Noise model generation method and noise model generation device

Abstract

To provide a noise model generation method and a noise model generation device which can enhance accuracy of a noise model.SOLUTION: A noise model generation method includes a transfer function calculation step of calculating a transfer function, a noise level waveform acquisition step of acquiring a noise level waveform of a frequency highly correlative to a trigger signal, and a noise model generation step of generating a noise model of an integrated circuit from the transfer function and the noise level waveform, wherein the noise level waveform acquisition step includes a measurement step of allowing the integrated circuit to output the trigger signal, and measuring the trigger signal and noise at a measurement point for each time; a conversion step of generating a noise level waveform indicating time change of the noise for each frequency, on the basis of the measurement point at the measurement point; a correlation value calculation step of calculating a correlation value indicating correlation between the noise level waveform for each of the frequencies and the time change of the trigger signal; and a filtering step of taking out the noise level waveform of the highly correlative frequency from the correlation value.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] Embodiments of the present invention relate to a method for generating a noise model and an apparatus for generating a noise model.

Background Art

[0002] Noise generated from an integrated circuit not only causes magnetic interference to other devices but also has an adverse effect on its own circuit operation. For the purpose of suppressing noise in an integrated circuit, various methods for generating a noise model have been proposed (for example, Patent Document 1). In recent years, it has been required to generate a noise model with higher accuracy than ever before.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] The problem to be solved by the present invention is to provide a method for generating a noise model and an apparatus for generating a noise model that can improve the accuracy of the noise model.

Means for Solving the Problems

[0005] The method for generating a noise model according to the embodiment is a method for generating a noise model of an integrated circuit. The method for generating a noise model includes a transfer function calculation step, a noise level waveform acquisition step, and a noise model generation step for generating a noise model. The transfer function calculation step calculates a transfer function from the inside of the integrated circuit to a measurement point on the electric circuit that constitutes the electric circuit including the integrated circuit. The noise level waveform acquisition step acquires a noise level waveform of a frequency highly correlated with the trigger signal from the actual measurement result of the noise when a trigger signal is output to the integrated circuit. The noise model generation step generates a noise model of the integrated circuit from the transfer function calculated in the transfer function calculation step and the noise level waveform acquired in the noise level waveform acquisition step. The noise level waveform acquisition step includes a measurement step, a conversion step, a correlation value calculation step, and a filtering step. The measurement step outputs a trigger signal to the integrated circuit and measures the trigger signal and the noise at the measurement point at each time. The conversion step generates a noise level waveform indicating the time change of the noise for each frequency based on the measurement value at the measurement point. The correlation value calculation step calculates a correlation value indicating the correlation relationship between the noise level waveform for each frequency and the time change of the trigger signal. The filtering step extracts a noise level waveform of a frequency having a high correlation relationship from the correlation values.

[0006] The apparatus for generating a nozzle model according to the embodiment implements the above-described method for generating a noise model.

Brief Description of the Drawings

[0007]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Best Mode for Carrying Out the Invention

[0008] Hereinafter, a method for generating a noise model of an embodiment and a device for generating a noise model will be described with reference to the drawings.

[0009] FIG. 1 is a schematic diagram of a noise model generation device 100 (hereinafter simply referred to as a generation device) of the present embodiment. The generation device 100 includes a power supply 20, an oscilloscope 30, a computer 40, and a stripline 50. An electric circuit 3 including an integrated circuit 1 to be calculated for the noise model is connected to the generation device 100.

[0010] The power supply 20 is connected to the power supply line (not shown) of the electric circuit 3 and supplies power to the electric circuit 3. The power supply 20 is connected to the computer 40 and is controlled by the computer 40. The oscilloscope 30 is connected to a noise measurement point (measurement point) P1 provided on the electric circuit 3 and an output measurement point P2, and measures the noise N and the trigger signal T of the electric circuit 3. The computer 40 is connected to the oscilloscope 30, controls the oscilloscope 30, and calculates a noise model at the observation point P0 inside the integrated circuit based on the output result of the oscilloscope 30.

