WO2010111857A1 - 亚阈值集成电路中抗工艺涨落的方法和体电位调制电路 - Google Patents
亚阈值集成电路中抗工艺涨落的方法和体电位调制电路 Download PDFInfo
- Publication number
- WO2010111857A1 WO2010111857A1 PCT/CN2009/073744 CN2009073744W WO2010111857A1 WO 2010111857 A1 WO2010111857 A1 WO 2010111857A1 CN 2009073744 W CN2009073744 W CN 2009073744W WO 2010111857 A1 WO2010111857 A1 WO 2010111857A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- potential modulation
- modulation circuit
- body potential
- mos device
- circuit
- Prior art date
Links
Classifications
-
- H—ELECTRICITY
- H03—ELECTRONIC CIRCUITRY
- H03C—MODULATION
- H03C1/00—Amplitude modulation
- H03C1/36—Amplitude modulation by means of semiconductor device having at least three electrodes
Definitions
- the invention relates to a method for resisting process fluctuations in a sub-threshold integrated circuit and an implementation circuit thereof, and belongs to the technical field of integrated circuits. Background technique
- CMOS integrated circuits In order to meet the urgent needs of low-power consumption of portable devices and the energy-saving needs of large-scale systems, low-voltage and low-power consumption have become the main directions for the development of CMOS integrated circuits (ICs) in the future.
- the threshold voltage cannot be reduced in proportion to the power supply voltage. Therefore, low-voltage, low-power analog integrated circuit design faces enormous challenges, and the design of operational amplifiers is the main bottleneck of low-voltage analog circuits.
- Subthreshold technology is an effective means to solve the problem of low operating voltage. In the sub-threshold circuit, some of the MOSFETs operate in the subthreshold region and are well suited for low voltage operating environments. Because of this, subthreshold technology is widely used in the field of analog integrated circuits.
- inverters instead of traditional operational amplifiers to achieve low-voltage high-performance switched-capacitor circuits has become a research hotspot.
- the inverters use a similar working mode as Class C amplifiers.
- the input transistors work in most of the time.
- the threshold area minimizes system power consumption.
- this type of inverter a Class-C Inverter.
- switched-capacitor integrators and sigma-delta modulators based on Class C inverters have been reported, for example, Youngcheol Chae and Gunhee Han, “Low Voltage, Low Power, Inverter-Based Switched- Capaci tor Delta- Sigma Modulator", IEEE Journal of Sol id - State Circuit, 2009, 44 (2): p. 458-472.
- the technical problem to be solved by the present invention is to provide a method for resisting process fluctuations of a sub-threshold integrated circuit, so as to overcome the influence of process fluctuations on the MOS device in the sub-threshold working area in the prior art. This causes a sub-threshold integrated circuit performance degradation or even a loss of functionality.
- Another technical problem to be solved by the present invention is to provide a body potential modulation circuit that realizes the above-described method for resisting process fluctuations.
- Another technical problem to be solved by the present invention is to provide a method for applying the above-described body potential modulation circuit to a class C inverter and data of the application effect thereof, thereby realizing a c-class inversion against process fluctuations.
- a method for applying the above-described body potential modulation circuit to a class C inverter and data of the application effect thereof thereby realizing a c-class inversion against process fluctuations.
- the inductive M0S device senses the trend of the target M0S device at different process angles and outputs it as a drain-source induced current
- the induced current signal outputted by the inductive MOS device is converted into a voltage signal by a current-to-voltage circuit, and the variation characteristic of the induced current is reflected on the voltage signal in real time;
- the voltage signal output from the current-to-voltage circuit is fed back to the body end of the target M0S device to form an inductive feedback loop for body potential modulation, which reduces the influence of process fluctuations on the performance parameters of the target device.
- the body potential modulation circuit of the present invention is an implementation circuit of the anti-process fluctuation method, comprising: a target MOS device, the object of the anti-process fluctuation method;
- the sensing M0S device is used to sense the parameter variation trend of the target MOS device at different process angles;
- the current-to-voltage circuit is configured to convert the induced current outputted by the sensing MOS device into an induced voltage, and feed the induced voltage to the target MOS device.
- the body end realizing the body potential modulation of the target MOS device;
- the body potential modulation circuit of the present invention can be classified into two types: a PM0S body potential modulation circuit and an NM0S body potential modulation circuit.
- the PM0S body potential modulation circuit is used to realize the process fluctuation of the PM0S tube in the subthreshold state, and includes a first PM0S tube M1, a second PMOS tube M2 and a first resistor R1, wherein the first PM0S tube M1 is a PM0S body
- the target MOS device of the potential modulation circuit, the second PMOS transistor M2 is an inductive MOS device of the PM0S body potential modulation circuit, and the first resistor R1 realizes the function of the current to voltage circuit in the PM0S body potential modulation circuit;
- the source end of the second PM0S tube M2 is connected to the body end, and the drain end is connected to one end of the first resistor R1 and the body end of the first PM0S tube M1, and the other end of the first resistor R1 is connected to the common mode voltage.
- the NM0S body potential modulation circuit is used to realize the process fluctuation of the NM0S tube in the subthreshold state, and includes a first ⁇ OS tube M3, a second ⁇ OS tube M4 and a second resistor R2, wherein the first ⁇ OS tube ( M3) is the target M0S device of the ⁇ OS body potential modulation circuit, and the second NM0S tube M4 is the induction of the NM0S body potential modulation circuit
- the MOS device, the second resistor R2 is a current-to-voltage circuit of the NMOS body potential modulation circuit; the source end of the second NM0S tube M4 is connected to the body end, and the drain end is respectively connected to one end of the second resistor R2 and the body of the first NM0S tube M3
- the terminals are connected, and the other end of the second resistor R2 is connected to the common mode voltage.
- the body potential modulation circuit of the present invention By applying the body potential modulation circuit of the present invention to a class C inverter, a class C inverter resistant to process fluctuations can be realized, and the class C inverter is based on the prior art class C inverter.
- the PM0S body potential modulation circuit and the NM0S body potential modulation circuit according to the present invention are added, wherein the prior art C-type inverter is used to implement an operational amplification function, and the PM0S and NM0S body potential modulation circuits are used to resist process fluctuations.
