TWI553495B - Method and apparatus for solving static operating points with PN junction characteristics - Google Patents

Description

Method and device for solving static working point with PN junction characteristics

The present invention relates to a method and apparatus for solving a static operating point of a nonlinear circuit, and more particularly to a method and apparatus for solving a static operating point of a nonlinear circuit having a PN junction characteristic.

The Simulation Program with Integrated Circuit Emphasis (HSPICE) requires DC analysis to establish the DC bias point of the circuit before performing any form. With this as a starting point, simulations of other properties such as transients, AC small signals, and noise can be performed. The calculation of the static working point of the circuit is essentially a mathematical solution to a nonlinear algebraic equation. In order to establish the DC analysis point of the circuit, HSPICE must solve a set of nonlinear equations describing the behavior of the circuit. The methods that can be solved numerically by the commonly used nonlinear algebraic equations are: Newton-Raphson Algorithm (NR) Law), continuation method and pseudo transient method. Although the basic principles of these methods are well known, how to implement algorithms with excellent performance and strong convergence for the characteristics of circuit simulation has always been the most troublesome problem for integrated circuit designers. In particular, the convergence problem of static operating points of circuits is the most difficult problem in circuit simulation.

In circuit simulation, the quiescent operating point is the basis for solving all other circuit characteristics. However, the static working point of nonlinear circuits is often difficult to find. The classical method for solving nonlinear equations is the N-R algorithm. The step of the N-R algorithm is to first give an appropriate initial value and then substitute it into the equation for iteration until the absolute value of the difference between the two successive solution vectors is less than a certain allowable error. The N-R algorithm may not converge under certain conditions, such as when the nonlinear equation is discontinuous, or the initial value used in the calculation process is inaccurate. When it is difficult to converge during the calculation process, HSPICE will increase the number of iteration operations or re-calculate after reducing the step size. However, these steps increase the simulation time, and after these steps the node voltage or current may still not converge to cause the analog to be interrupted. Due to the local convergence of the NR algorithm, the general circuit often cannot obtain the stable static working point by using the NR algorithm because it is not close enough to the initial state guess value of the real solution. Therefore, the convergence is the largest of these methods. problem.

The nonlinear component is one of the main reasons why HSPICE cannot converge. The following is a description of a PN junction component. If the circuit includes a PN junction component, its DC characteristics can be represented by a nonlinear current source i _D associated with the PN junction voltage across the VD:

During the solution process, HSPICE must establish the DC bias point of the circuit within a certain number of iterations, otherwise it will generate a non-convergence condition and interrupt the simulation. Since the IV characteristic curve of the PN junction element is exponentially rising, a small V _D variation in the circuit during the iteration process causes a large amount of current variation, making the circuit difficult to converge. In order to solve this problem, in HSPICE, certain measures have been taken for nonlinear components such as diodes, triodes, and MOS transistors, such as adding parallel conductance (gmin), PN junction limiting voltage, and the like.

If the HSPICE circuit includes a PN junction component, it includes a zero conductance region during reverse bias operation. This area creates a divide-by-zero error condition during circuit simulation. To avoid this problem, a gmin transducing element is connected in parallel to the PN junction at each PN junction in each HSPICE semiconductor component to avoid the next voltage generated during the iteration process from being far from the final solution.

However, current circuit designs often use components such as varistor, voltage-controlled current source (VCCS) to simulate PN junction characteristics, which are called PN-like components. The nonlinear resistance characteristics of such PN junction-like components hinder the convergence characteristics of HSPICE. In order to detect the components to enhance convergence by using the parallel conductance stepping or PN junction limiting voltage, it is necessary to propose a detection type PN junction component. The method is to identify the components and propose a simulation method for the PN junction components to improve the convergence of the circuit.