[0011] The electric circuit 3 of this embodiment includes an integrated circuit 1 and a substrate 2 on which the integrated circuit 1 is mounted. The integrated circuit 1 has a plurality of terminals connected to the substrate 2. The terminals of the integrated circuit 1 include a measurement terminal 1n that is a noise measurement target, an output terminal 1a that outputs a trigger signal, and a ground terminal 1g that is connected to the ground pattern 2g of the substrate 2. The measurement terminal 1n, which is a terminal of the integrated circuit 1, is connected to an observation point P0 inside the integrated circuit.

[0012] A plurality of circuit patterns are formed on the substrate 2. The circuit patterns include a ground pattern 2g, an output pattern 2a to which the output terminal 1a is connected, and a noise coupling path 2n to which the measurement terminal 1n is connected. When measuring radiated noise that propagates through space rather than conductive noise that propagates through a conductor, the noise propagation path from the observation point P0 can be used instead of the noise coupling path 2n, rather than on the substrate 2.

[0013] The electric circuit 3 is provided with a noise measurement point P1 for measuring the noise N of the integrated circuit 1 and an output measurement point P2 for measuring the trigger signal T. One end of a noise measurement wiring 31 is connected to the noise measurement point P1. The other end of the noise measurement wiring 31 is connected to the noise measurement channel Ch1 of the oscilloscope 30. One end of an output measurement wiring 32 is connected to the output measurement point P2. The other end of the output measurement wiring 32 is connected to the trigger signal measurement channel Ch2 of the oscilloscope 30.

[0014] In this embodiment, the noise measurement point P1 is provided on the noise coupling path 2n. That is, the noise measurement wiring 31 extending from the oscilloscope 30 is connected to the noise coupling path 2n. In this embodiment, a stripline 50 is provided in the path of the noise coupling path 2n. The ground pattern 2g of the substrate 2 is connected to the housing ground 52 of the stripline 50. The noise from the noise coupling path 2n is received by the septum 51 in the stripline 50 and output as noise N to the noise measurement wiring 31.

[0015] Note that, in the path of the noise coupling path 2n, it is not necessarily required to arrange the stripline 50. Further, the noise measurement point P1 may be arranged on another circuit pattern on the substrate 2. For example, the noise measurement point P1 may be provided on the output pattern 2a. In this case, it is preferable to input only the noise N obtained by filtering and removing the trigger signal output from the output pattern 2a into the trigger signal measurement channel Ch2 of the oscilloscope 30.

[0016] The output measurement point P2 of the present embodiment is provided on the output pattern 2a. However, the arrangement of the output measurement point P2 is not limited to the present embodiment, and it may be arranged at any position on the electric circuit 3 as long as the trigger signal T of the integrated circuit 1 can be output.

[0017] FIG. 2 is a block diagram of a method for generating a noise model according to the present embodiment. As shown in FIG. 2, the method for generating a noise model according to the present embodiment includes a transfer function calculation step S10, a noise level waveform acquisition step S20, and a noise model generation step S30.

[0018] The transfer function calculation step S10 is a step of calculating the transfer function H(f) from the observation point P0 inside the integrated circuit 1 to the noise measurement point P1 on the electric circuit 3. The transfer function is expressed as the impedance of the electric circuit 3. In the present embodiment, the transfer function calculation step S10 is performed by the computer 40. However, the device for performing the transfer function calculation step S10 is not limited to the present embodiment. When the characteristics of the noise measurement wiring 31 cannot be ignored, the characteristics of the noise measurement wiring 31 may be included in the transfer function H(f).

[0019] The transfer function calculation step S10 is performed by a circuit simulator implemented in the computer 40. Examples of the circuit simulator include SPICE (Simulation Program with Integrated Circuit Emphasis) that strictly considers physical characteristics. Note that the method for calculating the transfer function H(f) is not limited to simulation. For example, the transfer function H(f) may be obtained by applying probes to the exposed portion formed by exposing the output portion in the integrated circuit 1 and the noise measurement point P1, respectively. That is, the transfer function H(f) may be calculated by actual measurement. Further, the transfer function H(f) may be calculated by combining simulation and actual measurement.

