JPH03182968A - Circuit simulation method - Google Patents

Abstract

PURPOSE:To considerably shorten time required for the second and succeeding DC characteristic analysis by using the solution of initial DC characteristic analysis as the initial value of DC characteristic analysis in the second and succeeding analysis. CONSTITUTION:In addition to a simulation input data generation part 1, analysis part 2 and a simulation result display part 4, a simulation result extraction part 3 are included. The simulation result extraction part 3 includes a DC characteristic result extraction part 301, a storage part for setting voltage values for respective nodes 302 which extracts and stores DC working point incorporated in the result of DC characteristic analysis as the setting voltage values of respective nodes and an input data generation control switch part 303 giving the setting voltage values of respective nodes to a DC characteristic analysis part 201. The analyzed result of first DC analysis is used for the second and succeeding DC analysis. Thus, time required for the second and succeeding DC characteristic analysis can be shortened.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a circuit simulation method for simulating an electronic circuit and supporting its design, and in particular, a circuit simulation method for performing circuit simulation using an improved circuit simulator. Regarding.

[Prior Art] Electronic circuit design in LSI design, etc. is generally performed using "SP I CE", except for circuits for which configuration design techniques have been established.
Simulations are mainly carried out by computer systems using circuit simulators, such as software called It is a highly versatile CAD tool that allows a dedicated arithmetic processing device such as an analysis machine to analyze various electrical characteristics of the electronic circuit in response to related human power.

FIG. 4 is a diagram showing a circuit simulation system including an analysis machine that operates according to a conventional circuit simulator as a device constituted by a plurality of functional blocks. Therefore, in this diagram, multiple functions actually realized by the CPU of the analysis machine are shown as being realized by multiple different functional blocks.

Referring to the figure, the conventional circuit simulation system includes a simulation human data creation section 1, an analysis section 2, and a simulation result display section 4.

The simulation input data creation unit 1 includes a circuit diagram manual unit 101 that takes in a circuit diagram of an electronic circuit to be analyzed given by a user, and a circuit diagram input unit 101 that takes in a circuit diagram of an electronic circuit to be analyzed, and calculates connection relationships between electrical elements from the circuit diagram imported by the circuit diagram 101. and a netlist extraction unit 102 that extracts the characteristics of each element and creates a netlist showing them, and sets simulation conditions such as input signals to the electronic circuit and type of analysis according to the user's human power. , a simulation condition setting unit 103 that outputs a signal indicating a request to start simulation, and a DC characteristic analysis initial value storage unit 104 that stores a predetermined value or a value given by the user as an initial value in DC characteristic analysis to be described later. including.

The analysis unit 2 analyzes various data, that is, netlists, created by the simulation input data creation unit 1 in response to a request to start simulation from the simulation condition setting unit 103.

A DC characteristic analysis unit 201 that performs DC characteristic analysis based on simulation conditions and initial values in DC characteristic analysis, a simulation result receiving unit 202 that receives the analysis results from the DC characteristic analysis unit 201, and a simulation result receiving unit 202. Based on the received DC characteristic analysis results, an AC phase signal H-folding unit 203 and a transient angle Q-folding unit 204 perform other types of circuit analysis, in this case AC phase signal H analysis and transient analysis, which will be described later. including.

In the analysis section 2, the simulation result display section 4
The final analysis results are displayed on a display, printer, etc.

DC characteristic analysis is a circuit analysis that determines the DC operating point, element current, etc. after a sufficient period of time has elapsed in an electronic circuit driven only by a DC power source.

That is, in the DC characteristic analysis, the voltage at the connection point (hereinafter referred to as a node or node) between electric elements in the circuit and the current flowing through each element in a steady state after application of the power supply voltage are determined. For this reason, DC characteristic analysis includes AC small signal analysis, which analyzes the gain and phase frequency characteristics of a circuit (usually the output of a circuit is a transfer function), and AC small signal analysis, which analyzes the frequency characteristics of a circuit's gain and phase (usually the output of a circuit is a transfer function), and designation when a signal with a certain waveform is manually applied to a circuit. This is performed prior to various circuit analyzes such as transient analysis, which analyzes the time domain response characteristics at nodes that have been analyzed, and is used as basic data for these various analyses.

