Conducted emissions
Conducted emissions refer to electromagnetic noise generated by an electronic device that propagates along conductive paths, such as power cords, signal cables, or PCB traces, potentially coupling into external networks and causing interference with other equipment.[1] These emissions are a fundamental concern in electromagnetic compatibility (EMC), which ensures that devices operate without unacceptable degradation in their intended electromagnetic environment.[2] Unlike radiated emissions, which travel through free space as electromagnetic fields at higher frequencies, conducted emissions are predominant below 30 MHz, where the wavelength exceeds typical conductor lengths (e.g., around 10 meters at 30 MHz), making conduction the primary propagation mode.[3]Definition and Fundamentals
Definition of Conducted Emissions
Conducted emissions refer to the electromagnetic interference (EMI) generated by electronic devices that propagates along conductive pathways, such as power lines, signal cables, or ground planes, rather than through free space.[4] These emissions consist of unwanted noise components, typically in the form of voltage or current disturbances, that couple into conductors and can affect connected equipment or the power grid.[3] The basic principles of conducted emissions involve coupling mechanisms where electromagnetic energy transfers from a source to a conductor through direct galvanic connections, capacitive (electric field) coupling, or inductive (magnetic field) coupling.[5] This distinguishes conducted emissions from other forms of interference, as the energy is guided along physical media instead of radiating freely.[1] Harmonics from nonlinear loads, such as those in power supplies, represent a specific subset of conducted emissions that contribute to these disturbances.[6] The study of conducted emissions originated in early 20th-century efforts to mitigate interference in telegraph and radio systems, where electrical noise along wires disrupted communications.[7] These issues gained formal recognition in the electromagnetic compatibility (EMC) field following World War II, as the proliferation of electronic devices necessitated standardized approaches to control interference in civilian and military applications.[8] Conducted emissions are primarily characterized in the frequency spectrum from 150 kHz to 30 MHz, where high-frequency components are most relevant for regulatory testing.[9] Power quality issues like harmonics and flicker, which occur at lower frequencies below 150 kHz, are addressed under separate standards, while examples in the 150 kHz to 30 MHz range include switching noise from power supplies that generates broadband interference.[10]Distinction from Radiated Emissions
Conducted emissions propagate through physical conductors such as power lines, signal cables, and printed circuit board (PCB) traces, allowing electromagnetic interference (EMI) to travel directly along these pathways as conducted currents or voltages.[11] In contrast, radiated emissions propagate through free space as electromagnetic waves, detached from any physical connection and capable of affecting distant systems without direct contact.[12] This fundamental difference in propagation paths means that conducted emissions are confined to wired infrastructure, while radiated emissions can permeate environments broadly, necessitating distinct mitigation strategies in electromagnetic compatibility (EMC) design. The coupling mechanisms further delineate the two: conducted emissions couple via direct electrical connections or close-proximity effects like parasitic capacitance, enabling EMI to inject into adjacent circuits through shared conductors.[11] Radiated emissions, however, couple through antenna-like effects where device structures—such as enclosures, cables, or PCB edges—act as unintentional radiators, converting internal noise into propagating fields that induce currents in nearby receivers.[12] These mechanisms highlight why conducted emissions primarily threaten interconnected systems via conductive paths, whereas radiated emissions pose risks to wireless and unshielded environments through inductive or capacitive field interactions. Although both types of emissions can occur within the overlapping frequency range of 150 kHz to 30 MHz, their boundaries are defined by measurement approaches: conducted emissions are assessed at device interfaces like power ports using line impedance stabilization networks (LISNs), focusing on cable-conducted noise.