Figure 1 illustrates a schematic diagram of a process for solving a static operating point of a nonlinear circuit, in which the vertical axis represents current and the horizontal axis represents voltage, while the exponential of the PN-like junction exhibits an increasing characteristic. In the process of solving the static working point, it is assumed that the next step point B is entered after the initial point A. However, because of the exponential increasing characteristic of the PN-like junction, the gap between point A and point B is too large to be reached. convergence. As can be seen from Figure 1, due to the existence of such strong nonlinear characteristics, a slight change in the node voltage during the iterative process will result in a large increase in the output current, which is disadvantageous for the next NR iteration process. of. In order to avoid the non-linear characteristics of the nonlinear components in the circuit, the N-R algorithm fails, so it is very important to find the nonlinear components in the circuit and adjust them in advance.

The invention provides a method and a device for solving a static working point of a nonlinear circuit with a PN junction characteristic, which are mainly divided into two phases, the first phase is for detecting components with PN junction characteristics, and the second phase is for adjusting PN junctions. Characteristic component. Since the component with PN-like characteristics is one of the main reasons why HSPICE cannot converge, if it can be detected and adjusted in advance, the obstacle of HSPICE convergence can be eliminated in advance, and the speed and probability of HSPICE convergence can be accelerated.

The present invention provides a method for solving a quiescent operating point of a nonlinear circuit having a PN junction characteristic, comprising the steps of: detecting an element having an exponential change in voltage and current characteristics in the nonlinear circuit, and marking the PN junction characteristic as a class; Adjusting the components of the PN-like characteristic; and solving the adjusted static operating point of the nonlinear circuit by using a nonlinear equation.

The invention provides a device for solving a static working point of a nonlinear circuit with a PN junction characteristic, comprising a detecting unit, an adjusting unit and a calculating unit. The detecting unit detects an element whose exponential change in voltage and current characteristics in the nonlinear circuit is marked as having a PN-like characteristic. The adjustment unit adjusts the components labeled as PN-like characteristics. The computing unit solves the adjusted static operating point of the nonlinear circuit using a nonlinear equation.

The detecting unit of an embodiment of the invention comprises a nonlinear resistance detecting module and a VCCS detecting module. The varistor detection module adds a test voltage to both ends of the possible varistor and obtains a conductance value at each test voltage. If the set of conductance values has an exponential characteristic, the non-linear resistance is marked as having a PN-like characteristic. The VCCS detection module adds test voltages to both ends of a possible voltage control current source and obtains current values for each set of test voltages. If the set of current values has an exponential characteristic, the voltage controlled current source is marked as having a PN junction characteristic.

The technical features of the present invention have been briefly described above, and the detailed description below will be better understood. Other technical features constituting the subject matter of the patent application of the present invention will be described below. It is to be understood by those of ordinary skill in the art that the present invention may be practiced otherwise. It is to be understood by those of ordinary skill in the art that this invention is not limited to the scope of the invention.

In order to facilitate a better understanding of the spirit of the invention, the following description is further described in conjunction with the preferred embodiments of the invention. The direction discussed herein is a method and apparatus for solving a static operating point of a nonlinear circuit of a PN-like characteristic. In order to thoroughly understand the present invention, detailed steps and compositions will be set forth in the following description. It will be apparent that the implementation of the invention is not limited to the specific details familiar to those skilled in the circuit design. On the other hand, well-known components or steps are not described in detail to avoid unnecessarily limiting the invention. The preferred embodiments of the present invention are described in detail below, but the present invention is not limited by the scope of the invention, and the scope of the invention is not limited by the scope of the appended claims.

2 is a flow chart of solving a static operating point of a nonlinear circuit having a PN junction characteristic according to an embodiment of the present invention. At step 21, an element having an exponential change in voltage and current characteristics within the non-linear circuit is detected and labeled as having a PN-like characteristic. The component having the PN junction characteristic is the target to be detected by the present invention, so the primary object is to detect it. At step 22, the elements labeled as PN-like characteristics are adjusted. The adjustment method may be to make appropriate adjustments, such as limiting the voltage of the PN-like junction, adding parallel conductance (gmin), and the like. In step 23, a static operating point of the nonlinear circuit is solved using a nonlinear equation, such as calculating a Jacobi matrix after the voltage limitation, and performing NR iteration until the static working point of the circuit is obtained.