AC small signal analysis is performed by repeating the operation of creating a nodal equation, which will be described later, for each desired AC frequency and finding its solution.

Transient analysis generally involves converting time-dependent elements such as capacitors and inductors included in a circuit into a model represented by a resistor and power supply, and then analyzing the DC characteristics of the resulting nonlinear circuit. This is done by repeating the process at non-continuous discrete time intervals up to a desired time. This is equivalent to converting the time-domain circuit differential equation derived from the original circuit into a nonlinear difference equation at each discrete time and finding a solution. A time-dependent element is converted into a model represented by a resistance and a power source by numerically integrating a current/voltage relational expression representing its characteristics over adjacent discrete times.

The DC characteristic analysis method generally employs a method of solving nodal equations using nodal voltages, which are voltage reference variables, as variables. By the way, since the circuit to be simulated often includes nonlinear elements such as semiconductor elements such as diodes and transistors, the nodal equations to be solved are nonlinear equations. Therefore, the nodal equation is generally used as a Newton-Raphson method, which is frequently used as a numerical analysis method for nonlinear equations.
on) is solved by the method. The Newton-Loughran method gives a temporary solution to a nonlinear equation as an initial value, linearizes the nonlinear equation around this initial value, and then solves this linearized equation to solve the original nonlinear equation around the solution obtained. By linearizing, we create a new linear equation, solve it, and linearize the original nonlinear equation again around the solution obtained, until the resulting solution has a certain value, that is, This method is repeated until it converges on the true value.

When the Newton-Loughn method as described above is applied to the angle lr method of the nodal equation, the following operations are performed in actual circuit analysis.

By linearizing nonlinear elements included in a circuit to be simulated around a certain estimated operating point (a certain voltage and current value), nodal equations are created by regarding the circuit as a linear circuit.

This “certain operating point” is the “initial value of DC characteristic analysis” mentioned earlier.
It is. In other words, the nonlinear element is linearized so that the nodal equation to be solved is represented by a linear equation (this is called linear modeling of the nonlinear element). Then, a solution to this linear equation is obtained by a predetermined calculation. next,
The nonlinear elements are now linearized around this solution and new nodal equations are created. This series of operations is
It is repeated until the solution of the nodal equation converges on a certain value.

This focused value becomes the true value, that is, the analysis result of the DC characteristic analysis.

FIG. 5 is a processing flow diagram showing the processing performed by the simulation device shown in FIG. 4 when DC characteristic analysis is performed.

Referring to the figure, first set the data required for the simulation.
That is, the circuit netlist and simulation conditions are input (processing step 9).

Next, the initial value in the DC characteristic analysis is set to a predetermined value or a value given by the user (processing step 10). Thereafter, in processing step 1, DC characteristic analysis of the circuit manually performed is started.

First, the nonlinear elements included in the circuit are linearized around a certain estimated operating point set as an initial value for DC characteristic analysis in processing step S2 (processing step S2).
11). For example, when a nonlinear diode is included in the circuit, linearization of this diode will be described with reference to FIG.

FIG. 8 is a diagram for explaining linearization of a nonlinear diode, in which the horizontal axis represents the voltage v applied across the diode, and the vertical axis represents the current I flowing through the diode. It is assumed that the voltage/current characteristics of the diode have nonlinear characteristics as shown by curve (2) in FIG. 8, and that the current I is expressed by the following equation (in the following equation, C1 and C2 are constants).

(2) l-C1X (eCz-1) At this time, linear modeling of this diode around the estimated operating point A in the figure is performed as follows.

Therefore, at operating point A, ■=■.

■ Then, 1-C3x (V-Vo) 10I.

It is expressed as In other words, from the nonlinear characteristics of the diode shown by the curve ■, to the straight line ■
It is converted into a linear characteristic as shown in . This replaces the circuit to be analyzed with a linear circuit.