- the PM0S and NM0S input transistors in the prior art C-type inverter (51) are the first PMOS as the target MOS device in the PM0S body potential modulation circuit (52) and the NMOS body potential modulation circuit (53), respectively.
- the anti-process fluctuation method of the present invention can modulate the electrical parameters of the target MOS device in real time by sensing the body potential modulation of the feedback loop, and effectively reduce the target MOS device in the subthreshold state.
- the body potential modulation circuit of the present invention realizes the entire inductive feedback loop with fewer circuit components, and its introduction can effectively improve the inclusion of the C-type inverter without significantly increasing the circuit complexity and power consumption.
- the performance stability, consistency and product yield of the sub-threshold integrated circuits are therefore of high practical value.
- FIG. 2 is a circuit configuration diagram of a PM0S body potential modulation circuit of the present invention
- FIG. 3 is a circuit structural diagram of a NM0S body potential modulation circuit of the present invention.
- FIG. 4 is a circuit structural diagram of a prior art C-type inverter
- Fig. 5 is a structural diagram of a circuit of a class C inverter resistant to process fluctuations of the present invention.
- the method for resisting process fluctuations in a sub-threshold integrated circuit proposed by the present invention uses a target M0S device, an inductive MOS device, and a current-to-voltage circuit in the implementation process. among them:
- Target MOS device the object of resistance to the process fluctuation method.
- the body end of the target MOS device needs to be separately cited Out, the body potential is adjustable.
- both the PM0S device and the NM0S device can be used to separate the body end.
- Inductive M0S device used to sense the trend of parameters of the target M0S device at different process angles.
- the sense M0S device has the same type as the target M0S device, the layout is matched, and the working state is the same. Therefore, the process fluctuation of the sensed M0S device at any time is similar to that of the target M0S device, that is, the induced transconductance of the M0S device and the output current change trend are the same as the target M0S device.
- the inductive M0S device is capable of sensing changes in the parameters of the transconductance, output current, etc. of the target M0S device at different process angles.
- the current-to-voltage circuit is configured to convert the induced current (the output current of the inductive MOS device) into an induced voltage, and feed the induced voltage to the body end of the target MOS device to realize the body potential modulation of the target MOS device.
- the key to the process-resistant fluctuation method of the present invention lies in body potential modulation, and the following relationship exists between the threshold voltage and the body potential of the MOS device:
- the flow chart of the anti-process fluctuation method of the present invention is shown in Fig. 1. It includes a target MOS device 11, an inductive MOS device 12, a current-to-voltage circuit 13, and the like.
- the parameter fluctuation of the target MOS device 11 due to factors such as process fluctuations is sensed in real time to the sensing MOS device 12, and the induced output current of the MOS device 12 is fed back to the body end of the target MOS device 11 through the current converting voltage circuit 13 to realize an induction.
- the feedback loop effectively reduces the sensitivity of the target MOS device 11 to process fluctuations by body potential modulation.
- a current-to-voltage circuit 13 is designed to convert the induced output current I into a voltage signal V B and to change the trend of V B to I. UT2 is the same, so V B decreases with decreasing I.
- the voltage signal V B finally reaches the body terminal of the target PMOS device 11, and the absolute value of the threshold voltage of the target PMOS device 11 is reduced by the body potential modulation, the transconductance and the output current are increased, and the entire inductive feedback loop forms a negative feedback, which is effective.
- the ground weakens the effect of process fluctuations on the target PMOS device 11. For NMOS devices, the process of establishing negative feedback is similar.
- Target M0S device and the connection of the sensing M0S device body The target M0S device body terminal needs to be separately extracted to achieve adjustable body potential.
- the sensing M0S device needs to faithfully reflect the influence of the process fluctuation, so the body terminal is the conventional connection method: The sensing NM0S device body is connected to the low level, and the sensing PM0S device body is connected to the high level.
- the size of the sensing M0S device is not necessarily equal to the size of the target M0S device.
- the channel lengths of both are the same, and the channel width is appropriately proportioned.
- the anti-process fluctuation method of the present invention is generally used in a subthreshold integrated circuit.
- the PM0S body potential modulation circuit and the NM0S body potential modulation circuit of the present invention are shown in Figs. 2 and 3, respectively.
- the PM0S body potential modulation circuit is used to realize the resistance fluctuation of the PM0S tube in the subthreshold state. It consists of a target PM0S tube M1, an inductive PM0S tube M2 and a resistor R1. Assume that the target PM0S tube M1 is in the sub-threshold state, and the potentials of the gate, drain and source are all provided by the circuit where it is located, and the gate-source voltage (V eP -V DDH ) of the sensed PM0 tube M2 is set to work in the sub-threshold state. Therefore, M2 can sense the change trend of parameters such as transconductance and output current of M1 at different process angles.
- the resistor R1 converts the induced current signal (M2 output current) into a voltage signal V BP and then feeds back to the body terminal of M1 to form an inductive feedback loop for body potential modulation.
- the M2 source terminal potential V DDH determines the upper limit of the Ml body potential modulation range (V BP range), which can be set according to the actual application; the common mode voltage VTM determines the lower limit of the M1 body potential modulation range.
- the modulation of the voltage signal V BP at the M1 body end ie, adjusting the source voltage of M1 makes the transconductance and output current of M1 relatively uniform at different process angles.
- the NM0S body potential modulation circuit is used to realize the process fluctuation of the NM0S tube under the subthreshold state. It consists of a target NM0S tube M3, an inductive NM0S tube M4 and a resistor R2.
- the resistor R2 converts the induced output current into a voltage signal and feeds back to the body terminal of M3 for body potential modulation.
- the inductive ⁇ OS tube M4 operates in the same state as the target NMOS tube M3 (all operating in a subthreshold state).
- the M4 source potential GNDL determines the lower limit of the M3 body potential modulation range (V BN value range), and the common mode voltage VTM determines the upper limit of the M3 body potential modulation range.
- the absolute value of the M1 threshold voltage becomes large, which causes the transconductance to decrease when Ml operates in the subthreshold region, and the bandwidth decreases, and the output current reaches the minimum value. Since the sensed PM0 tube M2 can sense the current trend of M1, the induced output current of M2 will also reach the minimum value I. UT2 — ss . Therefore, the Ml body potential V BP (ie, V eM +I.