[0020] The computer 40 simulates the voltage V in (f) at the noise measurement point P1 with respect to the current I out (f) at the observation point P0 inside the integrated circuit 1. As shown in the following (Equation 1), the computer 40 performs a frequency analysis of the transfer function H(f) obtained by dividing the voltage V in (f) at the noise measurement point P1 by the current I out (f) at the observation point P0, and generates a transfer function H(f) with the frequency f as a variable.

[0021] [Number]

[0022] FIG. 3 is a diagram showing an example of the transfer function H(f) calculated in the transfer function calculation step S10. In FIG. 3, the horizontal axis represents the frequency, and the vertical axis represents the absolute value of the transfer function H(f).

[0023] As shown in FIG. 1, the noise level waveform acquisition step S20 is a step that can be performed in parallel with the transfer function calculation step S10. Therefore, either the transfer function calculation step S10 or the noise level waveform acquisition step S20 may be performed first, or they may be performed simultaneously.

[0024] The noise level waveform acquisition step S20 includes a measurement step S21, a conversion step S22, a correlation value calculation step S23, and a filtering step S24. The measurement step S21, the conversion step S22, the correlation value calculation step S23, and the filtering step S24 are performed in this order.

[0025] In the measurement step S21, the integrated circuit 1 is caused to output a trigger signal T, and the noise N is measured by the oscilloscope 30. As shown in FIG. 1, the trigger signal voltage Vt output from the integrated circuit 1 is output in the output pattern 2a of the substrate 2 and input to the trigger signal measurement channel Ch2 of the oscilloscope 30 via the output measurement wiring 32. Also, some noise N is generated in the electric circuit 3 regardless of the presence or absence of the output of the trigger signal T, and this noise N is input from the noise measurement point P of the substrate to the noise measurement channel Ch1 of the oscilloscope 30 via the noise measurement wiring 31. The oscilloscope 30 measures the trigger signal voltage Vt and the noise voltage at each time before and after the output of the trigger signal T.

[0026] In this specification, "measuring at each time" means measuring at a predetermined sampling time and recording the measured values. By measuring the trigger signal voltage Vt and the noise voltage at each time, the oscilloscope 30 records the time change of these measured values and draws a waveform with time on the horizontal axis and voltage on the vertical axis.

[0027] In this embodiment, the oscilloscope 30 is described as measuring the voltages of the trigger signal T and the noise N, but the oscilloscope 30 may measure the current values of the trigger signal T and the noise N. That is, in the measurement step S21, as long as the trigger signal T and the noise N are measured, the parameter to be measured may be either a voltage or a current value.

[0028] Figure 4 is a graph showing the time variations of the trigger signal voltage Vt and the noise voltage Vn measured by the oscilloscope 30. In Figure 4, the horizontal axis represents time and the vertical axis represents voltage. In the example shown in Figure 4, the integrated circuit 1 operates in a pattern of repeatedly outputting and stopping the trigger signal T at regular intervals.

[0029] In the example shown in Figure 4, the noise voltage Vn measured by the oscilloscope 30 increases at the timing when the trigger signal T is output. That is, it can be confirmed that noise N is output along with the output of the trigger signal T. However, the output of noise N during the output of the trigger signal T also has floor noise such as noise caused by the measuring equipment superimposed on it. For this reason, it is difficult to determine from Figure 4 which component of the output noise voltage Vn is caused by the output of the trigger signal T.

[0030] The conversion step S22 is a step of converting the waveform of the noise voltage Vn measured by the oscilloscope 30 and calculating the time variation of the noise voltage Vn for each frequency. In the following description, the time variation of the noise voltage Vn for each frequency is referred to as a "noise level waveform". That is, the conversion step is a step of generating a noise level waveform showing the time variation of the noise voltage Vn for each frequency based on the measured value at the noise measurement point P1. The conversion step S22 is performed in the computer 40 connected to the oscilloscope 30.