Referring again to FIG. 5, next, linear nodal equations are created for the linear circuit obtained as described above (processing step 512).

Next, a solution to the nodal equation created in processing step S4 is obtained by a predetermined calculation (processing step 5
13). Next, the solution obtained in processing step S13 is compared with the previous solution, and it is determined whether the solution has converged (processing step 514).

The determination result in processing step S14 is "N.

If so, that is, if the solution is not converged, the solution found in processing step 813 is set as a new estimated operating point (processing step 515). The process returns again to process step Sll, where the nonlinear element is linearized around the new estimated operating point, and process steps S12,
813. Then, the series of processing by 0 S14 is performed again. In this way, DC characteristic analysis is performed on the linear circuit obtained by linearizing the nonlinear elements included in the circuit around the estimated operating point one iteration before until the solution to the nodal equation converges. That is, by repeating the processing in steps 811 to S15, DC characteristic analysis using the Newton-Loughran method is realized.

If the determination result in processing step 81.4 is "yeS", that is, if the solution is converged, the DC characteristic analysis ends. The analysis result including this focused value is output as a simulation result to the simulation result receiving section 202 in FIG. 4. Furthermore, if the analysis to be performed is direct current characteristic analysis itself, the simulation result is output to the simulation result display section 4.

FIG. 6 is a processing flow diagram showing the processing performed by the simulation apparatus shown in FIG. 4 when AC small signal analysis is performed for a certain range of AC frequencies.

1. Referring to FIG. 1, first, as in the case of DC characteristic analysis, data necessary for simulation is entered manually (processing step 516). Next, processing step 81 in FIG.
DC characteristic analysis is performed by the same process as the series of processes in steps 1 to S15 (processing step 517). Next, in processing step 818, the characteristics of the nonlinear element are linearized with respect to the DC operating point of the circuit, which is the analysis result in processing step S17. )
. As a result, the circuit to be simulated is considered to be composed only of linear elements, that is, a linear circuit. Next, AC small signal analysis is performed on this linear circuit as follows.

First, the analysis start AC frequency fsRT inputted as a simulation condition is set to the AC frequency f (processing step 519). Next, the AC frequency is set to processing step S
Nodal equations of the circuit in the case of the initial value fSRT set in step 19 are created (processing step 520). Next, a solution to the equation created in processing step S20 is obtained by a predetermined calculation (processing step 521). In other words, the response of the circuit at the AC frequency fSRT is determined. Next, in processing step S22, it is determined whether the AC frequency f is the analysis end frequency manually input as a simulation condition.

If the determination result is "No", that is, if the analysis has not been performed for the entire AC frequency range given as the simulation condition, the value of the AC frequency f is changed to the value at that point in the minute frequency width Δf. is replaced with the sum of f+Δf (processing step 523). In other words,
The value of AC frequency f is set as the next frequency value to be analyzed. Next, the process returns to process step S20 again, and processes step 323 through process steps 320 to S22.
Analysis is performed on the new AC frequency set in .

By repeating the 3 series of processing in processing steps 320 to 823, the analysis for all AC frequency ranges given as simulation conditions is completed and when the determination result in processing step S22 becomes "yes", the analysis processing is performed. ends. This analysis result is output as a simulation result to the simulation result display section 4 in FIG. 4.

FIG. 7 is a processing flow diagram showing the processing performed by the simulation apparatus shown in FIG. 4 when a transient analysis is performed for the time domain from time 0 to a designated time. Referring to the figure, first, as in the case of the AC phase signal analysis described above, data necessary for simulation is input (processing step 524), and DC characteristic analysis is performed (processing step 525). Next, in order to replace time-dependent elements with a model expressed by a resistor and a power supply at each discrete time, we next set the initial conditions of the characteristic equations of these elements (current value and voltage value of the element when time t is O). ) is set to the value obtained by the DC characteristic analysis in processing step S5 (processing step 526). Next, the value of the discrete time t is replaced with a value larger than the value at that point by a minute time Δt (processing step 527).

