CN102045035B - A low-power, wideband, high-gain, high-slew-rate single-stage operational transconductance amplifier - Google Patents

本发明公布了一种低功耗宽带高增益高摆率单级运算跨导放大器，由恒定电流偏置依次串接差分输入级、负载电流镜传输输出级三部分构成，其中负载电流镜传输输出级由八个N型MOS管NM1至NM8构成。本发明彻底解决线性运放电路内在固有局限约束，实现静态、交流小信号和动态大信号下电路性能的全面改善和提高。

The invention discloses a single-stage operational transconductance amplifier with low power consumption, broadband, high gain, and high slew rate, which is composed of three parts: a constant current bias sequentially connected to a differential input stage, and a load current mirror transmission output stage, wherein the load current mirror transmits the output stage The stage consists of eight N-type MOS transistors NM1 to NM8. The invention completely solves the inherent limitations and constraints of the linear operational amplifier circuit, and realizes the overall improvement and improvement of the circuit performance under static state, AC small signal and dynamic large signal.

一种低功耗宽带高增益高摆率单级运算跨导放大器A low-power, wideband, high-gain, high-slew-rate single-stage operational transconductance amplifier

技术领域 technical field

本发明涉及一种兼容低功耗、高速、高精度特性的线性跨导单级运放电路，属于模拟集成运放技术领域，通过对负载电流镜非线性效应与线性传输比控制的兼容和集成，获得高效电流利用效率以及功耗约束下电路速度与精度的协调统一，实现电路静态、动态特性的全面提升。 The invention relates to a linear transconductance single-stage operational amplifier circuit compatible with low power consumption, high speed and high precision characteristics, belonging to the technical field of analog integrated operational amplifiers, through the compatibility and integration of the nonlinear effect of the load current mirror and the control of the linear transmission ratio , to achieve high current utilization efficiency and coordination and unification of circuit speed and precision under power consumption constraints, and to achieve a comprehensive improvement in the static and dynamic characteristics of the circuit.

背景技术 Background technique

SoC数模混合单芯片系统中，系统功能扩展与性能提高只有在低功耗的约束下才具有更现实的意义。现有技术中，电路速度或动态性能的改善通常以大电流驱动即功耗增加为代价，低功耗高速高精度相互制约的要求显著增加了线性运放电路设计和实现的难度。因此，低功耗高速运放电路设计与实现，需要突破现有线性结构的局限。 In the SoC digital-analog hybrid single-chip system, system function expansion and performance improvement have more realistic significance only under the constraint of low power consumption. In the prior art, the improvement of circuit speed or dynamic performance is usually at the cost of high current drive, that is, increased power consumption. The mutual constraints of low power consumption, high speed and high precision significantly increase the difficulty of linear operational amplifier circuit design and implementation. Therefore, the design and implementation of low-power high-speed operational amplifier circuits need to break through the limitations of the existing linear structure.

对于传统的单级OTA线性运算跨导电路(见图1)，在低电源电压限制下，难以充分利用Cascode结构的高阻输出特性实现高增益，通过降低输出电流提高阻抗和增益的方法又难以满足动态电流调节与驱动下的高速响应要求；相反，当N值固定后，通过增大静态电流I_B提高带宽，获得大摆率驱动下的高速响应，不但引起功耗的增加，同时导致增益的下降。因此，对于传统OTA结构，只能在静态电流I_B、输入差分对管跨导因子、差分对负载电流镜W/L比例因子N值三种参数之间进行优化选取，以平衡增益、带宽、摆率和功耗的需求，但同时满足各方需求则极为困难。为解决电路中的固有矛盾，需通过非线性效应和参数可配置控制突破原有线性电路结构的制约和局限。 For the traditional single-stage OTA linear operation transconductance circuit (see Figure 1), under the limitation of low power supply voltage, it is difficult to make full use of the high-impedance output characteristics of the Cascode structure to achieve high gain, and it is difficult to increase the impedance and gain by reducing the output current. Meet the requirements of dynamic current regulation and high-speed response under driving; on the contrary, when the value of N is fixed, the bandwidth is increased by increasing the quiescent current I _B to obtain high-speed response under large slew rate driving, which not only causes an increase in power consumption, but also leads to gain Decline. Therefore, for the traditional OTA structure, it can only be optimized among the three parameters of quiescent current I _B , input differential pair transistor transconductance factor, and differential pair load current mirror W/L proportional factor N value to balance gain, bandwidth, The requirements of slew rate and power consumption, but it is extremely difficult to meet the needs of all parties at the same time. In order to solve the inherent contradictions in the circuit, it is necessary to break through the constraints and limitations of the original linear circuit structure through nonlinear effects and parameter configurable control.

发明内容 Contents of the invention

本发明目的在于为了解决了常规运放电路中难以调和的电路系统速度、精度与功耗间的固有矛盾，提供一种低功耗宽带高增益高摆率单级运算跨导放大器。 The purpose of the present invention is to provide a single-stage operational transconductance amplifier with low power consumption, wide band, high gain and high slew rate in order to solve the inherent contradiction among the speed, precision and power consumption of the circuit system which is difficult to reconcile in the conventional operational amplifier circuit.

本发明一种低功耗宽带高增益高摆率单级运算跨导放大器，由恒定电流偏置依次串接差分输入级、负载电流镜传输输出级三部分构成，其中负载电流镜传输输出级由八个N型MOS管NM1至NM8构成，N型MOS管NM1的漏极分别接差分输入级的一个输出端和N型MOS管NM3、NM6、NM7的栅极，N型MOS管NM1的漏极分别接差分输入级的另一个输出端和N型MOS管NM4、NM5、NM8的栅极，N型MOS管NM1的源极分别接N型MOS管NM3、NM5的漏极，N型MOS管NM2的源极分别接N型MOS管NM4、NM6的漏极，N型MOS管NM3、NM4、NM5、NM6、NM7、NM8的源极分别连接接地，N型MOS管NM7漏极构成负载电流镜传输输出级的第一输出端，N型MOS管NM8漏极构成负载电流镜传输输出级的第二输出端。 The present invention is a single-stage operational transconductance amplifier with low power consumption, wide band, high gain and high slew rate, which is composed of three parts: constant current bias, differential input stage and load current mirror transmission output stage, in which the load current mirror transmission output stage consists of three parts: Eight N-type MOS transistors NM1 to NM8 are formed. The drain of N-type MOS transistor NM1 is respectively connected to an output terminal of the differential input stage and the gates of N-type MOS transistors NM3, NM6, and NM7. The drain of N-type MOS transistor NM1 The other output end of the differential input stage and the gates of N-type MOS transistors NM4, NM5, and NM8 are respectively connected, the source of N-type MOS transistor NM1 is respectively connected to the drains of N-type MOS transistors NM3 and NM5, and the N-type MOS transistor NM2 The sources of the N-type MOS transistors NM4 and NM6 are respectively connected to the drains, the sources of the N-type MOS transistors NM3, NM4, NM5, NM6, NM7, and NM8 are respectively connected to the ground, and the drain of the N-type MOS transistor NM7 constitutes a load current mirror transmission The first output end of the output stage, the drain of the N-type MOS transistor NM8 constitutes the second output end of the load current mirror transmission output stage.

所述恒定电流偏置由P型MOS管PM0构成，P型MOS管PM0的漏极接电源Vdd，P型MOS管PM0的源极接差分输入级。 The constant current bias is formed by a P-type MOS transistor PM0, the drain of the P-type MOS transistor PM0 is connected to the power supply Vdd, and the source of the P-type MOS transistor PM0 is connected to the differential input stage.

所述差分输入级由两个P型MOS管PM1、PM2构成，P型MOS管PM1、PM2的漏极接恒定电流偏置的输出端，P型MOS管PM1、PM2的源极接负载电流镜传输输出级的输入端。 The differential input stage is composed of two P-type MOS transistors PM1 and PM2, the drains of the P-type MOS transistors PM1 and PM2 are connected to the output end of the constant current bias, and the sources of the P-type MOS transistors PM1 and PM2 are connected to the load current mirror Transmit the input of the output stage.