[13] Radiated emissions, typically evaluated from 30 MHz upward but sometimes extending into lower frequencies in near-field conditions, are measured at standardized distances such as 3 meters or 10 meters using antennas to capture field strengths in open-area test sites or semi-anechoic chambers.[14] This overlap underscores the potential for emissions below 30 MHz to manifest in both forms, but the distinction lies in their evaluation—interface-based for conducted versus distant-field for radiated—ensuring comprehensive EMC compliance. In terms of EMC implications, conducted emissions directly compromise wired networks and power quality by injecting noise into shared infrastructure, often requiring filters at entry points to prevent upstream propagation.[15] Radiated emissions, by contrast, disrupt wireless communications and sensitive receivers in open environments, where enclosure resonances can amplify otherwise benign internal noise into failing levels; for instance, an electric vehicle camera module might pass conducted tests on its power leads but fail radiated limits at 222 MHz due to the vehicle's door acting as a resonant half-wave antenna, necessitating additional shielding like conductive tape to suppress the effect by over 20 dB.[16] Such scenarios illustrate how a device compliant in conducted domains may still violate radiated regulations, emphasizing the need for holistic EMC engineering that addresses both propagation modes.Sources and Mechanisms
Primary Sources in Electronic Systems
Switching power supplies (SMPS) represent one of the primary sources of conducted emissions in electronic systems due to their high-frequency switching operations, which generate broadband noise that couples into power lines. These devices, commonly used for efficient AC-DC and DC-DC conversion in computers, consumer electronics, and industrial equipment, produce both common-mode and differential-mode noise from rapid voltage and current transitions during switching cycles. For instance, unfiltered SMPS operating at frequencies around 50-100 kHz can exhibit conducted emission levels of 50–100 dBµV at 1 MHz on the AC mains, often exceeding regulatory limits without proper filtering.[17][18] Digital circuits, including clocks and microprocessors, contribute significantly to conducted emissions through harmonic generation that couples into power rails and signal lines. High-speed clock signals in integrated circuits, such as those in microcontrollers operating at 10-100 MHz, produce odd harmonics due to fast rise and fall times, which propagate as noise on supply voltages and ground planes. These harmonics can reach levels that interfere with compliance testing, particularly when the clock energy spreads to lower-order frequencies via parasitic capacitances.[19][20] Motors and drives, especially those employing pulse-width modulation (PWM) techniques, are another key source, as the switching actions in inverters create high dv/dt and di/dt transients that inject noise into power cables. In variable-speed AC motor drives, PWM operation at carrier frequencies of 5-20 kHz generates broadband conducted emissions, with common-mode currents flowing through motor cables and grounding paths. This is prevalent in industrial automation and electric vehicles, where unmitigated systems can produce emission spectra peaking in the 150 kHz to 30 MHz range.[21] RF transmitters also originate conducted emissions, primarily through modulation sidebands and harmonics that conduct back into the power supply lines if isolation is inadequate. In communication devices like wireless modules, the RF signal's envelope variations cause fluctuating currents on DC power rails, leading to conducted noise that can couple to AC mains via the power supply.[22] Environmental factors such as ground loops and poor cabling can amplify emissions from sources like LED drivers and USB interfaces by creating unintended current paths. Ground loops, arising from multiple grounding points with differing potentials, allow noise from switching in LED drivers—often operating at 100-500 kHz—to circulate and increase conducted levels on data or power lines. Similarly, unshielded or lengthy USB cables act as antennas for common-mode noise from host or peripheral switching, exacerbating emissions in interconnected systems.[23][24]Generation Mechanisms