- the voltage signal is fed back to the body end of M1, and the absolute value of the M1 threshold voltage is slightly lowered by the body potential modulation, Ml is The transconductance and output current increase during subthreshold operation, thereby achieving negative feedback modulation of the M1 parameter.
- Parameters such as UT2 and R1 can make the PM0S body potential modulation circuit generate a suitable V BP at different process angles, so that the transconductance and output current of the target PM0S tube M1 are consistent when working in the subthreshold region.
- the body potential modulation circuit of the present invention realizes the entire inductive feedback loop with fewer circuit elements, and the sensed MOSFETs M2 and M4 in both operate in the subthreshold region, and the power consumption is very low. Therefore, the introduction of the body potential modulation circuit can effectively reduce the adverse effect of the process fluctuation on the performance of the MOSFET without significantly increasing the circuit complexity and power consumption.
- Resistor Rl, R2 should use a resistor with a small tolerance.
- the V BP potential should not be too low, so as to prevent the source body junction of the M1 tube from being excessively positively biased, resulting in a significant increase in leakage current.
- the V BN potential It should not be too high, so as to avoid excessive positive deviation of the source of the M3 tube.
- Class C inverter technology is an emerging low-voltage analog circuit design technology that uses a Class C inverter instead of a conventional op amp to implement many low-voltage, low-power switched-capacitor circuits.
- a circuit configuration diagram of a prior art c-type inverter is shown in FIG.
- the C-type inverter supply voltage V DD is slightly lower than the sum of the threshold voltages of the two input transistors of the inverter.
- the circuit structure diagram of the C-type inverter resistant to process fluctuation of the present invention is as shown in FIG. 5, which adds the PM0S body potential modulation circuit 52 of the present invention to the prior art C-type inverter 51.
- the ⁇ OS body potential modulation circuit 53, the PMOS and NM0S input tubes of the prior art C-type inverter 51 are the target PMOS tube M1 and the target ⁇ OS tube M3 of the body potential modulation circuit, respectively.
- the prior art class C inverter 51 is used to implement an operational amplification function. It consists of PM0S input pipe Ml and NM0S input pipe M3. The body terminals of the inverter input tubes Ml and M3 are separately led out, and the body potential is adjustable.
- the PM0S body potential modulation circuit 52 and the NM0S body potential modulation circuit 53 of the present invention are respectively used for the process fluctuation of M1 and M3, so that the transconductance and the output current of Ml and M3 are relatively uniform at different process angles.
- the transconductance and output current of the input transistors M1 and M3 are directly related to the steady-state indexes such as the gain, bandwidth, and static power consumption of the entire inverter, so the introduction of the body potential modulation circuit 52 53 can be effective.