本发明通过小信号下起正反馈作用的交叉耦合对管结构与宽动态范围内有效的负载电流镜线性-非线性模式自适应配置结构的兼容与相互作用，有效解决了常规运放电路中难以调和的电路系统速度、精度与功耗间的固有矛盾。模式控制的关键是在静态条件下将负载电流镜偏置在线性-非线性临界工作模式下，确保静态低功耗；同时在交流小信号下，利用线性范围内交叉耦合对管对负载电流镜线性传输比的下降控制作用，实现高增益(A_V)倍增以确保高精度控制，带宽(p_-3dB和GBW)倍增以确保小信号高速及线性处理性能；在大信号下，虽然交叉耦合对管失效，但大的输入动态范围可自动将差分负载电流镜转入深度非线性工作模式，即通过电流传输的非线性效应实现压摆率(SR)的倍增，以提高瞬态响应速度。通过线性传输宽长比的配置以及非线性效应，能够彻底解决线性运放电路内在固有局限约束，实现静态、交流小信号和动态大信号下电路性能的全面改善和提高。 The present invention effectively solves the problems in conventional op-amp circuits through the compatibility and interaction of the cross-coupling that acts as positive feedback under small signals with the tube structure and the effective load current mirror linear-nonlinear mode self-adaptive configuration structure in a wide dynamic range. Reconciling the inherent tension between speed, accuracy and power consumption of circuitry. The key to mode control is to bias the load current mirror in the linear-nonlinear critical mode of operation under static conditions to ensure static low power consumption; at the same time, under small AC signals, use the cross-coupled tube-to-load current mirror in the linear range The reduction control function of the linear transfer ratio realizes the multiplication of high gain (A _V ) to ensure high-precision control, and the multiplication of bandwidth (p _-3dB and GBW) to ensure high-speed and linear processing performance of small signals; under large signals, although the cross-coupling pair Tube failure, but the large input dynamic range can automatically turn the differential load current mirror into a deep nonlinear working mode, that is, the nonlinear effect of current transmission realizes the multiplication of the slew rate (SR) to improve the transient response speed. Through the configuration of the linear transmission width-to-length ratio and the nonlinear effect, the inherent limitations of the linear operational amplifier circuit can be completely solved, and the overall improvement and enhancement of the circuit performance under static, AC small signal and dynamic large signal can be realized.

本发明具有在同等静态低功耗下，电路交流小信号增益、带宽以及大信号电压摆率的倍增。采用CSMC0.18mm标准CMOS工艺，在静态电流为29mA和30pF负载电容的条件下，低频增益为71.3dB，单位增益带宽为6.5MHz，大信号下的正向压摆率为+12.5V/ms，反向压摆率为-12.8V/ms。相比同等条件下的经典单级OTA结构，增益增加了24dB，带宽提高了9倍，电压摆率增大20倍。 The invention has the advantages of doubling the circuit AC small signal gain, bandwidth and large signal voltage slew rate under the same static low power consumption. Using CSMC0.18mm standard CMOS process, under the condition of quiescent current of 29mA and load capacitance of 30pF, the low frequency gain is 71.3dB, the unity gain bandwidth is 6.5MHz, and the forward slew rate under large signal is +12.5V/ms, The reverse slew rate is -12.8V/ms. Compared with the classic single-stage OTA structure under the same conditions, the gain is increased by 24dB, the bandwidth is increased by 9 times, and the voltage slew rate is increased by 20 times.

附图说明 Description of drawings

图1为传统运算跨导放大器OTA的原理电路图。 Figure 1 is a schematic circuit diagram of a traditional operational transconductance amplifier OTA.

图2为普通线性比例电流镜和Cascode线性电流镜的电路图。 Fig. 2 is the circuit diagram of ordinary linear proportional current mirror and Cascode linear current mirror.

图3为线性/非线性自适应电流镜的电路图。 Fig. 3 is a circuit diagram of a linear/nonlinear adaptive current mirror.

图4为非线性模式可配置OTA电路原理图。 Fig. 4 is a schematic diagram of a non-linear mode configurable OTA circuit.

图5为图4所示OTA的小信号波特图。 FIG. 5 is a small-signal Bode diagram of the OTA shown in FIG. 4 .

图6为图4所示OTA的大信号电压摆率。 Figure 6 shows the large-signal voltage slew rate of the OTA shown in Figure 4.

具体实施方式 Detailed ways

下面结合附图对发明的技术方案进行详细说明： Below in conjunction with accompanying drawing, the technical scheme of invention is described in detail:

本发明采用如下技术方案： The present invention adopts following technical scheme:

一种在低功耗约束下仍具有高精度高速响应的运算跨导放大电路(见图4)，其特征包括运算跨导放大器属于OTA单级增益电路结构；电路系统由恒定电流偏置、差分输入级、负载电流镜传输输出级三部分构成，其中PMOS差分对管采用固定尾电流偏置，两对对称的负载电流镜分别采用由三个NMOS管构成的线性-非线性模式动态可配置结构，同时增加一对交叉耦合对管对负载电流镜等效输入W/L调制控制的结构；电路输出级采用CMOS互补推挽的对称驱动结构。 An operational transconductance amplifier circuit with high precision and high speed response under the constraints of low power consumption (see Figure 4), its features include that the operational transconductance amplifier belongs to the OTA single-stage gain circuit structure; the circuit system is composed of constant current bias, differential The input stage, the load current mirror and the transmission output stage are composed of three parts, in which the PMOS differential pair tube adopts a fixed tail current bias, and the two pairs of symmetrical load current mirrors respectively adopt a linear-nonlinear mode dynamic configurable structure composed of three NMOS tubes. , while adding a pair of cross-coupled tube-to-load current mirror equivalent input W/L modulation control structure; the output stage of the circuit adopts a CMOS complementary push-pull symmetrical drive structure.

两种基于电路结构和工作模式的可配置控制：利用其在特定模式下所具有的良好互补性，相互配合以解决常规线性电路内在固有矛盾制约，全面提升电路的静态、交流和瞬态性能；一种是利用交叉耦合对管交流条件下的正反馈控制，实现电流镜线性传递系数在不同信号模式下的切换或配制，主要实现对电路交流小信号特性的改进；另一种是基于电流镜线性-非线性模式的动态配置，在保持低功耗的条件下主要实现对电路瞬态大信号特性的提升； Two configurable controls based on circuit structure and working mode: take advantage of their good complementarity in specific modes, cooperate with each other to solve the inherent contradictions and constraints of conventional linear circuits, and comprehensively improve the static, AC and transient performance of the circuit; One is to use the positive feedback control of the cross-coupled tube under AC conditions to realize the switching or configuration of the linear transfer coefficient of the current mirror in different signal modes, mainly to improve the characteristics of the AC small signal of the circuit; the other is based on the current mirror The dynamic configuration of linear-nonlinear mode mainly realizes the improvement of the transient large signal characteristics of the circuit under the condition of maintaining low power consumption;

差分输入级的结构和工作模式的设定：差分输入级采用经典三管差分结构，可以是三管PMOS差分对，也可以是与之镜像关系的NMOS差分对；差分对尾电流采用稳定的恒流偏置，差分对两输入管采用W/L相同的对称设计，确保静态时两差分对管各流过一半的尾电流，差分输出A、B两端电位相同，差分交流小信号驱动下两差分对管中的变化电流大小相同、极性相反，同时差分输出A、B两端保持差分信号的性质； The structure of the differential input stage and the setting of the working mode: the differential input stage adopts a classic three-tube differential structure, which can be a three-tube PMOS differential pair, or an NMOS differential pair with a mirror image relationship with it; the differential pair tail current adopts a stable constant Current bias, the two input tubes of the differential pair adopt the same symmetrical design of W/L to ensure that half of the tail current flows through each of the two differential pair tubes in static state, the potentials of the two ends of the differential output A and B are the same, and the differential AC small signal drives the two The changing currents in the differential pair tubes are the same in size and opposite in polarity, and at the same time, the nature of the differential signal is maintained at both ends of the differential output A and B;

电流镜线性-非线性模式可配置的结构和工作模式的设定：将四管宽摆幅Cascode电流镜结构(见图2)中输出支路上的M4管去除(或短路)，即可得到这种模式可配置电流镜结构(见图3)。具体的模式控制由电流镜输入电流I₁、输入支路各MOS管尺寸W/L以及Cascode偏置电压V_bn三种影响因素的相对关系决定。 Current mirror linear-nonlinear mode configurable structure and working mode setting: remove (or short-circuit) the M4 tube on the output branch of the four-tube wide-swing Cascode current mirror structure (see Figure 2), you can get this This mode can configure the current mirror structure (see Figure 3). The specific mode control is determined by the relative relationship among three influencing factors: the input current I ₁ of the current mirror, the size W/L of each MOS transistor in the input branch, and the bias voltage V _bn of the Cascode.

