation (CPWM) with the AEF design was proposed to enhance the EMI high-frequency mitigation effect. The simulation and experimental results show that the desired amplified gain of AEF can be reduced by CPWM, and thus, the EMI high-frequency inhibitory effect will be enhanced. This advanced technique uses chaotic sequences to modulate the switching pattern, creating a more uniform spectral distribution than traditional periodic modulation.
Chaotic modulation offers superior spectral spreading compared to simple frequency dithering, but requires more sophisticated control algorithms and may introduce additional output voltage ripple that must be filtered. The technique is most beneficial in applications where EMI margins are tight and conventional spread spectrum techniques prove insufficient.
Limitations and Considerations for Frequency Modulation
However, there are some situations in which spread spectrum simply can’t be used. For instance, if a customer is synchronizing one or many of their SMPS to an external signal, the spread spectrum feature of the device is turned off. The result is that the power at the fundamental switch-node frequency and the harmonics will be quite high, and additional filtering will be needed to meet EMI requirements.
Frequency modulation techniques can also introduce challenges in control loop design, as the varying switching frequency affects the converter’s small-signal response. Designers must ensure adequate phase margin and stability across the entire frequency modulation range. Additionally, some sensitive loads may be adversely affected by the frequency variation, requiring careful evaluation of system-level compatibility.
Soft-Switching Techniques for EMI Mitigation
Soft-switching techniques represent a fundamental approach to reducing EMI at its source by controlling the voltage and current waveforms during switching transitions. These methods minimize the overlap between voltage and current during switching events, dramatically reducing switching losses and high-frequency noise generation.
Zero-Voltage Switching (ZVS) Fundamentals
Zero voltage switching (ZVS) is a technique used in power electronics to minimize switching losses by ensuring that the voltage across the switching device becomes zero before it is turned on or off. ZVS is important in power electronics because it reduces switching losses, improves energy efficiency, and minimizes heat generation in switching circuits. It also helps to reduce electromagnetic interference (EMI) and noise.
The resonant soft switching technique is applied to ensure the voltage across or the current through the switching devices to be zero. Zero-voltage switching and zero-current switching are the two notable types under soft-switching topology to reduce dv/dt and di/dt, respectively. By reducing the rate of voltage change (dv/dt), ZVS significantly decreases the high-frequency components that contribute to both conducted and radiated EMI.
In addition to switching loss, ZVS also minimizes the switching noise during turn on and associated EMI. The smooth voltage transitions characteristic of ZVS operation eliminate the sharp edges that generate high-frequency harmonics in hard-switched converters.
ZVS Implementation Methods
This is accomplished by incorporating resonant circuits in conventional converter topologies so that zero voltage switching (ZVS) or zero current switching (ZCS) conditions can be created for the semiconductor switches. This type of switching is called soft switching in contrast to the PWM or the so-called hard switching.
The MOSFETs in a switching converter are driven to zero by parasitic inductance and the MOSFET output capacitance values (Coss), which create a resonant circuit with which switching events can be carefully timed. If the energy stored in the inductor or transformer can fully discharge Coss within the circuit’s dead-time, then you have achieved ZVS, no discrete resonant LC circuit is required. This approach leverages existing circuit parasitics, minimizing additional component count and cost.
The phase-shifted full-bridge (PSFB) converter achieves zero-voltage switching (ZVS) by leveraging the resonant interaction between the transformer’s leakage inductance (Llk) and the parasitic capacitances of the power switches (Coss). The key innovation lies in the controlled phase shift between the two half-bridge legs, which allows the converter to recycle energy stored in Llk for soft-switching transitions.
EMI Benefits of ZVS Operation
Compared to hard-switched full-bridge converters, PSFB-ZVS offers: Reduced switching losses (particularly at high frequencies), Lower electromagnetic interference (EMI) due to softer switching transitions. The EMI reduction stems from the elimination of capacitive discharge currents and voltage ringing that characterize hard-switched operation.
Two other advantages of ZVS are that it reduces the harmonic spectrum of any EMI (centering it on the switching frequency) and allows higher frequency operation resulting in reduced, easier-to-filter noise and the use of smaller filter components. By concentrating energy at the fundamental switching frequency rather than spreading it across numerous harmonics, ZVS simplifies filter design and reduces the required attenuation.
Reducing inrush also reduces strong bursts in the magnetic field which would be seen as radiated or conducted EMI at harmonics of the switching frequency. The controlled current transitions in ZVS converters prevent the sudden current steps that generate broadband electromagnetic noise.
Resonant Converter Topologies
Resonant converters (e.g., LLC resonant) can achieve ZVS quite easily as they already implement a resonant condition in their switching stage. LLC resonant converters have become increasingly popular in high-efficiency applications due to their inherent soft-switching characteristics across a wide load range.
High-frequency converters benefit from reduced electromagnetic interference (EMI) and improved efficiency, making them ideal for server power supplies and telecom rectifiers. The combination of high frequency operation and soft switching enables power densities that would be impossible with hard-switched topologies while maintaining excellent EMI performance.
