JP5219078B2 - Simulation method by numerical calculation of analog circuit and simulation method by numerical calculation of switching power supply circuit - Google Patents

Description

  The present invention relates to a simulation method by numerical calculation, and more particularly to a method of simulating frequency characteristics by numerical calculation in an analog circuit and a switching power supply circuit represented by a DC-DC converter in which an analog circuit and a digital circuit are mixed.

  As an analog electronic circuit simulation method, a circuit equation is solved using a Gaussian elimination method and an LU decomposition method, and a method using a Newton-Rahpson method is known for nonlinear circuit analysis. In order to simulate the dynamic characteristics of a circuit, the elements in the circuit are replaced with a current source and a voltage source that change every time step, and nonlinear circuit calculation is repeated in units of this time step. Also, a method called event-driven method is known as a digital circuit simulation method (Patent Document 1).

  In the above-mentioned Patent Document 1, in the analog / digital mixed circuit, the simulation of the analog part is usually made with the aim of high accuracy, and it is necessary to control the time interval, so that the calculation time becomes long. As a method for solving this problem, the transfer function of the analog circuit portion described by the S function is converted into a Z function having a sampling period having a period corresponding to the simulation accuracy, and the digital circuit network is converted based on the converted Z function. Forming and performing logic simulation is disclosed.

Further, a circuit simulation program called SPICE (registered trademark) is known as a program for simulating an analog circuit, and a program called MATLAB (registered trademark) is known as a numerical operation program.
Patent No. 3019368 (Prior Art, Problems to be Solved by the Invention, Means for Solving Problems)

  In the above-described simulation method of an analog circuit or an analog / digital mixed circuit, in order to perform a logic simulation of the operation of the analog circuit unit, the analog circuit is expressed by a transfer function by an S function, and the S function is converted into a Z function by digital. A technique for forming a circuit is used.

  However, in circuits that handle pulse waveforms, signal waveforms with sudden voltage changes or current changes, it may be difficult to express the circuit configuration with a transfer function, and it will change abruptly to perform transient analysis. There is a problem that the simulation time increases because it is necessary to set the analysis step time finely in order to finely analyze the portion. Further, in a circuit configuration in which a plurality of signal sources are connected, it may be difficult to express with a transfer function.

  Especially when analog circuits need to be simulated by numerical operations to analyze digital circuits, such as analog-digital mixed circuits, the operation analysis of analog circuits should be suitable for simulation by numerical operations. Desired. At this time, in the analog circuit part, it is required to reduce the time required to simulate the pulse waveform, the signal waveform having abrupt voltage change or current change, and the analog circuit configuration is a model suitable for numerical simulation It is required that

Hereinafter, this problem will be described using a DC-DC converter as an example. Generally, DC-DC converter converts the voltage pulse waveform by switching an input DC voltage V _in, different desired DC output voltage and the input voltage by smoothing the voltage pulse waveform by an inductor L and a capacitor C V _out is obtained.

のようになり、昇圧形（Ｖ_in＜Ｖ_out）の場合には

のようになる。 When the ratio t _B / T between the pulse time width t _B of the voltage pulse waveform and the repetition period T of the pulse waveform is represented by a duty ratio D, the DC output voltage V _out of the DC-DC converter is a step-down type (V _in > V _out )

In case of boost type (V _in <V _out )

become that way.

In operation of the formula (1) or formula (2), serves for feeding the input DC voltage V _in to the smoothing circuit as a switch to a voltage pulse waveform, the output transistor constituting the switching circuit is responsible. Even in the case where a large current value has to be supplied, since the voltage across the output transistor with the switch turned on is small, power consumption in the output transistor is small. For this reason, DC-DC converters are frequently used as converters with high power conversion efficiency.

However, the current supply capability required for the DC-DC converter always varies depending on the load condition. When the load is in the standby mode, the supplied current value is small, and when the load is active, the current value is large. When the current value varies, the DC output voltage _Vout also varies. Therefore, it is necessary to stabilize _Vout by applying negative feedback.

FIG. 12 is a block diagram of an LSI step-down DC-DC converter (V _in > V _out ). DC-DC converter shown in FIG. 12, the switching circuit 60 for switching an input DC voltage V _in, the output smoothing circuit 20 to form a DC output voltage V _out and smoothing the output of the switching circuit 60, the DC output voltage V _out an error amplifier 40, a current detector 50 for forming the current i _L flowing through the inductor 21 of the output smoothing circuit 20 that amplifies a voltage difference between a constant and (K) multiplied by voltage KV _out and the reference voltage V _ref the (delta), error The comparator 30 that compares the output voltage v _c of the amplifier unit 40 and the voltage value v _{cfb obtained} by adding a slope compensation current to the current detected by the current detector 50 and converting the voltage is fed back and the output v _f of the comparator 30 is fed back. The digital control circuit 10 controls the switching operation of the switching circuit 60. The digital control circuit 10 is driven by a periodic signal from an oscillator (OSC) 11.