- the sensitivity of each steady state index of the C-type inverter to the process fluctuation is reduced.
- the maximum deviation of gain, bandwidth and static power consumption is 27.8% 52.3% and 8%, compared with 28% of the prior art C-type inverters 435.8% and 577.4%, this
- the invention is greatly reduced by the influence of process fluctuations, which not only ensures sufficient gain and bandwidth, but also avoids unnecessary static power consumption, and the effect is very obvious; if only the extra positive level is introduced in the body potential modulation circuit, no negative power is introduced.
- the maximum deviation of the class inverters is 25.5% 287. % and 425.7%, which is also better than the prior art C-type inverters.
- the unity gain bandwidth index, the unity gain bandwidth of the prior art class C inverter that does not use the technique of the present invention is only about 5 MHz at the ss process angle, and cannot operate normally at high frequencies, and C using the present technology. Class inverters do not suffer from such serious problems under any circumstances.
- the anti-process fluctuation method of the invention realizes the target by sensing the body potential modulation of the feedback loop
- Real-time modulation of parameters such as transconductance and output current of the M0S device effectively reduces the sensitivity of the M0S device to process fluctuations in the subthreshold state.
- the body potential modulation circuit of the present invention realizes the entire inductive feedback loop with fewer circuit components, and its introduction can effectively improve the inclusion of the C-type inverter without significantly increasing the circuit complexity and power consumption.
- the performance stability, consistency and product yield of the sub-threshold integrated circuits are therefore of high practical value.
Landscapes
- Amplifiers (AREA)
- Semiconductor Integrated Circuits (AREA)
- Metal-Oxide And Bipolar Metal-Oxide Semiconductor Integrated Circuits (AREA)
Description
技术领域
本发明涉及一种在亚阈值集成电路中抗工艺涨落的方法及其实现电路, 属于集 成电路技术领域。 背景技术
为了满足便携式设备低功耗的迫切需求以及大型系统的节能需要, 低压低功 耗成为未来 CMOS集成电路 (IC ) 发展的主要方向。 然而, 考虑到 M0S器件漏电流 的影响, 其阈值电压不可能与电源电压同比例降低。 因此, 低压低功耗的模拟集成 电路设计面临着巨大的挑战,其中运算放大器的设计更是低压模拟电路的主要瓶颈 所在。亚阈值技术是解决低工作电压难题的一种有效手段。在亚阈值电路中, 部分 M0S管工作在亚阈值区域, 能够很好地适用于低电压的工作环境。 正因为如此, 亚 阈值技术在模拟集成电路领域得到了广泛的运用。
近年来, 用反相器代替传统的运算放大器来实现低压高性能开关电容电路已 成为研究热点,其中反相器采用与 C类放大器类似的工作方式,输入管在大多数时 间内都工作在亚阈值区,最大限度地降低了系统功耗。我们称此类反相器为 C类反 相器 (Class-C Inverter) 。 目前, 基于 C类反相器的开关电容积分器和∑-Δ调制 器等均已有所报道, 例如, Youngcheol Chae and Gunhee Han, "Low Voltage, Low Power, Inverter-Based Switched- Capaci tor Delta- Sigma Modulator" , IEEE Journal of Sol id— State Circuit, 2009, 44 (2) : p. 458— 472。
但是, 当 MOS器件工作在亚阈值区时, 相比于工作在正常导通状态, 其跨导、 输出电流等指标受不同工艺角的影响更为严重。因此,工艺涨落很有可能造成亚阈 值集成电路的性能下降甚至功能丧失,这极大程度地降低了包括 C类反相器在内的 亚阈值集成电路的性能稳定性、 一致性和产品良率, 最终影响其实用性。 发明内容
本发明要解决的技术问题是, 提供一种亚阈值集成电路抗工艺涨落的方 法, 以克服现有技术中处于亚阈值工作区的 M0S器件受工艺涨落影响较大, 进
而造成亚阈值集成电路性能下降甚至功能丧失的不足。
本发明要解决的另一技术问题是, 提供一种实现上述抗工艺涨落方法的体电 位调制电路。
本发明要解决的又一技术问题是, 提供一种将上述体电位调制电路应用于 c 类反相器中的方法及其应用效果的数据, 实现了一种抗工艺涨落的 c类反相器, 以 克服现有技术的 c类反相器抗工艺涨落能力较弱、 稳定性和实用性较差的不足。
本发明的抗工艺涨落方法步骤如下:
由感应 M0S器件感应目标 M0S器件在不同工艺角下的参数变化趋势, 并以漏 源感应电流形式输出;
通过电流转电压电路将感应 M0S器件输出的感应电流信号转变为电压信号, 将感应电流的变化特征实时地反映到该电压信号上;
将电流转电压电路输出的电压信号反馈到目标 M0S器件的体端, 形成感应反 馈环路, 用以体电位调制, 减弱工艺涨落对目标器件性能参数的影响。
