JP2014228386A - Noise source position estimation device and noise source position estimation program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain a noise source position estimation device capable of measuring a noise position irrespectively of a shape of a measurement target.SOLUTION: A measuring unit 10 acquires temporal changes in noise waveforms of a measurement target in a plurality of measurement locations in the same time domain. An arithmetic unit 20 operates an electromagnetic field analysis if the plurality of measurement locations are assumed as signal source positions and the temporally changing waveforms are assumed as excitation waveforms of the signal sources, detects a peak of an electromagnetic field intensity for an operation result, and determines a position at which the peak is detected as a noise source. A result display unit 50 displays thereon a position at which the noise source occurs.

Description

  The present invention relates to a noise source position estimation apparatus and a noise source position estimation program for estimating a generation source of noise generated in a measurement target such as a printed circuit board.

  For conventional noise source position estimation methods, especially for PCB power supply system noise, check that noise is detected by bringing a sensor such as a magnetic field probe close to the power supply pin of the IC that is a candidate noise source. The method of going was taken. Alternatively, as in Non-Patent Document 1 and Patent Document 1, a method of scanning the entire surface of the substrate with a sensor and searching for a strong noise portion or a noise propagation path has been used.

JP-A-5-281274

VCCI Kit Module Interference Wave Measurement Technical Standard V-A3 / 2009.04 Appendix 1

  However, in the conventional technique described in Patent Document 1, it is difficult to bring the probe close to the structure having pins on the back surface of the IC as in the BGA package. In addition, when noise propagates through the inner layer of a multilayer board, it is difficult to directly measure the noise signal, and only a limited number of measurement points, such as the edge of the board or the part drawn to the surface layer through the through hole, can be measured. Met.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to obtain a noise source position estimation apparatus capable of measuring a noise position regardless of the shape of a measurement target.

  The noise source position estimating apparatus according to the present invention is a noise source position estimating apparatus that estimates a source of noise generated in a measurement target. Calculates the electromagnetic field analysis when the measurement unit to be acquired and the signal source position for multiple measurement locations and the excitation waveform of the signal source for the waveform that changes over time, and detects the peak of the electromagnetic field intensity for the calculation result. And a calculation unit that determines a position where the peak is detected as a noise source.

  The noise source position estimation apparatus according to the present invention calculates an electromagnetic field analysis when a plurality of measurement points acquired in the same time domain are signal source positions, and a time-varying waveform is an excitation waveform of the signal source. On the other hand, since the peak of the electromagnetic field intensity is detected and the position where the peak is detected is determined as the noise source, the noise position can be measured regardless of the shape of the measurement target.

It is a block block diagram which shows the noise source position estimation apparatus by Embodiment 1 of this invention. It is explanatory drawing which shows the relationship between the noise source position estimation apparatus by Embodiment 1 of this invention, and a measuring object. It is a flowchart which shows operation | movement of the noise source position estimation apparatus by Embodiment 1 of this invention. It is explanatory drawing which shows the measurement of the noise waveform of the noise source position estimation apparatus by Embodiment 1 of this invention. It is explanatory drawing which shows the production | generation of the time reversal signal of the noise waveform of the noise source position estimation apparatus by Embodiment 1 of this invention. It is explanatory drawing which shows typically the electromagnetic field analysis which goes back the time of the noise source position estimation apparatus by Embodiment 1 of this invention. It is explanatory drawing which shows the electric field strength distribution immediately after the time reversal signal input of the noise source position estimation apparatus by Embodiment 1 of this invention. It is explanatory drawing which shows the electric field strength distribution at the time of the peak formation in the noise source position of the noise source position estimation apparatus by Embodiment 1 of this invention. It is explanatory drawing which shows generation | occurrence | production of the multipath by the boundary of the noise source position estimation apparatus by Embodiment 1 of this invention.