Next, the element characteristic equation for which the initial conditions were set in processing step S26 is numerically integrated over the minute time 111Δt from the value of t at the previous time to the value of t at the present time (processing step 528). This converts time-dependent elements into models that are represented only by resistances and power supplies.

Next, DC characteristic analysis is performed using the Newton-Loughn method for the circuit in which time-dependent elements are converted into a model represented only by resistances and power supplies (processing steps 829 to 532).

Next, the determination result in processing step S32 is “yes”.
In other words, when the solution to the nodal equation in processing step S30 at a certain discrete time t is found, this solution is stored (processing step 533). Next, it is determined whether or not the value of time t is the final time of the range of times for which the transient analysis is to be performed, given as a simulation, that is, whether or not the transient analysis has been completed (processing step 533). ).

If the above determination result is "NO", that is, if solutions have not yet been found for all discrete times in the time range in which the analysis should be performed, processing steps 827 to S3
The series of processes in step 4 are performed again.

By repeating a series of processes in processing steps 827 to S34, solutions are found for all discrete times within the time range, and processing steps S34
When the determination result in is "yes", the analysis process ends. The solution for each discrete time is output to the simulation result display section 4 as the final simulation result.

[Problems to be Solved by the Invention] As described above, in a circuit simulation method using a conventional circuit simulator, for a certain electronic circuit,
This also applies when multiple different types of circuit analysis are performed, or when various circuit analyzes are performed by changing setting conditions such as input signals, in other words, when the same circuit is analyzed under specific simulation conditions. In each case, DC characteristic analysis is first performed.

At that time, the initial values of DC characteristic analysis are set in advance,
Alternatively, it is a fixed value manually set by the user prior to starting the simulation.

On the other hand, DC characteristic analysis is performed by finding a convergence value (true solution to the nodal equation) through repeated calculations using the Newton-Loughfun method. Therefore, depending on the value set as the initial value, the number of repeated calculations may become extremely large, and the time required to obtain a focused value, that is, the time required for DC characteristic analysis, may become extremely long, or a solution may not be obtained. A phenomenon occurs. Therefore, if the values set as initial values for DC characteristic analysis are values that cause such a phenomenon, the time required for the DC characteristic analysis itself and various analyzes that use the results of DC characteristic analysis as basic data will be reduced. may become too long, or analysis may not be possible because a solution cannot be obtained within 7 hours.

For problems where no solution can be obtained, the user must calculate an appropriate initial value (a value that will always yield a solution).
This can be avoided by setting the initial value of DC characteristic analysis in advance. However, it is not easy to estimate in advance what value should be given as an initial value to shorten the time required for DC characteristic analysis.

Figure 9 shows the software “5P” as a circuit simulator.
VAX11/750 (
This is a diagram showing in table format the time required for various circuit analyzes and the number of iterative calculations in direct current characteristic analysis using the Newton-Loughn method when analysis is performed using the conventional circuit simulation method using .

Referring to FIG. 9(a), the number of elements is 3, including nonlinear elements.
When performing DC characteristic analysis on the circuit No. 00, the number of repeated calculations in DC characteristic analysis is 3752, and the data processing time (CPU time) on the 8 CPUs (Central Processing Unit) of the analysis machine for that calculation is It was confirmed through experiments that the time was 2 hours and 39 minutes. Next, when AC phase signal analysis is performed on the same circuit as this circuit, the number of iterative calculations for DC characteristic analysis is the same as before, 3752 times, as shown in FIG. 9(b). It was also confirmed through experiments that the data processing time (CPU time) in the CPU of the analysis machine for the entire AC small signal analysis including this DC characteristic analysis was 211318151 minutes.

As described above, conventional circuit simulation methods are not practical because they require a huge amount of time to analyze even when analyzing circuits with the same netlist.

Therefore, an object of the present invention is to provide a circuit simulation method that solves the above-mentioned problems and reduces analysis time.