本发明的体电位调制电路是所述抗工艺涨落方法的一种实现电路,它包括: 目标 M0S器件, 所述抗工艺涨落方法的作用对象;
感应 M0S器件, 用于感应目标 M0S器件在不同工艺角下的参数变化趋势; 电流转电压电路,用于将感应 M0S器件输出的感应电流转换成感应电压, 并将 该感应电压反馈到目标 M0S器件的体端, 实现目标 M0S器件的体电位调制;
本发明的体电位调制电路可分为 PM0S体电位调制电路和 NM0S体电位调制电路 两种类型。
PM0S体电位调制电路用于实现 PM0S管在亚阈值状态下的抗工艺涨落, 它包括 第一 PM0S管 Ml、 第二 PMOS管 M2和第一电阻 Rl, 其中, 第一 PM0S管 Ml为 PM0S 体电位调制电路的目标 M0S器件, 第二 PM0S管 M2为 PM0S体电位调制电路的感应 M0S器件, 第一电阻 R1实现 PM0S体电位调制电路中电流转电压电路的功能;
第二 PM0S管 M2 的源端与体端相连, 漏端分别与第一电阻 R1 的一端和第一 PM0S管 Ml的体端相连, 第一电阻 R1的另一端与共模电压相连。
NM0S体电位调制电路用于实现 NM0S管在亚阈值状态下的抗工艺涨落,它包括 第一匪 OS管 M3、第二匪 OS管 M4和第二电阻 R2, 其中,第一匪 OS管(M3 )为匪 OS 体电位调制电路的目标 M0S器件, 第二 NM0S管 M4为 NM0S体电位调制电路的感应
MOS器件, 第二电阻 R2为 NMOS体电位调制电路的电流转电压电路; 第二 NM0S管 M4 的源端与体端相连, 漏端分别与第二电阻 R2 的一端和第一 NM0S管 M3的体端相连, 第二电阻 R2的另一端与共模电压相连。
将本发明的体电位调制电路应用于 C类反相器中,可实现一种抗工艺涨落的 C 类反相器, 该 C类反相器在现有技术的 C类反相器的基础上增加了本发明所述的 PM0S体电位调制电路和 NM0S体电位调制电路,其中现有技术的 C类反相器用于实 现运算放大功能, 而 PM0S和 NM0S体电位调制电路用于抗工艺涨落, 现有技术的 C 类反相器 (51 ) 中的 PM0S和 NM0S输入管分别为所述 PM0S体电位调制电路 (52 ) 和 NM0S体电位调制电路 (53 ) 中作为目标 M0S器件的第一 PM0S管 (Ml ) 和第一 NM0S管 (M3 ) 。
本发明的优点和积极效果: 本发明所述的抗工艺涨落方法通过感应反馈环路 的体电位调制, 能够实时调制目标 M0S器件的电学参数, 并有效降低目标 M0S器件 在亚阈值状态下对于工艺涨落的敏感度。本发明所述的体电位调制电路以较少的电 路元件实现了整个感应反馈环路,它的引入能够在不明显增加电路复杂度和功耗的 情况下有效地提高包括 C类反相器在内的亚阈值集成电路的性能稳定性、一致性和 产品良率, 因而具有较高的实用价值。 附图概述
本发明的特征、 性能由以下的实施例及其附图进一步描述。
图 1为本发明的抗工艺涨落方法实施流程图;
图 2为本发明的 PM0S体电位调制电路的电路结构图;
图 3为本发明的 NM0S体电位调制电路的电路结构图;
图 4为现有技术的 C类反相器的电路结构图;
图 5为本发明的抗工艺涨落的 C类反相器电路结构图。 本发明的最佳实施方式
本发明提出的在亚阈值集成电路中抗工艺涨落的方法在实施过程中要用到目 标 M0S器件、 感应 M0S器件以及电流转电压电路等。 其中:
目标 M0S器件, 抗工艺涨落方法的作用对象。 目标 M0S器件的体端需单独引
出, 实现体电位可调。 在现在较为流行的三阱工艺中, PM0S器件和 NM0S器件均可 实现体端单独引出。
感应 M0S器件, 用于感应目标 M0S器件在不同工艺角下的参数变化趋势。 感 应 M0S器件与目标 M0S器件类型相同, 版图匹配, 且工作状态相同。 因此, 感应 M0S器件在任意时刻的工艺涨落程度与目标 M0S器件近似相同, 即感应 M0S器件跨 导、输出电流的变化趋势与目标 M0S器件相同。换句话说, 感应 M0S器件能够感应 目标 M0S器件在不同工艺角下的跨导、 输出电流等参数的变化。
电流转电压电路, 用于将感应电流(感应 M0S器件的输出电流)转换成感应电 压,并将该感应电压反馈到目标 M0S器件的体端,实现目标 M0S器件的体电位调制。
本发明的抗工艺涨落方法的关键在于体电位调制, M0S器件的阈值电压和体电 位之间存在如下关系:
其中 是 M0S管的源体电压, 。是 =0时的阈值电压, 是体效应系数, 是费米势。 由上式可知, 通过体电位调制 (调整^ ^ 可以改变 M0S器件的阈值 电压, 从而间接地改变 M0S器件的跨导和输出电流。
本发明的抗工艺涨落方法实施流程图如附图 1所示,它包括目标 M0S器件 11、 感应 M0S器件 12以及电流转电压电路 13等。 目标 M0S器件 11由于工艺涨落等因 素导致的参数波动被实时感应到感应 M0S器件 12上,感应 M0S器件 12的感应输出 电流通过电流转电压电路 13反馈到目标 M0S器件 11体端,实现一个感应反馈环路, 通过体电位调制有效降低目标 M0S器件 11对于工艺涨落的敏感度。
在本发明的抗工艺涨落方法中, 感应反馈环路是如何减弱工艺涨落对于目标 M0S器件的不利影响的, 我们以 PM0S器件为例进行说明。
假设在初始时刻工艺涨落导致目标 PM0S器件 11跨导和输出电流 I。UT1减小, 由于感应 PM0S器件 12能够感应目标 PM0S器件 11在不同工艺角下的参数变化,所 以感应 PM0S器件 12的感应输出电流 I。UT2随之减小。 设计一个电流转电压电路 13 将感应输出电流 I,转变为电压信号 VB, 并使 VB的变化趋势与 I。UT2相同, 所以 VB 随 I,减小而减小。 电压信号 VB最终到达目标 PM0S器件 11的体端, 并通过体电位 调制使目标 PM0S器件 11阈值电压绝对值减小,跨导和输出电流增大,整个感应反 馈环路形成一个负反馈, 有效地减弱工艺涨落对目标 PM0S器件 11的影响。
对于 NMOS器件而言, 负反馈的建立过程是类似的。
关于本发明的抗工艺涨落方法需要说明的是:
( 1 ) 目标 M0S器件和感应 M0S器件体端的接法。 目标 M0S器件体端需单独引 出, 实现体电位可调。而感应 M0S器件需要如实地反映工艺涨落的影响, 所以其体 端为常规接法: 感应 NM0S器件体端接低电平, 感应 PM0S器件体端接高电平。
( 2 ) 因为感应 M0S器件只需感应目标 M0S器件各参数的变化趋势, 所以感应 M0S器件的尺寸与目标 M0S器件的尺寸没必要完全相等。 在实际应用中, 出于芯片 面积、功耗和匹配精度的折衷考虑, 建议两者沟道长度一致, 沟道宽度取适当比例
(比如在 1/8至 1/20之间) 。
( 3 ) 由于 M0S器件处于亚阈值区时对工艺涨落极为敏感, 所以本发明的抗工 艺涨落方法一般用于亚阈值集成电路中。
本发明的 PM0S体电位调制电路和 NM0S体电位调制电路分别见图 2和图 3。
PM0S体电位调制电路, 用于实现 PM0S管在亚阈值状态下的抗工艺涨落。它由 目标 PM0S管 Ml、 感应 PM0S管 M2和电阻 R1组成。 假设目标 PM0S管 Ml处于亚阈 值状态, 其栅、 漏、 源三端电位均由其所在电路提供, 设置感应 PM0S管 M2的栅源 电压(VeP-VDDH)使其同样工作在亚阈值状态, 因此 M2能够感应 Ml在不同工艺角下 的跨导、 输出电流等参数的变化趋势。 电阻 R1将感应电流信号 (M2输出电流) 转 变为电压信号 VBP, 然后反馈到 Ml的体端, 形成感应反馈环路, 用以体电位调制。
