一个集成的节能电容式传感器数字接口电路外文翻译资料

 2022-05-19 22:24:56

An Integrated Energy-Efficient Capacitive Sensor Digital Interface Circuit

Hesham Omranlowast;, Muhammad Arsalan, Khaled N. Salama

Sensors Lab, Electrical Engineering Program

King Abdullah University of Science and Technology (KAUST) Thuwal, Kingdom of Saudi Arabia, 23955-6900

Abstract

In this paper, we propose an energy-efficient 13-bit capacitive sensor interface circuit. The proposed design fully relies on successive approximation algorithm, which eliminates the need for oversampling and digital decimation filtering, and thus low-power consumption is achieved. The proposed architecture employs a charge amplifier stage to acheive para- sitic insensitive operation and fine absolute resolution. Moreover, the output code is not affected by offset voltages or charge injection. The successive approximation algorithm is implemented in the capacitance-domain using a coarse-fine programmable capacitor array, which allows digitizing wide capacitance range in compact area. Analysis for the maximum achievable resolution due to mismatch is provided. The proposed design is insensitive to any reference voltage or current which translates to low temperature sensitivity. The oper- ation of a prototype fabricated in a standard CMOS technology is experimentally verified using both on-chip and off-chip capacitive sensors. Compared to similar prior work, the fabricated prototype achieves and excellent energy efficiency of 34 pJ/step.

Keywords: Capacitive sensor interface circuit; Capacitance-to-digital conversion (CDC);

Low-power; Energy-efficient; CMOS

lowast;Corresponding author at: King Abdullah University of Science and Technology (KAUST), Thuwal, Saudi Arabia, 23955-6900

Email address: hesham.omran@kaust.edu.sa (Hesham Omran)

Preprint submitted to Elsevier April 23, 2014

Introduction

Capacitive sensors find numerous applications in the consumer, medical, automotive, and industrial sectors [1]. The range of applications includes, but is not limited to, pressure sensors [2, 3], humidity sensors [4], tactile sensors [5], biological sensing microsystems

[6], and chemical detection of volatile organic compounds (VOCs) [7] which can be used as biomarkers for early non-invasive detection of lung cancer [8]. Although the type of application imposes different performance requirements on the sensor interface circuit, energy efficiency is always desirable.

Recent trends in capacitive sensing interface circuits favor direct digitization of the sensor

capacitance, rather than performing capacitance to voltage conversion and then digitizing the output voltage [9, 10, 11, 12]. Direct digitization offers less complexity, smaller area, and lower power consumption [9, 10, 11, 12]. In a “semi-digital” approach, the capacitance can be used to modulate the period or the pulse width of a digital signal [9, 12]. However, this approach requires a time-to-digital converter in order to provide digital output code,

e.g., a fast digital counter and a stable high frequency oscillator [12], which hinders its use in a low-power application. A more attractive approach for capacitance-to-digital conver- sion (CDC) is the use of ∆Sigma; modulators [10, 11, 13]. However, oversampling and digital decimation filtering required in ∆Sigma; architectures translate to large power consumption. In addition, ∆Sigma; interfaces suffer from limited capacitance range to avoid modulator overload

[10, 11]. In order to increase the capacitance range of ∆Sigma; interface circuits, successive approximation register (SAR) algorithm was proposed to adjust the modulator reference capacitor [10, 11]. However, the SAR step was only used for initial coarse calibration, while sensor digitization is still performed using the ∆Sigma; modulator.

A capacitance-to-digital converter (CDC) architecture that fully relies on SAR algorithm will

eliminate the need for oversampling, which will reduce power consumption and relax the re- quirements on the analog blocks. Moreover, digital output code is directly provided with no digital filtering required, which further reduces power consumption. A CDC that fully relies on SAR algorithm was proposed in [14], using an op-amp-less architecture. However, due

to the absence of the op-amp, the sensor capacitance is connected to a high impedance

node which leads to parasitic sensitive design. Moreover, as no charge amplifier stage is present, the change in voltage (∆V ) that needs to be resolved by the comparator will be inversely proportional to the total sum of the capacitances of the parasitic capacitors, the capacitive sensor, and the SAR capacitor array, which will result in poor absolute resolution (more than 60 f F for the design in [14]), in addition to sensitivity to noise, charge injection,

and offset voltage. When the CDC circuit is connected to off-chip capacitive sensor, the parasitic capacitor can be very large leading to degradation of the circuit performance. Fur- thermore, connecting the CDC circuit to off-chip capacitive sensor using a high impedance node makes it highly susceptible to noise coupling.