[Means for Solving the Problems] In order to achieve the above objects, the circuit simulation method according to the present invention includes the steps of performing DC characteristic analysis under certain initial conditions for an electric single circuit to be analyzed; The method includes a step of storing the analysis result of this DC characteristic analysis, and a step of further performing a DC characteristic analysis of this circuit under different conditions and using the stored analysis result as an initial value.

[Operation] Since the circuit simulation method according to the present invention includes the steps described above, when analyzing the same circuit under different conditions, for the second and subsequent DC analysis,
The analysis result of the first DC analysis is used as the initial value. Therefore, a value close to the true value is set as an initial value for the second and subsequent DC characteristic analyses, so that the time required for the second and subsequent DC characteristic analyzes is reduced.

[Embodiment] FIG. 1 is a schematic block diagram showing an example of a circuit simulation apparatus that implements the circuit simulation method according to the present invention.

Referring to the figure, this circuit simulation apparatus 0 includes a simulation result extraction section 3 in addition to a simulation human data creation section 1, an analysis section 2, and a simulation result display section 4, unlike the conventional circuit simulation apparatus.

The circuit simulation manual data creation section 1 includes a circuit diagram manual section 101 having the same functions as conventional ones, a netlist extraction section 102, a simulation condition setting section 103, and a DC characteristic analysis initial value storage section 104. Upon receiving a simulation request from the simulation condition setting section 103, it is determined whether the simulation condition setting in the simulation condition setting section 103 is for the second or subsequent time for the same circuit, and according to the determination result, the Input data creation control switch unit 303
and a determination unit 105 that controls the operation of the controller.

The internal configuration of the analysis section 2 is similar to the conventional one shown in FIG. However, the initial value for DC characteristic analysis given to the DC characteristic analysis section 201 is output from the simulation result extraction section 3, unlike the conventional case.

The simulation result extraction unit 3 includes a DC characteristic analysis result extraction unit 301 that extracts the DC characteristic analysis result received by the simulation result receiving unit 202, and a DC characteristic analysis result included in the DC characteristic analysis result extracted by the DC characteristic analysis result extraction unit 301. A set voltage value storage unit 302 of each node that extracts and stores the DC operating point, which is set as a set voltage value of each node in the initial state of the circuit to be analyzed; It includes an input data creation control switch section 303 that switches between the set voltage value of each node and the initial value stored in the initial value storage section 104 for DC characteristic analysis and provides the same to the DC characteristic analysis section 201 .

This circuit simulation device operates under the control of an improved circuit simulator.

FIG. 2 is a process flow diagram showing the process performed by the circuit simulation apparatus shown in FIG. 1 according to the improved circuit simulator.

Referring to FIG. 2, data necessary for the simulation is first taken in by the simulation data creation section 1 as in the conventional case (processing step S1). Next, it is determined whether the simulation condition setting in processing step S1 has been performed for the first analysis of the circuit to be analyzed (processing step S2).

Generally, multiple types of analyzes are performed successively on one circuit. In other words, the simulation condition setting unit 303 issues multiple simulation requests for one circuit. Therefore, specifically, in processing step S2, it is determined whether the simulation request is the first for the circuit.

If the result of the determination is “yes”, that is, if the circuit to be analyzed is different from the previous circuit to be analyzed, the determination unit 10
5 causes the control switch section 103 to output the value stored in the initial value storage section 104 for DC characteristic analysis. Therefore, in this case, a predetermined value stored in the initial value storage section 104 or a constant value given by the user prior to the start of simulation 3 is used as the initial value for DC characteristic analysis, as in the past. Once the settings have been made, a DC analysis is performed using the Newton-Loughn method (processing step S4). The specific processing contents in processing step S4 are the same as the series of processing in processing steps Sll to S15 in FIG.

Next, the DC operating point of the circuit obtained as a result of the DC characteristic analysis in processing step S4 is stored as the set voltage value of each node in the set voltage value storage unit 302 of each node (processing step S5).