Conducted emissions arise primarily through two distinct coupling mechanisms: differential-mode and common-mode. Differential-mode coupling occurs when noise currents flow in opposite directions between the two conductors of a circuit, such as the phase and neutral lines in a power supply, effectively representing the normal operational current imbalance that propagates along the line.[25] In contrast, common-mode coupling involves noise currents flowing in the same direction on both conductors relative to ground, often resulting from unintended return paths through parasitic elements or the earth ground.[26] These can be visualized in a simple diagram where, for differential-mode, arrows indicate opposing currents (I_D = (I_P - I_N)/2) between phase (I_P) and neutral (I_N) wires, while for common-mode, parallel arrows show equal currents (I_C = (I_P + I_N)/2) on both relative to ground.[1] The physical principles underlying these emissions stem from rapid changes in voltage (high dV/dt) or current (high dI/dt) within electronic circuits, which introduce high-frequency components into the signal spectrum. Such transients, common in switching operations, generate displacement currents through parasitic capacitances for dV/dt effects, primarily driving common-mode noise via electric field coupling, and magnetic field coupling for dI/dt effects, which dominate differential-mode noise through loop inductances.[27] For instance, in switching power supplies, square-wave-like voltage transitions decompose into a series of harmonics via Fourier analysis, where the waveform's sharp edges produce significant energy at odd multiples of the fundamental frequency, extending the spectrum into the MHz range and contributing to conducted emissions.[28] To illustrate, consider a periodic square wave of amplitude A and period T, defined as f(t) = A for $0 \leq t < T/2 and f(t) = -A for T/2 \leq t < T. The Fourier series representation is given by: f(t) = \sum_{n=1,3,5,\dots}^{\infty} A_n \sin\left(\frac{2\pi n t}{T}\right), where the harmonic amplitudes A_n for odd n are approximated as: A_n = \frac{4A}{\pi n}. This approximation arises from the Fourier coefficients, specifically the sine terms b_n, since the square wave is an odd function with no cosine components. The coefficient b_n is derived as: b_n = \frac{2}{T} \int_0^{T} f(t) \sin\left(\frac{2\pi n t}{T}\right) dt = \frac{4}{T} \int_0^{T/2} A \sin\left(\frac{2\pi n t}{T}\right) dt, evaluating the integral using \int \sin(kt) dt = -\frac{1}{k} \cos(kt), where k = 2\pi n / T: b_n = \frac{4A}{T} \left[ -\frac{T}{2\pi n} \cos\left(\frac{2\pi n t}{T}\right) \right]_0^{T/2} = \frac{4A}{T} \left( -\frac{T}{2\pi n} [\cos(\pi n) - \cos(0)] \right). For odd n, \cos(\pi n) = -1 and \cos(0) = 1, yielding \cos(\pi n) - \cos(0) = -2, so: b_n = \frac{4A}{T} \left( -\frac{T}{2\pi n} (-2) \right) = \frac{4A}{\pi n}. Thus, A_n = b_n, confirming the amplitude decay as $1/n, which concentrates higher-frequency energy and amplifies conducted emissions in the conducted frequency bands (e.g., 150 kHz to 30 MHz).[29][30] Parasitic effects in printed circuit boards (PCBs) further facilitate the injection of these emissions into conductive paths. Unintended capacitances between adjacent traces or layers couple high dV/dt noise capacitively, creating displacement currents that inject common-mode emissions onto power lines, while loop inductances from trace lengths and vias introduce series impedance that resonates with capacitances, amplifying differential-mode currents at high frequencies.[31] For example, a via's parasitic inductance (typically 0.5–1 nH) mismatches trace impedance, causing reflections and noise propagation along the board, directly contributing to conducted emission levels measurable at line impedance stabilization networks (LISNs).[32]Measurement and Compliance
Testing Methods
Testing for conducted emissions involves establishing a controlled environment to measure noise propagated through power lines or interconnecting cables of the equipment under test (EUT). The primary setup configuration utilizes a Line Impedance Stabilization Network (LISN), which is inserted between the power source and the EUT to provide a standardized 50 Ω impedance across the frequency range of interest, isolating the measurement from variations in the power supply's impedance.[33] For AC mains testing, a single-phase LISN typically includes two channels for line and neutral, while DC testing may require individual LISNs for positive and return lines if the cable length exceeds 20 cm.[33] The standard 50 µH/50 Ω LISN model features a series inductor of 50 µH followed by a 50 Ω resistor to ground, often with a parallel 5 Ω resistor (selectable via jumper) for low-frequency stabilization and a 1 µF capacitor shunting high frequencies below 5 MHz to the RF output port; this configuration ensures repeatable measurements by diverting conducted noise to the measurement instrument while blocking external interference.