EMI Issues in Power Converters - Radiated Emissions

qwqwqw2088

EMI Issues in Power Converters - Radiated Emissions [Copy link]

Radiated electromagnetic interference (EMI) is a problem that arises dynamically in a specific environment and is related to the parasitic effects, circuit layout and component arrangement inside the power converter and the overall system in which it operates. Therefore, from the design engineer's perspective, the problem of radiated EMI is usually more challenging and complex, especially when multiple DC/DC power stages are used on the system motherboard. It is important to understand the basic mechanisms of radiated EMI as well as the measurement requirements, frequency ranges and corresponding restrictions. This article focuses on these aspects and shows the results of the radiated EMI measurement setup and two DC/DC buck converters.

Near Field Coupling

Figure 1 provides an overview of the basic EMI coupling modes between a noise source and a victim circuit. In particular, inductive or H-field coupling requires a high di/dt, time-varying current source and two magnetically coupled loops (or parallel wires with a return path). On the other hand, capacitive or E-field coupling requires a high dv/dt, time-varying voltage source and two metal plates in close proximity. Both mechanisms are near-field coupling, where the noise source and victim circuit are very close and can be measured using a near-field sniffer.

查看详情

Figure 1: EMI coupling patterns

For example, modern power switches, especially Gallium Nitride (GaN) and Silicon Carbide (SiC) based transistors, have low output capacitance COSS _and low gate charge QG _, and can switch at very high dv/dt and di/dt slew rates. The potential for H-field and E-field coupling and crosstalk between adjacent circuits is high. However, as the mutual inductance or capacitance decreases and the spacing of the coupling structures increases, the near-field coupling is significantly reduced.

Far-field coupling

A typical electromagnetic (EM) wave propagates as a combination of an E field and an H field. The field structure near the radiating antenna source is a complex three-dimensional pattern. Analyzing further from the radiating source, the EM wave in the far-field region consists of E-field and H-field components that are orthogonal to each other and to the propagation direction. Figure 2 shows this plane wave, which represents the primary benchmark for radiated EMI, subject to various radiated standards.

查看详情

Figure 2: Electromagnetic plane wave propagation

The wave impedance shown in Figure 3 is equal to the ratio of the electric field strength to the magnetic field strength. The E and H components in the far-field region are in phase, so the far-field impedance is resistive and can be calculated using the plane wave solution to Maxwell's equations (shown in Equation 1):

查看详情If λ is the wavelength and F is the desired frequency, Equation 2 generally represents the boundary between the near-field and far-field regions:

查看详情However, this boundary is not a precise criterion but serves only to indicate the general transition region (depicted as l/16 to 3l in Fig. 3) where the field evolves from a complex distribution morphology to a plane wave.

查看详情

Figure 3: Wave impedance in the near and far field regions of Maxwell's laws

Since most antennas are designed to detect and respond to E-fields, the radiated electromagnetic waves are often referred to as being vertically or horizontally polarized, depending on the direction of the E-field. Antennas measuring E-fields should generally be oriented in the same plane as the propagating E-fields to detect maximum field strength. Therefore, radiated EMI test standards often describe measurements when the receiving antenna is mounted in vertical and horizontal polarizations.

Radiated EMI in Industrial and Multimedia Equipment

Table 1 lists the Class A and Class B radiated emission limits for unintentional radiators as specified by the Federal Communications Commission (FCC) Part 15, Subpart B. In addition, Section 15.109(g) of the specification allows the use of the radiated emission limits specified by the International Special Committee on Radio Interference (CISPR) 22 (described in Table 2) when the measurement method specified by the American National Standards Institute (ANSI) C63.4-2014 is used. The limits specified in Tables 1 and 2 are for frequencies below 1 GHz, using the CISPR quasi-peak (QP) detector function with a resolution bandwidth (RBW) of 120 kHz. The limits specified in Tables 3 and 4 are for frequencies above 1 GHz, using peak (PK) and average (AVG) detectors and a receiver with a resolution bandwidth of 1 MHz.