As ZCS or ZVS allows for almost zero turn-off/turn-on losses, the switching frequency of QRCs can be very large (megahertz), allowing for reduction in the size of the passive elements of the converter. This capability to operate at very high frequencies without excessive losses or EMI makes soft-switching techniques essential for next-generation compact power supplies.
Zero-Current Switching (ZCS)
Zero-Current Switching (ZCS) is a soft-switching technique where the power semiconductor device is turned on or off precisely when the current through it crosses zero. This eliminates switching losses associated with hard-switched converters, particularly in high-frequency applications. The principle relies on shaping the current waveform using resonant components (L and C) to ensure the current naturally decays to zero before the device transitions. The key advantage of ZCS is the reduction in di/dt and turn-off losses, making it suitable for high-power and high-frequency converters.
While ZVS is generally preferred for voltage-source converters and applications with capacitive switching losses, ZCS excels in current-source topologies and situations where turn-off losses dominate. The choice between ZVS and ZCS depends on the specific converter topology, semiconductor device characteristics, and application requirements.
Challenges and Limitations of Soft Switching
ZVS requires precise timing control and is highly dependent on load conditions. Light-load operation may fail to maintain sufficient energy in the resonant tank to achieve zero-voltage transitions. Adaptive dead-time control or auxiliary circuits are often employed to extend the ZVS range across varying loads.
Almost all soft switching topologies have potential operating regimes where ZVS is lost. When soft switching is lost, the converter reverts to hard switching with attendant increases in losses and EMI. Designers must ensure that the converter maintains soft switching across the expected range of operating conditions or implement protection mechanisms to handle hard-switching events.
Many conventional LLC converters, for example, use resonant ZVS to achieve high efficiencies at high and full loads for a given input voltage. Unfortunately, they suffer as the load reduces or the input voltage changes, and hard switching losses start to become more prevalent. This load-dependent behavior requires careful design optimization and may necessitate hybrid control strategies that adapt to operating conditions.
Comprehensive EMI Filter Design Strategies
While optimizing switching frequency and implementing soft-switching techniques reduce EMI at the source, effective filtering remains essential for achieving regulatory compliance. A well-designed EMI filter attenuates both common-mode and differential-mode noise across the required frequency range while minimizing size, cost, and impact on power supply performance.
Passive EMI Filter Topologies
A pi filter is used in most applications. The pi filter has an advantage in coupling with the LISN, effectively increasing the order of the filter, and works well in most SMPSs where a large bus cap is connected to the output of the filter. The pi configuration, consisting of input capacitor, series inductor, and output capacitor, provides effective attenuation while maintaining good impedance matching to the source.
Choose the EMI filter topology based on the required attenuation levels. Single-stage LC filters may suffice for applications with moderate EMI requirements, while multi-stage filters become necessary when dealing with high switching frequencies, stringent regulatory limits, or challenging operating environments.
Three-stage filtering proves effective for soft-switching converters: First stage: Low-ESR ceramic capacitors (100nF-1μF) at switching devices; Second stage: LC filter with damped resonance (Q ≈ 1) Third stage: Feedthrough capacitors for >50 MHz suppression. This multi-stage approach addresses different frequency ranges with appropriately selected components, optimizing both performance and cost.
Common-Mode vs. Differential-Mode Filtering
Effective EMI filter design requires separate consideration of common-mode and differential-mode noise paths. Common-mode chokes, which present high impedance to common-mode currents while allowing differential-mode currents to pass freely, form the cornerstone of common-mode filtering. These components use coupled inductors wound on high-permeability cores to achieve large inductance values in compact packages.
Differential-mode filtering typically employs series inductors and shunt capacitors arranged in LC or multi-stage configurations. The inductor values required for differential-mode filtering are generally smaller than those for common-mode filtering, but the current-carrying requirements may be more demanding. Careful selection of core materials and winding techniques ensures that the inductors maintain their performance across the required frequency range without saturating under normal operating conditions.
Y-capacitors (line-to-ground) primarily address common-mode noise, while X-capacitors (line-to-line) target differential-mode noise. Safety regulations strictly limit Y-capacitor values to prevent excessive leakage current, requiring designers to balance EMI performance against safety requirements. X-capacitors face fewer restrictions but must be selected to handle the full line voltage plus any transients.
Active EMI Filtering
CM currents are one of the main causes of EMI generation, and many articles have optimized active filters in terms of attenuating CM noise. Active EMI filters use power electronics to inject cancellation currents that oppose the noise currents, effectively reducing EMI without the size and weight penalties of large passive components.
Active filters excel in applications where space and weight are critical constraints, such as aerospace and electric vehicle systems. They can provide frequency-dependent attenuation that adapts to the noise spectrum, potentially offering superior performance compared to passive filters of equivalent size. However, active filters introduce additional complexity, cost, and potential failure modes that must be carefully evaluated.
Hybrid approaches combining passive and active filtering elements can optimize the trade-off between performance, size, and cost. The passive components handle the bulk of the filtering at lower frequencies where they are most effective, while active circuits target specific problematic frequency ranges or provide adaptive attenuation based on operating conditions.
