14 nm SOI FinFET器件单粒子瞬态的复合双指数电流源模型

单粒子瞬态(SET)的电路仿真通常是注入双指数电流源来模拟,然而,纳米FinFET器件的SET采用单个双指数电流源模拟会带来较大误差。TCAD仿真结果较准确,但耗时较长,为了较为准确地电路仿真SET,提出了一种SET的复合双指数电流源模型。利用TCAD对电学特性校准的14 nm SOI FinFET器件的SET进行仿真,通过分析瞬态电流波形,对比双指数模型特点,提取特征参数,并利用遗传算法对模型参数进行优化处理,得到了关于线性能量转移(LET)的复合双电流源参数的解析模型。利用此复合双指数电流源模型与TCAD得到的瞬态电流波形、峰值和收集电荷量进行对比检验。结果显示,本文模型得到的SET电流波形与TCAD的基本吻合,与TCAD相比,模型的峰值电流的平均误差和最大误差分别为3.00%、5.06%;收集电荷量的平均误差和最大误差分别为4.02%、7.17%。     

负偏置温度不稳定性导致单粒子瞬态脉冲的压缩与展宽

摘要：本文基于3D TCAD 器件模拟,研究了130nm体硅工艺下,负偏置温度不稳定性（NBTI）对单粒子瞬态（SET）脉冲的影响。研究结果表明：当粒子轰击高输入反相器的PMOS管时,NBTI能够导致所产生的SET脉冲的宽度和幅度随时间不断压缩,当粒子轰击低输入反相器的NMOS管时,NBTI能够导致所产生的SET脉冲的宽度和幅度随时间不断展宽。基于研究结果,本文首次提出：NBTI对粒子轰击NMOS管所产生的SET脉冲的影响已经十分严重,在进行抗辐照加固设计时必须考虑NBTI所造成的影响。     

Buffer单元单粒子效应及其若干影响因素研究

基于标准0.13μm工艺使用Sentaurus TCAD软件采用3D器件/电路混合模拟方式仿真了buffer单元的单粒子瞬态脉冲。通过改变重离子的入射条件,得到了一系列单粒子瞬态电流脉冲(SET)。分析了LET值、入射位置、电压偏置等重要因素对SET峰值和脉宽的影响。研究发现,混合模式仿真中的上拉补偿管将导致实际电路中SET脉冲的形状发生明显的变化。     

超深亚微米工艺下的电路级耦合SET脉冲注入

超深亚微米工艺下在电路模拟器中使用独立电流源方法的单粒子瞬态(single event transient,SET)脉冲注入与实验结果有很大误差.作者提出了一种基于二维查找表的耦合电流源注入的方法,并且基于开源的SPICE代码实现.该方法的计算结果与器件/电路混合模拟接近,而其计算时间远小于混合模拟.该法与SPICE集成,可以引入实验测量数据,适合于大规模组合电路的SET错误率分析.     

超深亚微米工艺下的电路级耦合SET脉冲注入

超深亚微米工艺下在电路模拟器中使用独立电流源方法的单粒子瞬态(single event transient, SET)脉冲注入与实验结果有很大误差. 作者提出了一种基于二维查找表的耦合电流源注入的方法,并且基于开源的SPICE 代码实现. 该方法的计算结果与器件/电路混合模拟接近,而其计算时间远小于混合模拟. 该法与SPICE 集成,可以引入实验测量数据,适合于大规模组合电路的SET 错误率分析.     

TCAD结合SPICE的单粒子效应仿真方法

为解决传统集成电路抗单粒子加固设计中存在的不足,利用TCAD及SPICE软件,探索出一种单粒子效应仿真与电路抗辐射加固设计相结合的方法。该方法通过TCAD软件的器件建模、仿真单粒子效应对器件的影响,得出器件在单粒子辐射条件下的3个关键参数。利用SPICE软件将此参数转化为模拟单粒子效应的扰动源,进而指导电路抗单粒子效应的加固设计工作。通过对一款SRAM的加固设计及辐射试验对比,证明了该方法的正确性和有效性,同时也为以后单粒子效应设计加固提供了依据。     