For a given measurement distance, the Class B residential or domestic application limits are typically tighter than the Class A commercial or industrial application limits, typically by 6dB to 10dB. Also note that Tables 1 and 2 also include an inverse linear distance (1/d) scaling factor of 20 dB/dec used per 15.31(f)(1) to normalize the limits for 3m and 10m antenna measurement distances to determine compliance. For example, if the antenna is placed at 3 meters instead of 10 meters to stay within the test equipment boundaries, the limit is adjusted by approximately 10.5dB.

Table 1: Radiated emission field strength QP limits for the 30MHz to 1GHz range as specified in 47 CFR 15.109(a) and (b)

查看详情

Table 2: Radiated emission field strength QP limits for the 30MHz to 1GHz range as specified in 47 CFR 15.109(g)/CISPR 22/32

查看详情

Table 3: Radiated emission field strength limits for the 1 GHz to 6 GHz range as specified in 47 CFR 15.109(a) and (b)

查看详情

Table 4: Radiated emission field strength limits for the 1 GHz to 6 GHz range as specified in 47 CFR 15.109(g)/CISPR 22/32

查看详情

Figure 4 shows a graph of the Class A and Class B limits when the antenna distance is 3m. The FCC-compliant design includes a battery-powered gas sensor implementation using Bluetooth low energy technology, available from Texas Instruments (TI). Users can download the FCCA Class Compliance Report for this design, which lists the radiated emission test data and graphs for easy reference.

查看详情

Figure 4: FCC Part 15 and CISPR 22 Class A and Class B radiated emissions limits (using QP and AVG detectors for conditions below and above 1 GHz, respectively)

The radiated EMI test procedure involves placing the equipment under test (EUT) and supporting equipment on a non-conductive turntable (0.8m above the reference ground plane) in a semi-anechoic chamber (SAC) or open area test site (OATS) as defined in CISPR 16-1, as shown in Figure 5. The EUT is set up 3m from the receiving antenna mounted on the antenna tower.

Using the PK detector pre-scan function of a calibrated broadband antenna (a combination of a biconical and log-periodic antenna, or Bilog antenna), radiated emissions are tested from 30MHz to 1GHz along both horizontal and vertical antenna polarizations. This exploratory test determines the frequencies of all significant emissions. After performing this test, the QP detector is used to check for relevant trouble spots and the final compliance measurements are recorded.

During the test, the RBW of the EMI receiver was set to 120kHz. The antenna was configured for horizontal and vertical polarization (rotated 90° relative to the ground plane) and the height was adjusted from 1m to 4m above the ground plane to maximize the field strength reading for each test frequency while taking into account ground reflections. During the measurement, the EUT on the turntable was rotated between 0 and 360°, so that the azimuth angle between the antenna and the EUT was changed to obtain the maximum field strength reading depending on the orientation of the EUT. The antenna was located in the far field region of the EUT, corresponding to a 3m antenna distance and a frequency of 15.9MHz.

查看详情

Figure 5: Radiated emission measurement setup for FCC Part 15 and CISPR 22/32

A PK detector pre-scan can be performed using a horn antenna for frequencies above 1GHz, and then an AVG detector can be used when approaching the limit. The RBW of the EMI receiver is set to 1MHz. The antenna direction is clear, so there is no need to perform a height scan, and reflections from the ground plane and the chamber walls are unlikely to cause interference. However, the radiated emissions of the EUT at these frequencies are more directional, so the turntable is rotated 360 degrees again to determine the antenna polarization direction for maximum response. According to Table 5, the upper range of the measurement frequency changes with the highest internal frequency of the EUT.