在偏置电压V_bn和电流I₁均固定的条件下，减小具有相同尺寸的M1和M3两管的W/L, 将使电流镜由线性性质向非线性性质转化，相反，增加W/L将使电流镜由非线性性质向线性性质转化；通过合适的W/L设计, 能够将此电流镜根据需要设定在所需的线性或非线性工作模式。差分对在静态条件下对差分负载输入驱动电流恒定且V_bn恒定，利用这一点特性即W/L的调节来设置在静态条件下负载电流镜的模式。 Under the condition that the bias voltage V _bn and the current I ₁ are fixed, reducing the W/L of the two tubes M1 and M3 with the same size will transform the current mirror from linear to nonlinear. On the contrary, increasing W/ L will transform the current mirror from nonlinear to linear; through appropriate W/L design, the current mirror can be set in the required linear or nonlinear working mode as required. The differential pair has a constant input drive current and V _bn to the differential load under static conditions. This feature, that is, the adjustment of W/L, is used to set the mode of the load current mirror under static conditions.

在偏置电压V_bn和输入管器件尺寸W/L均固定的条件下，增加输入电流I₁将使电流镜由线性性质向非线性性质转化，相反，减小输入电流I₁将使电流镜由非线性性质向线性性质转化；因此，通过输入电流的变化，可以使电流镜实现不同传输性质之间的转化。差分对在交流小信号小条件下对差分负载输入驱动电流有变化，但V_bn恒定，W/L经过静态模式设定后不再变化，利用这一点特性即输入电流的变化实现在交流小信号条件下负载电流镜模式的动态自适应调节； Under the condition that the bias voltage V _bn and the size of the input tube device W/L are fixed, increasing the input current I ₁ will make the current mirror change from linear to nonlinear, on the contrary, reducing the input current I ₁ will make the current mirror Transform from nonlinear property to linear property; therefore, through the change of input current, the current mirror can realize the transformation between different transmission properties. The differential pair has a change in the input drive current of the differential load under the condition of small AC small signal, but V _bn is constant, and W/L will not change after being set in static mode. Using this characteristic, the change of input current is realized in AC small signal Dynamic adaptive adjustment of load current mirror mode under certain conditions;

差分对负载模式可配置电流镜，在静态低功耗要求下，静态偏置在线性-非线性临界工作点或略偏向非线性的工作模式，在差分小信号状态下差分负载一对电流镜将分别向线性和非线性两个模式偏转，进入大信号输入状态后，其中一个负载电流镜通常能够转移入非线性工作模式； The current mirror can be configured in the differential pair load mode. Under the static low power consumption requirement, the static bias is at the linear-nonlinear critical operating point or a slightly non-linear operating mode. In the differential small signal state, the differential load pair of current mirrors will Deflection to linear and non-linear modes respectively, after entering the large signal input state, one of the load current mirrors can usually be transferred to the non-linear working mode;

交叉耦合对管结构对差分对负载电流镜输入管等效W/L的调节控制：在静态模式下使等效W/L增加，电流传输线性系数降低；在交流小信号模式下等效W/L减小，电流传输线性系数增加；在瞬态大信号下等效W/L不变，电流传输线性系数与没有交叉耦合对管的结构相同。 The cross-coupled pair tube structure adjusts and controls the equivalent W/L of the differential pair load current mirror input tube: in the static mode, the equivalent W/L is increased, and the linear coefficient of current transmission is reduced; in the AC small signal mode, the equivalent W/L When L decreases, the linear coefficient of current transmission increases; under the condition of large transient signal, the equivalent W/L remains unchanged, and the linear coefficient of current transmission is the same as that of the structure without cross-coupling pair of tubes.

电路工作模式的设置：在静态低功耗约束下，负载电流镜应偏置在线性-非线性临界工作点模式，或略偏向非线性电流镜模式。该电路适合较高的电源电压和较小的运放尾电流，产生的动态变化范围足够使大信号下负载电流镜进入深度非线性模式，并在高共模电平输入下实现摆率最大提升。 Setting of circuit working mode: Under static low power consumption constraints, the load current mirror should be biased in the linear-nonlinear critical operating point mode, or slightly biased towards the nonlinear current mirror mode. This circuit is suitable for higher supply voltages and smaller op amp tail currents, and the resulting dynamic range is sufficient to enable the load current mirror to enter deep nonlinear mode under large signals and achieve the largest increase in slew rate at high common-mode level inputs .

图2和图3给出了线性、线性-非线性自适应两类电流镜结构的对比，M₁与M₂管的W/L之比均为m:1。对于线性电流镜，电流传输关系为I₂=I₁/m，即不论I₁如何变化，只要M₂管维持饱和恒流区工作状态不变，I₂与I₁始终保持以上线性关系不变。如果在不同的信号状态下能够实现对m值的控制或调节，即静态直流下使m增大以降低电流传输，交流小信号下使m减小以增大电流传输，即线性传输比m值在不同状态下的动态配置，就能够解决传统固定传输比的线性电流镜的缺陷，满足电路交直流特性的共同需求。 Figure 2 and Figure 3 show the comparison of linear and linear-nonlinear adaptive current mirror structures. The W/L ratio of M ₁ and M ₂ tubes are both m:1. For a linear current mirror, the current transmission relationship is I ₂ =I ₁ /m, that is, no matter how I ₁ changes, as long as the M ₂ tube maintains the working state of the saturated constant current region, the above linear relationship between I ₂ and I ₁ will always remain unchanged . If it is possible to control or adjust the value of m under different signal states, that is, increase m under static direct current to reduce current transmission, and reduce m under small AC signals to increase current transmission, that is, the linear transmission ratio m value The dynamic configuration in different states can solve the defects of the traditional linear current mirror with fixed transmission ratio, and meet the common requirements of the AC and DC characteristics of the circuit.

对于自适应模式可配置电流镜，静态条件下M₁管在特定的输入电流I₁及V_bn电压偏置下，限定在饱和恒流区与线性电阻区的临界工作点，此时电流传输保持线性性质；动态条件下随着I₁的增加，M₃管V_GS3随之增加，因V_DS1=V_bn-V_GS3，则在固定V_bn偏置下V_DS1下降，强制M₁管进入线性电阻区，增加的电流以及线性工作区下的M₁管将使其栅压显著提高，若M₂仍维持初始饱和恒流区状态，输出电流I₂明显增加，即原来的线性电流传输转变为非线性电流倍增传输。 For the adaptive mode configurable current mirror, under the static condition, under the specific input current I ₁ and V _bn voltage bias, the M ₁ tube is limited to the critical operating point in the saturated constant current region and the linear resistance region, and the current transmission is maintained at this time Linear property; under dynamic conditions, with the increase of I ₁ , the V _GS3 of the M ₃ tube will increase accordingly, because V _DS1 =V _bn -V _GS3 , then V _DS1 will drop under the fixed V _bn bias, forcing the M ₁ tube to enter linearity In the resistance area, the increased current and the _M1 tube in the linear working area will significantly increase the gate voltage. If _M2 still maintains the initial saturated constant current area state, the output current _I2 will increase significantly, that is, the original linear current transmission will be transformed into Nonlinear current multiplying transfer.

将以上m值及模式可配置的电流镜代替模式和m值均固定的线性电流镜作为OTA电路的差分负载，利用输入差分信号控制输入驱动电流I₁，使其在静态或小信号条件下设置在线性-非线性临界模式，通过线性或近似线性电流传输维持低功耗与高增益特性；进入动态大信号动状态后，在增大的输入电流驱动下，自动转入深度非线性电流传输模式以大幅提高输出电流，实现摆率以及瞬态响应速度的显著提高。在此基础上，进一步采用交叉耦合对管对负载电流镜的有效W/L进行调节，获得不同信号模式下可配置的线性传输系数m，实现交流特性的改善和提高。非线性与m值可配置两类控制机制的协调与配合，最终实现电路综合性能的全面改善和提高。 Use the above current mirror with configurable m value and mode instead of the linear current mirror with fixed mode and m value as the differential load of the OTA circuit, and use the input differential signal to control the input drive current I ₁ so that it can be set under static or small signal conditions In the linear-nonlinear critical mode, maintain low power consumption and high gain characteristics through linear or approximately linear current transmission; after entering the dynamic large-signal dynamic state, it will automatically transfer to the deep nonlinear current transmission mode driven by the increased input current In order to greatly increase the output current, the slew rate and the transient response speed are significantly improved. On this basis, the cross-coupling is further used to adjust the effective W/L of the tube-to-load current mirror, and the configurable linear transmission coefficient m under different signal modes is obtained to realize the improvement and enhancement of the AC characteristics. The coordination and cooperation of the two types of control mechanisms, nonlinear and m-value configurable, finally realize the overall improvement and improvement of the comprehensive performance of the circuit.