基于 PDSOI工艺的抗 SET锁相环设计

基于0．35μm PDSOI工艺设计了一款输出频率范围为700M Hz-1．0GHz的锁相环电路,利用Sentaurus TCAD工具对其进行单粒子瞬变（SET ）混合模拟仿真,确定其SET敏感部件并建立SET分析模型,分析了SET与锁相环系统参数之间的关系．通过增加由一个感应电阻、一对互补运算放大器和互补SET电流补偿晶体管组成的限流电路并利用多频带结构降低了VCO的增益,显著提升了锁相环的抗SET性能．仿真结果表明,CP中发生SET后VCO控制电压Vc的波动峰值、锁相环的恢复时间以及输出时钟的错误脉冲数明显降低,分别为未加固锁相环的43．9％、49．7％和29．1％,而辐射加固前后 VCO的基本结构变化不大,其SET轰击前后无明显变化．     

基于PDSOI单粒子翻转物理机制的SPICE模型研究

通过研究半导体器件单粒子翻转的物理机制,利用Synopsys TCAD工具对基于中国科学院微电子所开发的0.35μm部分耗尽SOI器件进行单粒子翻转的模拟,讨论了器件模拟物理模型的选择,验证了理论分析的正确性,并对重离子撞击引起的瞬态电流过程进行分析.分析表明单粒子翻转存在两个放电阶段,第一阶段过量电子漂移扩散电流组成激增电流部分;第二阶段部分耗尽SOI器件寄生三极管放电机制以及过量空穴放电机制引起的缓慢电流放电＂尾部＂.结合激增电流的物理意义,提出合理的数学模型,推导出描述此电流的一维解析解;对于缓慢衰减的＂尾部＂电流,提出子电路模型,并基于SPICE三极管模型进行参数提取,着重讨论了单粒子翻转的敏感参数.最后给出了以反相器为例的SPICE模拟与TCAD模拟在瞬态电流,输出节点电荷收集,LET阈值的对比结果,验证了SPICE模型的合理性和精确性.     

VDMOS等效电路的SPICE模型

分析了VDMOS器件中存在的各种寄生效应以及这些寄生效应对器件性能的影响,在此基础上建立了VDMOS等效电路的SPICE模型.通过MEDICI数值分析软件,模拟VDMOS在不同偏置条件下的电压、电流、电容特性,从而提取出VDMOS等效电路模型参数.并用SPICE软件对等效电路仿真,进行了直流分析和瞬态分析,得到等效电路的电学特性曲线图.仿真的结果与MEDICI器件模拟工具模拟结果相互比较,具有较好的一致性.     

NBTI效应导致SET脉冲在产生与传播过程中的展宽

本文研究了负偏置温度不稳定性(NBTI)对单粒子瞬态(SET)脉冲产生与传播过程的影响.研究结果表明:NBTI能够导致SET脉冲在产生与传播的过程中随时间而不断展宽.本文还基于工艺计算机辅助设计模拟软件(TCAD)进行器件模拟,提出了一种在130nm体硅工艺下,计算SET脉冲宽度的解析模型,并结合NBTI阈值电压退化的...     

A Study of PN Junction Diffusion Capacitance of MOSFET in Presence of Single Event Transient

In this paper, a 65 nm MOSFET 3D structure is built based on Technology Computer Aided Design (TCAD) 3D device simulation software, and the single-event transient (SET) effect in 65 nm CMOS inverter is analyzed using TCAD-HSPICE mixed-mode simulation based on heavy ion model. The formation and function of the PN junction diffusion capacitance in the Metal-Oxide-Semiconductor (MOS) device are discussed by analyzing the drain and substrate voltage characteristics of the device under the SET effect. Then the sub-circuit structure of this device for SET is established, and the mechanism of the diffusion capacitance of PN junction during the heavy ion action process is verified comparing with the results of sub-circuit HSPICE simulation results and the TCAD-HSPICE simulation results. Finally, A sub-circuit model is provided, to support circuit-level simulation of single-event effects.     

Research on the effect of single-event transient of an on-chip linear voltage regulator fabricated on 130 nm commercial CMOS technology

We report single-event transient (SET) responses of an on-chip linear voltage regulator in 130 nm commercial standard CMOS technology by heavy ion experiments at first. Responses can be distinguished by the load current. When the light load current was applied, the negative SET on the output of the regulator larger than 200 mV was not observed, while the positive SETs that are larger than 400 mV and last for about 200 ns were observed. By comparison, when the heavy load current was applied, both positive and negative SETs that are larger than 400 mV and last for several hundred ns were observed. Next, the mechanism behind the phenomenon is analysed and then verified by the post-layout SPICE circuit simulation. It is demonstrated that the input voltage, load current and the load capacitance are key elements in determining the severity of SET. Finally, the most sensitive node is located by analysis and SPICE circuit simulation, which lies in the output of the amplifier inside of the bandgap reference (BGR). This result is a primary consideration in the development of the hardening technique.     