[0031] In the conversion step S22 of the present embodiment, a noise level waveform for each frequency is generated by performing a short-time Fourier transform on the waveform of the noise voltage Vn. That is, the result of performing a short-time Fourier transform on the noise voltage Vn(t) is the noise level waveform Vn(f,t). By performing a short-time Fourier transform on the noise voltage Vn, it becomes possible to extract a period with high noise while retaining time information and analyze the frequency of the noise. In particular, as a conversion method, by performing a short-time Fourier transform, it becomes possible to analyze the frequency spectrum with a certain resolution.

[0032] In this embodiment, it is preferable that the short-time Fourier transform is performed every time within 1 ms. By setting the resolution of the short-time Fourier transform to 1 ms or less, frequency spectrum analysis can be performed with sufficient resolution.

[0033] Note that in the conversion step S22, wavelet transform may be performed instead of the short-time Fourier transform. However, when performing wavelet transform, since the time interval does not necessarily become constant, there is a possibility that the resolution with respect to the time axis becomes low in a desired section.

[0034] FIG. 5 is a graph showing a noise level waveform generated in the conversion step S22. In FIG. 5, the horizontal axis represents time, the vertical axis represents frequency, and the color shading represents the absolute value of the noise level (voltage).

[0035] FIG. 6 is a graph showing the noise level waveform at the first frequency F1 in FIG. 5, FIG. 7 is a graph showing the noise level waveform at the second frequency F2 in FIG. 5, and FIG. 8 is a graph showing the noise level waveform at the third frequency F3 in FIG. 5. In FIGS. 6 to 8, the horizontal axis represents time, and the vertical axis represents the absolute value of Vn(f,t) which is the short-time Fourier transform.

[0036] As shown in FIGS. 5 and 6, the first frequency F1 is the frequency of the noise N output in conjunction with the output of the trigger signal T. As shown in FIGS. 5 and 7, the second frequency F2 is the noise N output independently of the output of the trigger signal T. Further, as shown in FIGS. 5 and 8, the third frequency F3 is considered to be floor noise.

[0037] The correlation value calculation step S23 is a step of calculating a correlation value Cn(f) indicating the correlation between the noise level waveform Vn(f,t) for each frequency and the time change of the trigger signal voltage Vt(t). The conversion step S22 is performed in the computer 40.

[0038] The following (Equation 2) shows the correlation function C(f) between the noise level waveform Vn(f,t) and the trigger signal voltage Vt. Also, the following (Equation 3) shows the correlation value Cn(f) obtained by dimensionlessizing (Equation 2).

[0039] [Number]

[0040] [Number]

[0041] FIG. 9 is a graph showing the correlation value Cn(f) between the noise level waveform of the present embodiment obtained in the correlation value calculation step S23 and the trigger signal T. In FIG. 9, the horizontal axis represents the frequency, and the vertical axis represents the absolute value of the correlation value Cn(f). In FIG. 9, the closer the absolute value of the correlation value Cn(f) is to 1, the higher the correlation, and the closer the absolute value of the correlation value Cn(f) is to 0, the lower the correlation.

[0042] The filtering step S24 is a step of extracting the noise level waveform of a frequency having a high correlation based on the correlation value Cn(f) calculated in the correlation value calculation step S23. The filtering step S24 is performed in the computer 40.

[0043] As shown in FIG. 9, in the present embodiment, for the frequency at which the absolute value of the correlation value Cn(f) is 0.6 or more, it is regarded as having a correlation between the noise level waveform and the trigger signal T. That is, in the present embodiment, the threshold for determining the presence or absence of correlation is set to 0.6. The threshold is appropriately set in advance in consideration of the balance such as the accuracy of the noise model and the calculation time.