图4给出的本发明电路原理图，包含两种特殊结构相互作用及其效应对电流传输控制的影响，并由此带来运放电路性能的改善与提高，以下分七个方面进行详细分析和阐述。 The schematic diagram of the circuit of the present invention shown in Fig. 4 includes the influence of the interaction of two special structures and their effects on the control of current transmission, and thus brings about the improvement and improvement of the performance of the operational amplifier circuit. The following seven aspects are analyzed in detail and elaboration.

1、正反馈交叉耦合结构对线性电流传输系数的控制 1. Control of linear current transfer coefficient by positive feedback cross-coupling structure

NM5、NM6两管构成的交叉耦合结构，仅在小信号线性范围内起正反馈调节作用，改变电流传输的有效线性比值N_eff。NM5与NM7以及NM6与NM8的宽长比之比均为K:1。设I_B为差分对管的固定尾电流偏置，静态下输入差分对两个支路电流均为I_B/2，且a、b两结点电压相等，各负载MOS管中的静态电流与其W/L成正比。设N_T=N+K，N_S=N-K, 经典设计中综合各方性能需求一般选择N_T=1，静态下N_eff=N_T, 输出支路电流为I_o=I_B/(2N_T)，电路总静态电流达I_Q=(1+N_T ^-1)I_B??2I_B。静态功耗的降低同时也有利于输出阻抗的提高和低频增益的增大。 The cross-coupling structure composed of NM5 and NM6 tubes only plays the role of positive feedback adjustment in the linear range of small signals, changing the effective linear ratio N _eff of current transmission. The width-to-length ratios of NM5 and NM7 as well as NM6 and NM8 are all K:1. Let I _B be the fixed tail current bias of the differential pair tube. Under static conditions, the current of the two branches of the input differential pair is I _B /2, and the voltages of the two nodes a and b are equal. The quiescent current in each load MOS tube is equal to W/L is directly proportional. Assume N _T =N+K, N _S =NK, in the classic design, N _T =1 is generally selected for comprehensive performance requirements of all parties, N _eff =N _T under static conditions, and the output branch current is I _o =I _B /(2N _T ), the total quiescent current of the circuit reaches I _Q =(1+ _NT ^-1 )I _B ??2I _B . The reduction of static power consumption is also conducive to the improvement of output impedance and the increase of low frequency gain.

交流小信号差分输入驱动下，差分负载a、b两点电位变为差分信号，在此差分信号驱动下NM5与NM6交叉耦合对管满足交流小信号电流的互补驱动特性，使负载电流镜的等效线性传输比减小为N_eff=N_S，传输电流增大。由于电路结构的线性性质，同时在小信号工作条件下能够维持这种线性性质，则直流电流传输比与交流电流传输比相同，如下式： Driven by the AC small signal differential input, the potentials of the differential load points a and b become differential signals. Driven by this differential signal, the cross-coupled pair of NM5 and NM6 meets the complementary driving characteristics of the AC small signal current, so that the load current mirror is equal The effective linear transfer ratio decreases to N _eff =N _S , and the transfer current increases. Due to the linear nature of the circuit structure and the ability to maintain this linear property under small-signal operating conditions, the DC current transfer ratio is the same as the AC current transfer ratio, as follows:

                               (1)

(1)

直流状态下N_eff(DC)=N_T，交流状态下N_eff(AC)=N_S，N_eff(DC)>>N_eff(AC)。 N _eff (DC)= _NT in DC state, N _eff (AC)=N _S in AC state, N _eff (DC)>>N _eff (AC).

2、模式可配置线性-非线性自适应电流镜的状态因子 2. The mode can configure the state factor of the linear-nonlinear adaptive current mirror

NM1、NM3、NM7与NM2、NM4、NM8分别构成两组对称的线性-非线性自适应电流镜，其中由宽长比决定的线性电流传输比为N:1，而输入支路中两管采用W/L相同的对称设计。根据NM3或NM4 Cascode输入管工作状态的不同，可将电流镜偏置在不同的模式状态下。 NM1, NM3, NM7 and NM2, NM4, NM8 respectively constitute two sets of symmetrical linear-nonlinear adaptive current mirrors, in which the linear current transmission ratio determined by the width-to-length ratio is N:1, and the two tubes in the input branch adopt W/L same symmetrical design. Depending on the working state of the NM3 or NM4 Cascode input tube, the current mirror can be biased in different mode states.

以NM1、NM3、NM7三管组成的可配置电流镜为例，分析电流镜的状态判定与工作模式设定。设该电流镜的输入和输出电流分别为I_in与I_o，当三管均工作在饱和恒流源工作区时，为线性电流镜，其交直流信号传输性质相同并满足线性传输规律；当NM3管进入线性电阻区而其余两管仍保持饱和恒流特性，则为非线性电流镜，其交直流信号传输特性分离且分别满足不同的非线性传输关系。NM3管的工作状态可由其柵电压与Cascode偏置电压V_bn之间的相对关系决定，设D_M3=V_GS3-V_TN、D_Mb=V_bn-V_TN，定义a_V=D_M3/D_Mb，NM3管的电阻或恒流状态以及由此决定的电流镜传输性质均可由a_V的大小加以区分，因此称其为电路的状态因子或电流镜的非线性因子。当NM3进入线性电阻区后，在忽略NM1管衬偏效应的前提下，有： Taking the configurable current mirror composed of three transistors NM1, NM3, and NM7 as an example, the state judgment and working mode setting of the current mirror are analyzed. Assuming that the input and output currents of the current mirror are I _in and I _o respectively, when the three tubes all work in the saturated constant current source working area, it is a linear current mirror, and its AC and DC signal transmission properties are the same and satisfy the linear transmission law; when When the NM3 tube enters the linear resistance region and the other two tubes still maintain the saturated constant current characteristic, it is a nonlinear current mirror, whose AC and DC signal transmission characteristics are separated and satisfy different nonlinear transmission relationships respectively. The working state of the NM3 tube can be determined by the relative relationship between its gate voltage and the Cascode bias voltage V _bn , set D _M3 =V _GS3 -V _TN , D _Mb =V _bn -V _TN , define a _V =D _M3 /D _Mb , the resistance or constant current state of the NM3 tube and the transmission properties of the current mirror determined by it can be distinguished by the size of a _V , so it is called the state factor of the circuit or the nonlinear factor of the current mirror. When NM3 enters the linear resistance area, under the premise of ignoring the lining offset effect of NM1 tube, there are:

                  (2)

(2)

设归一化的V_DS3为h=V_DS3/D_Mb，在k₁=k₃的对称设计下，得到h及其在a_V>>1的深线性电阻区下的近似,，即有： Let the normalized V _DS3 be h=V _DS3 /D _Mb , under the symmetric design of k ₁ =k ₃ , get h and its approximation in the deep linear resistance region of a _V >>1, namely:

                  (3)

(3)

很明显，a_V与h近似为反相关系，a_V=h=1/2对应于NM3管线性电阻区与饱和恒流区的临界状态点，在此临界模式下电流镜仍保持线性传输性质；当a_V>1/2或h<1/2后，NM3管进入非线性电阻区，电流镜由线性传输性质转变为非线性传输，a_V越大或h越小，非线性效应越强。作为状态模式或非线性程度的指示因子a_V和h，与输入电流I_in的关系为： Obviously, a _V and h are approximately inversely related, and a _V = h = 1/2 corresponds to the critical state point between the linear resistance area and the saturated constant current area of the NM3 tube. In this critical mode, the current mirror still maintains the linear transmission property ; When a _V >1/2 or h<1/2, the NM3 tube enters the nonlinear resistance area, and the current mirror changes from linear transmission property to nonlinear transmission. The larger a _V or the smaller h, the stronger the nonlinear effect . As the indicator factors a _V and h of the state mode or degree of nonlinearity, the relationship with the input current I _in is:

                           (4)

(4)

当a_V>>1时，近似有??h/??a_V??0。利用(2)式给出的线性电流模型，得到： When a _V >>1, there is approximately ??h/??a _V ??0. Using the linear current model given by (2), we get:

                  (5)

(5)

以上关系说明，状态因子a_V将随差分负载电流镜的输入电流而改变，a_V值在1/2前后发生的变化对应于负载电流镜传输性质的变化，而a_V>1/2后在数值上的变化，体现非线性程度强弱的改变。因此，只有利用输入电流的变化特性，才能有效调节负载电流镜的工作模式，从而构成该类型电流镜工作模式自适应动态配置的基本原理。 The above _relationship shows that the state factor a _V will change with the input current of the differential load current mirror _. The change in value reflects the change of the degree of nonlinearity. Therefore, only by using the changing characteristics of the input current can the working mode of the load current mirror be effectively adjusted, thus constituting the basic principle of adaptive dynamic configuration of the working mode of this type of current mirror.