The output voltage V _out constant (K) multiplied by the voltage difference between the voltage KV _out and the reference voltage V _ref E a (delta) was amplified by the error amplifier 40, and sends its output voltage v _c to the comparator 30. When the output voltage V _out constant (K) multiplied by voltage KV _out is greater than the voltage of the reference voltage V _ref, the output voltage v _c of the error amplifier 40 becomes small.

On the other hand, the current i _L flowing through the inductor 21 (L) of the output smoothing circuit 20 matches the current flowing through the output transistor Mp or the output transistor Mn of the switching circuit 60. Since the change of the current value of the current i _L cannot be sufficiently detected by the error amplifier section 40, the change of the current value of the current i _L is detected by the current detector 50, further subjected to the slope compensation 51, and input to the comparator 30. .

As described above, the voltage information (V _out ) of the DC output voltage and the current information (i _L ) of the inductor are fed back to the digital control circuit 10 through the comparator 30, and the digital control circuit 10 outputs the output of the switching circuit 60. The on period of the transistor Mp or Mn is controlled. As a result, even if the DC output voltage (V _out ) changes or the output current (i _L ) changes and the DC output voltage (V _out ) changes, the DC output voltage (V _out ) is always kept constant. Be drunk.

  In the circuit of FIG. 12, the analog circuit of the error amplifier section 40 mainly handles waveforms with a gradual change, but other analog circuits such as the output smoothing circuit 20 and the current detector 50 have a pulse waveform or a waveform with abrupt changes. Is dealing.

  Here, if a signal waveform with a sudden voltage change or current change is transiently analyzed using a circuit simulation tool of an analog circuit, the circuit simulation tool analyzes the detailed and precise analysis in a part where there is a sudden change. Set the step time very finely. For this reason, the simulation time required for the transient analysis increases rapidly.

  Furthermore, since the analog circuit simulation tool is a transistor-based simulation tool, the analysis is performed in a form including all elements included in each circuit configuration of FIG.

  As described above, in order to analyze the pulse-like operation, it is necessary to set the time step of the analysis very finely, and it is necessary to perform analysis based on each element constituting the circuit, so the circuit scale is large. Due to the two factors of increasing, there is a problem that when the transient analysis of the circuit as shown in FIG.

In FIG. 12, a pulse-like signal waveform is handled and many switches and digital circuit functions are included. Therefore, AC analysis using a transfer function based on a small signal model used in an analog simulator cannot be performed. When obtaining the open loop characteristics, a low resistance of about 50Ω is connected between the DC output voltage V _out and the negative input terminal of the error amplifier section 40, and the negative input terminal side of the error amplifier section 40 is connected to both ends of the low resistance. A transient analysis is performed by giving a signal such that At this time, since the input signal component appears superimposed on the DC output voltage _Vout , the input signal component that has passed through the open loop included in the transient analysis result is extracted.

  However, the above operation can only provide an open loop response to a signal of one frequency. In order to obtain the frequency characteristics of the open loop, it is necessary to input signals of various frequencies to the open loop and collect response data for each frequency. The analysis time required for one transient analysis is long, and further, when obtaining the frequency characteristic, it is necessary to repeat this operation, and there is a problem that it cannot be obtained unless a large amount of time is spent.

  SUMMARY OF THE INVENTION Accordingly, it is an object of the present invention to solve the above-described problems and to simulate a transient analysis of a pulse signal or a signal waveform accompanied by a sudden waveform change by numerical calculation in an analog circuit or a circuit in which an analog circuit and a digital circuit are mixed. The purpose is to shorten the analysis time.

  The first aspect of the present invention simulates the operation of an analog circuit or a circuit in which an analog circuit and a digital circuit are mixed by numerical calculation, and simulates the operation of a switching power supply circuit in which an analog circuit and a digital circuit are mixed by numerical calculation. And a second aspect.

  A first aspect of the present invention is a method for simulating an operation of an analog circuit or a circuit in which an analog circuit and a digital circuit are mixed by numerical operation. The analog circuit is divided into a plurality of circuit blocks, and the divided circuit blocks In FIG. 2, a circuit block that performs numerical operation is modeled in two forms, and the time change of the output of the circuit block is obtained by solving the modeled circuit block by numerical operation. This makes it possible to perform simulation by numerical computation of transient analysis of pulse waveforms and signal waveforms with sudden waveform changes in analog circuits or circuits in which analog circuits and digital circuits coexist, and shorten analysis time. .