[0044] As shown in FIG. 9, the range of the frequency at which the absolute value of the correlation value Cn(f) is 0.6 or more is defined as the bandwidth fw. Also, the center frequency of the bandwidth fw is defined as the center frequency fc. In the present embodiment, the center frequency fc is a value in the vicinity of the first frequency F1 shown in FIG. 5.

[0045] In the filtering step S24 of this embodiment, noise N with a low correlation value Cn (not exceeding the threshold) is removed using the band-pass filter B(f). First, the bandwidth Bw is calculated from the bandwidth fw using (Equation 4) shown below. Further, the generated bandwidth Bw and the center frequency fc are substituted into (Equation 5) below to generate the band-pass filter B(f). FIG. 10 shows the characteristic diagram of the band-pass filter B(f) of this embodiment.

[0046]

Number

[0047]

Number

[0048] As shown in (Equation 6) below, the generated band-pass filter B(f) and the noise level waveform Vn(f,t) are multiplied to filter the noise level waveform. In the following description, the filtered noise level waveform is referred to as the extracted noise level waveform Vnn(f,t). That is, in the filtering step S24, the extracted noise level waveform Vnn(f,t) is obtained.

[0049]

Number

[0050] FIG. 11 is a diagram showing the extracted noise level waveform Vnn(f,t) of this embodiment. In FIG. 11, the horizontal axis represents time, the vertical axis represents frequency, and the color shading represents the absolute value of the noise level (voltage). Compared with FIG. 5, FIG. 11 shows that the noise N outside the vicinity of the first frequency F1 with a high correlation with the trigger signal T has been removed.

[0051] Through the above measurement step S21, conversion step S22, correlation value calculation step S23, and filtering step S24, the noise level waveform acquisition step S20 is completed. In this way, the noise level waveform acquisition step S20 is a step of acquiring a noise level waveform (i.e., the extracted noise level waveform Vnn(f,t)) of a frequency highly correlated with the trigger signal T from the actual measurement result of the noise N when the integrated circuit 1 is output with the trigger signal T.

[0052] The noise model generation step S30 is a step of generating a noise model of the integrated circuit 1 from the transfer function H(f) calculated in the transfer function calculation step S10 and the noise level waveform Vnn(f) acquired in the noise level waveform acquisition step S20. In the noise model generation step S30, for example, the noise current I(f) can be calculated as a noise model based on (Equation 7) shown below. The noise model generation step S30 is performed in the computer 40.

[0053]

Equation

[0054] The noise model I(f) obtained in the noise model generation step S30 of the present embodiment can be used for noise evaluation of the integrated circuit 1.

[0055] Each configuration described in the present embodiment is an example, and may be recombined with other configurations as appropriate. For example, in the present embodiment, the oscilloscope 30 is used in the measurement step S21 to measure the trigger signal T and the noise N. However, the measuring instrument used in the measurement step S21 is not limited to the oscilloscope 30 as long as it can measure the time changes of the trigger signal T and the noise N.

[0056] In the conventional method for generating a noise model, the measured value of the noise serving as the basis for analysis did not take into account the temporal change in the noise with respect to the input of the trigger signal T. For this reason, the generated noise model was affected by the noise that occurred independently of the trigger signal T, resulting in a low accuracy of the noise model. In addition, there was a problem that the correlation between the type of the trigger signal T and the generated noise could not be accurately grasped, and it was difficult to lead to fundamental measures for noise suppression even when a noise model was generated.