M2源端电位 VDDH决定 Ml体电位调制范围 (VBP取值范围) 的上限, 可根据实际应用 设定; 共模电压 V™决定 Ml体电位调制范围的下限。根据本发明所述的抗工艺涨落 方法, 通过电压信号 VBP在 Ml体端的调制作用 (即调节 Ml的源体电压) , 使得 Ml 的跨导和输出电流在不同的工艺角下较为一致。
NM0S体电位调制电路, 用于实现 NM0S管在亚阈值状态下的抗工艺涨落。它由 目标 NM0S管 M3、 感应 NM0S管 M4和电阻 R2组成。 电阻 R2将感应输出电流转变为 电压信号 ^并反馈到 M3的体端, 用以体电位调制。 类似地, 感应匪 OS管 M4与目 标 NM0S管 M3工作状态相同 (均工作于亚阈值状态) 。 M4源端电位 GNDL决定 M3 体电位调制范围(VBN取值范围)的下限, 共模电压 V™决定 M3体电位调制范围的上 限。 通过电压信号 ^在 3体端的调制作用, 使得 M3的跨导和输出电流在不同的 工艺角下较为一致。
本发明的 PMOS体电位调制电路的工作原理具体说明如下:
当工艺角为 tt ( typical-typical ) 时, 设 M2的感应输出电流为 I。UT2 tt, 调节 I。UT2 tt (与 M2尺寸、 源端电位 VDDH等有关) 以及 R1等参数使得 Ml体端电位 VBP (即 VCM+IOUT2_«R1 ) ^ VDD (假设未引入体电位调制电路时 Ml体端电位为电源电压 VDD ) , 电路进入典型 (tt) 工作状态。
当工艺角为 ss ( slow-slow) 时, Ml阈值电压的绝对值变大, 导致 Ml在亚阈 值区工作时跨导减小, 带宽降低, 此时输出电流达到最小值。 由于感应 PM0S管 M2 能够感应到 Ml的电流变化趋势, 所以 M2的感应输出电流也将达到最小值 I。UT2— ss。 因此 Ml体端电位 VBP (即 VeM+I。UT2— SS R1 ) 〈 VDD, 将该电压信号反馈到 Ml的体端, 通 过体电位调制使 Ml阈值电压的绝对值略为降低, Ml在亚阈值区工作时跨导和输出 电流增大, 从而实现了对 Ml参数的负反馈调制。
当工艺角为 ff (fast-fast) 时, Ml阈值电压的绝对值变小, 导致 Ml跨导增 大, 此时 M2的输出电流达到最大值 I。UT2 ff。 此时 VBP (即 V„+I。UT2— ff Rl ) > VDD, 将其 反馈到 Ml的体端,使 Ml阈值电压的绝对值提高,跨导和输出电流减小,功耗降低。 需要注意的是, 由于 M2源端电位 VDDH决定了 VBP的上限, 所以只有当 VDDH>VDD时, Ml 在 ff工艺角下的体电位调制才有效。
综上所述, 通过调节 M2的感应输出电流 I。UT2以及 R1等参数, 可使 PM0S体电 位调制电路在不同工艺角下均产生较为合适的 VBP, 使目标 PM0S管 Ml在亚阈值区 工作时跨导和输出电流较为一致。
本发明的体电位调制电路以较少的电路元件实现了整个感应反馈环路, 且其 中的感应 M0S管 M2和 M4均工作在亚阈值区, 功耗非常低。 因此, 体电位调制电路 的引入能够在不明显增加电路复杂度和功耗的情况下有效减弱工艺涨落对 M0S 管 性能的不利影响。
关于本发明的体电位调制电路在实际运用中需要注意的是:
( 1 ) 电阻 Rl、 R2应选用允差较小的电阻。
(2 )在 PM0S体电位调制电路中, VBP电位不宜过低, 以免 Ml管的源体结过度 正偏而导致漏电流明显增大; 同样地, 在 NM0S体电位调制电路中, VBN电位不宜过 高, 以免 M3管体源结过度正偏。
(3 ) 在 PM0S体电位调制电路中, M2源端电位 VDDH—般大于或等于目标 PM0S
管 Ml的电源电压 VDD, 同样地, 在 NM0S体电位调制电路中, M4源端电位 GNDL小于 或等于目标 NM0S管 M3的地电位 GND。 取 VDDH大于 VDD和 /或 GNDL小于 GND, 需要增 加额外偏置电平, 但此时调节范围增大, 调节效果明显。 如果不增加额外电平(取 VDD„= VDD和 /或 GNDL = GND ) , 则调节效果会受到相应影响。 具体可参见实施例二 中表 1和表 2的数据。
( 4 ) 为避免提供过多的偏置电平, 在 PM0S体电位调制电路中, M2栅端电位 VCP可选用电源电压 VDD (当 VDDH > VDD ) 或共模电压 VCM (当 V = VDD ) ; 而在 NM0S体 电位调制电路中, M4栅端电位 V™可选用 GND (当 GNDL〈 GND )或 VeM(当 GNDL = GND ) 以下对本发明应用实例中的 C类反相器进行进一步描述。
C类反相器技术是一种新兴的低压模拟电路设计技术,采用 C类反相器代替传 统的运算放大器能够实现许多低压低功耗开关电容电路。现有技术的 c类反相器的 电路结构图如附图 4所示。 C类反相器供电电压 VDD略低于反相器两输入管阈值电 压之和,假设 C类反相器两输入管阈值电压近似相等,输入共模电压 VeM=VDD/2即可 使两输入管 Ml和 M3均处于亚阈值区,此时 C类反相器具有极低的功耗和较高的增 益, 但其对工艺涨落极为敏感。
本发明的抗工艺涨落的 C类反相器的电路结构图如附图 5所示, 它在现有技 术的 C类反相器 51的基础上增加了本发明的 PM0S体电位调制电路 52和匪 OS体电 位调制电路 53,现有技术的 C类反相器 51的 PM0S和 NM0S输入管分别为体电位调 制电路的目标 PM0S管 Ml禾卩目标匪 OS管 M3
现有技术的 C类反相器 51用于实现运算放大功能。它由 PM0S输入管 Ml和 NM0S 输入管 M3组成。 反相器输入管 Ml和 M3的体端均单独引出, 体电位可调。
本发明的 PM0S体电位调制电路 52和 NM0S体电位调制电路 53分别用于 Ml和 M3的抗工艺涨落, 使 Ml和 M3在不同工艺角下跨导和输出电流较为一致。 在 C类 反相器中, 输入管 Ml和 M3的跨导和输出电流与整个反相器的增益、 带宽、静态功 耗等稳态指标有直接关系, 所以引入体电位调制电路 52 53能够有效地降低 C类 反相器各稳态指标对于工艺涨落的敏感度。
在亚阈值区不同工艺角下,现有技术的 C类反相器和本发明中抗工艺涨落的 C 类反相器的增益、 带宽以及静态功耗等性能偏差情况见表 1。 其中电源电压 VDD为 1. 2 V, GND为 0V, Ml宽长比为 180 /0. 35 , M3的宽长比为 60 /0. 35
M2和 M4宽度分别为 Ml和 M3的 1/12, 反相器的负载电容均取 5 pF 表 1: 不同工艺角下 C类反相器性能最大偏差情况对比
注: 以上四种情况下 C类反相器相位裕度均超过 90°, 且在各工艺角下偏差范围均小于 4%。 故略。
由表 1可知, 当引入额外正、 负电平, 即 VDDH =1. 8 V GNDL= -0. 6 V, 本发明 中抗工艺涨落的 C类反相器在亚阈值区不同工艺角下的增益、带宽和静态功耗的最 大偏差分别为 27. 8% 52. 3%和 8%, 相比于现有技术 C类反相器的 28% 435. 8%和 577. 4%, 本发明受工艺涨落的影响大大减小, 既保证了足够的增益和带宽, 又避免 了无谓的静态功耗, 效果十分明显; 若在体电位调制电路中只引入额外正电平, 不 引入负电平 (这种现象在目前集成电路设计中非常普遍) , 即 VDDH = 1. 8 V GNDL = 0 V, 则抗工艺涨落的 C类反相器的最大偏差范围分别为 29. 1% 169. 3%和 81. 9%; 若在体电位调制电路中既不引入额外正电平, 也不引入负电平, 即 VDDH = 1. 2 V GNDL = 0 V, 则抗工艺涨落的 C 类反相器最大偏差范围分别为 25. 5% 287. %和 425. 7%, 也好于现有技术的 C类反相器。尤其是单位增益带宽指标, 未使用本发明 技术的现有技术 C类反相器的单位增益带宽在 ss工艺角下仅仅为 5MHz左右,在高 频下无法正常工作,而使用本发明技术的 C类反相器在任何情况下都不会发生这样 严重的问题。