Thereafter, a circuit analysis (in this case, AC phase signal analysis or transient analysis) according to the type of analysis to be performed, given as data in processing step S1, is performed (processing step S8), and the processing ends. The analysis results obtained by the analysis in processing step S8 are displayed on the simulation result display section 4 as the final simulation results.
is output to. Note that if the analysis to be performed is DC characteristic analysis itself, processing step S5 becomes the fourth and final processing step, and the DC characteristic analysis result in processing step S4 is output to the simulation result display section 4 as the final simulation result. be done.

In this way, if the circuit to be analyzed is different from the previous one, analysis is performed using the same processing as in the past. However, unlike the conventional method, the set voltage value of each node obtained by DC characteristic analysis is stored in a predetermined storage section 302.

The determination result in processing step S2 is "N.

In other words, if the circuit to be analyzed is the same as the previous circuit to be analyzed, the determination unit 105 determines that the control switch unit 303
Then, the value stored in the set voltage value storage section 302 of each node is output. Therefore, in this case, unlike in the past, the values stored in the set voltage value storage unit 302 of each node are set as the initial values for the DC characteristic analysis (processing step S6), and the DC analysis using the Newton-Loughfun method is performed. is carried out (processing step S7). The processing contents in processing step S7 are also the same as the series of processing in processing steps 811 to S15 in FIG. 5, as well as in processing step S4. The values stored in the set voltage value storage section 302 of each node are the solutions obtained by previously performing DC characteristic analysis on the same circuit as the current circuit to be analyzed. Therefore, if the current simulation conditions are the same as those in the previous DC characteristics Ml analysis, the set voltage value of each node, which is the solution of the previous DC characteristics Wl analysis, is set as the initial value for the current DC characteristics analysis. This means that this initial value becomes the solution for the current DC characteristic analysis and is used in processing step S.
This means that the number of iterative calculations in the Newton-Loughn method performed in 7 is only one. Current r1:
, even if the simulation conditions of The solution of the analysis is considered to be close to the solution of the DC characteristics analysis described above. Therefore, in such a case, using the value stored in the set voltage value storage unit 302 of each node as the initial value for the current DC characteristic analysis will reduce the iterative calculation for the current DC characteristic analysis. It means to start from a value close to the true value.

Therefore, the number of iterative calculations performed by the Newton-Loughn method in processing step S7 is reduced compared to the conventional method. In any case, the number of iterative calculations in processing step S7 is smaller than in the past. That is, in processing step S7, DC characteristic analysis is performed in a shorter time than conventionally.

Next, using the analysis results obtained by the DC characteristic analysis in processing step S7 as basic data, a circuit analysis is performed according to the type of analysis to be performed, which was determined as data in processing step S1 (processing step S8), all processing ends. The analysis result obtained by the processing in processing step S8 is output to the simulation result display section 4 as the final simulation result. Note that if the specified analysis is DC characteristic analysis itself, processing step S7 becomes the final processing step, and the analysis result in processing step S7 is output to the simulation result display section 4 as the final simulation result. Ru.

The specific processing content in processing step S8 is the series of processing in processing steps 818 to S23 in FIG. 6 (in the case of AC small signal analysis) or processing step 82 in FIG.
This is the same as the series of processes in steps 6 to S34 (in the case of transient angle q analysis).

Figure 3 shows the circuit simulation device of this embodiment, using software obtained by partially modifying the manual control part of the software "5PICE" as a circuit simulator, and using a VAXll/750 as an analysis machine. FIG. 7 is a diagram showing in table format the time required for various circuit analyzes and the number of iterative calculations for DC characteristic analysis using the Newton-Loughfun method when various analyzes are performed.

We performed DC characteristic analysis on the same circuit used in the experiment using the conventional circuit simulation method, which yielded the data shown in Figure 9, and then changed some of the specification values in the circuit. If the DC characteristic analysis is performed again after the DC characteristic analysis has been completed, the number of iterations for the subsequent DC characteristic analysis will be 11, and the CPU time will be only 2 minutes, as shown in Figure 3 (a). Confirmed by experiment. Furthermore, if the same circuit is analyzed using the DC characteristic analysis itself or the analysis results obtained therefrom as basic data, and then an AC small signal analysis is performed, Figure 3 (
As shown in b), it was experimentally confirmed that the number of iterative calculations for DC characteristic analysis in this AC small signal analysis was 11 times, and the CPU time was only 4 minutes.