  A first form of modeling of a circuit block for performing a numerical operation is applied to a circuit block that requires a transient analysis for analyzing a temporal change in output. In this form of modeling, the voltage and current in the circuit block are used as variables, and this block is described using a state equation including a differential variable obtained by differentiating this variable to form a calculation block for numerical calculation. The circuit block is modeled by

  In the modeled circuit block, the time change of the output is obtained by solving the differential variable by numerical calculation.

  Since the state equation is described including a differential variable, the time change can be obtained by numerical calculation. Moreover, since it is a numerical calculation, it is possible to simulate a transient analysis of a pulse signal or a signal waveform accompanied by a sudden waveform change by numerical calculation without repeating a huge analysis step.

  The second form of circuit block modeling for performing numerical operations is applied to a circuit block that does not require the transient analysis. This form of modeling applies the theory of superposition by electrical circuits, and communicates when a voltage source or current source exists alone for a voltage source and / or current source that is input to an analog active element of a circuit block. The function is described from the voltage source or current source to the analog active element by the S function, and the sum of the product of the transfer function of this circuit block and the voltage of the voltage source or the current of the current source described by the S function is obtained. The sum of products is described as an input signal of an analog active element by the principle of superposition. In addition, the transfer function of the analog active element is described by an S function, and the overall transfer function of the circuit block is described by the product of the transfer function of the input signal and the transfer function of the analog active element, and an arithmetic block for numerical calculation The circuit block is modeled by this arithmetic block, and the Laplace inverse transform is performed on the total transfer function to obtain the time change of the circuit block output.

  The second aspect of the present invention is a method of simulating the operation of a switching power supply circuit in which an analog circuit and a digital circuit are mixed by numerical calculation.

  The switching power supply circuit to which the simulation method of the present invention is applied includes a switching circuit that switches an input DC voltage and outputs a voltage pulse, an output smoothing circuit that smoothes the voltage pulse and supplies it to a load, and an output voltage of the output smoothing circuit A difference voltage between a constant multiple of the reference voltage and the reference voltage is used as an input signal, the input signal is amplified, and a low-pass voltage is output, and a limiter current is fed back from the output terminal to the input terminal, and an output A current detection circuit that detects a current flowing through the smoothing circuit, converts the current into a voltage and feeds it back to the comparator, and a digital control circuit that controls the switching period of the switching circuit.

  In the above switching power supply circuit, an output smoothing circuit, an error amplifier, a current detection circuit, and a digital control circuit are modeled, and a time change of the signal is obtained by numerical calculation.

  The output smoothing circuit uses the output voltage and load current of the output transistor constituting the switching circuit as input signals, the output voltage of the output smoothing circuit and the current flowing through the inductance element included in the output smoothing circuit as output signals, and the output smoothing circuit. Form a calculation block that performs numerical calculations by describing the output smoothing circuit using a state equation including a differential variable obtained by differentiating the voltage across the capacitor element and the current flowing through the inductance element. The switching circuit and the output smoothing circuit are modeled using the calculated operation block, and the time change of the current flowing through the inductance element and the output DC voltage of the output smoothing circuit is obtained in this modeling.

  According to this modeling, integrating a differentiated variable in a state equation corresponds to finding the time change of the variable. Transient analysis of pulse waveforms and signal waveforms with rapid waveform changes can be simulated by numerical calculation without repeating.

  The error amplifier includes a transfer function from an input signal voltage source to the input terminal of the error amplifier, a transfer function from an output signal voltage source to the input terminal of the error amplifier, and a reference voltage source to the input terminal of the error amplifier. A transfer function when a transfer function and a transfer function from a limiter current source representing a current flowing in a diode limiter connected between the input terminal and the output terminal to the input terminal of the error amplifier exist independently. Is described by the S function, and the sum of the product of each transfer function and the input signal voltage source, output signal voltage source, reference voltage source, limiter current source voltage or current source current described by the S function is obtained. Is described as the input signal of the error amplifier by the superposition principle.

  Further, an error amplifier transfer function is described by an S function, and an arithmetic block for performing a numerical operation is described by describing the overall transfer function of the error amplifier by the product of the transfer function of the input signal and the transfer function of the error amplifier. Then, the error amplifier is modeled, and in this modeling, the Laplace inverse transform is performed on the total transfer function to determine the time change of the output of the error amplifier.

  Here, since the signal waveform of the error amplifier is not accompanied by a sudden waveform change, it can be modeled by a transfer function. In the present invention, modeling is performed by combining transfer functions of a plurality of external elements. More specifically, the superposition theory is used to obtain the sum of the products of the transfer function and the transfer function of the input signal when each exists alone.