C类反相器在不同工艺角下的性能最大偏差汇总简表如表 2所示。 不同工艺角下 C类反相器性能最大偏差汇总简表
静态功耗 增益 单位增益带宽 c类反相器 偏差 范围 偏差 范围 偏差 范围
范围 缩小 范围 缩小 范围 缩小 现有技术 577. 4 % 28. 8 % 435. 8 % 本发明
(V =1· 8V 8 % 98. 6 % 27. 8 % 3. 5 % 52. 3 % 88. 0 % GNDL=-0. 6V)
本发明
(V =1· 8V 81. 9 % 85. 8 % 29. 1 % -1. 0 % 169. 3 % 61. 2 % GNDL=0V)
本发明
(V =1· 2V 425. 7 % 26 % 25. 5 % 11. 5 % 287. 0 % 34 % GNDL=0V)
虽然本发明的描述结合了特定的实施例, 但是本领域普通技术人员应该理解 本发明并不限于在此描述的实施例,并可以进行各种修改和变化而不背离本发明的 精神和范围。 工业应用性
本发明所述的抗工艺涨落方法通过感应反馈环路的体电位调制, 实现对目标
M0S器件跨导、 输出电流等参数的实时调制, 有效降低了 M0S器件在亚阈值状态下 对于工艺涨落的敏感度。本发明所述的体电位调制电路以较少的电路元件实现了整 个感应反馈环路,它的引入能够在不明显增加电路复杂度和功耗的情况下有效地提 高包括 C类反相器在内的亚阈值集成电路的性能稳定性、一致性和产品良率, 因而 具有较高的实用价值。
Claims
1. 一种在亚阈值集成电路中抗工艺涨落的方法,其特征在于:包括以下步骤: 由感应 M0S器件感应目标 M0S器件在不同工艺角下的参数变化趋势, 并以漏 源感应电流形式输出;
通过电流转电压电路将感应 M0S器件输出的感应电流信号转变为电压信号, 将感应电流的变化特征实时地反映到该电压信号上;
将电流转电压电路输出的电压信号反馈到目标 M0S器件的体端, 形成感应反 馈环路, 用以体电位调制, 减弱工艺涨落对目标器件性能参数的影响。
2. 一种体电位调制电路, 用于实现权利要求 1所述的抗工艺涨落方法, 其特 征在于: 该体电位调制电路包括:
目标 M0S器件, 所述抗工艺涨落方法的作用对象;
感应 M0S器件, 用于感应目标 M0S器件在不同工艺角下的参数变化趋势; 电流转电压电路,用于将感应 M0S器件输出的感应电流转换成感应电压, 并将 该感应电压反馈到目标 M0S器件的体端, 实现目标 M0S器件的体电位调制;
该体电位调制电路可分为 PM0S体电位调制电路和 NM0S体电位调制电路两种类 型。
3. 根据权利要求 2所述的体电位调制电路, 其特征在于:
所述 PM0S体电位调制电路用于实现 PM0S管在亚阈值状态下的抗工艺涨落, 它包括第一 PM0S管 (Ml ) 、 第二 PMOS管 (M2 )和第一电阻 (R1 ) , 其中, 所述第 一 PM0S管 (Ml ) 为所述 PMOS体电位调制电路的目标 MOS器件, 所述第二 PM0S管 (M2 ) 为 PM0S体电位调制电路的感应 M0S器件, 所述第一电阻 (R1 ) 实现 PM0S 体电位调制电路中电流转电压电路的功能;
所述第二 PM0S管 (M2 ) 的源端与体端相连, 漏端分别与所述第一电阻 (R1 ) 的一端和所述第一 PM0S管 (Ml ) 的体端相连, 第一电阻 (R1 ) 的另一端与共模电 压相连。
4. 根据权利要求 2所述的体电位调制电路, 其特征在于:
所述 NM0S体电位调制电路用于实现 NM0S管在亚阈值状态下的抗工艺涨落, 它包括第一 NM0S管 (M3) 、 第二 NM0S管 (M4) 和第二电阻 (R2) , 其中,所述第 一 NM0S管 (M3) 为所述 NM0S体电位调制电路的目标 M0S器件, 所述第二 NM0S管 (M4) 为 NM0S体电位调制电路的感应 M0S器件, 所述第二电阻 (R2) 实现 NM0S 体电位调制电路中电流转电压电路的功能;
所述第二 NM0S管 (M4) 的源端与体端相连, 漏端分别与所述第二电阻 (R2) 的一端和所述第一 NM0S管 (M3) 的体端相连, 第二电阻 (R2) 的另一端与共模电 压相连。
5. 一种抗工艺涨落的 C类反相器, 它包括:
现有技术的 C类反相器 (51) , 用于实现运算放大功能;
其特征在于: 它还包括权利要求 3、 4所述的 PM0S体电位调制电路 (52) 和 NM0S体电位调制电路 (53) ;
现有技术的 C类反相器(51) 中的 PM0S和 NM0S输入管分别为所述 PM0S体电 位调制电路(52)和 NM0S体电位调制电路 (53) 中作为目标 M0S器件的第一 PM0S 管 (Ml) 和第一 NMOS管 (M3) 。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2011517743A JP5275462B2 (ja) | 2009-04-03 | 2009-09-04 | サブスレッショルド集積回路におけるプロセスばらつき防止方法を実現するボディ電位変調回路及びプロセスばらつきを防止するc型インバータ |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2009103013271A CN101510769B (zh) | 2009-04-03 | 2009-04-03 | 体电位调制器和c类反向器 |
CN200910301327.1 | 2009-04-03 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2010111857A1 true WO2010111857A1 (zh) | 2010-10-07 |
Family
ID=41003045
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2009/073744 WO2010111857A1 (zh) | 2009-04-03 | 2009-09-04 | 亚阈值集成电路中抗工艺涨落的方法和体电位调制电路 |
Country Status (3)
Country | Link |
---|---|
JP (1) | JP5275462B2 (zh) |
CN (1) | CN101510769B (zh) |
WO (1) | WO2010111857A1 (zh) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117688888A (zh) * | 2024-02-04 | 2024-03-12 | 电子科技大学(深圳)高等研究院 | 一种知识和数据双驱动的多任务电路参数进化方法 |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101510769B (zh) * | 2009-04-03 | 2011-07-06 | 浙江大学 | 体电位调制器和c类反向器 |
CN101692603B (zh) * | 2009-04-21 | 2012-05-09 | 浙江大学 | 增益自举型c类反向器及其应用电路 |
CN102355261B (zh) * | 2011-08-05 | 2013-12-11 | 清华大学 | 一种适用于高速模数转换器的电压缓冲器 |
CN102394634B (zh) * | 2011-09-21 | 2013-01-09 | 浙江大学 | 数模混合控制体偏置型c类反相器 |