3 is a flow chart of detecting a PN junction characteristic element of an embodiment of the present invention. First determine the components that may exhibit PN-like characteristics as nonlinear resistors and voltage controlled current sources (VCCS). A test voltage is then added to both ends to test the current or resistance characteristics of the component.

At step 31, a test voltage is applied to both ends of the varistor and the conductance values at each set of test voltages are determined. At step 32, it is determined whether the set of conductance values has an exponential characteristic. If the answer is yes, go to step 33, otherwise end the test. For a varistor, if the test conductance increases exponentially with increasing test voltage, it can be judged that the varistor has a PN-like junction characteristic. At step 33, the varistor is marked as having a PN-like characteristic.

At step 34, a test voltage is applied to both ends of the voltage controlled current source, and current values at each set of test voltages are obtained. At step 35, it is determined whether the set of current values has an exponential characteristic. If the answer is yes, go to step 36, otherwise end the test. For VCCS, if the test current increases exponentially with the increase of the test voltage, it can be judged that the VCCS has a PN-like characteristic. At step 36, the voltage controlled current source is marked as having a PN junction characteristic.

After detecting the relevant components with PN-like characteristics, we can do related processing in the next N-R iteration process, such as PN-like junction voltage limiting, adding parallel conductance (gmin). These measures are beneficial to limit the component current or resistance within a certain range and to increase the degree of linearization of the component by parallel conducting.

4 is a schematic diagram of an apparatus for solving a quiescent operating point of a nonlinear circuit having a PN junction characteristic according to an embodiment of the present invention, which includes a detecting unit 41, an adjusting unit 42, and a calculating unit 43. The detecting unit 41 detects an element whose voltage and current characteristics in the nonlinear circuit are exponentially changed and is marked as having a PN-like characteristic. The adjusting unit 42 adjusts an element labeled as a PN-like junction characteristic, for example, as a PN-like junction voltage limit, adding a parallel conductance (gmin) or the like. The calculating unit 43 solves the static working point of the nonlinear circuit by using a nonlinear equation, for example, calculating the Jacobi matrix after the voltage limitation, and performing NR iteration until the static working point of the circuit is obtained.

FIG. 5 is a schematic diagram of a detecting unit 41 according to an embodiment of the present invention, which includes a nonlinear resistance detecting module 51 and a VCCS detecting module 52. The varistor detecting module 51 adds a test voltage to both ends of the possible varistor and obtains a conductance value at each set of test voltages. If the set of conductance values has an exponential characteristic, for example, the test conductance increases exponentially with an increase in the test voltage, it is judged that the varistor has a PN-like junction characteristic, and the varistor is marked as having a PN-like characteristic. The VCCS detection module 52 adds test voltages to both ends of a possible voltage control current source and obtains current values for each set of test voltages. If the current value of the group has an exponential characteristic, for example, the test current increases exponentially with an increase in the test voltage, it is judged that the VCCS has a PN-like junction characteristic, and the voltage control current source is marked as a PN-like characteristic.

The steps of the present invention are divided into two phases, the first phase is to detect components having PN junction characteristics, and the second phase is to adjust components of PN junction characteristics.

In the first stage, the test voltage is added to both ends of the nonlinear resistor to obtain the conductance value under each group of test voltages, and then according to a certain pattern recognition rule, it is judged whether the set of test conductance values has an exponential curve characteristic. If it has the characteristic of an exponential curve, it is determined that the non-linear resistance is a PN-like characteristic resistance and is marked. Secondly, the present invention utilizes a test voltage applied across the VCCS to obtain a current value at each set of test voltages. According to a certain pattern recognition rule, it is judged whether the set of test current values have the characteristics of an exponential curve. If there is a characteristic of the exponential curve, it is determined that the VCCS is a PN-like VCCS, which is marked.

