Low-voltage DRAM sensing scheme with offset-cancellation sense amplifier

A novel bitline sensing scheme is proposed for low-voltage DRAM to achieve low power dissipation and compatibility with low-voltage CMOS. One of the major obstacles in low-voltage DRAM is the degradation of data-retention time due to low signal level at the memory cell, which requires power-consuming refresh operations more frequently. This paper proposes an offset-cancellation sense-amplifier scheme (OCSA) that improves data-retention time significantly even at low supply voltage. It also improves die efficiency, because the proposed scheme reduces the number of sense amplifiers by supporting more cells in each sense amplifier. Measurements show that the data-retention time of the proposed scheme at 1.5-V supply voltage is 2.4 times of the conventional scheme at 2.0 V.     

Adaptive static and dynamic noise margin improvement in minimum-sized 6T-SRAM cells

We present a novel SRAM technique for simultaneously enhancing the static and dynamic noise margins in six transistor cells implemented with minimum size devices using a design for manufacturability constrained layout. During each access, the word-line voltage (V_WL) is internally reduced with respect to the cell and bit-line voltages that are maintained at nominal V_DD. A specific V_WL can be determined for each memory region, thus allowing for an adaptive approach. The benefits and drawbacks of the technique on the overall memory performance are thoroughly investigated through both simulations and experimental data. Simulations results show that this technique expands the read margin without an appreciable increase of memory area. Specifically, an improvement of 52.6% in static noise margin and a 24.5% in critical charge (parameter used to account for the dynamic stability) has been achieved with a V_WL reduction of 20%. The impact of variability on SNM is reduced, while both read and write delay increase by a specific amount that should be considered as a tradeoff when setting the word-line voltage value. A 16Kbit SRAM test chip including the proposed technique has been fabricated in a 65 nm CMOS technology. Silicon measurements confirm that the proposed technique improves cell stability during READ, which allows operating at relatively low values of V_WL with a small impact on read time.     

Gate leakage current reduction in IP3 SRAM cells at 45 nm CMOS technology for multimedia applications

We have presented an analysis of the gate leakage current of the IP3 static random access memory (SRAM) cell structure when the cell is in idle mode (performs no data read/write operations) and active mode (performs data read/write operations), along with the requirements for the overall standby leakage power, active write and read powers. A comparison has been drawn with existing SRAM cell structures, the conventional 6T, PP, P4 and P3 cells. At the supply voltage, V_DD = 0.8 V, a reduction of 98%, 99%, 92% and 94% is observed in the gate leakage current in comparison with the 6T, PP, P4 and P3 SRAM cells, respectively, while at V_DD = 0.7 V, it is 97%, 98%, 87% and 84%. A significant reduction is also observed in the overall standby leakage power by 56%, the active write power by 44% and the active read power by 99%, compared with the conventional 6T SRAM cell at V_DD = 0.8 V, with no loss in cell stability and performance with a small area penalty. The simulation environment used for this work is 45 nm deep sub-micron complementary metal oxide semiconductor (CMOS) technology, t_ox = 2.4 nm, V_thn = 0.22 V, V_thp = 0.224 V, V_DD = 0.7 V and 0.8 V, at T = 300 K.     

A low-noise gain-tunable amplifier for large array biopotential recording systems

This paper presents a low-noise gain-tunable biopotential amplifier that is designed based on a folded-cascode structure. Sub-threshold and self-biasing techniques are employed to achieve a low-noise and low-power amplification. With a bias-current tuning block, the gain of the proposed biopotential amplifier can be precisely adjusted. Designed in a standard 0.13 μm CMOS process, the proposed amplifier provides a 5.9 kHz bandwidth and 30.1 dB gain with 732 nW power. The input-referred noise over the entire bandwidth is 4.3 μV _rms , equivalent to a noise-efficiency factor of 2.48.     

An experimental DRAM with a NAND-structured cell

An experimental 256-Mb dynamic random access memory using a NAND-structured cell (NAND DRAM) has been fabricated. The NAND-structured cell has four memory cells connected in series, which reduces the area of isolation between the adjacent cells and also reduces the bit-line contact area. The cell area per bit measures 0.962 μm², using 0.4-μm CMOS technology, which is 63% in comparison with the conventional cell. In order to reduce the die size, time division multiplex sense-amplifier (TMS) architecture, in which a sense amplifier is shared by four bit lines, has been newly introduced. The chip area is 464 mm², which is 68% compared with the DRAM using the current cell structure. The data can be accessed by a fast-block-access mode up to 512 bytes as well as a random access mode. Typical 112-ns access time of the first data in a block and 30-ns serial cycle time are achieved     

Improving Yield and Defect Tolerance in Subthreshold CMOS Through Output-Wired Redundancy

This paper presents simulations of three different implementations of the minority-3 function, with special focus on mismatch analysis through statistical Monte Carlo-simulations. The simulations clearly favors the minority-3 Mirrored gate, and a gate-level redundancy scheme, where identical circuits with the same input drive the same output-node, is further explored as a means of increasing fault- and defect-tolerance. Important trade-offs between supply voltage, redundancy and yield are revealed, and V _DD = 175 mV is suggested as a minimum useful operating voltage, combined with a redundancy factor of 2, in 90 nm CMOS.     