Table 5: Maximum measurement frequency of radiated emissions (based on the highest frequency of the EUT internal clock source)

查看详情

Radiated emission tests calibrate the electric field strength in decibels per microvolt (dB/mV) per meter. The antenna factor (AF) is the ratio of the electric field generated by the antenna plane (mV/m) to the voltage measured by a spectrum analyzer (SA) or a swept EMI receiver (dB/mV). In general, the corrected emission level is derived from Equation 3, which takes into account the AF, cable loss (CL), attenuator and RF limiter loss factors (AL), and amplifier pre-gain (AG).

查看详情

A photo of the LMR16030 60V/3A buck converter radiated emissions test setup and results are shown in Figure 6. The measurement conditions were 24V input, 5V output, 3A load current, and 400kHz switching frequency.

查看详情

Figure 6: CISPR 22 radiated EMI test: Photo of the test setup (a); radiated EMI results for horizontally and vertically polarized antennas (b)

Radiated EMI in Automotive Systems

Although shielded cables can attenuate interference effects in automotive systems, EMI can be "effectively" coupled in susceptible circuits through crosstalk. Due to field line coupling effects, radiated emissions can also cause radiated immunity issues in signal interconnects for relatively small vehicles with dense power distribution and signals passing through cable bundles. For these reasons, evaluating EMI performance has become a key concern for automotive engineers when designing and testing electric vehicles.

UNECE Regulation No. 10 and CISPR 25

CISPR 12 and CISPR 25 are international standards that provide limits and procedures for radio disturbance measurements to protect on-board and off-board receivers in vehicles, respectively. CISPR 25 is specifically applicable at the automotive level, but also to all electronic assemblies (ESAs) used in vehicles. Compared to other standards, CISPR 25 is often used as a basis for automotive manufacturers and their suppliers to define product specifications, but not as a benchmark for assessing compliance and adherence. This difference has emerged in the United Nations Economic Commission for Europe (UNECE) Article 10 regulation since the repeal of the EU Electric Vehicle EMC Directive.

CISPR 25 defines several methods and limit classes for emissions measurements from vehicle components, taking into account both broadband (BB) and narrowband (NB) sources. Figure 7 illustrates the 5-category limit values for components/modules using PK and AVG detectors. The measurements are made on receivers in vehicles operating in the broadcast and mobile service bands. The lowest measurement frequency concerns the European longwave (LW) broadcast band from 150kHz to 300kHz, and the highest frequency is 2.5GHz (taking into account Bluetooth transmissions).

查看详情

Figure 7: CISPR 25 Class 5 radiated limits for components/modules measured with peak and average detectors (linear frequency scale) using the absorber-lined shielded enclosure (ALSE) method

The RBW of the scanning receiver is 9kHz for detection below 30MHz and 120kHz for detection above 30MHz. The exceptions are the GPS L1 Civilian (1.567GHz to 1.583GHz) and Global Navigation Satellite System (GLONASS) L1 (1.591GHz to 1.613GHz) bands. In these two bands, a RBW of 9kHz and a maximum step size of 5kHz are required to detect the corresponding NB transmissions using only the AVG detector.

Antenna systems for CISPR 25

The measurements were performed using a linearly polarized E-field antenna with a nominal output impedance of 50Ω. Table 6 and Figure 8 show the antennas recommended by CISPR 25 to improve the consistency of results provided by different laboratories.

Table 6: According to CISPR 25, it is recommended to use an electric field antenna; the biconical antenna and the log-periodic antenna have overlapping frequencies, and the Bilog antenna covers the frequency range of both antennas.

查看详情

查看详情

Figure 8: CISPR 25 compliant measurement antenna

For low frequency measurements, passive/active rod monopole antennas with counterpoise are used. Biconical and log-periodic dipole array (LPDA) antennas typically cover the frequency ranges of 30MHz to 200MHz and 200MHz to 1GHz, respectively. Finally, double-ridged horn antennas (DRHA) are typically used from 1GHz to 2.5GHz. Broadband Bilog antennas are larger in size than biconical or log-periodic antennas and are sometimes used to cover the frequency range of 30MHz to 1GHz.

