scispace - formally typeset
Search or ask a question

Showing papers by "Jun Fan published in 2008"


Journal ArticleDOI
TL;DR: In this article, the via-plate capacitance for a via transition to a multilayer printed circuit board is evaluated analytically in terms of higher order parallel-plate modes.
Abstract: The via-plate capacitance for a via transition to a multilayer printed circuit board is evaluated analytically in terms of higher order parallel-plate modes. The Green's function in a bounded coaxial cavity for a concentric magnetic ring current is first derived by introducing reflection coefficients for cylindrical waves at the inner and outer cavity walls. These walls can be perfect electric conductor (PEC)/perfect magnetic conductor(PMC) or a nonreflective perfectly matched layer. By further assuming a magnetic frill current on the via-hole in the metal plate, an analytical formula is derived for the via barrel-plate capacitance by summing the higher order modes in the bounded coaxial cavity. The convergence of the formula with the number of modes, as well as with the radius of the outer PEC/PMC wall is discussed. The analytical formula is validated by both quasi-static numerical methods and measurements. Furthermore, the formula allows the investigation of the frequency dependence of the via-plate capacitance, which is not possible with quasi-static methods.

137 citations


Proceedings ArticleDOI
18 Nov 2008
TL;DR: In this article, the impact of ground vias as a function of the number of the ground Vias, their locations, and the size of the plane pair is discussed in a block-by-block physics-based equivalent circuit modeling approach.
Abstract: When a high-speed signal transits through a via that penetrates a plane pair, parallel-plane resonances can cause additional insertion loss for the signal. To eliminate this via-plane coupling, ground vias are added adjacent to the signal via. This paper discusses the impact of the ground vias as a function of the number of the ground vias, their locations, and the size of the plane pair. A block-by-block physics-based equivalent circuit modeling approach is used in the study. The underlying physics of the phenomenon and the design implications are also discussed in the paper.

22 citations


Proceedings ArticleDOI
01 Sep 2008
TL;DR: In this paper, a fast and accurate way to assess the impact of signal vias where the return current must change reference planes, including the effect of a return current via placed various distances from the signal via, is discussed.
Abstract: This paper discusses a fast and accurate way to assess the impact of signal vias where the return current must change reference planes, including the effect of a return current via placed various distances from the signal via. The equivalent circuit includes capacitance of the via anti-pad as well as cavity effects.

19 citations


Proceedings ArticleDOI
14 Oct 2008
TL;DR: In this article, the authors used a segmentation approach combined with a via capacitance model and a plane-pair cavity model to study the noise coupling between signal and power/ground nets.
Abstract: Signal vias are often used to move a signal from one PCB layer to another. As a result, these vias can penetrate power/ground plane pair and cause noise coupling (crosstalk) between signal and power/ground nets. This paper studies the noise coupling mechanism using a segmentation approach combined with a via capacitance model and a plane-pair cavity model. Noise coupling from signal to power/ground, and vice versa, is demonstrated in the modeling results.

19 citations


Proceedings ArticleDOI
14 Oct 2008
TL;DR: In this paper, an approach using combined near-field and far-field measurements is proposed, which consists of recording signals from a near field probe and from an antenna in the far field using a high speed oscilloscope and analyzing the relationship between them via different post processing methods.
Abstract: The evaluation of a product in terms of radiated emissions involves identifying the noise sources. Spectrum analyzer (SA) measurements alone are unable to identify noise sources when multiple sources are responsible for emissions at a particular frequency. In this paper, an approach using combined near-field and far-field measurements is proposed. This method consists of recording signals from a near field probe and from an antenna in the far-field using a high speed oscilloscope and analyzing the relationship between them via different post processing methods. The noise source can be identified by varying the location of near-field probe and searching for the probe signal that best correlates to the far field signal. A variety of post processing methods have been employed in this work. The short term fast Fourier transform (STFFT) is used to visualize the time dependence of the frequency content. Envelope correlation, coherence factor, and cross-correlation methods are further explained and tested for their ability to identify possible sources of emission problems.

14 citations


Proceedings ArticleDOI
14 Oct 2008
TL;DR: In this article, a 26-layer printed circuit board including several test sites has been analyzed and a physics-based circuit model has been assembled in a spice-based simulation tool.
Abstract: A 26-layer printed circuit board including several test sites has been analyzed. All the sites have a transition from coupled microstrips to coupled striplines through signal vias. Differential measurements have been performed on some of these test sites to estimate the effect on S-parameters and eye diagrams due to via and antipad radius variation, and different lengths of via stub. The focus of this paper is on a test site with a transition from top to the sixth layer. At the same time, a physics based circuit model has been assembled in a spice-based simulation tool and a full-wave model has been generated as well. The paper will show that the process of modeling can require a series of adjustments to get reasonable results. A brief discussion about possible issues associated with fabrication tolerances is presented in the last chapter.