[33][34] Essential equipment includes an EMI receiver or spectrum analyzer capable of scanning from 150 kHz to 30 MHz with resolution bandwidths of 9 kHz or 10 kHz, quasi-peak and average detectors, and a dynamic range better than -100 dBm.[34] Transient limiters and attenuators (e.g., 10 dB) protect the receiver input, while calibration standards such as signal generators verify system accuracy before testing.[34] Antennas may be used for validation to correlate conducted noise with potential radiated emissions, ensuring the setup does not inadvertently introduce airborne interference.[35] The measurement process begins with preparing the test site: the EUT is placed on a non-conductive table 0.8 m above a horizontal ground plane, with a vertical ground plane 0.4 m behind it, and the LISN bonded to the ground plane.[34] The EUT's power cord is connected to the LISN output without coiling, laid flat to minimize inductive coupling, and the LISN RF port is linked to the EMI receiver via a 50 Ω cable, often with an attenuator.[34] The receiver is configured for an initial peak detector scan from 150 kHz to 30 MHz using a 1 MHz resolution bandwidth to identify potential emissions quickly.[36] Peaks exceeding thresholds are then re-measured with narrower bandwidth (9 kHz) and quasi-peak or average detectors: the quasi-peak detector applies a charge-discharge weighting to simulate human perception of interference, while the average detector provides a linear time average for steady-state noise assessment.[36][37] Measurements are performed on each power line (e.g., line-to-ground and neutral-to-ground) with the EUT operating in its worst-case mode, and the process is repeated for all relevant ports.[35] Common challenges in conducted emissions testing include ambient noise interference, which requires subtraction by measuring the baseline spectrum with the EUT powered off and deducting it from EUT results to isolate device-generated emissions.[38] Repeatability issues often arise from variations in cable positioning or ground plane contacts, leading to discrepancies between test site and laboratory results; for instance, a 2-3 dB shift in peak levels can occur due to minor setup differences, necessitating strict adherence to calibration and environmental controls.[38] Unlike radiated emissions testing, which employs antennas in open-area or semi-anechoic chambers to capture electromagnetic fields, conducted testing isolates line-conducted noise through the LISN without requiring field-strength measurements.[35]Key Standards and Regulations
Conducted emissions are regulated by several key international and regional standards to ensure electromagnetic compatibility in electronic systems. The International Electrotechnical Commission (IEC) through its Comité International Spécial des Perturbations Radioélectriques (CISPR) subcommittee develops primary standards, including CISPR 11 for industrial, scientific, and medical (ISM) equipment, which covers emissions from 9 kHz to 400 GHz, and CISPR 32 for information technology equipment (ITE) and multimedia devices, focusing on emissions from multimedia sources. In the United States, the Federal Communications Commission (FCC) enforces limits under Part 15 Subpart B for unintentional radiators such as digital devices. In Europe, EN 55032 serves as the harmonized standard equivalent to CISPR 32, applying to multimedia equipment under the EMC Directive 2014/30/EU. These standards specify limits for conducted emissions, measured as radio-frequency voltage on power ports using a Line Impedance Stabilization Network (LISN), typically in the 150 kHz to 30 MHz range with quasi-peak (QP) and average (AV) detectors. For Class B devices—intended for residential, commercial, or light-industrial environments—the QP limit at 150 kHz is 66 dBµV on AC mains ports, with limits decreasing logarithmically. Class A limits for industrial or heavy-commercial settings are higher, such as 79 dBµV QP at 150 kHz under CISPR 11. DC power ports often follow similar profiles but may have adjusted thresholds. The following table summarizes representative QP and AV limits for Class B conducted emissions on AC mains ports (150 kHz to 30 MHz), as defined in CISPR 32, EN 55032, and FCC Part 15:| Frequency Range (MHz) | QP Limit (dBµV) | AV Limit (dBµV) |
|---|---|---|
| 0.15–0.50 | 66 to 56* | 56 to 46* |
| 0.50–5.00 | 56 | 46 |
| 5.00–30.0 | 60 | 50 |