Radiated EMI Testing Using ALSE

Figures 9, 10 and 11 show typical setups using the CISPR 25 ALSE method (also known as the antenna method) to measure radiated emissions over the frequency range specified in Table 6.

The EUT and cable harness are placed on a non-conductive dielectric material (relative dielectric constant εr is low, not higher than 1.4) 50mm above the ground plane. The length of the harness parallel to the front of the ground plane is 1.5m, and the total length of the test harness between the EUT and the load simulator does not exceed 2m. The long section of the test harness is parallel to the ground plane towards the edge of the antenna, 100mm away from the edge. The ground plane requirements are a minimum width and length of 1m and 2m respectively, or 200mm under the entire device, whichever is greater. Based on the near-to-far field conversion given by Equation 2 and a 1m antenna distance, the frequency must be below 48MHz when measuring in the near-field region of the EUT.

查看详情

Figure 9: CISPR 25 radiated emissions measurement setup for a monopole whip antenna (150kHz to 30MHz)

查看详情

Figure 10: CISPR 25 radiated emission measurement setup for a biconical antenna (30 MHz to 300 MHz) or a log-periodic antenna (200 MHz to 1 GHz)

查看详情

Figure 11: CISPR 25 radiated emission measurement setup for horn antennas (above 1 GHz)

The horn antenna was aligned with the EUT and the other antennas were placed at the midpoint of the harness. All measurements were performed with an antenna distance of 1 meter. Measurements in the frequency range of 150kHz to 30MHz were performed only for vertical antenna polarization. Scans in the frequency range of 30MHz to 2.5GHz were performed for both horizontal and vertical polarizations.

As mentioned previously, the antenna voltage detected by the EMI receiver combined with the AF produces the electric field strength at the antenna location. Note that separate AFs are available for horizontal and vertical polarizations, so each polarization can be measured using the corresponding AF value.

Radiated EMI pre-compliance testing and results

Figure 12 shows a photo of the radiated emissions test setup for the LM53635-Q1 automotive grade synchronous buck converter [9]. The EUT was powered from the automotive battery with both the positive and negative supply lines connected to a line impedance stabilization network (LISN). The output was 3.3V at a 3.5A resistive load. The switching frequency was 2.1MHz, above the AM band required for many automotive systems, with spread spectrum frequency modulation (SSFM) enabled. Figures 13 to 16 show the measurement results for passing CISPR 25 Class 5 limits using various test antennas.

查看详情

Figure 12: Photo of CISPR 25 pre-compliance measurement setup

查看详情

Figure 13: Radiated emission results: 150kHz to 30MHz, whip antenna, vertical polarization

查看详情

Figure 14: Radiated emission results: 30MHz to 300MHz, biconical antenna, horizontal and vertical polarization

查看详情

Figure 15: Radiated emissions results: 200MHz to 1GHz, log-periodic antenna, horizontal and vertical polarization

查看详情

Figure 16: Radiated emission results: 1 GHz to 2.5 GHz, horn antenna, horizontal polarization

in conclusion

Radiated emissions affect the EMI characteristics of power converters at high frequencies [10]. The upper frequency limit of radiated testing extends to 1 GHz or even higher (depending on the specification), which is much higher than conducted emissions. Although not as straightforward as conducted emissions testing, radiated emissions measurements are essential for compliance testing and can easily become a bottleneck in the product development process.

For automotive applications, the cable bundle is primarily a radiating structure at low frequencies due to its length. The measured radiated emission profile is primarily derived from the common-mode current in the connected cables, driven by near-field electrical coupling between the printed circuit board (PCB) and the cables. I will discuss radiated EMI mitigation techniques in a later section of this article.

alan000345

What a comprehensive sharing, even the formulas are given so clearly, thank you for sharing, it is really useful.

led2015

Radiation, interference, waveform, and testing give us a headache. It is not easy to fully understand them.