In this work, an integrated 13 minus; bit SAR CDC that addresses the previous limitations while

maintaining excellent energy efficiency is presented. The proposed CDC employs a charge amplifier stage, which results in insensitivity to parasitic capacitors, insensitivity to charge injection and offset voltages, fine absolute resolution, and immunity to noise coupling. The SAR algorithm is performed in the capacitance-domain using a coarse-fine programmable capacitor array (PCA), which enables digitizing a wide range of capacitance in a compact

area. The proposed CDC is insensitive to the value of any reference voltage or current, which translates to very small t

全文共22501字,剩余内容已隐藏,支付完成后下载完整资料


一个集成的节能电容式传感器数字接口电路

Hesham Omranlowast;, Muhammad Arsalan, Khaled N. Salama

传感器实验室,电气工程计划

阿卜杜拉国王科技大学(KAUST)沙特阿拉伯王国Thuwal,23955-6900

摘 要

在本文中,我们提出了一个高能效的13位电容式传感器接口电路。所提出的设计完全依赖于逐次逼近算法,该算法消除了对过采样和数字抽取滤波的需要,从而实现了低功耗。所提出的架构采用电荷放大器级来实现寄生不灵敏操作和精确的绝对分辨率。而且,输出代码不受偏移电压或电荷注入的影响。逐次逼近算法使用粗略精细的可编程电容器阵列在电容域内实现,该电容器阵列允许在紧凑区域数字化宽电容范围。分析由于失配导致的最大可实现分辨率。所提出的设计对任何参考电压或电流都不敏感,这转化为低温度敏感性。采用标准CMOS技术制造的原型的操作通过片上和片外电容式传感器进行了实验验证。与之前类似的工作相比,制造的原型达到了34 pJ / step的卓越能效。

关键词:电容式传感器接口电路;电容数字转换(CDC);低电量;高效节能; CMOS

*通讯作者:阿卜杜拉国王科技大学(KAUST),沙特阿拉伯Thuwal,23955-6900

电子邮件地址:hesham.omran@kaust.edu.sa(Hesham Omran)

1.简介

电容式传感器在消费者,医疗,汽车和工业部门有广泛的应用[1]。应用范围包括但不限于压力传感器[2,3],湿度传感器[4],触觉传感器[5],生物传感微系统[6]和化学检测挥发性有机化合物(VOCs)[7]可作为早期非侵入性检测肺癌的生物标志物[8]。尽管应用类型对传感器接口电路提出了不同的性能要求,但始终需要提高能效。

电容式传感接口电路的最新趋势有利于传感器的直接数字化电容,而不是执行电容到电压的转换,然后数字化输出电压[9,10,11,12]。直接数字化的复杂度较低,面积较小,功耗较低[9,10,11,12]。在“半数字”方法中,电容可用于调制数字信号的周期或脉冲宽度[9,12]。但是,这种方法需要一个时间数字转换器来提供数字输出代码,例如快速数字计数器和稳定的高频振荡器[12],这阻碍了它在低功耗应用中的使用。对于电容数字转换(CDC)而言,更有吸引力的方法是使用Delta;Sigma;调制器[10,11,13]。然而,Delta;Sigma;体系结构所需的过采样和数字抽取滤波转换为大功耗。此外,Delta;Sigma;接口受限于电容范围,以避免调制器过载[10,11]。为了增加Delta;Sigma;接口电路的电容范围,提出了逐次逼近寄存器(SAR)算法来调整调制器参考电容[10,11]。但是,SAR步骤仅用于初始粗略校准,而传感器数字化仍然使用Delta;Sigma;调制器执行。