Exact analytical model of single electron transistor for practical IC design

This paper presents a new exact analytical model for single electron transistor (SET) applicable for circuit simulation. It has been developed based on orthodox theory of single electronics using master equation where a scheme has been suggested to determine the most probable occupied electron states. The proposed model is more flexible and is valid for single or multi-gate, symmetric or asymmetric devices and can also consider the background charge effect. It can be used for large drain-source voltage range whatever the device is biased under symmetric or asymmetric bias conditions. SET characteristics produced by the proposed model have been verified against widely accepted single electron circuits Monte Carlo simulator SIMON and show a good agreement. Moreover, the model has been implemented in a widely used commercial circuit simulator SPICE to enable simulation with conventional electronic elements and a single electron inverter has been simulated and verified with SIMON results.     

65 nm体硅工艺NMOS中单粒子多瞬态效应的研究

针对NMOS场效应晶体管由重离子辐射诱导发生的单粒子多瞬态现象,参考65 nm体硅CMOS的单粒子瞬态效应的试验数据,采用TCAD仿真手段,搭建了65 nm体硅NMOS晶体管的TCAD模型,并进一步对无加固结构、保护环结构、保护漏结构以及保护环加保护漏结构的抗单粒子瞬态效应的机理和能力进行仿真分析。结果表明,NMOS器件的源结和保护环结构的抗单粒子多瞬态效应的效果更加明显。     

A physical charge-based model for non-fully depleted SOI MOSFET'sand its use in assessing floating-body effects in SOI CMOS circuits

A new model for the non-fully depleted (NFD) SOI MOSFET is developed and used to study floating-body effects in SOI CMOS circuits. The charge-based model is physical, yet compact and thus suitable for device/circuit simulation. Verified by numerical device simulations and test-device measurements, and implemented in (SOI)SPICE, it reliably predicts floating-body effects resulting from free-carrier charging in the NFD/SOI MOSFET, including the purportedly beneficial supra-ideal sub-threshold slope due to impact ionization and a saturation current enhancement due to thermal generation. SOISPICE CMOS circuit simulations reveal that the former effect is not beneficial and could be detrimental, but the latter effect can be beneficial, especially in low-voltage applications, when accompanied by a dynamic floating-body effect that effectively reduces static power. The dynamic floating-body effects are hysteretic, however, and hence exploitation of the beneficial ones will necessitate device/circuit design scrutiny aided by physical models such as the one presented herein     

Single-Event Effects Analysis of a Pulse Width Modulator IC in a DC/DC Converter

Alpha particles, neutrons and laser-beam test results on an integrated pulse width modulation (PWM) controller operating in a DC/DC converter are presented in this paper. The PWM is fabricated on a 600-nm Bi-CMOS technology. Single-Event Transient (SET) derived from a bandgap circuit was amplified by a filter capacitor in the propagation path. Finally, a constant 6-??s SET pulse was observed on PGOOD pin which is a supervisory signal. This glitch caused system resets. Pulsed laser technology was adopted to locate the origin of the SET. 3D TCAD and circuit simulation tools were used to analyze the root cause. System and circuit level hardening approaches to mitigate the SET are also presented.     

SET-based nano-circuit simulation and design method using HSPICE

This paper presents a simulation and design method for complementary SET-based nano-circuits from a practical circuit design point of view. HSPICE behavioral implementation of modified Lientschnig's SET model based on the orthodox theory and the Birth-Death Markov chain is demonstrated and verified with Coulomb characteristics. It shows reduced CPU time, improvement of accuracy, and more compatibility with other SPICE softwares on both Windows and Unix platforms. The proposed design methodology presents how to build static CMOS-like SET circuits, and demonstrates that conventional CMOS circuit design methodologies are all applicable to SET circuit designs based on the methodology. HSPICE simulation results show that, for 1 MΩ junction resistance, the power consumption of a SET NAND2 gate is less than 0.3 pW, and the propagation delay for a SET XOR2 gate is 29.8 ns while driving a 10 aF load.