Hark-Byeong Park

Radio-Frequency Interference Estimation Using Equivalent Dipole-Moment Models and Decomposition Method Based on Reciprocity

In modern electronic products, the noise from high-speed digital parts is likely to interfere with nearby receivers, causing radio-frequency interference (RFI) issues. In this paper, the equivalent dipole-moment models and a decomposition method based on reciprocity theory are proposed being used together to estimate the coupling from the noise source to the victim antennas. The dipole-moment models are extracted from the near fields of the noise source by solving the inverse problem. The tangential electromagnetic fields on a Huygenss surface, which enclose the victim antenna, can be calculated from these equivalent dipole-moment models. Then, the victim antenna only is treated as a radiator. The tangential electromagnetic fields from the radiating antenna on the same Huygenss surface can be obtained. With these two groups of the fields on the Huygenss surface, the reciprocity theory is applied to estimate the coupling from the noise source to the victim antenna. This method is validated by full-wave simulations and measurements of a simple printed circuit board. The proposed method provides convenience to estimate RFI issues in the early design stage and saves the time of RFI simulation and measurements.

A Simple Method of Estimating the Radiated Emission From a Cable Attached to a Mobile Device

When a mobile device is connected to cables for charging power or transmitting data, the radiated emission from the attached cables, which are typically effective electromagnetic interference (EMI) antennae at certain frequencies, can cause serious system-level EMI problems. The measurement of system-level radiation during compliance and precompliance tests is not only a time-consuming task, but also requires expensive facilities such as a semianechoic chamber. This paper proposes a simple method of predicting far-field radiation from cables attached to mobile devices at the early stage of the design and development phase without using an EMI chamber. The method combines radiation characterization of a simple box-source-cable geometry using full-wave simulations with the measurement of the real common-mode current flowing through the cable. The proposed method was applied to mobile phones to estimate the far-field radiated emissions, which were compared with the measurement results. The accuracy of the predicted results was evaluated using the feature selective validation technique, indicating good agreement and correlation.

A Magnetic-Field Resonant Probe With Enhanced Sensitivity for RF Interference Applications

High-sensitive field probes are highly desirable for radio-frequency (RF) interference studies, where ultralow noise levels are of interest. By incorporating an LC resonant circuit in a differential-loop probe, together with a Marchand balun, a magnetic-field probe with enhanced sensitivity is developed. Its equivalent circuit model and design methodology are established. The design is validated by measurements. The measured relative sensitivity in terms of |S21| of the proposed probe increases by approximately 8.63 dB at the resonant frequency of 1.575 GHz compared to that of a conventional design. The advantage of the proposed probe is validated through its application in the measurement of a microstrip trace and a real-world cell phone design.

Near-field coupling estimation by source reconstruction and Huygens's equivalence principle

This study is to estimate the near-field coupling in mixed digital/RF circuit design for modern high speed electronic systems. The noise source IC is first modelled by physics-based dipole moment model with data obtained from a near-field scanning plane. The victim RF antenna is modelled in full-wave simulation tool and the noise IC is further modelled as Huygenss equivalent source. The tangential fields on the Huygenss box can be calculated by dipole moment model with negligible multiple scattering effect assumption. The noise coupling then can be obtained by surface integration of Poynting vector at the RF antenna receiving port. A full-wave simulation model is first studied to demonstrate the method and a noisy clock buffer IC with victim patch antenna is measured for further validation.

Estimation of Power Switching Current by Chip-Package-PCB Cosimulation

This paper presents a methodology to estimate power switching current on printed circuit boards (PCBs) through chip-package-PCB cosimulation. A macromodel for a timing controller chip running pseudo H-pattern data was generated from transistor-level simulations. The macromodel consists of a passive impedance network and internal switching activity of the chip. Power delivery network models for package and PCB were produced as a RLCG netlist and S-parameter touch stone files, respectively, using commercial tools. It is found that comparison between the simulated and measured impedances of the chip and package shows excellent agreement up to 300 MHz. Also, the simulated and measured impedances of the PCB match well in terms of magnitude and resonance frequency up to 3 GHz. Moreover, the results of power switching current from cosimulation and measurement show good agreement within 5 dB difference at major harmonic frequencies of 20 MHz data and 80 MHz clock patterns up to 1 GHz.