A 1.8-ns access, 550-MHz, 4.5-Mb CMOS SRAM

An ultrahigh-speed 4.5-Mb CMOS SRAM with 1.8-ns clock-access time, 1.8-ns cycle time, and 9.84-μm² memory cells has been developed using 0.25-μm CMOS technology. Three key circuit techniques for achieving this high speed are a decoder using source-coupled-logic (SCL) circuits combined with reset circuits, a sense amplifier with nMOS source followers, and a sense-amplifier activation-pulse generator that uses a duplicate memory-cell array. The proposed decoder can reduce the delay time between the address input and the word-line signal of the 4.5-Mb SRAM to 68% of that of an SRAM with conventional circuits. The sense amplifier with nMOS source followers can reduce not only the delay time of the sense amplifier but also the power dissipation. In the SRAM, the sense-amplifier activation pulse must be input into the sense amplifier after the signal from the memory cell is input into the sense amplifier. A large timing margin required between these signals results in a large access time in the conventional SRAM. The sense-amplifier activation pulse generator that uses a duplicate memory-cell array can reduce the required timing margin to less than half of the conventional margin. These three techniques are especially useful for realizing ultrahigh-speed SRAM's, which will be used as on-chip or off-chip cache memories in processor systems     

PMOS-switching dual-path charge pump in standard twin-well CMOS technology

In this article a new charge pump circuit is presented, which is feasible for implementation with the standard twin-well CMOS process. The proposed charge pump employs PMOS-switching dual charge-transfer paths and a simple two-phase clock. Since charge transfer switches are fully turned ON during each half of the clock cycle, they transfer charges completely from the present stage to the next stage without suffering threshold voltage drop. During one clock cycle, the pump transfers charges twice through two pumping paths which are operating alternately. Test chips have been fabricated in a 0.35-μm twin-well CMOS process. The output voltage of a 4-stage charge pump with each pumping capacitor of 7.36 pF measures 6.7 V under a 1.5 V power supply and 20 MHz clock frequency. It can supply a maximum load current of about 180 μA. Although the proposed circuit exhibits somewhat inferior performances against triple-well charge pumps using additional mask and process steps, it shows at least 60% higher voltage gain at V _DD = 0.9 V, approximately 10% higher peak power efficiency at V _DD = 1.5 V, much larger output current drivability and faster initial output rising than traditional twin-well charge pumps. This new pumping efficient circuit is suitable for design applications with a low-cost standard twin-well CMOS process.     

A 29-ns 64-Mb DRAM with hierarchical array architecture

A 29-ns (RAS access time), 64-Mb DRAM with hierarchical array architecture has been developed. For consistent high yields and high speed, a CMOS segment driver circuit is used as a hierarchical word line scheme. To achieve high speed, precharge signal (PC) drivers for equalizing the bit lines pairs, and shared sense amplifier signal (SHR) drivers are distributed in the array. To enhance sense amplifiers speed in low array voltage, an over driven sense amplifier is adopted. A hierarchical I/O scheme with semidirect sensing switch is introduced for high speed data transfer in the I/O paths. By combining these proposed circuit techniques and 0.25-μm CMOS process technologies with phase-shift optical lithography, an experimental 64-Mb DRAM has been designed and fabricated. The memory cell size is 0.71×1.20 μm ², and the chip size is 15.91×9.06 mm². A typical access time under 3.3 V power supply voltage is 29 ns     

A 60-ns 16-Mbit CMOS DRAM with a transposed data-line structure

Low-noise, high-speed circuit techniques for high-density DRAMs (dynamic random-access memories), as well as their application to a single 5-V 16-Mb CMOS DRAM with a 3.3-V internal operating voltage for a memory array, are described. It was found that data-line interference noise becomes unacceptably high (more than 25% of the signal) and causes a serious problem in 16-Mb DRAM memory arrays. A transposed data-line structure is proposed to eliminate the noise. Noise suppression below 5% is confirmed using this transposed data-line structure. A current sense amplifier is also proposed to maintain the data-transmission speed in common I/O lines, in spite of a reduced operating voltage and increased parasitic capacitance loading in the memory array. A speed improvement of 10 ns is achieved. Using these circuit techniques, a 16-Mb CMOS DRAM with a typical RAS access time of 60 ns was realized     