CN104579315B (zh) * | 2014-12-03 | 2017-06-23 | 浙江大学 | 同时实现高增益和宽输出摆幅的c类反相器 |
CN106921349B (zh) * | 2017-03-02 | 2020-10-09 | 中国电子科技集团公司第二十四研究所 | 基于反相器结构的放大器 |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1497725A (zh) * | 2002-10-21 | 2004-05-19 | 松下电器产业株式会社 | 半导体集成电路装置 |
JP2007004581A (ja) * | 2005-06-24 | 2007-01-11 | Seiko Epson Corp | レギュレータ回路 |
CN101510769A (zh) * | 2009-04-03 | 2009-08-19 | 浙江大学 | 采用体电位调制器的c类反向器 |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0680824B2 (ja) * | 1984-05-29 | 1994-10-12 | 株式会社東芝 | Mosトランジスタ閾値の自動設定装置 |
JP4138047B2 (ja) * | 1997-07-15 | 2008-08-20 | 富士通株式会社 | 内部電圧生成回路 |
JP2001203568A (ja) * | 2000-01-24 | 2001-07-27 | Nec Ic Microcomput Syst Ltd | バッファ装置 |
JP4221274B2 (ja) * | 2003-10-31 | 2009-02-12 | 株式会社東芝 | 半導体集積回路および電源電圧・基板バイアス制御回路 |
JP4620522B2 (ja) * | 2005-05-11 | 2011-01-26 | 旭化成エレクトロニクス株式会社 | 半導体回路 |
JP2009064860A (ja) * | 2007-09-05 | 2009-03-26 | Renesas Technology Corp | 半導体装置 |
-
2009
- 2009-04-03 CN CN2009103013271A patent/CN101510769B/zh not_active Expired - Fee Related
- 2009-09-04 JP JP2011517743A patent/JP5275462B2/ja not_active Expired - Fee Related
- 2009-09-04 WO PCT/CN2009/073744 patent/WO2010111857A1/zh active Application Filing
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1497725A (zh) * | 2002-10-21 | 2004-05-19 | 松下电器产业株式会社 | 半导体集成电路装置 |
JP2007004581A (ja) * | 2005-06-24 | 2007-01-11 | Seiko Epson Corp | レギュレータ回路 |
CN101510769A (zh) * | 2009-04-03 | 2009-08-19 | 浙江大学 | 采用体电位调制器的c类反向器 |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117688888A (zh) * | 2024-02-04 | 2024-03-12 | 电子科技大学(深圳)高等研究院 | 一种知识和数据双驱动的多任务电路参数进化方法 |
CN117688888B (zh) * | 2024-02-04 | 2024-05-17 | 电子科技大学(深圳)高等研究院 | 一种知识和数据双驱动的多任务电路参数进化方法 |
Also Published As
Publication number | Publication date |
---|---|
JP5275462B2 (ja) | 2013-08-28 |
CN101510769B (zh) | 2011-07-06 |
JP2011528170A (ja) | 2011-11-10 |
CN101510769A (zh) | 2009-08-19 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2010111857A1 (zh) | 亚阈值集成电路中抗工艺涨落的方法和体电位调制电路 | |
CN105700601B (zh) | 一种ldo线性稳压器 | |
CN103457553B (zh) | 增益和摆率增强型放大器 | |
CN106774595A (zh) | 一种用于低压差线性稳压器的过流保护电路 | |
CN101976095B (zh) | 一种基于发射极电流补偿的高精度带隙基准源电路 | |
TWI351591B (en) | Voltage generating apparatus | |
CN105912066B (zh) | 一种低功耗高psrr的带隙基准电路 | |
WO2020156588A1 (zh) | 电压基准源电路及低功耗电源系统 | |
TWI724312B (zh) | 能隙電壓參考電路 | |
US7733182B2 (en) | Hybrid class AB super follower | |
CN101471632A (zh) | 环路增益可控的自偏置低压运算跨导放大器电路 | |
CN104216455B (zh) | 用于4g通信芯片的低功耗基准电压源电路 | |
CN104753481B (zh) | 差动运算放大器以及带隙参考电压产生电路 | |
CN107402594A (zh) | 实现高电源电压转变的低功耗低压差线性稳压器 | |
CN206627849U (zh) | 动态密勒补偿的cmos低压差线性稳压器及电子设备 | |
CN107066024A (zh) | 一种低功耗高精度非带隙基准电压源 | |
CN104881070A (zh) | 一种适用于mems应用的超低功耗ldo电路 | |
CN101839941A (zh) | 信号感测放大器 | |
TWI355800B (en) | Dual supply amplifier | |
TW202042507A (zh) | 共模電壓準位轉移及鎖定電路 | |
CN102981537A (zh) | 一种带反馈电路的高压稳压电路 | |
CN115586810A (zh) | 一种稳定cascode电流镜工作状态的偏置网络 | |
CN103135642B (zh) | 一种环路补偿电路 | |
CN105739586B (zh) | 一种电流基准源电路 | |
CN110445482B (zh) | 一种低功耗高摆率的比较器 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 09842513 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2011517743 Country of ref document: JP Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 09842513 Country of ref document: EP Kind code of ref document: A1 |