  In the current detection circuit, a detection transistor that flows a predetermined ratio of current to the current flowing through the switching transistor of the switching circuit is connected in parallel to the switching transistor, and is described by the product of the current flowing through the detection transistor and the resistance. An operation block for performing an operation is formed, and the current detection circuit is modeled using the formed operation block.

  In this modeling, a current is detected by performing a numerical operation using a product of a predetermined ratio and a resistance as a constant current-voltage conversion coefficient.

  By modeling each analog circuit described above, the operation of each analog circuit of the switching circuit, the output smoothing circuit, the error amplifier, and the current detection circuit can be simulated by numerical calculation.

  In the switching power supply circuit, the control operation of the digital control circuit that controls the switching period of the switching circuit that switches the input DC voltage and outputs the voltage pulse is simulated by numerical calculation. The difference from the output voltage of the error amplifier is compared with zero to output a logical value, and the control operation of the digital control circuit that controls the switching period by the logical value of the comparison result is described by a Boolean algebra. An arithmetic block to be performed is formed, and the digital control circuit is modeled using the arithmetic block. In this modeling, the switching period is controlled by Boolean algebra.

  According to the present invention, in a circuit in which an analog circuit or an analog circuit and a digital circuit are mixed, a transient analysis of a pulse signal or a signal waveform accompanied by a sudden waveform change can be simulated by numerical calculation, and the analysis time can be shortened. Can do. As a result, the frequency characteristics can be obtained in a short time.

  Hereinafter, embodiments of the present invention will be described in detail based on examples. FIG. 1 is a diagram for explaining a block configuration formed by modeling a switching power supply circuit in the present invention.

In FIG. 1, the switching power supply circuit includes the digital control circuit 10, the switching circuit 60, the output smoothing circuit 20, the error amplifier unit 40, the current detector 50, and the comparator 30 of the DC-DC converter shown in FIG. Consists of modeled circuit blocks. This circuit block includes a digital control circuit block 1, a switching circuit block 6, an output smoothing circuit block 2, an error amplifier block 4, and a current detection circuit block 5. In FIG. 1, the digital control circuit block 1 includes a comparison block 3 that obtains a differential voltage v _f by subtracting the output voltage v _cfb of the current detection circuit block 5 and the output voltage v _c of the error amplifier block 4. Yes.

  Hereinafter, the blocking of each part of the switching circuit block 6, the output smoothing circuit block 2, the error amplifier block 4, the digital control circuit block 1, and the current detection circuit block 5 will be described.

  First, the switching circuit block 6 and the output smoothing circuit block 2 will be described. FIG. 2 is a diagram showing an output transistor (Mp, Mn), an output smoothing circuit 20, and a load portion of the switching circuit 60 of the switching power supply circuit to which the present invention is applied.

In the output smoothing circuit 20 shown in FIG. 12, the inductor L actually has a resistance component R _L , and the capacitor C also has a resistance component R _c . In practice, a circuit that divides the DC output voltage _Vout with a resistor is included in the input of the error amplifier section 40, so that _Rf is connected as this resistor. Here, the load is indicated by a current source _Iout .

がなりたつため、

という状態方程式が得られる。 In Figure 2, applying Kirchhoff's law and formulating it,

Because

The state equation is obtained.

の形で表すことができ、変数ｘの時間変化つまりｖ_C（ｔ），ｉ_L（ｔ）の時間変化を計算し、Ｖ_out（ｔ），ｉ_L（ｔ）の時間変化を求めている。ただしｘ'は時間ｔによるｘの微分を示している。ここで、ｘ、ｙ、ｕは以下で表される。 Equation (4) is

The time change of the variable x, that is, the time change of v _C (t), i _L (t) is calculated to obtain the time change of V _out (t), i _L (t). . However, x 'has shown the differentiation of x by time t. Here, x, y, and u are expressed as follows.

Therefore, obtaining the variable y corresponds to a transient analysis of i _L (t) and V _out (t). A, B, C, and D corresponding to the coefficients of equation (5) are constants that can be easily calculated from the circuit constants in equation (4), and the time-varying waveforms of v _c (t) and i _L (t) Is also considered as an ideal change waveform, so speeding up of numerical calculations is expected.

  On the other hand, when analog circuit simulation software (for example, a program called SPICE (registered trademark)) is used, it is necessary to finely set the time step width in the simulation when the change is abrupt, which increases the processing time. .

  FIG. 3 shows an input example of a program called MATLAB (registered trademark) for numerically calculating the equation (4).