Filter Component Selection and Parasitics
There are larger mismatches at higher frequency noise peaks, but these are of lower importance because the DM conducted EMI filter size is mainly determined by lower frequency noise spikes. Some of this discrepancy is due to the accuracy of inductor and capacitor parasitic models, including PCB layout parasitic values. Component parasitics become increasingly important at higher frequencies, where they can significantly degrade filter performance.
Capacitor equivalent series resistance (ESR) and equivalent series inductance (ESL) limit high-frequency attenuation, while inductor self-resonance and parasitic capacitance create impedance peaks that can actually amplify noise at certain frequencies. Designers must account for these non-ideal characteristics when selecting components and predicting filter performance.
Implement enough headroom for the parasitic ESR of the EMI filter. The resistance of filter components contributes to power dissipation and can impact efficiency, particularly in high-current applications. Low-ESR capacitors and low-DCR inductors minimize these losses while maintaining filtering effectiveness.
Filter Design Tools and Validation
Please note the LTpowerCAD filter tool is an estimation tool, providing an initial design point for EMI filters. Nothing can replace a real lab test of a prototype supply board for truly accurate EMI data. While simulation tools provide valuable guidance during the design phase, empirical testing remains essential for validating EMI performance.
Pre-compliance testing using near-field probes and spectrum analyzers allows designers to identify EMI issues early in the development process, when changes are less costly. These measurements help pinpoint noise sources, evaluate the effectiveness of mitigation techniques, and guide iterative design improvements. Final compliance testing in accredited EMI test facilities provides the definitive assessment of whether the design meets regulatory requirements.
PCB Layout Optimization for EMI Reduction
PCB layout exerts profound influence on EMI performance, often determining whether a design achieves compliance without requiring additional filtering or shielding. Careful attention to current loop areas, trace routing, grounding strategy, and component placement can reduce EMI by 10-20 dB or more compared to poorly optimized layouts.
Minimizing High-Frequency Current Loops
Minimize high-current loop areas and parasitic capacitance of nodes with high dv/dt. The area enclosed by high-frequency current loops directly determines the magnitude of radiated emissions, following the relationship that radiated field strength is proportional to loop area, current magnitude, and the square of frequency.
The origin of an electromagnetic field is typically a current containing high-frequency harmonics that are flowing in a loop. Reducing the loop area, the di/dt, or the current peak amplitude will help reduce radiated noise. Identifying the critical current loops—particularly those carrying switching currents with fast edges—and minimizing their area represents one of the most effective EMI reduction techniques available to PCB designers.
Reducing switching power-loop parasitic inductances and minimizing input trace routing contribute to a lower EMI signature. Input pinout optimization also reduces switch-node ringing, output voltage noise, and EMI. The input decoupling loop, which carries the full switching current with fast transitions, demands particular attention. Placing high-frequency ceramic capacitors as close as possible to the switching devices, with wide, short traces or planes for connections, minimizes loop inductance and associated EMI.
Strategic Component Placement
Component placement should prioritize minimizing the length and area of traces carrying high-frequency switching currents. Power switches, input capacitors, and freewheeling diodes should be placed in close proximity to minimize the switching loop. Output filter components should be positioned to create compact current paths while maintaining adequate clearance from sensitive analog circuitry.
The packaging of SMPS ICs is an incredibly important factor in its EMI performance. Packages optimized for EMI have reduced power-loop parasitic inductance, which reduces switch-node ringing. Modern power IC packages increasingly incorporate features specifically designed to minimize EMI, such as optimized pin arrangements, integrated decoupling capacitors, and flip-chip interconnections that eliminate bond wire inductance.
Current flowing through a copper trace generates a magnetic field, which results in an overall increase in EMI noise measurements. The HotRod package pinout is designed to have the input current loop split onto either side of the device. This symmetrical input and ground pins create an equal and opposite magnetic field, which provides a self-containing effect on the magnetic field and further reduces EMI. This principle of using symmetrical current paths to create field cancellation can be applied at both the package and PCB level.
Grounding and Layer Stackup
A solid, continuous ground plane provides a low-impedance return path for high-frequency currents and serves as a reference for signal integrity. Multi-layer boards with dedicated ground planes significantly outperform two-layer designs in EMI performance, though they increase manufacturing cost. The ground plane should extend under all critical components and traces, with minimal interruptions that could force currents to take circuitous paths.
Layer stackup configuration affects both EMI and signal integrity. Placing high-speed signal layers adjacent to ground planes provides natural shielding and controlled impedance. Power planes should be paired with ground planes to minimize power distribution network impedance and provide decoupling capacitance through the plane-to-plane capacitance.
Via placement and stitching techniques help maintain ground plane continuity and provide low-inductance connections between layers. Ground vias should be placed liberally around the perimeter of the board and near high-frequency components to minimize ground impedance and prevent ground plane resonances that could amplify EMI.