“Higher-κ” dielectrics for advanced silicon microelectronic devices: A combinatorial research study

Combinatorial methodology is used to rapidly screen suitable ternary higher-κ dielectrics for future complementary metal oxide semiconductor (CMOS), and dynamic random access memory (DRAM) devices. Dielectric constant (κ) and leakage current (L_C) were mapped from capacitance–voltage (C–V) and current–voltage (I–V) measurements. HfO₂–TiO₂–Y₂O₃ library films, made by pulsed laser deposition (PLD), have been characterized. We found a band of compositions in the middle of the HfO₂–TiO₂–Y₂O₃ phase diagram that have dielectric constants in the range of 50–80, with reasonably low leakage currents, that are therefore promising for these applications.     

Underlap channel metal source/drain SOI MOSFET for thermally efficient low-power mixed-signal circuits

In this paper it has been shown that employing an underlap channel created by varying the lateral doping straggle in dopant-segregated Schottky barrier SOI MOSFET not only improves the scalability but also suppresses the self-heating effect of this device. Although in strong inversion region the reduced effective gate voltage due to voltage drop across the underlap lengths reduces the drive current, in weak/moderate inversion region defined at I_D=5 μA/μm and V_DS=0.5 V the analog figures of merit such as transconductance, transconductance generation factor and intrinsic gain of the proposed underlap device are improved by 15%, 35% and 20%, respectively over the conventional overlap channel structure. In addition to this, at V_DD=0.5 V the gain-bandwidth product in a common-source amplifier based on proposed underlap device is improved by ～20% over an amplifier based on the conventional overlap channel device. The mixed-mode device/circuit simulation results of CMOS inverter, NAND and the NOR gates based on these devices also show that at V_DD=0.5 V the switching energy, static power dissipation and the propagation delay in the case of proposed underlap device are reduced by ～10%, ～35% and ～25%, respectively, over the conventional overlap device. Thus, significant improvement in analog figures of merit and the reduction in digital design metrics at lower supply voltage show the suitability of the proposed underlap device for low-power mixed-signal circuits.     

BiCMOS circuit technology for high-speed DRAMs

A BiCMOS circuit technology featured by a novel bit-line sense amplifier has been developed. The bit-line sense amplifier is composed of a BiCMOS differential amplifier, the impedance-converting means featured by the CMOS current mirror circuit or the clocked CMOS inverter between the bit line and the base node of the BiCMOS differential amplifier, and a conventional CMOS flip-flop. This technology can reduce the access time to half that of a conventional CMOS DRAM access time. Applied to a 1-kb DRAM test chip, a new BiCMOS circuit technology was successfully verified. Furthermore, the sensitivity and area penalty of the new BiCMOS bit-line sense amplifier and future applications to megabit DRAMs are discussed     

A technique to mitigate impact of process,voltage and temperature variations on design metrics of SRAM Cell

This paper presents a technique for designing a variability aware SRAM cell. The architecture of the proposed cell is similar to the standard 6T SRAM cell with the exception that the access pass gates are replaced with full transmission gates. The paper studies the impact of V_t (threshold voltage) variation on most of the design metrics of SRAM cell. The proposed design achieves 1.4× narrower spread in I_READ at the expense 1.2× lower I_READ at nominal V_DD. It offers 1.3× improvements in T_RA (read access time) distribution at the expense of 1.2× penalty in read delay. The proposed bitcell offers 1.1× tighter spread in T_WA (write access time) incurring 1.3× longer write delay. It shows 180 mV of SNM (static noise margin) and is equally stable in hold mode. It offers 1.3× higher RSNM (100 mV) compared to 6T (75 mV). It exhibits improved SINM (static current noise margin) distribution at the expense of 1.6× lower WTI (write trip current). It offers 1.05× narrower spread in standby power. Thus, comparative analysis based on Monte Carlo simulation exhibits that the proposed design is capable of mitigating impact of V_t variation to a large extent.     

SRAM devices and circuits optimization toward energy efficiency in multi-Vth CMOS

Minimum-energy-driven circuit design is highly required in numerous emerging applications such as mobile electronics, wireless sensor nodes, implantable biomedical devices, etc. Due to high computing capability requirements in such applications, SRAMs play a critical role in energy consumption. This paper presents SRAM energy analysis utilizing multi-threshold (multi-V_th) voltage devices and various circuit techniques for power reduction and performance improvement, and suggests optimal device combinations for energy efficiency improvement. In general, higher-V_th devices are preferred in the cross-coupled latches and the write access transistors for reducing leakage current while lower-V_th devices are desired in the read port for implementing higher performance. However, excessively raised V_th in the write paths, i.e. the cross-coupled latches and the write access transistors, leads to slower write speed than read, quickly nullifying improved energy efficiency. In this work, the energy efficiency improvement of 6.24× is achieved only through an optimal device combination in a commercial 65 nm CMOS technology. Employing power reduction and performance boosting techniques together with the optimal device combination enhances the energy efficiency further up to 33×.     