Usually, such symbol notation is used for program input. In order to calculate the state variable equation in the form of equation (5), it is necessary to specify the external input U and the output y. In this case, V _LX (t) and I _out (t) are designated as the external input U, and V _out (t) and i _L (t) are designated as the output y. The state variable x and the coefficients A, B, C, and D are separately designated by type input.

Next, the error amplifier block 4 will be described. 4 is a diagram showing an error amplifier unit 40 of a switching power supply circuit to which the present invention is applied, and FIG. 5 is a diagram showing a circuit block in which the error amplifier unit 40 is modeled. Here, V _out is the output DC voltage, v _c is the output of the error amplifier 46, v _i is the input of the error amplifier 46. Since the error amplifier circuit of FIG. 4 is an analog circuit, it is appropriate to express it by a transfer function in modeling.

Here, when the transfer function of the error amplifier section 40 is expressed by v _c / V _out , it is necessary to describe the frequency characteristic of the error amplifier 46 by including it in the transfer function. By defining the error amplifier 46 independently and describing an LPF (low pass filter) connected to the output of the error amplifier 46, the degree of freedom of simulation can be increased. Therefore, here, the input v _i of the error amplifier 46 is used to superimpose, and thereafter, the voltage amplification part of the error amplifier (represented by the voltage amplification degree (−k1) in FIG. 5) and the low-pass filter part. (LPF) is represented by a connected configuration.

という形で記述される。 The input v _i applies the superposition theory to the DC output voltage V _out , the output v _c of the error amplifier 46, the reference voltage V _ref of the reference voltage source 45, and the current I _lim flowing through the limiter 47,

It is described in the form.

In Expression (7), the transfer function H _out (s) is a voltage v _c (s) other than the DC output voltage V _out (s), a reference voltage V _ref (s), a current I _lim (s) flowing through the limiter, and the like. This is a transfer function showing the influence of only the DC output voltage V _out (s) on the input v _i (s), with the values of the voltage source and the current source of zero being zero. Here, each transfer function, voltage, and current are described by an S function. Similarly to the transfer function H _out (s), the transfer function H _c (s) from the output section of the error amplifier 46 and the transfer function H _lim (s) of the limiter 47 are set to zero when the corresponding voltage source or current source is zero. Seeking only by input. The transfer function H _out (s), the transfer function H _c (s), and the transfer function H _lim (s) thus obtained are each expressed by the following equations.

However, a _{Z CR = R eff + 1 /} sC eff. The M _lim 47 is a limiter for such voltage v _c of the output terminal of the error amplifier 46 does not become large. A resistor 41 (R _f1 ) and a resistor 42 (R _f2 ) are resistors that divide the DC output voltage V _out , and a resistor 48 (R _f3 ) is an input resistor connected to the input terminal of the error amplifier 46. The source 45 (V _ref ) is a reference voltage source. The capacitor 43 (C _err ) and the resistor 44 (R _err ) constitute a feedback path for the error amplifier 46.

の電流が流れる。ただしβ_limはリミタ４７のトランジスタの定数（ゲート係数と呼ばれる）である。 Here, when the limiter 47 operates,

Current flows. Here, β _lim is a constant of the transistor of the limiter 47 (called a gate coefficient).

  FIG. 5 is a diagram in which the error amplifier section 40 is modeled on MATLAB (registered trademark) using the equation (7). The error amplifier 46 can be modeled by a subordinate connection of a voltage amplification part (voltage amplification degree -K1) and a low-pass filter part (LPF). The other part is an example configured according to the equation (7).

Next, the digital control circuit block 1 will be described. FIG. 6 is a diagram showing a circuit block obtained by blocking the digital control circuit 10 of the switching power supply circuit to which the present invention is applied. FIG. 6 represents the digital control circuit 10 in FIG. 12 based on Boolean algebra. The digital control circuit block 1 includes NOT (1b), AND (1c), x (product) (1e), S-R flip-flop (1d), and a comparison result> = 0 function (1a). The output is indicated by V _LX . These are basic circuits of a digital circuit even in an actual circuit.

  Since the OSC (oscillator) only has a trigger function, not only OSC but also an arbitrary repetitive pulse waveform can be used.

Further, the function of the comparator 30 in FIG. 12 can be modeled by subtraction. Note that the structure of modeling by subtraction is not shown in FIG. The comparison operation 1a in FIG. 6 is a component that compares the output v _f of the function of the comparator 30 with zero. FIG. 7 shows signals of the respective parts shown in FIG. The comparison result shown in FIG. 7C is a comparison result between the output v _f of the comparator 30 and zero, and the period during which the output Q of the RS flip-flop is turned on is controlled according to this logical value. a period in which the voltage V _in is output is controlled, the output voltage V _LX of the switching transistor of the switching circuit is controlled.