Trace Routing Best Practices
High dv/dt nodes, particularly the switch node in buck converters or the drain connection in flyback converters, should be kept as small as possible and routed away from sensitive circuits. These nodes couple capacitively to nearby conductors, injecting noise that can propagate throughout the system. Minimizing the copper area of these nodes and providing adequate clearance to other traces reduces capacitive coupling.
Differential signal pairs should be routed with matched lengths and tight coupling to maximize common-mode rejection. Guard traces or ground fills between sensitive analog signals and noisy digital or power traces provide isolation and reduce crosstalk. However, guard traces must be properly terminated to ground at multiple points to be effective; floating guard traces can actually worsen coupling in some cases.
Return current paths deserve careful consideration, as high-frequency currents naturally follow the path of least impedance, which is typically directly beneath the signal trace. Interruptions in the return path force currents to detour, increasing loop area and EMI. When signals must cross split planes or change layers, closely spaced stitching vias should provide a low-impedance return path.
Shielding Techniques
A copper shield placement below the DM filter may help. When all else fails, add copper layers to the top and bottom of the PCB plus two vertical copper shields on either side of the filter. In (b), the filter is atop a double-copper-layer (top and bottom) PCB and two vertical copper shields, with 3.5 mm between each of the filter components. Shielding provides a last line of defense when other EMI mitigation techniques prove insufficient, though it adds cost and complexity.
Effective shielding requires complete enclosure with proper grounding at multiple points. Gaps or seams in the shield can actually create slot antennas that radiate more effectively than the unshielded circuit. Shield grounding must provide low impedance at the frequencies of concern, which may require multiple ground connections or conductive gaskets at shield seams.
Local shielding of specific components or circuit sections can be more cost-effective than full enclosure. Metal cans over switching regulators, shielded inductors, and grounded copper pours surrounding sensitive circuits all contribute to EMI reduction. The effectiveness of local shielding depends on proper grounding and ensuring that the shield does not create new coupling paths or resonances.
Advanced Slew Rate Control Techniques
Controlling the rate of voltage and current change during switching transitions provides a direct method of reducing high-frequency harmonic content and EMI. While slower transitions increase switching losses, the trade-off often proves worthwhile when EMI compliance is challenging or when the benefits of reduced filtering outweigh the efficiency penalty.
Gate Drive Optimization
A higher RBOOT resistance yields slower switch-node rise times. Accurately controlling the rise time of the switch-node voltage makes it possible to precisely control the switch-node harmonics’ roll-off frequency, which effectively improves the noise amplitude measurement. Gate resistor selection provides a simple, low-cost method of adjusting switching speed to balance EMI and efficiency.
The gate drive circuit determines how quickly the MOSFET or IGBT transitions between on and off states. Stronger gate drivers with lower source impedance produce faster switching, reducing switching losses but increasing EMI. Weaker gate drivers slow the transitions, reducing high-frequency harmonics at the cost of increased switching losses and potentially higher device temperatures.
In some applications, true slew-rate control can eliminate the need for shielding and common-mode chokes, which further reduces total solution size. Depending on the application, there is a slight trade-off in that efficiency, but true slew-rate control can yield a solution that is CISPR 25 Class 5-compliant. Advanced gate drivers with integrated slew rate control provide precise adjustment of switching speed, enabling optimization for specific applications and operating conditions.
Snubber Circuits
Snubber circuits dampen voltage and current oscillations during switching transitions, reducing both EMI and stress on switching devices. RC snubbers across switching devices absorb energy from parasitic resonances, preventing the voltage ringing that generates high-frequency noise. The resistor dissipates the energy while the capacitor provides a low-impedance path for high-frequency currents.
RCD snubbers add a diode to prevent the snubber from interfering with normal circuit operation, activating only during switching transients. These circuits prove particularly effective in flyback converters and other topologies with significant leakage inductance. Proper snubber design requires careful selection of component values to achieve adequate damping without excessive power dissipation.
Active snubbers use controlled switches to recover and recycle the energy that would otherwise be dissipated in passive snubbers, improving efficiency while maintaining EMI benefits. These more complex circuits find application in high-power systems where the energy savings justify the additional cost and control complexity.
Diode Selection and Reverse Recovery
Diode reverse recovery characteristics significantly impact EMI in many power converter topologies. When a diode transitions from conducting to blocking, stored charge must be removed, creating a reverse recovery current spike with fast di/dt that generates high-frequency noise. Standard recovery diodes exhibit abrupt, hard recovery with severe current spikes, while soft recovery diodes provide more gradual transitions with reduced EMI.
Schottky diodes, with their majority carrier conduction mechanism, exhibit minimal reverse recovery and represent the preferred choice for high-frequency applications where their voltage rating suffices. Silicon carbide Schottky diodes extend this advantage to higher voltage applications, combining negligible reverse recovery with high temperature capability and low forward voltage drop.
Synchronous rectification, replacing diodes with actively controlled MOSFETs, eliminates reverse recovery issues entirely while improving efficiency. The MOSFET’s body diode may still conduct briefly during dead time, but proper timing control minimizes this conduction and its associated recovery effects. Synchronous rectification has become standard in low-voltage, high-current applications and increasingly common across a wide range of power levels.