3、状态因子的动态范围及其控制 3. Dynamic range and control of state factors

忽略衬偏效应，则固定D_Mb偏置满足D_Mb=D_M1+V_DS3=D_M1+hD_Mb约束，得到D_Mb=D_M1/(1-h)关系, 该式对不同的输入电流均可适用，其中与初始静态电流I_in0对应的参数为D_M1,0和h₀，而h₀则对应于初始设置的a_V0。若动态变化的电流I_in=dI_in0，则有D_M1=D_M1,0d^1/2，根据电流变化下D_Mb恒定的条件，得到d^1/2=(1-h)/(1-h₀)，再结合(3)式给出的h=f(a_V)函数关系，解得：

     (6) Neglecting the lining offset effect, the fixed D _Mb bias satisfies the constraint of D _Mb =D _M1 +V _DS3 =D _M1 +hD _Mb , and the relationship of D _Mb =D _M1 /(1-h) is obtained. Applicable, wherein the parameters corresponding to the initial quiescent current I _in0 are D _M1,0 and h ₀ , and h ₀ corresponds to the initially set a _V0 . If the dynamically changing current I _in =dI _in0 , then there is D _M1 =D _M1,0 d ^1/2 , and according to the condition that D _Mb is constant under the current change, d ^1/2 =(1-h)/(1- h ₀ ), combined with the h= f (a _V ) function relationship given by (3), the solution is:

(6)

非线性因子a_V与输入电流动态变化范围d、初始静态下的非线性因子a_V0控制的h₀因子有关。若初始静态选择线性-非线性临界模式点a_V0=h₀=1/2，在此约束下解得a_V=(1+d/2-d^1/2)/(2-d^1/2)。当该电流镜用于固定尾电流差分结构时因d=2，则a_V=1，即在差分负载输入电流2倍的动态变化下，电流镜非线性因子由初始临界模式下的a_V0=1/2，增大到非线性模式下的a_V=1，非线性效应强度不够。 The nonlinear factor a _V is related to the input current dynamic range d and the h ₀ factor controlled by the nonlinear factor a _V0 in the initial static state. If the initial static selection of the linear-nonlinear critical mode point a _V0 =h ₀ =1/2, the solution under this constraint is a _V =(1+d/2-d ^1/2 )/(2-d ^1/2 ). When the current mirror is used in a fixed tail current differential structure, because d=2, then a _V =1, that is, under the dynamic change of 2 times the input current of the differential load, the nonlinear factor of the current mirror is changed from a _V0 = in the initial critical mode 1/2, increasing to a _V =1 in the nonlinear mode, the nonlinear effect is not strong enough.

当电流镜用于固定尾电流差分负载时，由于输入电流2倍的动态范围不足以产生足够大的非线性因子，因此采用动态尾电流控制以增加d因子成为增强非线性强度十分有效的方法，在以上理想条件下当电流动态变化接近4倍时，a_V????，可以实现任意强度的非线性。实际上，V_GS3,max由特定电路结构和电源电压限定，则实际a_V,max存在上限范围。而在固定尾电流条件下，为提高动态下的a_V因子，还可通过减小h₀而实现。根据h与a_V的反相关系，直接增加a_V0，即将电流镜静态下偏置在弱非线性区，可以通过功耗的适度增加，即a_V0的小幅增加来换取a_V的大幅提高。 When the current mirror is used for a fixed tail current differential load, since the dynamic range of twice the input current is not enough to generate a large enough nonlinear factor, it is very effective to enhance the nonlinear strength by using dynamic tail current control to increase the d factor. Under the above ideal conditions, when the dynamic change of the current is close to 4 times, a _V ????, nonlinearity of any intensity can be achieved. In fact, V _GS3,max is limited by a specific circuit structure and power supply voltage, and there is an upper limit for the actual a _V,max . However, under the condition of fixed tail current, in order to improve the a _V factor under dynamic conditions, it can also be realized by reducing h ₀ . According to the inverse relationship between h and a _V , increasing a _V0 directly, that is, statically biasing the current mirror in the weak nonlinear region, can exchange for a large increase in a _V through a moderate increase in power consumption, that is, a small increase in a _V0 .

4、模式可配置线性-非线性自适应电流镜传输特性 4. Mode configurable linear-nonlinear adaptive current mirror transfer characteristics

根据NM3和NM7管的饱和电流模型，输出I_o=(1/2)k₇D_M7 ², D_M7=D_M3, k₁=k₃, 当NM3管进入深度线性电阻区后，其V_DS3减小使该管过驱动D_M3增加，最终决定输出电流的提高，即： According to the saturation current model of NM3 and NM7 tubes, output I _o =(1/2)k ₇ D _M7 ² , D _M7 =D _M3 , k ₁ =k ₃ , when NM3 tube enters the deep linear resistance region, its V _DS3 Decrease makes the overdrive D _M3 of the tube increase, and finally determines the increase of the output current, that is:

           (7) (7)

以上关系描述了直流大信号传输特性，1/N为W/L有关的线性传递系数，b_DC体现出非线性效应的贡献，即非线性效应对线性传递系数的倍增因子，并有： The above relationship describes the transmission characteristics of DC large signals, 1/N is the linear transfer coefficient related to W/L, b _DC reflects the contribution of nonlinear effects, that is, the multiplication factor of nonlinear effects to linear transfer coefficients, and has:

                    (8)

(8)

在a_V=h=1/2的临界模式下，b_DC=1，随着a_V因子增大，倍增因子b_DC明显增加。电流镜非线性效应体现的另一个方面，是直流大信号与交流小信号的传输系数发生分离，两者不在不再是线性条件下的相同量。但交直流特性间存在不可割裂的关联性，利用对直流传输方程的偏导处理，并利用(5)式关系，得到的交流传输方程为： In the critical mode where a _V = h = 1/2, b _DC = 1, and the multiplication factor b _DC increases obviously as the a _V factor increases. Another aspect of the nonlinear effect of the current mirror is that the transmission coefficients of the large DC signal and the small AC signal are separated, and the two are not the same amount under the condition that they are no longer linear. However, there is an inseparable relationship between the AC and DC characteristics. Using the partial derivative treatment of the DC transfer equation and using the relationship in (5), the AC transfer equation is obtained as:

            (9) (9)

交流传输特性与直流传输特性具有相同的形式，其中线性传输系数仍为1/N, 交流传输系数为b_AC。同样，在a_V>>1的强非线性条件下，交流倍增因子近似为： The AC transfer characteristic has the same form as the DC transfer characteristic, where the linear transfer coefficient is still 1/N and the AC transfer coefficient is b _AC . Likewise, under the strong nonlinear condition of a _V >>1, the AC multiplication factor is approximately:

                             (10) (10)

当强非线性效应出现后，交流传输系数明显增加，增加的速度超过直流，同时交直流的强非线性倍增量将超过线性倍增系数。 When the strong nonlinear effect appears, the AC transmission coefficient increases significantly, and the increase speed exceeds that of DC, and at the same time, the strong nonlinear multiplier of AC and DC will exceed the linear multiplication coefficient.

5、运放关键性能参数变化规律 5. Variation law of key performance parameters of operational amplifiers

两种物理效应引起交直流电流传输特性的变化，是导致电路交直流特性参数产生变化的根源，根据图4结构，静态下a、b两点电位相同，电流镜线性传输比为1/N_T，则静态功耗为： The two physical effects cause the change of AC and DC current transmission characteristics, which is the root cause of the change of the AC and DC characteristic parameters of the circuit. According to the structure in Figure 4, the potentials of points a and b are the same under static conditions, and the linear transmission ratio of the current mirror is 1/N _T , then the static power consumption is:

                  (11)

(11)

式中I_B为差分对尾电流，I_o为输出支路电流。在静态点附近的交流小信号条件下，a、b两点转化为差分信号，由于交叉耦合对管的补偿作用，电流镜线性传输比切换为1/N_S, 则运放小信号跨导为： In the formula, I _B is the differential pair tail current, and I _o is the output branch current. Under the condition of AC small signal near the static point, the two points a and b are converted into differential signals. Due to the compensation effect of the cross-coupling on the tube, the linear transfer ratio of the current mirror is switched to 1/ _NS , and the small signal transconductance of the operational amplifier is :

          (12)

(12)

