Electrical ground looks simple on a schematic; unfortunately, the actual performance of a circuit is dictated by its layout （and by its printed-circuit-board）. When the ground node moves, system performance suffers ...Electrical ground looks simple on a schematic; unfortunately, the actual performance of a circuit is dictated by its layout （and by its printed-circuit-board）. When the ground node moves, system performance suffers and the system radiates electromagnetic interferences. But the understanding of the physics of ground noise can provide an intuitive sense for reducing the problem. Ground bounce can produce transients with amplitudes of volts; most often changing magnetic flux is the cause; in this work, the authors use a Finite-Difference Time-Domain to begin to understand such phenomena. Additionally, predicting substrate cross-talks in mixed-signal circuits has become a critical issue to preserve signal integrity in future integrated systems. Phenomena that involve parasitic signal propagation into the substrate are discussed. A simple methodology to predict the substrate cross-talk and some associated tools are presented. Finally, the authors indicate a stochastic method which could grasp both outer or inner RF （Radio-Frequency） radiations and substrate parasites.展开更多
The proposed DAC consists of a unit current-cell matrix for 8MSBs and a binary-weighted array for 4LSBs,trading-off between the precision,speed,and size of the chip.In order to ensure the linearity of the DAC,a double...The proposed DAC consists of a unit current-cell matrix for 8MSBs and a binary-weighted array for 4LSBs,trading-off between the precision,speed,and size of the chip.In order to ensure the linearity of the DAC,a double Centro symmetric current matrix is designed by the Q2 random walk strategy.To achieve better dynamic performance,a latch is added in front of the current switch to change the input signal,such as its optimal cross-point and voltage level.For a 12bit resolution,the converter reaches an update rate of 300MHz.展开更多
This paper presents a 2GS/s 10bit CMOS digital-to-analog converter （DAC） that consists of two unit current-cell matrixes for 6MSBs and 4LSBs, respectively, trading off between the precision and size of the chip. Cur...This paper presents a 2GS/s 10bit CMOS digital-to-analog converter （DAC） that consists of two unit current-cell matrixes for 6MSBs and 4LSBs, respectively, trading off between the precision and size of the chip. Current mode logic （CML） is used to ensure high speed,and a double centro-symmetric current matrix is designed by the Q^2 random walk strategy in order to ensure the linearity of the DAC. The DAC occupies 2.2mm × 2.2mm of die area and consumes 790mW with a single 3.3V power supply.展开更多
Programmable Logic Array (PLA) is an important building circuit of VLSI chips and some of the FPGA architectures have evolved from the basic PLA architectures. In this letter, a dynamic and static mixed PLA with singl...Programmable Logic Array (PLA) is an important building circuit of VLSI chips and some of the FPGA architectures have evolved from the basic PLA architectures. In this letter, a dynamic and static mixed PLA with single-phased clock is presented. Combining both dynamic and static design style rather than introducing additional interface-buffers overcomes the racing problem, thereby saves the chip area. Besides inheriting the advantages of dynamic circuit-low power dissipation and compact structure, this approach also provides high-speed operation.展开更多
文摘Electrical ground looks simple on a schematic; unfortunately, the actual performance of a circuit is dictated by its layout （and by its printed-circuit-board）. When the ground node moves, system performance suffers and the system radiates electromagnetic interferences. But the understanding of the physics of ground noise can provide an intuitive sense for reducing the problem. Ground bounce can produce transients with amplitudes of volts; most often changing magnetic flux is the cause; in this work, the authors use a Finite-Difference Time-Domain to begin to understand such phenomena. Additionally, predicting substrate cross-talks in mixed-signal circuits has become a critical issue to preserve signal integrity in future integrated systems. Phenomena that involve parasitic signal propagation into the substrate are discussed. A simple methodology to predict the substrate cross-talk and some associated tools are presented. Finally, the authors indicate a stochastic method which could grasp both outer or inner RF （Radio-Frequency） radiations and substrate parasites.
文摘The proposed DAC consists of a unit current-cell matrix for 8MSBs and a binary-weighted array for 4LSBs,trading-off between the precision,speed,and size of the chip.In order to ensure the linearity of the DAC,a double Centro symmetric current matrix is designed by the Q2 random walk strategy.To achieve better dynamic performance,a latch is added in front of the current switch to change the input signal,such as its optimal cross-point and voltage level.For a 12bit resolution,the converter reaches an update rate of 300MHz.
文摘This paper presents a 2GS/s 10bit CMOS digital-to-analog converter （DAC） that consists of two unit current-cell matrixes for 6MSBs and 4LSBs, respectively, trading off between the precision and size of the chip. Current mode logic （CML） is used to ensure high speed,and a double centro-symmetric current matrix is designed by the Q^2 random walk strategy in order to ensure the linearity of the DAC. The DAC occupies 2.2mm × 2.2mm of die area and consumes 790mW with a single 3.3V power supply.
基金Supported by the Commission of Science Technology and Industry for National Defense and the National Natural Science Foundation of China (No. 90307011)
文摘Programmable Logic Array (PLA) is an important building circuit of VLSI chips and some of the FPGA architectures have evolved from the basic PLA architectures. In this letter, a dynamic and static mixed PLA with single-phased clock is presented. Combining both dynamic and static design style rather than introducing additional interface-buffers overcomes the racing problem, thereby saves the chip area. Besides inheriting the advantages of dynamic circuit-low power dissipation and compact structure, this approach also provides high-speed operation.