完全依赖SAR算法的电容数字转换器(CDC)架构将会消除过采样的需要,这将降低功耗并放宽模拟模块的要求。此外,数字输出码直接提供,不需要数字滤波,这进一步降低了功耗。完全依赖于SAR算法的CDC在[14]中被提出,使用无运算放大器的架构。但是,由于没有运算放大器,传感器电容连接到高阻抗导致寄生敏感设计的节点。此外,由于不存在电荷放大器级,因此需要由比较器解决的电压变化(Delta;V)将与寄生电容器,电容式传感器和SAR电容器的电容总和成反比阵列,这将导致不良的绝对分辨率([14]中的设计超过60 fF),除了对噪声敏感,电荷注入,和偏移电压。当CDC电路连接到片外电容式传感器时,寄生电容可能非常大,导致电路性能下降。此外,使用高阻抗节点将CDC电路连接到片外电容式传感器,使其非常容易受到噪声耦合的影响。在这项工作中,一个集成的13位SAR CDC解决了以前的局限性 介绍了保持优秀的能源效率。所提出的CDC采用电荷放大器级,其对寄生电容不敏感,对电荷注入和偏移电压不敏感,精确的绝对分辨率以及抗噪声耦合。 SAR算法使用粗略精细的可编程电容阵列(PCA)在电容域内执行,该算法能够在一个紧凑型区。建议的CDC对任何参考电压或电流的值都不敏感,这意味着非常小的温度敏感度。由于电路操作对寄生电容不敏感,因此可以与片上和片外电容式传感器以及多路复用电容式传感器阵列一起使用。拟议的SAR CDC采用0.35mu;m标准CMOS技术实施。该原型能够以2.75 f F分辨率对16 pF电容范围进行数字化,同时仅占用0.07 mm2的芯片面积。制造的原型的操作通过使用片上和片外电容式传感器进行实验验证。实现的能效为34 pJ / step,优于最近公布的使用相同技术和电源电压实现的电容时间和Delta;Sigma;CDC [10,这表明了所提出的架构的优点。

本文的其余部分安排如下。在第2节讨论,由于分析电路非理想性,所提出的电路的操作是分离的,。第3节介绍了可编程电容阵列(PCA)。分析和计算由于失配导致的PCA分辨率和动态范围的限制。 第4节给出了系统描述和操作,数字接口的实验结果。第5节给出了制造原型的实验结果。

  1. 连续逼近CDC

2.1.Circuit操作

所提出的SAR CDC电路的原理图如图1所示.CSENS是未知的电容式传感器和CREF是一个实现为二进制加权可编程电容阵列(PCA)的参考电容。 U 1是带有PMOS输入不平衡差分对的两级米勒补偿运算放大器,U 2是比较器。 电路操作分为两个阶段; 预充电阶段和评估阶段。 在预充电阶段(CLK = 1,CLKB = 0),运算放大器(U 1)作为一个单位增益工作缓冲器,即 i.e., VX = VREF 。 感应电容上的电荷(CSENS)由下式给出

Q = CSENS times; VREF (1)

接下来,在评估阶段(CLK = 0,CLKB = 1),CSENS两端的电压差为零,电荷重新分配给参考电容(CREF)和反馈电容(CF)。负载则为

Q=CREFtimes;VREF (VREFminus;Vo)times;CF (2)

由于电荷是守恒的,从(1)和(2)中,运算放大器的输出由下式给出

(3)

因此,比较器(U2)的差分输入由下式给出幅度

并且比较器的输出由下式给出

(4)