According to this configuration, the output voltage V _LX can be obtained only by a simple logical operation. When calculating using the SPICE (registered trademark) circuit simulation program used for analog circuit simulation, the analysis is performed along the time axis of the clock waveform. Therefore, every time the waveform changes sharply, the time is divided finely. It is necessary to make calculations, which results in enormous calculation time.

On the other hand, according to this configuration, an ideal waveform can be considered and processed unlike an actual waveform, and V _LX is required to control output transistors on and off with a simple function. Therefore, there is an effect that an error is hardly affected.

  Next, the current detection circuit block 5 will be described. FIG. 8 is a diagram showing a current detection circuit 50 of a switching power supply circuit to which the present invention is applied. FIG. 8 shows a circuit that copies the current flowing through the inductor 21 of FIG. 12 and converts it as a control voltage of the current feedback loop.

In FIG. 8, the PMOS transistor M _pc (513) is the same as the output transistor Mp except that the element size is different. Since the terminal voltage of the c _f1 and c _f2 equals amplifier A1 and (514) by the action of the _M pcfb (515), the ratio of the current element size flowing to M _p in a state in which the current i _L flows M _pc ( 513). The current flowing through M _pc (513) flows through the resistor R _cfb (516) and is converted into the control voltage v _cfb .

で表される。 In FIG. 8, the amplifier is treated as an element having a voltage amplification degree A1, and frequency characteristics are not considered. However, since the transistor M _pc operates in the linear region, its output conductance g _dsmpc affects the conversion ratio. Also affects the conversion ratio transconductance g _Mcfb transistor _M pcfb. These can be said to be a possible effect from the circuit configuration. Given their impact,

It is represented by

Actually, the influence of the switches SW1 (511) and SW2 (512) is taken into consideration. When the output transistor M _p is on, the switch SW1 (511) is on and the SW2 (512) is off (1 in FIG. 9). In the other cases, SW1 is turned off and SW2 is turned on (in the case of 2 in FIG. 9), and two states are conceivable, so the model diagram is as shown in FIG.

Figure 9 shows a circuit block that models the current detection circuit, the current detection circuit block 5, the transconductance g _Dsmp transistor M _p (5a), the transfer switch SW conductance _g dsssw (5b), the resistor R _cfb ( 5d, 5e), magnification (5c), etc.

_Here , g _dssw represents the conductance of the switches SW1 and SW2, i _ref represents a constant current value, and _x10 represents a function of multiplying by 10. Further, if K is obtained in advance, v _cfb can be easily calculated for the cases 1 and 2 in FIG.

  Next, an example of the simulation is shown using FIGS. 10 and 11, and the effect of the present invention is shown.

FIG. 10 shows a circuit configured by the normal SPICE (registered trademark) circuit simulation program (hereinafter referred to as a SPICE (registered trademark) circuit) and the numerical calculation program according to the present invention. When the power supply of the circuit in FIG. 12 is raised from 0 V to 3 V using two types of programs of a circuit (hereinafter referred to as a MATLAB circuit) that is configured using matlab (MATLAB (registered trademark)) The result of having simulated the change of DC output voltage _Vout of this is shown.

In FIG. 10, V _in is raised at time t = 0 to 20 μS, the time interval of transient analysis is set to 1 nS, and the change in the DC output voltage V _out is observed up to 100 μS. The simulation conditions are that the OSC, the reference voltage, and the reference current are in an ideal state, the injection signal is not applied, V _in = 3 V, V _out = 2 V, V _ref = 0.5 V, and the output current is 100 mA. .

  FIG. 10 shows that the result of the SPICE (registered trademark) circuit and the result of the MATLAB (registered trademark) circuit are substantially the same.

  In the transient analysis, the SPICE (registered trademark) circuit required about 3,600 seconds of CPU time, but the MATLAB (registered trademark) circuit took about 25 seconds. This means that the method of the present invention can achieve a speed increase of 100 times or more.

FIG. 11 is a diagram showing frequency characteristics of the circuit of FIG. In FIG. 12, a low resistance of about 50Ω is connected between the DC output voltage V _out and the negative (−) terminal of the error amplifier section 40 in FIG. 12, and the negative input terminal side of the error amplifier section 40 is positive at both ends of the low resistance. And a signal component appearing in the DC output voltage _Vout is observed. That is, the open loop characteristics are evaluated by equivalently cutting the loop. FIG. 11A shows the amplitude characteristics, and FIG. 11B shows the phase characteristics.