[0057] The method for generating the noise model of this embodiment is a method for generating the noise model of integrated circuit 1. Also, as shown in FIG. 2, the method for generating the noise model includes a transfer function calculation step S10, a noise level waveform acquisition step S20, and a noise model generation step S30. The transfer function calculation step S10 is a step of calculating the transfer function H(f) from the inside of the integrated circuit 1 to the noise measurement point P1 on the electric circuit 3 that constitutes the electric circuit 3 including the integrated circuit 1. The noise level waveform acquisition step S20 is a step of acquiring a noise level waveform (extracted noise level waveform Vnn(f,t)) of a frequency highly correlated with the trigger signal T from the measurement result of the noise N when the trigger signal T is output to the integrated circuit 1. The noise model generation step S30 is a step of generating the noise model I(f) of the integrated circuit 1 from the transfer function H(f) calculated in the transfer function calculation step S10 and the noise level waveform Vn(f,t) acquired in the noise level waveform acquisition step S20. The noise level waveform acquisition step S20 includes a measurement step S21, a conversion step S22, a correlation value calculation step S23, and a filtering step S24. The measurement step S21 is a step of outputting the trigger signal T to the integrated circuit 1 and measuring the trigger signal T and the noise N at the noise measurement point P1 for each time. The conversion step S22 is a step of generating a noise level waveform Vn(f,t) indicating the time change of the noise N for each frequency based on the measurement value at the noise measurement point P1. The correlation value calculation step S23 is a step of calculating a correlation value Cn(f) indicating the correlation relationship between the noise level waveform Vn(f,t) for each frequency and the time change of the trigger signal T. The filtering step S24 is a step of extracting a noise level waveform (extracted noise level waveform Vnn(f,t)) of a frequency with a high correlation relationship from the correlation value Cn(f).

[0058] According to the above configuration, in the noise level waveform acquisition step S20, only the noise N output with respect to the trigger signal T is extracted as the extracted noise level waveform Vnn(f,t), and the noise model I(f) is generated in the noise model generation step S30. Therefore, the influence of the noise N irrelevant to the trigger signal T can be suppressed, and the noise model I(f) can be generated, and the accuracy of the generated noise model I(f) can be improved.

[0059] Furthermore, according to the above configuration, for the integrated circuit 1 having various operation modes, it is also possible to output various trigger signals T corresponding to each operation mode and separately calculate the noise models I(f) for the respective operation modes. In this case, by analyzing the noise models I(f) of each operation mode, it is possible to identify and correct the circuits that are the sources of noise in the integrated circuit 1 driven in each operation mode, which can be used to reduce the generation of noise.

[0060] In the method for generating a noise model of the present embodiment, the conversion step S22 is a step of performing a short-time Fourier transform on the waveform of the noise at the noise measurement point P1.

[0061] According to the above configuration, it is possible to extract the frequency spectrum of the noise while retaining the time information. Therefore, in the correlation value calculation step S23, it is possible to easily obtain the correlation with the trigger signal T (i.e., the correlation value Cn(f)) by comparing with the time information of the trigger signal T. Also, by performing the conversion step using the short-time Fourier transform, the time resolution can be made constant, and analysis with a desired resolution can be performed.

[0062] In the method for generating a noise model of the present embodiment, the filtering step S24 is a step of extracting the noise level waveform Vn(f,t) at frequencies where the correlation value Cn(f) exceeds the threshold value.

[0063] According to the above configuration, in the filtering step S24, it is possible to easily extract the noise level waveform Vn(f,t) based on the threshold value.

[0064] In the method for generating a noise model of the present embodiment, the measurement step S21 is a step of measuring noise in the stripline 50 connected to the substrate 2 on which the integrated circuit 1 is mounted.

[0065] According to this configuration, the noise N radiated from the integrated circuit 1 can be directly measured using the stripline 50, and the measurement accuracy of the noise N at the noise measurement point P1 can be improved.

[0066] The noise model generation device 100 of the present embodiment implements the above-described noise model generation method.

[0067] According to this configuration, an accurate noise model I(f) of the integrated circuit 1 can be generated and used for circuit design or modification of the integrated circuit 1.

[0068] According to at least one of the embodiments described above, only the noise N output in response to the trigger signal T is extracted to generate the noise model I(f), so that a method for generating a noise level for improving the accuracy of the generated noise model I(f) can be provided.