Application of dipole-moment model in EMI estimation

This paper used magnetic near fields to extract the dipole-moment model to represent the real radiation source. This method prevents the measurement of electric fields so that the scanning time and points are saved significantly. These equivalent dipole-moment models can take the place of the real radiation source in the full-wave simulation tool. The electromagnetic interference between the real source and victim structures are well predicted in the simulation by the dipole-moment model. This is validated by a numerical example in this paper.

Measurement validation for radio-frequency interference estimation by reciprocity theorem

This paper presents the measurement validation of reciprocity theorem method for near-field coupling estimation. The overall problem is decomposed into two parts, the first part is called forward problem, and the second part is called the reverse problem. For forward problem, the noise source IC is modelled by physics-based dipole moment model with data obtained from a near-field scanning plane, then the tangential E and H fields on a Huygenss box enclosing the victim antenna are calculated by analytical expression. In reverse problem, the victim RF antenna is modelled in full-wave simulation tool and the tangential E and H field are obtained by simulation. With tangential E and H field obtained in forward problem and reverse problem, the coupled noise power is then estimated by reciprocity theorem. The estimated noise coupling power is compared with measured power at the victim antenna port with IC excited. The difference is within 5dB which is acceptable for engineering practice.

An EMI Evaluation Method for Integrated Circuits in Mobile Devices

This paper presents an electromagnetic interference (EMI) evaluation method for ICs in mobile devices. The evaluation method consists of a noise measurement method for ICs and a qualification specification, which has a direct correlation with system-level radiated emissions. The 1-Ω direct coupling method, which belongs to the IEC 61967 standard, is adopted to measure the conducted emission from a specially designed low-voltage differential signaling loop-back test IC. To devise the specification for the IC-level EMI evaluation, a definite correlation between the conducted emission from the IC and the radiated emission from a test system, which consists of a printed circuit board with the IC and an attached power cable, was obtained as a radiation transfer function (RTF). By combining the RTF with the system-level regulation specification provided by CISPR or FCC, an EMI evaluation specification for the test IC was derived. For several test cases, we measured the conducted emissions from the IC to assess the noise level and pass/fail statement. Compared with the radiated emissions from the test system, it was observed that there is a meaningful correlation in terms of the emission peak level and the pass/fail decision. The proposed methodology can be applied to component-level EMI assessment at the early design stage of modern high-speed mobile devices and will be very helpful in reducing system-level EMI problems and design failures in advance.

A Novel Shielding Effectiveness Matrix of Small Shield Cans Based on Equivalent Dipole Moments for Radio-Frequency Interference Analysis

A definition of a shielding effectiveness (SE) matrix for small shield cans used to suppress intra-system noise coupling in mobile devices is proposed. Assuming shield cans enclosing a noise source can be modeled as equivalent dipole moments along with integrated circuits (ICs), a new SE matrix for radio-frequency interference analysis is defined as a ratio of equivalent dipole moments with and without a shield can. The equivalent dipole moments are experimentally extracted using a GTEM cell and two types of probes (monopole and loop), which mimic well the radiations of ICs. Furthermore, the estimation of suppression of noise coupling to an antenna using the proposed SE matrix is successfully validated through numerical simulations.

An Evaluation Method for Radiated Emissions of Components and Modules in Mobile Devices

In this paper, a component-level electromagnetic interference (EMI) evaluation method for multifunctional mobile devices is suggested. The proposed method is based on the extraction of the transfer function between the component-level and the system-level radiated emissions. To demonstrate the practical correlation between the near-field emission from components or modules and far-field radiation from systems, measurement-based transfer functions are extracted by testing various mobile phones integrating camera modules as a target component. A specification for the component-level EMI evaluation is developed with a statistical transfer function and the system-level EMI regulation of CISPR. The proposed evaluation method is experimentally verified using 86 real mobile phones with a good correlation between the component-level and the system-level evaluation results.