14 citations


Proceedings ArticleDOI
19 May 2008
TL;DR: In this article, a circuit model for decoupling capacitors including both parasitic inductances and via capacitances is derived and a closed-form expression for via barrel-plate capacitance is derived.
Abstract: A closed-form expression for via barrel-plate capacitances is derived. This results in a more accurate equivalent circuit model for decoupling capacitors including both parasitic inductances and via capacitances. Multilayer power distribution network (PDN) are then analyzed by incorporating the circuit models of both vias and parallel plane pair. Circuit simulator is used to evaluate the coupling properties among various locations in the multilayer PDN structures.

11 citations


Proceedings ArticleDOI
14 Oct 2008
TL;DR: In this paper, the authors studied the noise coupling mechanism caused by a differential pair of vias penetrating power/ground plane pair using a physics-based via-plane model combined with transmission line models for traces.
Abstract: Due to the increase in board density, routing traces on different layers becomes a widely used strategy. Through-hole vias are often used to connect these traces. Those vias that penetrate power/ground plane pairs could cause noise coupling between signal and power/ground nets. At the same time, the need for clean signal transmitted to receivers results in a wide use of differential signals. This paper studies the noise coupling mechanism caused by a differential pair of vias penetrating power/ground plane pair using a physics-based via-plane model combined with transmission line models for traces. A 26-layer printed circuit board with a pair of differential vias have been modeled. The simulated results clearly demonstrate the impact of ground vias and via stubs on noise coupling.

9 citations


Proceedings ArticleDOI
30 Dec 2008
TL;DR: In this article, an admittance matrix is derived for the via ports at the top/bottom surfaces of the via holes, where the magnetic frill currents are expressed as cylindrical waves.
Abstract: Parallel plate modes are excited by the magnetic frill currents in the via holes (anti-pads). These modes are expressed as cylindrical waves. Multiple scattering of these modes among vias as well as from the edge boundaries of the plate pair are rigorously considered with the addition theorem of the cylindrical waves. An admittance matrix is derived for the via ports at the top/bottom surfaces of the via holes. Good agreement has been found between numerical simulations and the algorithm presented.

9 citations


Proceedings ArticleDOI
01 Sep 2008
TL;DR: In this paper, a step-by-step process to design the decoupling strategy for the charge supply is described, where the distance from the decouple capacitors to the ground plane, the number of decoupled capacitors, and the inductance associated with the connection of the decouppling capacitor to the power and ground reference planes will all influence how much charge is delivered.
Abstract: This paper's goal is to help designers go through a step-by-step process to design the decoupling strategy for the charge supply. The distance to the decoupling capacitors, the number of decoupling capacitors, and the inductance associated with the connection of the decoupling capacitor to the power and ground-reference planes will all influence how much charge is delivered.

8 citations


Proceedings ArticleDOI
01 Sep 2008
TL;DR: In this paper, the authors extended the cavity model with segmentation method to the analysis of radio frequency interference (RFI) problems in a shielding enclosure with printed circuit boards (PCBs).
Abstract: Cavity model with segmentation method is extended to the analysis of radio frequency interference (RFI) problems in a shielding enclosure with printed circuit boards (PCBs) . Sixteen different Green's functions, instead of one in the conventional segmentation method developed for PCB cavities, are introduced to describe the fields in various cavities formed by enclosure walls and PCB copper planes. Both horizontal and vertical connections among these cavities are achieved by enforcing the boundary conditions along their common interfaces. Numerical examples demonstrate the efficiency and accuracy of the method by compared with full wave simulations.

Proceedings ArticleDOI
14 Oct 2008
TL;DR: In this paper, the link path is designed on a block-by-block basis, based on the physics-based model of each block, which relates performance to geometry and makes design improvement and optimization possible.
Abstract: In high-speed data communication systems, the complexity of link path between transmitters and receivers present a challenge for designers to maintain an acceptable bit error rate. An approach is presented in this paper to design the link path on a block-by-block basis. The unique advantage of this approach lies on the physics-based model of each block, which relates performance to geometry and makes design improvement and optimization possible. An example link path involving a backpanel is investigated using the approach. The via stubs and the dielectric materials in the backpanel are demonstrated to be critical factors for link performance in certain situations.