In the second stage of the adjustment process, when the NR iteration is solved for the static working point, the PN-like junction voltage is limited to the component labeled as having the PN-like characteristic, and the parallel conductance (gmin) is added. After that, the Jacobi matrix after the voltage limitation is calculated, and the NR iteration is performed until the static working point of the circuit is obtained.

Since the present invention detects and pre-adjusts components having PN-like characteristics that are unfavorable for HSPICE convergence, the obstacle of HSPICE convergence can be eliminated in advance, and the speed and probability of HSPICE convergence can be accelerated.

The technical contents and technical features of the present invention have been disclosed as above, and those skilled in the art can still make various substitutions and modifications without departing from the spirit and scope of the invention. Therefore, the scope of the present invention is not to be construed as being limited to the details of the embodiments disclosed herein

21~23. . . step

31~36. . . step

41. . . Detection unit

42. . . Adjustment unit

43. . . Computing unit

51. . . Nonlinear resistance detection module

52. . . VCCS detection module

Figure 1 illustrates a schematic diagram of a process for solving a static operating point of a nonlinear circuit;

2 is a flow chart of solving a static operating point of a nonlinear circuit having a PN junction characteristic according to an embodiment of the present invention;

3 is a flow chart of detecting a PN junction characteristic component of an embodiment of the present invention;

4 is a schematic diagram of an apparatus for solving a static operating point of a nonlinear circuit having a PN junction characteristic according to an embodiment of the present invention; and

Figure 5 is a schematic illustration of a detection unit in accordance with an embodiment of the present invention.

31~36. . . step

Claims (14)

A method for solving a static operating point of a nonlinear circuit having a PN-like characteristic includes the following steps: using a computer simulation to detect an element whose exponential change in voltage and current characteristics in the nonlinear circuit is marked as having a PN-like characteristic, Each of the components is subjected to a simulated varying voltage to determine whether the current response varies exponentially with the varying voltage; the component labeled as a PN-like characteristic is adjusted; and the nonlinear equation is used to solve the adjustment The static operating point of the nonlinear circuit. The method of claim 1, wherein the component having the PN junction characteristic comprises a non-linear resistor and a voltage controlled current source. The method of claim 1 wherein the adjusting comprises limiting the PN junction or adding a parallel conductance. The method of claim 3, wherein the parallel conductance is an input and an output of the PN junction. The method of claim 2, wherein the varistor and the voltage controlled current source are elements that are characteristic of the analog PN junction. According to the method of claim 1, it is used in a circuit simulation environment of HSPICE. The method of claim 1, wherein the nonlinear equation is using a Newton-Raphson method. A device for solving a static operating point of a nonlinear circuit having a PN-like characteristic, comprising: A detecting unit uses a computer simulation to detect an element whose exponential change in voltage and current characteristics in the non-linear circuit is marked as having a PN-like characteristic, wherein each element of the element is subjected to an analog variable voltage to determine its current Whether the reaction changes exponentially with the fluctuating voltage; an adjusting unit adjusts the component marked as a PN-like characteristic; and a calculating unit solves the adjusted static operating point of the nonlinear circuit by using a nonlinear equation. The device of claim 8, wherein the component having the PN junction characteristic comprises a non-linear resistor and a voltage controlled current source. The device of claim 8, wherein the adjustment unit limits or adds parallel conductance to the PN-like junction. The device of claim 10, wherein the parallel conductance is an input and an output of the PN junction. The device of claim 9, wherein the varistor and the voltage controlled current source are elements used to simulate PN junction characteristics. The apparatus of claim 8 which is used in a circuit simulation environment of HSPICE. The apparatus of claim 8, wherein the nonlinear equation is a Newton-Raphson method.