  In FIG. 11, the dark solid line represents a simulation value using MATLAB (programmed by the method of the present invention), and the thin solid line represents an actual measurement obtained from an IC chip in which the circuit of FIG. 12 was actually prototyped using a 0.35 μm CMOS process. Represents a value. Further, a cross indicates a simulation value by SPICE (registered trademark). 13 shows characteristics when the circuit of FIG. 12 is operated with a 5 MHz clock.

  According to FIG. 11, it is observed that the simulation values of the amplitude characteristic and the phase characteristic are in good agreement with the actual measurement values. However. Since the period of the clock signal is 5 MHz, the characteristics are only valid up to 2.5 MHz. Moreover, since the analysis by SPICE (registered trademark) requires at least one hour / one point, only four points are shown in FIG.

  In FIG. 11, the time for calculating the overall frequency characteristics required about 25 minutes in the simulation on MATLAB (registered trademark).

  As described above, according to the comparison in FIG. 11, if the numerical calculation program of MATLAB (registered trademark) is used by applying the method of the present invention, the frequency characteristics of the circuit in FIG. Simulation can be performed at high speed with accuracy.

  The simulation method by numerical operation of the present invention can be used for designing a switching power supply represented by a DC-DC converter, and can perform circuit simulation in a short time with the same accuracy as the conventional method. Therefore, the design efficiency can be greatly improved.

  The simulation method according to the present invention can be applied to a circuit for processing a pulse waveform, a signal waveform having a sudden voltage change or a current change, in addition to being used for designing a switching power supply represented by a DC-DC converter. It can be applied to simulation of a circuit in which analog circuits and digital circuits are mixed.

It is a figure for demonstrating the block structure formed by modeling the switching power supply circuit of this invention. It is a figure which shows the part of the output transistor, output smoothing circuit, and load of the switching circuit of the switching power supply circuit to which this invention is applied. It is an example of the input of the program for calculating numerical values. It is the figure which showed the error amplifier part of the switching power supply circuit to which this invention is applied. It is the figure which showed the circuit block which modeled the error amplifier part by this invention. It is the figure which showed the circuit block which modeled the digital control circuit by this invention. It is a figure which shows the signal of each part of the circuit block of a digital control circuit. It is the figure which showed the current detection circuit of the switching power supply circuit to which this invention is applied. It is the figure which showed the circuit block which modeled the current detection circuit by this invention. It is the figure which performed the transient analysis of the same circuit structure by the program which applied the method by this invention on MATLAB (trademark), and the conventional SPICE (trademark) program, and compared execution time. It is the figure which compared the frequency characteristic of a circuit with the program which applied the method by this invention on MATLAB (trademark), the conventional SPICE (trademark) program, and the measured value made as an experiment. It illustrates a block diagram of an LSI buck DC-DC converter (V _in> V _out).

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Digital control circuit block 2 Output smoothing circuit block 3 Comparison block 4 Error amplifier block 5 Current detection circuit block 6 Switching circuit block 10 Digital control circuit 20 Output smoothing circuit 21 Inductor 30 Comparator 40 Error amplifier part 45 Reference voltage source 46 Error amplifier 47 Limiter 50 Current detector 51 Slope compensation 60 Switching circuit

Claims (3)