Topology Selection for Inherent EMI Advantages
Different power converter topologies exhibit inherently different EMI characteristics based on their switching patterns, current waveforms, and voltage stresses. Selecting a topology with favorable EMI properties can simplify compliance and reduce filtering requirements, though topology choice must also consider efficiency, cost, complexity, and other application-specific factors.
Continuous vs. Discontinuous Conduction Mode
Continuous conduction mode (CCM) operation maintains current flow through the inductor throughout the switching cycle, producing relatively smooth current waveforms with lower high-frequency content. Discontinuous conduction mode (DCM) allows the inductor current to reach zero during each cycle, creating current pulses with sharper edges and higher harmonic content. From an EMI perspective, CCM generally proves superior, though DCM offers advantages in certain applications such as power factor correction.
Boundary conduction mode (BCM) or critical conduction mode operates at the boundary between CCM and DCM, switching when the inductor current reaches zero. This variable-frequency operation provides some of the benefits of both modes while introducing frequency variation that can help spread the EMI spectrum. However, the wide frequency variation in BCM can complicate filter design and may create audible noise in some applications.
Isolated vs. Non-Isolated Topologies
Isolated topologies using transformers provide galvanic isolation between input and output, which can reduce common-mode noise coupling and simplify EMI filter design in some applications. The transformer’s leakage inductance can be leveraged for soft switching in resonant and quasi-resonant designs. However, transformer parasitics also create challenges, including interwinding capacitance that couples high-frequency noise across the isolation barrier.
Choose topology that creates low EMI (e.g. QR flyback for lower powers, LLC for higher power applications). Quasi-resonant flyback converters combine the simplicity and low cost of the flyback topology with valley switching that reduces EMI. LLC resonant converters achieve excellent efficiency and EMI performance through inherent soft switching across a wide load range.
Non-isolated buck, boost, and buck-boost converters offer simplicity and high efficiency but may require more extensive filtering to achieve EMI compliance. The direct connection between input and output provides a path for both differential-mode and common-mode noise, requiring careful filter design to meet conducted emission limits.
Multi-Phase and Interleaved Converters
Multi-phase converters operate multiple switching stages in parallel with phase-shifted timing, distributing the current among several inductors and switches. This approach reduces current ripple in the input and output capacitors, which can significantly decrease both conducted and radiated EMI. The interleaved switching also creates partial cancellation of ripple currents, with the degree of cancellation depending on the number of phases and the precision of phase alignment.
Interleaved operation spreads the switching energy across multiple frequency components, potentially reducing peak emissions at any single frequency. However, the multiple switching events also create more opportunities for EMI generation, and careful layout becomes even more critical to prevent coupling between phases. The EMI benefits of interleaving are most pronounced when the phases are precisely synchronized and the layout maintains symmetry among the parallel paths.
Resonant and Quasi-Resonant Topologies
Another positive side effect of quasi-resonant switching is that because switching is triggered when a valley is detected, rather than at a fixed frequency, a degree of frequency jitter is introduced, spreading the RF emission spectrum and further reducing EMI. Soft switching at zero voltage Quasi-resonant switching is a good technique for improving voltage-converter efficiency. The inherent soft switching and frequency variation of quasi-resonant converters provide EMI advantages without requiring additional control complexity.
Full resonant converters, including series resonant, parallel resonant, and LLC topologies, achieve soft switching through resonant tank circuits that shape the voltage and current waveforms. These converters can operate at very high frequencies while maintaining excellent efficiency and low EMI. The sinusoidal current waveforms characteristic of resonant operation contain far less harmonic content than the square waves of hard-switched PWM converters.
The primary disadvantages of resonant converters include increased complexity, higher component count, and more challenging control design. The resonant tank components must be carefully selected and matched to achieve the desired performance across the operating range. Despite these challenges, resonant topologies increasingly dominate applications where high efficiency, high power density, and excellent EMI performance are required.
Measurement and Diagnostic Techniques
Effective EMI mitigation requires accurate measurement and diagnosis of noise sources and coupling paths. Pre-compliance testing during development enables iterative improvement before formal compliance testing, reducing development time and cost. Understanding measurement techniques and interpreting results guides design decisions and validates the effectiveness of mitigation strategies.
Conducted Emission Measurement
To quantify conducted input EMI, a line impedance stabilization network (LISN) is placed at the regulator’s input, providing a standard input source impedance. The LISN serves dual purposes: it presents a defined impedance to the device under test across the measurement frequency range, and it isolates the test setup from variations in the AC mains impedance that would otherwise affect measurement repeatability.
Conducted emission measurements typically employ both quasi-peak and average detectors, as specified by regulatory standards. Quasi-peak detection weights repetitive signals more heavily than random noise, reflecting the subjective annoyance factor of different interference types. Average detection provides a measure of the mean power in the emissions. Both measurements must comply with applicable limits for the product to pass certification.