式中g_m1,2为差分对跨导，G_{m_ref}= g_m1,2/N_T为图1常规结构参考运放在相同静态功耗下的跨导，则改进电路跨导的提高来自尺寸配置效应形成的线性比例因子a_AC=N_T/N_S与非线性交流传输效应产生的b_AC因子的共同作用，即两因子的乘积决定跨导总的倍增因子。单级稳定系统的单位增益带宽为GBW=G_m/C_L，因此带宽GBW与G_m具有完全相同的倍增因子。在电流传输和跨导变化的共同影响下，低频增益的变化为： In the formula, g _m1,2 is the transconductance of the differential pair, G _{m_ref} = g _m1,2 /N _T is the transconductance of the conventional structure reference operational amplifier in Figure 1 under the same static power consumption, and the improvement of the improved circuit transconductance comes from the size configuration The linear proportional factor a _AC = _NT / _NS formed by the effect and the b _AC factor generated by the nonlinear AC transmission effect work together, that is, the product of the two factors determines the total multiplication factor of the transconductance. The unity-gain bandwidth of a single-stage stable system is GBW=G _m / _CL , so the bandwidth GBW has exactly the same multiplication factor as G _m . Under the combined effect of current transfer and transconductance change, the low frequency gain changes as:

         (13)

(13)

式中l为MOS管沟道调制因子，I_in为差分负载电流镜的输入电流，b_AC、b_DC分别为静态点附近小信号范围内电流镜非线性传输引起的交流和直流倍增因子。当静态工作点设置在线性模式下，则交流小信号下负载电流镜仍然比较接近线性性质，不但取值较接近1并有b_AC??b_DC，致使电路低频直流增益的倍增因子主要由a_AC因子决定。 In the formula, l is the channel modulation factor of the MOS tube, I _in is the input current of the differential load current mirror, b _AC and b _DC are the AC and DC multiplication factors caused by the nonlinear transmission of the current mirror in the small signal range near the static point, respectively. When the static operating point is set in the linear mode, the load current mirror is still relatively close to the linear property under the AC small signal, not only the value is closer to 1 but also has b _AC ??b _DC , so that the multiplication factor of the low-frequency DC gain of the circuit is mainly determined by a _AC factor decides.

当输入差分信号超出动态范围后，一方面因交叉耦合对管中的电流为零而失效，另一方面总有一个差分负载电流镜在大信号驱动下而注入全部的尾电流，通常可使该电流镜进入强非线性工作模式，由此产生的输出压摆率为： When the input differential signal exceeds the dynamic range, on the one hand, it fails because the current in the cross-coupling pair tube is zero, and on the other hand, there is always a differential load current mirror driven by a large signal to inject all the tail current, which can usually make the The current mirror enters a strongly nonlinear mode of operation, and the resulting output slew rate is:

     (14)

(14)

式中SR_ref=I_{o_ref}/C_L为参考OTA电路的输出压摆率, 电流镜由直流静态点切换到大信号模式后交叉耦合对管因电流为零而作用失效，其有效W/L在两种模式下的变换所产生的线性倍增因子为a_DY=N_T/N，b_DY则为大信号下NM3管进入深线性电阻区后非线性电流镜的直流传输倍增因子b_DC。 Where SR _ref =I _{o_ref} /C _L is the output slew rate of the reference OTA circuit. After the current mirror is switched from the DC static point to the large signal mode, the cross-coupling pair tube is invalid because the current is zero, and its effective W/L is at The linear multiplication factor generated by the transformation in the two modes is a _DY = _NT /N, and b _DY is the DC transmission multiplication factor b _DC of the nonlinear current mirror after the NM3 tube enters the deep linear resistance region under large signal.

很明显，运放各类交直流参数指标的变化均来自两类倍增因子的贡献，一种是负载电流镜W/L在不同模式下切换带来的尺寸线性因子，该因子在不同的信号模式变化下取值不同，但与具体电流大小无关，在特定模式间的切换为特定的线性常数；另一种是电流镜进入非线性传输模式后带来的非线性因子，该因子随实际电流大小而变化，表现出随电路状态 Obviously, the changes of various AC and DC parameters of the operational amplifier come from the contribution of two types of multiplication factors. One is the size linear factor caused by the switching of the load current mirror W/L in different modes. This factor is different in different signal modes. The value is different when the current mirror changes, but it has nothing to do with the specific current. The switching between specific modes is a specific linear constant; the other is the nonlinear factor brought by the current mirror entering the nonlinear transmission mode, which varies with the actual current. vary, exhibiting a function of circuit state

连续变化的性质。                                                                                                                      The nature of continuous change.                       

6、运放最优工作模式设置 6. The optimal working mode setting of the op amp

选择不同的偏置电压V_bn(如图4所示)，理论上可将OTA运放电路偏置在三种不同的静态工作点模式。其中，静态工作点若选择典型线性电流传输模式，即a_V<<1/2, 则以上各类倍增因子均为1，电路退化为传统结构，性能无任何改进，因此没有选择该模式的现实意义；相反，静态工作点若选择典型线性电流传输模式，即a_V>>1/2，b_DC的增加将使静态功耗明显增加，此类模式因不符合低功耗设计要求，无法成为最佳工作模式。因此，为满足运放电路低功耗要求，获得与参考结构相同的功率消耗，只能选取a_V=1/2、b_DC=1的静态偏置条件，这对应于负载电流镜静态工作点选择在线性-非线性传输的临界模式，这一模式或在其附近的工作点模式即对应为系统工作的最佳模式。 By choosing different bias voltage V _bn (as shown in Figure 4), the OTA operational amplifier circuit can be biased in three different quiescent operating point modes in theory. Among them, if the typical linear current transmission mode is selected for the static operating point, that is, a _V <<1/2, the above-mentioned multiplication factors are all 1, and the circuit degenerates into a traditional structure without any improvement in performance, so there is no reality in choosing this mode On the contrary, if the static operating point chooses a typical linear current transmission mode, that is, a _V >>1/2, the increase of b _DC will significantly increase the static power consumption. This type of mode cannot be used because it does not meet the design requirements of low power consumption. Best working mode. Therefore, in order to meet the low power consumption requirements of the operational amplifier circuit and obtain the same power consumption as the reference structure, only the static bias conditions of a _V = 1/2 and b _DC = 1 can be selected, which corresponds to the static operating point of the load current mirror Select the critical mode of linear-nonlinear transmission, this mode or the operating point mode near it corresponds to the best mode for the system to work.

在此模式下，因有b_DC??b_AC??1, 与参考运放电路相比静态功耗保持不变，跨导、带宽和增益的倍增因子均为a_AC、摆率的倍增因子略低于2b_DY。以上结果表明，在最优的低功耗模式下，尺寸线性因子a_AC主要实现对小信号交流特性的倍增，大信号直流深度非线性因子b_DY主要实现对宽摆幅动态特性的倍增，两种效应之间有良好的互补特性，在其共同作用下能够改善电路各个方面的性能。 In this mode, since there is b _DC ??b _AC ??1, the static power consumption remains unchanged compared with the reference op amp circuit, and the multiplication factors of transconductance, bandwidth and gain are all multiplication factors of a _AC and slew rate slightly lower than 2b _DY . The above results show that in the optimal low power consumption mode, the size linear factor a _AC mainly realizes the multiplication of small-signal AC characteristics, and the large-signal DC deep nonlinear factor b _DY mainly realizes the multiplication of wide-swing dynamic characteristics. There are good complementary characteristics between these effects, and the performance of all aspects of the circuit can be improved under their joint action.

以上临界工作点模式的设定由静态尾电流I_B、差分对负载电流镜输入管W/L、Cascode管偏置电压V_bn三者共同决定。为克服NM1管衬偏效应的影响，V_bn应略大于不考虑衬偏时的理论预测值，增加的部分应补偿开启电压因衬偏而带来的变化量DV_TN??zV_DS3。在深度非线性传输模式下，因V_DS3很小，衬偏效应的影响很弱，衬偏效应对a_V以及各相关倍增因子的影响可以忽略。 The setting of the above critical operating point mode is jointly determined by the static tail current I _B , the input transistor W/L of the differential pair load current mirror, and the bias voltage V _bn of the Cascode transistor. In order to overcome the influence of the lining offset effect of the NM1 tube, V _bn should be slightly larger than the theoretical prediction value when the lining offset is not considered, and the increased part should compensate for the variation DV _TN ??zV _DS3 of the turn-on voltage due to the lining offset. In the deep nonlinear transmission mode, because V _DS3 is very small, the influence of the lining offset effect is very weak, and the influence of the lining offset effect on a _V and related multiplication factors can be ignored.