A method of simulating the operation of an analog circuit by numerical calculation,
Divide analog circuit into multiple circuit blocks,
In the divided circuit block,
For circuit blocks that require transient analysis to analyze temporal changes to obtain output,
Using the voltage and current in the circuit block as variables and describing them using a state equation that includes a differential variable obtained by differentiating the variables, an operation block for numerical operation is formed, and the circuit block is modeled by the operation block. Find the time change of the output of the circuit block by solving the state equation including variables by numerical calculation,
For circuit blocks that do not require transient analysis,
With respect to the voltage source and / or current source input to the analog active element of the circuit block, the transfer function when the voltage source or current source exists alone is represented by the S function from the voltage source or current source to the analog active element. A sum of products of the voltage source voltage or the current source current described by the transfer function and the S function, and describing the sum of the products as an input signal of the analog active element by superposition
The transfer function of the analog active element is described as an S function,
The overall transfer function of the circuit block is described by the product of the transfer function of the input signal and the transfer function of the analog active element to form an operation block for numerical operation, and the circuit block is modeled by the operation block. , By calculating the time change of the output of the circuit block by inversely transforming the overall transfer function,
A simulation method by numerical calculation of an analog circuit, wherein the process for obtaining the time change is executed by a computer . A switching circuit that switches input DC voltage and outputs a voltage pulse; an output smoothing circuit that smoothes the voltage pulse and supplies the voltage pulse to a load;
An error amplifier that takes a differential voltage between an output voltage of the output smoothing circuit and a reference voltage as an input signal, amplifies the input signal, outputs a low-pass voltage, and feeds back a limiter current from the output terminal to the input terminal. ,
Detecting the current flowing in the output smoothing circuit, converting the current into a voltage and outputting the current; and
Comparing the error amplifier output voltage and the output voltage of the current detection circuit,
A method of simulating the operation of a switching power supply circuit comprising a digital control circuit for controlling a switching period of the switching circuit by numerical calculation,
The output smoothing circuit uses an output voltage and a load current of an output transistor constituting the switching circuit as input signals,
The output voltage of the output smoothing circuit and the current flowing through the inductance element included in the output smoothing circuit are output signals,
An arithmetic operation that describes the output smoothing circuit using a state equation that includes a differential variable obtained by differentiating the variable, using the voltage across the capacitive element included in the output smoothing circuit and the current flowing through the inductance element as variables. Forming a block and modeling the switching circuit and the output smoothing circuit using the operation block,
Obtain the time change of the current flowing through the inductance element and the output DC voltage of the output smoothing circuit,
The error amplifier is
A transfer function from the input signal voltage source to the input of the error amplifier;
A transfer function from the output signal voltage source to the input of the error amplifier;
A transfer function from a reference voltage source to the input of the error amplifier;
For the transfer function from the limiter current source representing the current flowing in the diode limiter connected between the input end and the output end to the input end of the error amplifier, the transfer function when each exists independently is expressed as S function. Describe,
The sum of the product of each of the transfer functions and the input signal voltage source, the output signal voltage source, the reference voltage source, the voltage of the limiter current source or the current of the current source described by the S function is obtained. Described as an error amplifier input signal by reason,
The transfer function of the error amplifier is described as an S function,
An arithmetic block for performing a numerical operation is described by describing a total transfer function of the error amplifier by a product of the S function of the input signal and the transfer function of the error amplifier, and the error amplifier is modeled using the arithmetic block. ,
Obtain the time change of the output of the error amplifier by inversely transforming the overall transfer function Laplace,
The current detection circuit includes:
A detection transistor for flowing a current of a predetermined ratio with respect to the current flowing through the switching transistor is connected in parallel to the switching transistor,
A calculation block that performs numerical calculation by describing the product of the current flowing through the detection transistor and the resistance is formed, and a current detection circuit is modeled using the calculation block.
A current is detected by performing a numerical calculation using a product of the predetermined ratio and the resistance as a constant value current-voltage conversion coefficient ,
A switching power supply characterized in that the processing for obtaining the time change is executed by a computer, and the operation of each analog circuit of the switching circuit, the output smoothing circuit, the error amplifier, and the current detection circuit is simulated by numerical operation. A simulation method by numerical computation of a circuit. A switching circuit that switches input DC voltage and outputs a voltage pulse;
An output smoothing circuit that smoothes the voltage pulse and supplies the load to a load;
An error amplifier that takes a differential voltage between an output voltage of the output smoothing circuit and a reference voltage as an input signal, amplifies the input signal, outputs a low-pass voltage, and feeds back a limiter current from the output terminal to the input terminal. ,
A current flowing in the output smoothing circuit is detected, the current is converted into a voltage, and the current detection circuit;
In a switching power supply circuit comprising a digital control circuit for controlling a switching period of the switching circuit,
The difference voltage between the input signal voltage and the reference voltage is used as an input signal, the input signal is voltage amplified, the low-pass voltage is output, and the frequency characteristics of the error amplifier that feeds back the limiter current from the output terminal to the input terminal are numerical values. A method of simulating by computation,
A transfer function from the input signal voltage source to the input of the error amplifier;
A transfer function from the output signal voltage source to the input of the error amplifier;
A transfer function from a reference voltage source to the input of the error amplifier;
For the transfer function from the limiter current source representing the current flowing in the diode limiter connected between the input end and the output end to the input end of the error amplifier, the transfer function when each exists independently is expressed as S function. Describe,
The sum of the product of each of the transfer functions and the input signal voltage source, the output signal voltage source, the reference voltage source, the voltage of the limiter current source or the current of the current source described by the S function is obtained. Described as an error amplifier input signal by reason,
The transfer function of the error amplifier is described as an S function,
An arithmetic block for performing a numerical operation is described by describing a total transfer function of the error amplifier by a product of the S function of the input signal and the transfer function of the error amplifier, and the error amplifier is modeled using the arithmetic block. ,
Obtain the time change of the output of the error amplifier by inversely transforming the overall transfer function Laplace,
A simulation method by numerical calculation of an error amplifier, wherein the process for obtaining the time change is executed by a computer .

Images