Pre-compliance measurements can be performed with less expensive equipment and in non-ideal test environments, providing guidance during development. However, final compliance testing must be conducted in accredited test facilities with calibrated equipment and proper test chambers to ensure accurate, legally defensible results.
Radiated Emission Measurement
Radiated emission testing measures the electromagnetic fields generated by the device under test at specified distances, typically 3 or 10 meters depending on the standard. Testing occurs in anechoic chambers or open-area test sites that minimize reflections and external interference. Measurements are performed with calibrated antennas across the required frequency range, typically 30 MHz to 1 GHz or higher for some applications.
Perform first radiated EMI test to identify potential “candidates” for radiated fields. Early radiated emission testing, even in non-ideal environments, helps identify problem areas that require attention. Near-field scanning with magnetic and electric field probes can pinpoint specific circuit areas or components responsible for excessive radiation, guiding targeted mitigation efforts.
Near-Field Probing and Diagnosis
Near-field probes enable non-invasive measurement of magnetic and electric fields in close proximity to the circuit under test. These tools prove invaluable for identifying EMI sources, evaluating the effectiveness of layout changes, and optimizing component placement. Magnetic field probes (H-field) respond to current loops and are particularly useful for identifying high di/dt paths, while electric field probes (E-field) respond to voltage gradients and high dv/dt nodes.
Three-dimensional probes with orthogonal sensing elements provide omnidirectional sensitivity, simplifying the scanning process by eliminating the need to orient the probe for maximum response. One-dimensional probes offer better spatial resolution and directional selectivity, helping to precisely locate noise sources once problem areas have been identified with broader scanning.
Time-domain measurements with oscilloscopes synchronized to the switching frequency reveal the temporal characteristics of EMI, showing how noise correlates with specific switching events. Frequency-domain measurements with spectrum analyzers display the harmonic content and identify which frequencies exceed limits. Combining time and frequency domain analysis provides comprehensive understanding of EMI behavior and guides effective mitigation.
Simulation and Modeling
EMI simulation tools enable prediction of conducted and radiated emissions during the design phase, before hardware is available for testing. These tools model the switching waveforms, parasitic elements, and coupling paths to estimate the EMI spectrum. While simulation accuracy depends on the quality of component models and the completeness of parasitic extraction, even approximate predictions provide valuable guidance for design decisions.
Duan et al. modeled the EMI source and coupling path of the conducted noise of a full SiC three-phase inverter, and predicted the conducted noise at the power port by using time-domain simulation and fast Fourier transform. In the frequency range of 10 kHz–30 MHz, the simulation results were basically consistent with the measured results. Accurate modeling enables virtual design iteration, reducing the number of physical prototypes required and accelerating development.
Electromagnetic field solvers can predict radiated emissions based on PCB layout and enclosure geometry, identifying potential radiation hot spots before fabrication. These tools require significant computational resources and expertise to use effectively, but provide insights that are difficult or impossible to obtain through measurement alone. The combination of simulation and measurement provides the most comprehensive approach to EMI analysis and mitigation.
Regulatory Standards and Compliance Requirements
Power electronic systems must comply with electromagnetic compatibility (EMC) regulations that vary by region, application, and product category. Understanding applicable standards and their requirements guides design decisions and ensures that products can be legally sold in target markets. Compliance testing represents a significant cost and schedule factor in product development, making early attention to EMI critical for project success.
International EMC Standards
Many industries are using systems that require increasingly careful control of transmitted electromagnetic signals. To this end, there are a number of published clear standards on EMI. The International Electrotechnical Commission (IEC) publishes CISPR standards that form the basis for EMC regulations in many countries. CISPR 11 covers industrial, scientific, and medical equipment, while CISPR 22 (now superseded by CISPR 32) addresses information technology equipment.
The European Union enforces EMC requirements through the EMC Directive, which mandates compliance with harmonized standards such as EN 55032 for emissions and EN 55024 for immunity. Products must bear the CE mark to indicate conformity with applicable directives before they can be sold in EU member states. The United States regulates EMC through the Federal Communications Commission (FCC), with Part 15 covering unintentional radiators including most power electronic equipment.
Automotive applications face particularly stringent EMC requirements due to the safety-critical nature of vehicle systems and the harsh electromagnetic environment. CISPR 25 specifies limits and test methods for automotive components, with different classes reflecting varying levels of stringency. Meeting Class 5 requirements, the most stringent level, demands careful attention to all aspects of EMI mitigation.
Emission Limits and Classes
EMC standards typically define different emission limit classes based on the intended operating environment. Class A limits apply to equipment intended for industrial environments, where higher emission levels are tolerated due to greater separation from sensitive receivers and the presence of other industrial equipment. Class B limits, approximately 10 dB more stringent, apply to equipment intended for residential environments where protection of broadcast reception and other sensitive uses is paramount.
Conducted emission limits typically apply from 150 kHz to 30 MHz, covering the frequency range where power line conducted interference is most problematic. Radiated emission limits extend from 30 MHz to 1 GHz or higher, addressing far-field radiation that can interfere with radio communications and other wireless services. Some standards include additional requirements for specific frequency ranges or applications, such as protection of cellular bands or GPS frequencies.