由于采用固定尾电流偏置，动态下电流变化2倍，因此在以上静态临界点模式下，状态因子由静态时的a_V0=1/2增大到动态大信号下的a_V=1，直流传输的非线性倍增因子对应地由b_DC=1增大到b_DY=2，实际的输出摆率倍增幅度仍偏小，远离电路允许的上限范围。为此，在工作模式的定量设置上可略作调整，静态下将电流镜偏置在较弱的非线性，以适度牺牲静态功耗为代价，换取大信号下a_V和b_DY的提高，改善大信号瞬态驱动特性。在具体实现方法上，可在原有基础上适当降低D_Mb偏置电压，或在D_Mb偏压不变的条件下适当增加NM3、NM4两管的W/L。 Due to the use of fixed tail current bias, the dynamic current changes by 2 times, so in the above static critical point mode, the state factor increases from a _V0 = 1/2 in the static state to a _V = 1 in the dynamic large signal, DC The nonlinear multiplication factor of transmission increases from b _DC =1 to b _DY =2 correspondingly, and the actual output slew rate multiplication range is still relatively small, which is far away from the upper limit range allowed by the circuit. For this reason, the quantitative setting of the working mode can be slightly adjusted, and the current mirror is biased at a weaker nonlinearity under static conditions, at the expense of moderately sacrificing static power consumption, in exchange for the improvement of a _V and b _DY under large signals, Improve large signal transient drive characteristics. In terms of specific implementation methods, the _DMb bias voltage can be appropriately reduced on the original basis, or the W/L of the NM3 and NM4 transistors can be appropriately increased under the condition that the _DMb bias voltage remains unchanged.

7、倍增因子的上限范围 7. The upper limit range of the multiplication factor

在最佳工作模式下，为获得最优的电路交直流和大信号特性需要尽可能提高各类倍增因子的上限范围，但因受实际电路结构和工作条件的限制，各类倍增因子均有特定的上限，缓解这些限制对发挥电路的最大潜能有重要作用。 In the best working mode, in order to obtain the optimal circuit AC, DC and large signal characteristics, it is necessary to increase the upper limit range of various multiplication factors as much as possible, but due to the limitation of actual circuit structure and working conditions, various multiplication factors have specific , and alleviating these limitations is important for realizing the maximum potential of the circuit.

交流小信号参数倍增因子a_AC的上限由运放带宽小于次极点p₂的条件限定，输出高阻大电容决定了输出极点为系统主极点，电路内部的1/g_m型低阻负载结点构成系统次极点。在图4各结点位置中，d、e两结点因电容相对较小导致极点频率较高，而差分对负载电流镜a、b两结点处因寄生栅电容最大且等效跨导最小而成为频率最低的次极点，并且因结构对称而使两个次极点频率相等。 AC small signal parameter multiplication factor a The upper limit of _AC is limited by the condition that the operational amplifier bandwidth is smaller than the sub-pole p ₂ , the output high-impedance large capacitance determines that the output pole is the main pole of the system, and the 1/g _m -type low-impedance load node inside the circuit constitute the sub-pole of the system. In the position of each node in Figure 4, the pole frequency of the two nodes d and e is relatively small due to the relatively small capacitance, while the two nodes of the differential pair load current mirror a and b have the largest parasitic gate capacitance and the smallest equivalent transconductance And become the sub-pole with the lowest frequency, and the frequency of the two sub-poles is equal due to the symmetry of the structure.

以a点小信号等效电容C_a为例，该电容应包含NM3、NM7、NM6三管的栅电容以及PM1、NM1两管的漏电容，若以上各MOS管的W/L越大，同时从降低失配方面考虑L取值越大，则大的器件面积形成的C_a电容也越大，极点频率p_a,b越小。在60°相位裕度的约束下，电路带宽GBW越小，允许的a_AC倍增因子上限峰值也越小。在最佳模式下，a、b两结点的交流输出阻抗近似为r_a,b??1/g_{m_Ns}, 其中g_{m_Ns}为静态工作点下差分对负载电镜的流交流小信号跨导，由此形成的寄生次极点为p_a,b=g_{m_Ns}/C_a, 带宽GBW=G_m/C_L。 Taking the small-signal equivalent capacitance C _a at point a as an example, the capacitance should include the gate capacitance of NM3, NM7, and NM6 and the drain capacitance of PM1 and NM1. If the W/L of each of the above MOS transistors is larger, at the same time From the perspective of reducing mismatch, the larger the value of L, the larger the capacitance C _a formed by the large device area, and the smaller the pole frequency p _a,b . Under the constraint of 60° phase margin, the smaller the circuit bandwidth GBW is, the smaller the upper limit peak value of the allowable a _AC multiplication factor is. In the best mode, the AC output impedance of the two nodes a and b is approximately r _a,b ??1/g _{m_Ns} , where g _{m_Ns} is the current AC small signal transconductance of the differential pair load electron microscope at the static operating point, given by The formed parasitic secondary pole is p _a,b =g _{m_Ns} /C _a , and the bandwidth GBW=G _m / _CL .

考虑到两个重合的低频次极点，系统稳定应满足4GBW<p_a,b的限制。从差分对参数考虑，通过降低差分对管W/L可使输入跨导和参考运放的带宽降低，则允许的交流倍增因子增大；从差分负载电流镜考虑，通过增大N_T值可使电流传输系数及参考运放的带宽减小，同样使允许的交流倍增因子增大。根据以上带宽约束调节，结合差分对管与其负载管增益因子k_PM12=k_NT的对称或近似相同的条件，可得： Considering the two coincident low-frequency sub-poles, the system stability should satisfy the limit of 4GBW<p _a,b . Considering the parameters of the differential pair, by reducing the W/L of the differential pair tube, the input transconductance and the bandwidth of the reference op amp can be reduced, and the allowable AC multiplication factor increases; considering the differential load current mirror, by increasing the _NT value, the Reducing the current transfer coefficient and the bandwidth of the reference op amp also increases the allowable AC multiplication factor. According to the above bandwidth constraint adjustment, combined with the condition that the differential pair tube and its load tube gain factor k _PM12 =k _NT are symmetrical or approximately the same, it can be obtained:

                   (15) (15)

在深亚微米工艺条件，寄生电容C_a可控制在几百皮法的范围以内，则输出电容与C_a电容之比将达到2个数量级以上，由此允许的交流倍增因子上限达到a_AC>10以上。 In deep submicron process conditions, the parasitic capacitance C _a can be controlled within the range of several hundred picofarads, and the ratio of output capacitance to C _a capacitance will reach more than 2 orders of magnitude, thus allowing the upper limit of the AC multiplication factor to reach a _AC > 10 or more.

输出摆率受到的各种制约因素，其中之一是输入摆率的限制，当输出摆率增加到超出电路内部摆率时电路的速度将由内部结点电容的充放电响应决定，增加输出电容的充放电速度已无意义。因此内部结点充放电的最大摆率，对应于输出最大摆率的上限值，由此得到大信号下b_DY因子的上限为： Various constraints on the output slew rate, one of which is the limitation of the input slew rate, when the output slew rate increases beyond the internal slew rate of the circuit, the speed of the circuit will be determined by the charge and discharge response of the internal node capacitance, increasing the output capacitance The charging and discharging speed is meaningless. Therefore, the maximum slew rate of internal node charging and discharging corresponds to the upper limit of the maximum output slew rate, and thus the upper limit of the _bDY factor under large signals is:

                            (16) (16)

此外，差分对共源端c点共模电平对a、b两点电压最大值的限制，对输出电流和输出摆率同样产生重要影响。c点电位随差分输入共模电平V_com而变化，最高增加到使尾电流进入线性电阻-饱和恒流的临界点，即V_c,max=V_DD-D_PM0, 其中D_PM0为尾电流管的过驱动电压, 实际电位为V_c=V_TP+D_PM1,max, 其中D_PM1,max为动态大信号下差分对管的最大过驱动电压，即对应差分输入的动态范围。因c点电位将在V_c??V_c,max范围内变化，且V_a,b??V_c, 则输出管的最大过驱动电压为V_c-V_TN; 而静态时输出管的过驱动电压为a_VD_Mb, 并且在最佳模式下a_V??1/2。考虑到输出驱动电流与输出饱和管过驱动电压的平方成正比，而且输出电流的变化来自电流镜工作模式由临界线性切换到非线性，则受V_c电平限制的b_DY范围为：  In addition, the limitation of the common-mode level of point c on the common source end of the differential pair to the maximum value of the voltages at points a and b also has an important impact on the output current and output slew rate. The potential of point c changes with the differential input common-mode level V _com , and the maximum increases to the critical point where the tail current enters the linear resistance-saturation constant current, that is, V _c,max =V _DD -D _PM0 , where D _PM0 is the tail current The overdrive voltage of the tube, the actual potential is V _c =V _TP +D _PM1,max , where D _PM1,max is the maximum overdrive voltage of the differential pair tube under dynamic large signal, that is, the dynamic range corresponding to the differential input. Because the potential of point c will change within the range of V _c ??V _c,max , and V _a,b ??V _c , the maximum overdrive voltage of the output transistor is V _c -V _TN ; and the overdrive voltage of the output transistor in static state The driving voltage is a _V D _Mb , and in the best mode a _V ??1/2. Considering that the output drive current is proportional to the square of the overdrive voltage of the output saturation tube, and the change of the output current comes from the switching of the current mirror operating mode from critical linear to nonlinear, the range of b _DY limited by the V _c level is:

              (17) (17)

由上式看出，增加电源电压并采用静态低功耗偏置以降低D_Mb过驱动电压，能够提高摆率倍增因子的峰值水平。在V_com??V_DD-(V_TP+D_PM1,max+D_PM0)范围内，增加V_com可以提高允许的最大输出直流倍增因子，而实际产生的倍增因子，与差分对负载电流镜输入电流动态范围内的a_V因子有关，输入电流动态范围越宽，a_V变化越大，非线性效应越强，则实际产生的非线性倍增因子b_DY越大。 It can be seen from the above formula that increasing the power supply voltage and adopting static low power consumption bias to reduce the overdrive voltage of D _Mb can increase the peak level of the slew rate multiplication factor. In the range of V _com ??V _DD -(V _TP +D _PM1,max +D _PM0 ), increasing V _com can increase the maximum allowable output DC multiplication factor, and the actual multiplication factor, and the differential pair load current mirror input The a _V factor in the dynamic range of the current is related. The wider the dynamic range of the input current, the greater the change of a _V , the stronger the nonlinear effect, and the larger the actual nonlinear multiplication factor _bDY .

在最佳模式下，应合理配置静态尾电流I_B与过驱动偏置D_MB的相对关系，通过合理设置偏置管与差分负载电流镜中各MOS管的W/L, 可以使实际输出的摆率倍增因子能够达到其特定共模输入电平下所对应的理论上限。 In the optimal mode, the relative relationship between the static tail current I _B and the overdrive bias D _MB should be reasonably configured. By setting the W/L of each MOS transistor in the bias transistor and the differential load current mirror reasonably, the actual output The slew rate multiplication factor can reach its theoretical upper limit for a specific common-mode input level.

8、实际电路设计及其性能改善 8. Actual circuit design and performance improvement

对于图4电路结构，典型工作条件为V_DD=3.3V, C_L=30pF, I_B=10mA, 提供V_bn的偏置支路10mA, 静态下偏置在a_V=1/2临界工作点模式, 各MOS管近似0.2V的过驱动电压。 For the circuit structure in Figure 4, the typical operating conditions are V _DD =3.3V, C _L =30pF, I _B =10mA, the bias branch that provides V _bn is 10mA, and the static bias is at a _V =1/2 the critical operating point mode, the overdrive voltage of each MOS transistor is approximately 0.2V.

实际参数设计选取N_T=1, N_S=0.1, a_AC=N_T/N_S=10, a_DY=N_T/N=1/0.55=1.82。采用CSMC 0.18mm标准CMOS工艺，V_TN=0.5V, V_TP=0.6V, a、b结点寄生电容近似为C_a<0.2pF, C_L/C_a>100。 The actual parameter design selects N _T =1, N _S =0.1, a _AC =N _T /N _S =10, a _DY =N _T /N=1/0.55=1.82. Using CSMC 0.18mm standard CMOS process, V _TN =0.5V, V _TP =0.6V, a, b node parasitic capacitance is approximately C _a <0.2pF, C _L /C _a >100.

图5为电路交流特性仿真结果，在60°相位裕度约束下，实际带宽相比参考运放提高9倍，理论预测为10倍；实际低频增益提高22dB, 理论预测为20dB。因此，对于交流特性，理论模型计算与实际电路仿真结果十分吻合。 Figure 5 shows the simulation results of the AC characteristics of the circuit. Under the 60° phase margin constraint, the actual bandwidth is 9 times higher than the reference op amp, and the theoretical prediction is 10 times; the actual low-frequency gain is 22dB higher, and the theoretical prediction is 20dB. Therefore, for the AC characteristics, the theoretical model calculations are in good agreement with the actual circuit simulation results.

图6为由运放电路构成电压跟随器的大信号动态响应特性，从图中延迟数据可求出由仿真测试得到的输出摆率，同样对参考结构进行类似的测试，两者比较发现，摆率倍增达到20倍，按理论计算出非线性因子为a_V=3.2，b_DC=10，a_DY=1.82, 摆率倍增达到18.6倍, 同样与仿真结果吻合。 Figure 6 shows the large-signal dynamic response characteristics of a voltage follower composed of an operational amplifier circuit. From the delay data in the figure, the output slew rate obtained by the simulation test can be obtained. A similar test is also carried out on the reference structure. A comparison between the two shows that the swing rate The rate multiplication reaches 20 times. According to the theoretical calculation, the nonlinear factors are a _V = 3.2, b _DC = 10, a _DY = 1.82, and the slew rate multiplication reaches 18.6 times, which is also consistent with the simulation results.

测试中输入信号共模电平为1.5V, 该电平限制的b_DC??40，实际产生的参数倍增均远离其理论上限，尤其在高电源和高输入共模电平下。最后，表1给出了传统OTA(图1)与本发明OTA(图4)性能仿真结果的对比。 The common-mode level of the input signal is 1.5V in the test, and the b _DC of this level limit is ??40. The actual parameter multiplication is far away from its theoretical upper limit, especially at high power supply and high input common-mode level. Finally, Table 1 shows the comparison of performance simulation results of the traditional OTA (FIG. 1) and the OTA of the present invention (FIG. 4).

表1 性能对比(C_L=30pF) Table 1 Performance comparison ( _CL =30pF)

参数parameter 传统OTA结构Traditional OTA structure 本发明OTA结构OTA structure of the present invention 新旧结构对比Comparison of old and new structures 工作电压(V)Working voltage (V) 3.33.3 3.33.3 工作电压相同same working voltage 静态电流(mA)Quiescent current (mA) 3030 2929 静态电流相同same quiescent current 低频增益(dB)Low frequency gain(dB) 49.449.4 71.371.3 增益提高22 dBGain increased by 22 dB 单位增益带宽(Hz)Unity Gain Bandwidth (Hz) 714k714k 6.5M6.5M 带宽增加9倍9x increase in bandwidth 相位裕度(deg)Phase Margin(deg) 89.489.4 60.660.6 减小但满足要求Reduced but meets the requirements 正摆率(V/ms)Positive Slew Rate(V/ms) +0.62+0.62 +12.5+12.5 提高20倍20 times higher 负摆率(V/ms)Negative Slew Rate(V/ms) -0.65-0.65 -12.8-12.8 提高20倍20 times higher PSRR(dB)PSRR(dB) 28.610kHz28.610kHz 7210kHz7210kHz 增加40 dB以上Increase over 40 dB

9、总结 9. Summary

利用输入电流的大小控制动态自适应电流镜的状态和工作模式，能够最大程度地兼容静态、交流小信号和动态大信号的性能。静态时较小的输入电流将可配置结构偏置在线性电流镜模式以求低功耗，交流大信号下变化下因输入电流动态范围增大而带来可配置电流镜强非线性效应。 Using the magnitude of the input current to control the state and working mode of the dynamic self-adaptive current mirror can maximize the compatibility with static, AC small-signal and dynamic large-signal performance. The small input current in static state biases the configurable structure in the linear current mirror mode for low power consumption, and the configurable current mirror has a strong nonlinear effect due to the increase of the dynamic range of the input current under large AC signal changes.

本发明电路充分利用了时变非线性电流镜组合结构产生的效应，即在不同工作模式下利用交叉耦合对管对W/L配置实现的时变特性，结合对线性-非线性自适应电流镜的非线性模式控制，实现了在静态低功耗条件下差分电流传输的倍增，克服了线性单模式运放电路难以克服的内在局限，全面提升了电路的综合性能，理论模型、仿真结果和相关实验结果均验证的新电路可行性与优越性。 The circuit of the present invention makes full use of the effect produced by the time-varying nonlinear current mirror combination structure, that is, the time-varying characteristics realized by the cross-coupling pair of tube-to-W/L configurations are utilized in different operating modes, combined with the linear-nonlinear adaptive current mirror The non-linear mode control realizes the multiplication of differential current transmission under the condition of static low power consumption, overcomes the inherent limitations that are difficult to overcome in linear single-mode operational amplifier circuits, and comprehensively improves the comprehensive performance of the circuit. Theoretical models, simulation results and related The experimental results have verified the feasibility and superiority of the new circuit.