As can be seen from the above, in this embodiment, when multiple analyzes are performed on the same circuit, the solution of the first DC characteristic analysis is used as the initial value for the DC characteristic analysis in the second and subsequent analyses. Therefore, the time required for the second and subsequent DC characteristic analyzes is significantly reduced compared to the conventional method.

In this example, when multiple analyzes with different simulation conditions are performed on the same or equivalent circuit in succession, the DC characteristic solution obtained in the first analysis in the series of analyzes will be used in the second and subsequent analyses. It was used for this purpose. but,
If the circuit to be analyzed this time is the same as any of the circuits analyzed previously, use the results of the previous DC characteristic analysis of that same circuit as the initial value and perform the desired analysis of the circuit to be analyzed this time. Good too. In such a case, for example, the set voltage value storage section 30 of each node
2, stores a plurality of DC characteristic analysis results of the circuits analyzed so far, stores the netlist extracted by the netlist extraction section 102, and reads out the netlist of the circuit to be analyzed up to the previous time from the storage section. , it is only necessary to determine whether the circuit to be analyzed this time is the same circuit as the circuit that was the target of analysis up to the previous time, from the netlist manually created this time. Then, based on this determination result, the DC characteristic analysis result of the same circuit as the current circuit to be analyzed may be used to determine the initial value of the current analysis.

In addition, in the data created by the human data creation unit 1, if a part of the circuit to be analyzed is treated as a model that operates as one functional element, the internal information of this modeled circuit part is The set voltage value of each node is also stored in the storage unit 302.

In this way, according to this embodiment, circuit analysis with the same accuracy as before can be performed much more accurately than before by simply changing the program for the human control part without changing the algorithm of the existing circuit simulator at all. It can be done in a short time.

[Effects of the Invention] As described above, according to the present invention, when multiple analyzes with different simulation conditions are performed on the same circuit, the analysis accuracy is maintained at the same level as before, and the analysis accuracy is improved from the second time onwards. The time required is reduced to about 1 compared to the conventional method. As a result, L
It becomes possible to design SI etc. extremely efficiently.

[Brief explanation of drawings]

FIG. 1 is a schematic block diagram of a circuit simulation device operated by an improved circuit simulator, showing one embodiment of the present invention, and FIG. 2 is a schematic block diagram of a circuit simulation device shown in FIG. Figure 4 is a schematic block diagram of a circuit simulation device operated by a conventional circuit simulator, and Figure 5 is a diagram showing the required time etc. in a table format. 6 is a process flow diagram showing the processing performed by the circuit simulation device shown in FIG. 6. FIG. 6 is a process flow diagram showing the processing performed by the circuit simulation device shown in FIG. 7 is a processing flow diagram showing a series of processes performed by the circuit simulation apparatus shown in FIG. 4 when transient analysis is performed by the conventional circuit simulation method, and FIG. 8 is a process flow diagram showing the conventional circuit simulation method. FIG. 9 is a diagram illustrating the time required for various circuit analyzes according to the method in a table format, and FIG. 9 is a diagram for explaining linear modeling of a nonlinear element. In the figure, 1 is a simulation human data creation section;
2 is an analysis section, 3 is a simulation result 2 extraction section, 4 is a simulation result display section, 105 is a discrimination section, 301 is a DC characteristic analysis result extraction section, 302 is a set voltage value storage section for each node, 303 is an input data creation section The control switch section 51-534 is a processing step. In each figure, the same numbers indicate the same or equivalent parts.

Claims (1)

[Claims of Claims] A circuit simulation method for analyzing the characteristics of an electronic circuit, comprising: performing DC characteristic analysis of the circuit to be analyzed under certain initial conditions; and storing the analysis results of the DC characteristic analysis. A circuit simulation method, further comprising performing DC characteristic analysis on the electronic circuit under different conditions and using the stored analysis results as initial values.