Margin to the regulatory limits provides insurance against unit-to-unit variation, component tolerance, and aging effects. Designs that barely pass compliance testing in the laboratory may fail in production or field use due to these variations. Experienced designers target margins of 6 dB or more to ensure robust compliance across production volumes and product lifetime.
Immunity Requirements
In addition to limiting emissions, EMC standards require that equipment operate correctly in the presence of external electromagnetic disturbances. Immunity testing subjects the device to various types of interference, including electrostatic discharge (ESD), electrical fast transients (EFT), surge, conducted RF, and radiated RF. The equipment must continue to operate within specifications or fail gracefully without damage or safety hazards.
Immunity requirements often drive design decisions that also benefit emissions performance. Robust filtering, careful grounding, and shielding that protect against external interference also reduce emissions. However, some immunity mitigation techniques, such as transient voltage suppressors and surge protection devices, have minimal impact on emissions performance and must be addressed separately.
System-level immunity testing evaluates the complete product in its intended configuration, including cables, enclosures, and interconnected equipment. Component-level immunity testing during development helps ensure that individual subsystems meet requirements, but cannot replace system-level validation. The interaction between subsystems and the effects of realistic cable routing and grounding can only be fully evaluated at the system level.
Practical Design Examples and Case Studies
Examining specific design examples illustrates how the principles and techniques discussed above are applied in real-world applications. These case studies demonstrate the iterative nature of EMI mitigation and the importance of systematic approaches to achieving compliance.
High-Frequency Buck Converter Optimization
A 12V to 1.5V, 10A synchronous buck converter operating at 500 kHz presents typical EMI challenges for modern point-of-load applications. The high switching frequency enables use of a compact 1 µH inductor and small ceramic capacitors, but creates significant conducted and radiated emissions. Initial testing reveals conducted emissions exceeding Class B limits by 15 dB at the fundamental switching frequency and lower harmonics.
The first mitigation step involves PCB layout optimization, minimizing the input current loop by placing the input capacitors directly adjacent to the high-side MOSFET with wide, short traces. This change alone reduces emissions by 8 dB at the fundamental frequency. Adding a small RC snubber across the high-side MOSFET dampens voltage ringing, providing an additional 4 dB reduction at higher harmonics where ringing contributes significantly to emissions.
Implementing spread spectrum frequency modulation with ±3% deviation spreads the energy at the fundamental and harmonics, reducing peak emissions by 6-8 dB. The combination of layout optimization, snubbing, and frequency spreading achieves compliance with 3-5 dB margin without requiring input filtering beyond the decoupling capacitors already present for functional purposes. This example demonstrates how multiple small improvements combine to achieve significant overall EMI reduction.
Flyback Converter EMI Mitigation
An offline flyback converter for a 65W USB-C power adapter operates at 65 kHz under full load, increasing to over 100 kHz at light load in quasi-resonant mode. The wide frequency variation provides some spectral spreading, but conducted emissions still exceed Class B limits by 10-12 dB in the 150-500 kHz range. Common-mode emissions dominate, driven by the high dv/dt at the drain of the primary MOSFET coupling through transformer interwinding capacitance.
A two-stage EMI filter is implemented, with the first stage consisting of a common-mode choke and Y-capacitors to address the dominant common-mode noise. The second stage adds differential-mode filtering with an X-capacitor and small differential-mode inductor. The filter achieves 30-40 dB attenuation in the critical frequency range, providing comfortable margin to the limits.
Radiated emissions testing reveals excessive radiation from the transformer and primary-side circuitry. Adding a grounded copper shield on the PCB beneath the transformer and optimizing the snubber circuit across the primary MOSFET reduces radiated emissions by 8-10 dB, achieving compliance. The final design meets all requirements with a compact, cost-effective solution suitable for high-volume production.
LLC Resonant Converter for Server Power Supply
A 1.5 kW LLC resonant converter for server applications operates at 100-200 kHz depending on load, achieving >95% efficiency through soft switching across the full operating range. The inherent soft switching provides excellent EMI performance, with conducted emissions 15-20 dB below Class A limits without input filtering. However, the high power level and compact design create thermal challenges that must be addressed without compromising EMI performance.
Measurements on a 500kHz LLC converter showed 15dB EMI reduction after implementing: The design achieved CISPR 32 Class B compliance without additional shielding. This example demonstrates how topology selection and careful implementation can achieve excellent EMI performance with minimal filtering, reducing cost and complexity while meeting stringent efficiency targets.
The resonant tank components require careful selection to maintain soft switching across the load range while minimizing circulating currents that increase conduction losses. High-quality film capacitors with low ESR and inductors with optimized core materials ensure that the resonant tank operates efficiently without generating excessive heat. The result is a compact, high-efficiency power supply that easily meets all EMC requirements.
Future Trends in EMI Mitigation
The continuing evolution of power electronics technology drives ongoing development of new EMI mitigation techniques and approaches. Understanding emerging trends helps designers prepare for future challenges and opportunities in power system design.