[0069] The method for manufacturing a semiconductor device and the semiconductor device of the embodiment include the following appended aspects. (Appended Note 1) A method for generating a noise model of an integrated circuit, a transfer function calculation step of configuring an electric circuit including the integrated circuit and calculating a transfer function from inside the integrated circuit to a measurement point on the electric circuit; a noise level waveform acquisition step of acquiring a noise level waveform of a frequency highly correlated with the trigger signal from an actual measurement result of noise when a trigger signal is output to the integrated circuit; and a noise model generation step of generating a noise model of the integrated circuit from the transfer function calculated in the transfer function calculation step and the noise level waveform acquired in the noise level waveform acquisition step. The noise level waveform acquisition step includes a measurement step of outputting a trigger signal to the integrated circuit and measuring the trigger signal and the noise at the measurement point at each time; and a conversion step of generating the noise level waveform indicating the time change of the noise for each frequency based on the measurement value at the measurement point. A correlation value calculation step of calculating a correlation value indicating a correlation between the noise level waveform for each frequency and the temporal change of the trigger signal; A filtering step of extracting the noise level waveform of the frequency having a high correlation from the correlation value, and having: A method for generating a noise model. (Appendix 2) The conversion step is a step of performing a short-time Fourier transform on the waveform of the noise at the measurement point. The method for generating a noise model according to Appendix 1. (Appendix 3) The filtering step is a step of extracting the noise level waveform of the frequency at which the correlation value exceeds a threshold value. The method for generating a noise model according to Appendix 1 or 2. (Appendix 4) The measurement step is a step of measuring the noise in a stripline connected to a substrate on which the integrated circuit is mounted. The method for generating a noise model according to any one of Appendices 1 to 3. (Appendix 5) Implementing the method for generating a noise model according to any one of Appendices 1 to 4, A device for generating a nozzle model.

[0070] Although some embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the gist of the invention. These embodiments and their modifications are included in the scope and gist of the invention, and are also included in the invention described in the claims and the equivalent scope thereof.

Explanation of Signs

[0071] 1...Integrated circuit, 2...Substrate, 3...Electrical circuit, 50...Strip line, 100...Generator, N...Noise, P1...Noise measurement point (measurement point), S10...Transfer function calculation step, S20...Noise level waveform acquisition step, S21...Measurement step, S22...Conversion step, S23...Correlation value calculation step, S24...Filtering step, S30...Noise model generation step, T...Trigger signal

Claims

1. A method for generating a noise model of an integrated circuit, comprising: a transfer function calculation step of configuring an electrical circuit including the integrated circuit and calculating a transfer function from the inside of the integrated circuit to a measurement point on the electrical circuit; a noise level waveform acquisition step of acquiring a noise level waveform of a frequency highly correlated with a trigger signal from an actual measurement result of noise when a trigger signal is output to the integrated circuit; a noise model generation step of generating a noise model of the integrated circuit from the transfer function calculated in the transfer function calculation step and the noise level waveform acquired in the noise level waveform acquisition step; wherein the noise level waveform acquisition step includes: a measurement step of outputting a trigger signal to the integrated circuit and measuring the trigger signal and the noise at the measurement point at each time; a conversion step of generating the noise level waveform indicating the time change of the noise for each frequency based on the measurement value at the measurement point; a correlation value calculation step of calculating a correlation value indicating a correlation relationship between the noise level waveform for each frequency and the time change of the trigger signal; and a filtering step of extracting the noise level waveform of the frequency having a high correlation relationship from the correlation value. A method for generating a noise model.

2. The conversion step is a step of performing a short-time Fourier transform on the waveform of the noise at the measurement point. The method for generating a noise model according to Claim 1.

3. The filtering step is a step of extracting the noise level waveform of the frequency at which the correlation value exceeds a threshold value. The method for generating a noise model according to Claim 1.

4. The measurement step is a step of measuring the noise in a stripline connected to a substrate on which the integrated circuit is mounted. The method for generating a noise model according to Claim 1.

5. Implementing the method for generating a noise model according to any one of Claims 1 to 4, A device for generating a nozzle model.

Images