基于VCMP,SAR逻辑使用二进制搜索逐次逼近算法改变PCA数字输入来增加或减少CREF。为了测试的灵活性,该逻辑在片外实现。在转换周期结束时, CREF的值在PCA的1 LSB误差范围内与 CSENS 匹配。转换时间是Ntimes;TCLK,,其中N是PCA的位数,TCLK是转换时钟(CLK)的周期。 VREF的值设置为1.45 V,大致位于运算放大器和比较器的共模输入范围的中间。但是,电路操作对VREF 的值不敏感。 VREF 的任何变化或漂移都会同样影响两者 U 1和U 2,因此它将被取消并且(4)将保持有效。无论其绝对值如何,比较器的数字输出由Delta;Vo的符号决定。 VREF中的任何漂移或变化都会改变Delta;Vo的绝对值,但不会改变其值标志,因为符号由差异决定(CREF minus; CSENS)。因此,变化在VREF 中不会影响电路功能,因为它不会改变输出比较。

图1:提出的SAR CDC电路原理图。

对于SAR ADC,只需要一个有源模块(比较器),因为输入信号是缓冲模拟电压。然而,对于直接将电容转换为数字输出的SAR CDC,由于若干原因,运算放大器是必要的。首先,电容式传感器(CSENS)与大型寄生电容器相关联,因此需要将其放置在两个低阻抗节点之间,这是通过运算放大器虚拟地实现的。CSENS CREF端子处的寄生电容或者接地,VREF或者U1的虚拟接地反相输入,这些都是低阻抗节点。因此,寄生电容不影响电路操作。其次,运放作为电荷放大器,提高了接口的绝对分辨率(LSB)。在不使用运算放大器(即无运算放大器架构)的情况下,在评估阶段,电荷将在所有电容之间重新分配,因为输出节点(Vo = VX)是浮动节点。因此,Delta;Vo将与连接到节点Vx的所有电容的总和成反比,即其中CP X 是节点VX处的寄生电容,而不是与小反馈电容器成反比(CF),其中总和(CSENS CREF CP X)比CF大几个数量级。此外,得益于电荷放大器,输出代码不会受到偏移电压和电荷注入的影响,这将在下一小节中介绍。第三,通过U1虚拟地,连接到片外传感器的节点VX是低阻抗节点,因此不容易产生噪声耦合。

2.2.电路非理想

之前的分析假定理想的操作。在本节中,我们将研究运算放大器,比较器和反馈开关的非理想性。首先,考虑运算放大器 (Vos,opminus;amp)的偏移电压,在预充电阶段,CSENS, CREF ,和CF上的电荷将由

qCSENS= CSENS times; (VREF Vos,opminus;amp) (5a)

qCREF= CREF times; Vos,opminus;amp (5b)

qCF= 0 (5c)

接下来,在评估阶段,负载由下列公式得出:

qCSENS= CSENS times; Vos,opminus;amp (6a)

qCREF= CREF times; (VREF Vos,opminus;amp) (6b)

qCF= CF times; (VREF Vos,opminus;amp minus; Vo) (6c)

由于(5)和(6)两个阶段的总充电量保持不变,运算放大器的输出将由下式给出

(7)

这意味着运算放大器的失调电压不会被放大,而是出现在输出端。通过考虑比较器偏移电压(Vos,cmp),比较器数字输出将由下式给出

(8)

VDelta;Vo由下式给出

(9)

而且,由于反馈开关,当电路从预充电阶段移动到评估阶段时,即在CLK的下降沿,电荷被注入到CF中。从而,

(8)将被修改为

(10)

其中Verr是由下式给出的总误差电压

Verr = Vos,tot Vinj (11)

其中Vos,tot = Vos,opminus;amp Vos,cmp ,cmp是总偏移电压,Vinj是开关电荷注入引起的误差电压。 Vinj是反馈电容(CF)的函数,随着CF的减小而增加。运算放大器和比较器的失调电压是两个独立的随机变量。统计蒙特卡洛模拟用于生成它们的分布,捕获不匹配和工艺变化的影响。图(2)描述了由运算

全文共7186字,剩余内容已隐藏,支付完成后下载完整资料


资料编号:[12099],资料为PDF文档或Word文档,PDF文档可免费转换为Word

原文和译文剩余内容已隐藏,您需要先支付 30元 才能查看原文和译文全部内容!立即支付

以上是毕业论文外文翻译,课题毕业论文、任务书、文献综述、开题报告、程序设计、图纸设计等资料可联系客服协助查找。