A fully on-chip 1-μW capacitor-free low-dropout regulator with adaptive output stage

A fully on-chip 1-μW fast-transient response capacitor-free low-dropout regulator (LDO) using adaptive output stage (AOS) is presented in this paper in standard 0.13-μm CMOS process. The AOS circuit is proposed to deliver extra four times of output current of the operational amplifier at medium to heavy load to extend the bandwidth of the LDO and enhance the slew rate at the gate of the power transistor. And the AOS circuit is shut off at light load to reduce the quiescent current and maintain the stability without requiring area-consuming on-chip capacitor. Meanwhile, the proposed AOS circuit introduces V_OUT offset at medium to heavy load to counteract the V_OUT drop, which is caused by I_LOAD increase. Hence, transient performances of LDO and V_OUT drop between light load and full load are improved significantly with 1.1-μA quiescent current at light load. From the post simulation results, the LDO regulates the output voltage at 0.7 V from a 0.9-V supply voltage with a 100-mA maximum load current. The undershoot, the overshoot and the recovery time of the proposed LDO with I_LOAD switching from 50 μA to 100 mA in 1 μs are about 130 mV, 130 mV and 1.5 μs, respectively. And the V_OUT drop between light load and full load reduces to 0.16 mV.     

OTA based on CMOS inverters and application in the design of tunable bandpass filter

A new operational transconductance amplifier (OTA) builds with CMOS inverters only is proposed in this paper. Simulations with typical BSIM3V3 parameters of a 0.35 μm CMOS process have shown a 3.56 GHz gain-bandwidth product under 2.5 V supply voltage. The corresponding total harmonic distortion is equal to 0.46% for 2 V peak–peak differential output voltage. At the same supply voltage, the circuit can provided at each output a voltage swing of 2.25 V peak–peak. From V_DD = 2 V to V_DD = 2.5 V the differential transconductance varies from 72 to 108.4 μΩ⁻¹. The corresponding common mode rejection ratio and the total power consumption are always lower than −31 dBc and 800 μW, respectively. Typical application of a biquad filter is proposed to illustrate the circuit capabilities.     

Standby supply voltage minimization for deep sub-micron SRAM

Suppressing the leakage current in memories is critical in low-power design. By reducing the standby supply voltage (V_DD) to its limit, which is the data retention voltage (DRV), leakage power can be substantially reduced. This paper models the DRV of a standard low leakage SRAM module as a function of process and design parameters, and analyzes the SRAM cell stability when V_DD approaches DRV. The DRV model is verified using simulations as well as measurements from a 4 KB SRAM chip in a 0.13 μm technology. Due to a large on-chip variation, DRV of the 4 KB SRAM module ranges between 60 and 390 mV. Measurements taken at 100 mV above the worst-case DRV show that reducing the SRAM standby V_DD to a safe level of 490 mV saves 85% leakage power. Further savings can be achieved by applying DRV-aware SRAM optimization techniques, which are discussed at the end of this paper.     

小容量相变存储器芯片版图设计与验证分析

设计了基于1T1R结构的16 kb相变存储器(PCRAM)芯片及其版图。芯片包括存储阵列、外围读写控制电路、纠错电路(ECC)、静电防护电路(ESD)。版图上对纳米存储单元(1R)与CMOS工艺的融合作了优化处理,给出了提高存储单元操作电流热效率的具体方法。1R位于顶层金属(TM)和二层金属(TM-1)之间,包含存储材料以及上下电极,需要在传统CMOS工艺基础上添加掩膜版。读出放大器采用全对称的差分拓扑结构,大大提升了抗干扰能力、灵敏精度以及读出速度。针对模块布局、电源分配、二级效应等问题,给出了版图解决方案。采用中芯国际130 nm CMOS工艺流片,测试结果显示芯片成品率(bit yield)可达99.7%。     

A four-level storage 4-Gb DRAM

A 4-Gb DRAM with multilevel-storage memory cells has been developed. This large memory capacity is achieved by storing data at four levels, each corresponding to two-bit-data storage in a single memory cell. The four-level storage reduces the effective cell size by 50%. A sense amplifier using charge coupling and charge sharing was developed for the four-level sensing and restoring. The sense amplifier uses a hierarchical bit-line scheme and operates in a time-sharing mode, thus reducing the sense amplifier area. A 4-Gb DRAM fabricated using 0.15-μm CMOS technology measures 986 mm². The memory cell is 0.23 μm². Its capacitance of 60 fF is achieved by using a high-dielectric-constant material BST