Integration and Advanced Packaging
Increasing integration of power conversion functions into single packages reduces parasitic inductances and capacitances that contribute to EMI. Power modules that integrate switches, drivers, and passive components in optimized 3D structures achieve EMI performance that would be difficult or impossible with discrete implementations. Advanced packaging technologies such as embedded die, flip-chip interconnections, and integrated passives enable this integration while maintaining thermal performance and manufacturability.
Modern implementations integrate ZVS with wide-bandgap devices (GaN, SiC) to push efficiency boundaries in high-frequency power conversion. The combination of advanced semiconductors and optimized packaging enables power converters that operate at multi-MHz frequencies while maintaining excellent efficiency and EMI performance. These developments enable dramatic reductions in size and weight for applications ranging from consumer electronics to electric vehicles and renewable energy systems.
Digital Control and Adaptive Techniques
Digital controllers enable sophisticated EMI mitigation strategies that adapt to operating conditions in real time. Adaptive frequency modulation can optimize the modulation pattern based on the measured EMI spectrum, concentrating mitigation efforts where they are most needed. Adaptive dead-time control maintains soft switching across varying load and line conditions, maximizing the EMI benefits of resonant and quasi-resonant operation.
Machine learning algorithms may eventually enable predictive EMI mitigation, learning the EMI characteristics of specific designs and automatically adjusting control parameters to minimize emissions. While still largely in the research phase, these techniques promise to simplify EMI compliance and enable more aggressive optimization of efficiency and power density.
Active EMI Cancellation
Active EMI cancellation techniques that inject anti-phase noise currents to cancel conducted emissions are becoming more practical as the required power electronics become smaller and less expensive. These techniques can provide 20-30 dB of additional attenuation in targeted frequency ranges, potentially eliminating the need for bulky passive filters. The challenge lies in achieving sufficient bandwidth, accuracy, and stability to maintain cancellation across varying operating conditions and component tolerances.
Hybrid approaches combining passive filtering for broadband attenuation with active cancellation for specific problematic frequencies may offer the best balance of performance, cost, and reliability. As active filter technology matures and costs decrease, these techniques will likely see increasing adoption in applications where size and weight are critical constraints.
Standardization and Harmonization
Ongoing efforts to harmonize EMC standards across regions and applications simplify compliance for products sold in multiple markets. The transition from product-specific standards to more generic standards based on operating environment rather than product category reflects this trend. However, some applications, particularly automotive and aerospace, will likely maintain specialized requirements reflecting their unique safety and reliability needs.
Tightening emission limits in response to increasing electromagnetic spectrum congestion and proliferation of wireless services will continue to challenge power electronics designers. Techniques that provide comfortable margins to current limits may barely meet future requirements, driving continued innovation in EMI mitigation. Proactive design practices that exceed current requirements provide insurance against future regulatory changes and ensure long product lifetimes.
Conclusion and Best Practices Summary
Optimizing switching frequencies for reduced EMI in power electronic systems requires a comprehensive, systematic approach that addresses EMI generation, propagation, and coupling at every level of the design. Success depends on understanding the fundamental mechanisms of EMI generation, applying appropriate mitigation techniques, and validating performance through careful measurement and testing.
Key best practices include selecting switching frequencies that balance component size, efficiency, and EMI performance for the specific application; implementing soft-switching techniques where feasible to reduce EMI at the source; employing frequency modulation strategies to spread spectral energy and reduce peak emissions; designing PCB layouts that minimize high-frequency current loops and parasitic coupling; selecting appropriate filter topologies and components to attenuate conducted emissions; and validating designs through pre-compliance testing early in development.
Addressing emerging EMI issues in modern power electronic device-based converters is essential for ensuring safe and reliable operations. Through strategic design optimization and the implementation of EMI mitigation strategies, modern converters can seamlessly be integrated into diverse applications, offering improved EMI performance as a hallmark of their versatility.
The iterative nature of EMI mitigation demands early attention to electromagnetic compatibility in the design process. Addressing EMI as an afterthought inevitably leads to costly redesigns, schedule delays, and suboptimal solutions. Integrating EMC considerations from the earliest conceptual design phases through final validation ensures efficient development and robust products that meet all requirements with adequate margin.
As power electronics continue to evolve toward higher frequencies, higher power densities, and more demanding applications, EMI mitigation will remain a critical design challenge. The techniques and principles discussed in this article provide a foundation for addressing these challenges, but successful implementation requires experience, attention to detail, and willingness to iterate toward optimal solutions. By combining theoretical understanding with practical measurement and systematic optimization, engineers can design power electronic systems that achieve excellent performance while meeting all electromagnetic compatibility requirements.
For additional information on power electronics design and EMI mitigation techniques, consult resources from organizations such as the Institute of Electrical and Electronics Engineers (IEEE), the International Electrotechnical Commission (IEC), and the Power Sources Manufacturers Association (PSMA). These organizations provide standards, technical publications, and educational resources that support ongoing professional development in power electronics and electromagnetic compatibility.