Proceedings ArticleDOI
Norman Smith1, Jun Fan, J. Andresakis, Y. Fukawa, M. Harvey, James L. Knighten1 
27 May 2008
TL;DR: This paper utilizes simulated as well as measured product data to compare the performance of the standard design to one using various types of buried capacitance layers with a reduced number of SMT decoupling capacitors.
Abstract: Embedding capacitive layers inside the printed circuit board (PCB) have demonstrated the ability to reduce the number of surface mount technology (SMT) chip decoupling capacitors on the PCB surface as well as greatly improve the performance of the power distribution system. Many systems today utilize this technology, but most public information is limited to data on test vehicles or emulators. This paper utilizes simulated as well as measured product data to compare the performance of the standard design to one using various types of embedded capacitor layers with a reduced number of SMT decoupling capacitors. A methodology is provided that can be utilized for other designs.

Proceedings ArticleDOI
14 Oct 2008
TL;DR: In this article, a hybrid method is proposed for radio frequency interference prediction of a metal enclosure with an aperture on the top wall, where the structure is divided into several segments, and the fields in rectangular segments are described by cavity model, the segments with apertures are modeled by the commercial finite element solver (HFSS).
Abstract: A hybrid method is proposed for radio frequency interference (RFI) prediction of a metal enclosure with an aperture on the top wall. The structure is divided into several segments. While the fields in rectangular segments are described by cavity model, the segments with apertures are modeled by the commercial finite element solver (HFSS). Tangential field continuities along the common boundaries of different segments are enforced by the voltages and currents of boundary ports. Good agreement has been achieved between the hybrid method and full wave simulation.

Proceedings ArticleDOI
14 Oct 2008
TL;DR: This paper presents an approach to quantify channel performance using TWDP (transmitter waveform and dispersion penalty) with frequency-domain S-parameter data, obtained either from measurements or simulations.
Abstract: This paper presents an approach to quantify channel performance using TWDP (transmitter waveform and dispersion penalty) with frequency-domain S-parameter data. TWDP is initially defined to characterize the performance of a transmitter in optical links. The same concept has been extended to quantify channel performance as well, especially in high-speed copper links. This paper focuses on channel characterization. Instead of using time-domain oscilloscope measurements as defined in the original approach, a new method is proposed by using the frequency-domain S-parameter data, obtained either from measurements or simulations. A parametric study on TWDP with respect to bit rate, number of samples per bit, rise/fall time, etc., is also presented with discussions.

Proceedings ArticleDOI
18 Nov 2008
TL;DR: In this paper, the cavity model and the segmentation technique were used to investigate the noise coupling mechanism from a printed circuit board (PCB) inside a compact-size enclosure, which is an effective path for high-speed digital noise to be coupled to the RF receivers inside the same enclosure.
Abstract: The power distribution network in a printed circuit board (PCB) inside a compact-size enclosure is an effective path for high-speed digital noise to be coupled to the RF receivers inside the same enclosure, causing RF interference (RFI) issues. This noise coupling mechanism from PCB to shielding enclosure is investigated in this paper using the cavity model and the segmentation technique. In this approach, the structure of an enclosure with a PCB inside is divided into cavities with both horizontal and vertical connections. Modeling result agrees well with full wave simulations, and the simulation time is considerably reduced. Furthermore, the relationship among the noise coupling, the PCB-related resonances, and the enclosure-related resonances is studied as well.


Proceedings ArticleDOI
14 Oct 2008
TL;DR: In this article, a fast method based on a cavity formulation can be used to estimate the internal noise coupling mechanisms inside the enclosure, assuming that only the TMz0 mode exists inside an enclosure, and the electric field along the z-direction is constant.
Abstract: Modern electronic systems may use mixed RF/digital technologies to achieve various functionalities, which leads to various intra-system interference problems including the RF interference from noisy digital circuits to sensitive RF receivers, especially when the overall system is contained in a metal enclosure. A fast method based on a cavity formulation can be used to estimate the internal noise coupling mechanisms inside the enclosure. This method assumes that only the TMz0 mode exists inside the enclosure, i.e., the electric field along the z-direction is constant. The cavity formulation fails in the region adjacent to an aperture in an enclosure wall, since the aperture introduces higher order modes. The developed closed-form expressions compute the Ez-field variation along the z-direction. Thus, they can be used to estimate the breakpoint where the cavity method is no longer effective.