Embodiment 1 FIG.
FIG. 1 is a block diagram showing a noise source position estimating apparatus according to Embodiment 1 of the present invention.
The noise source position estimation apparatus shown in FIG. 1 includes a measurement unit 10, a calculation unit 20, an analysis setting / model input unit 30, a threshold value input unit 40, and a result display unit 50. The measurement unit 10 measures the time change of the noise waveform to be measured at a plurality of measurement locations at the same time, and includes sensors 11 a to 11 n and a memory 12. Each of the sensors 11a to 11n is a sensor for detecting a change in a noise waveform to be measured. Note that n represents an arbitrary number, and the number of sensors is not particularly limited as long as it is a plurality of sensors. The memory 12 is a storage unit for storing noise waveform signals from these sensors 11a to 11n.

  The calculation unit 20 calculates an electromagnetic field analysis when a plurality of measurement locations are signal source positions and a time-varying waveform is an excitation waveform of the signal source, and detects the peak of the electromagnetic field intensity from the calculation result. The position where the peak is detected is determined as a noise source, and includes a time reversal signal calculation unit 21, an operation initial setting unit 22, an electromagnetic field calculation unit 23, an electromagnetic field calculation end determination unit 24, and a peak detection unit 25. Yes. The time reversal signal calculation unit 21 is a calculation unit that performs time reversal on the noise waveform signal from the measurement unit 10. The calculation initial setting unit 22 is a processing unit that performs initial setting in the electromagnetic field calculation unit 23 based on a setting input from the analysis setting / model input unit 30. The electromagnetic field calculation unit 23 is a calculation unit that executes time-domain electromagnetic field analysis (electromagnetic field simulation). The electromagnetic field calculation end determination unit 24 is a processing unit that performs end determination in the electromagnetic field calculation unit 23 based on the electromagnetic field analysis cutoff time information input from the analysis setting / model input unit 30. The peak detection unit 25 uses the threshold value input from the threshold value input unit 40 to detect a peak from the calculation result in the electromagnetic field calculation unit 23, and outputs this to the result display unit 50 as a noise position. Part.

  The analysis setting / model input unit 30 is an input unit for inputting an initial setting value in the electromagnetic field calculation unit 23, an abort time in the electromagnetic field calculation end determination unit 24, and the threshold value input unit 40 is a peak detection unit. An input unit for inputting a threshold value when the unit 25 performs peak detection, and the result display unit 50 are display units for displaying a noise position.

  The calculation unit 20 in the noise source position estimation apparatus is realized using a computer, and the time reversal signal calculation unit 21 to the peak detection unit 25 include software corresponding to each function and a CPU for executing these software, It consists of hardware such as memory.

  FIG. 2 is an explanatory diagram illustrating the relationship between the measurement target and the noise source position estimation apparatus. In the present embodiment, it is assumed that the noise generation source position in the parallel plate is estimated from the noise signal transmitted in the parallel plate. In FIG. 2, two flat plates 201a and 201b form a parallel plate, and the noise source position estimation apparatus measures a noise source inside these parallel plates.

  On these parallel plates, four voltage probes 202a, 202b, 202c, 202d are connected to the device body 200 of the noise source position estimation device via cables 203a, 203b, 203c, 203d, respectively. It corresponds to the sensors 11a to 11n in FIG. In FIG. 2, four voltage probes 202a, 202b, 202c, and 202d are provided. However, the number is not particularly limited to four as long as there are a plurality of voltage probes. The voltage probes 202a, 202b, 202c, and 202d are connected to a position where the voltage between parallel plates such as a substrate end can be measured. The measurement point does not need to be near the noise source as in the prior art, and does not need to be along the noise propagation path.

Hereinafter, the operation of the noise source position estimation apparatus according to the present embodiment will be described.
FIG. 3 is a flowchart showing the operation of the noise source position estimation apparatus.
When performing noise source position estimation processing, first, a noise waveform for a measurement target is measured (step ST1).
FIG. 4 is an explanatory diagram showing measurement of a noise waveform. In FIG. 4, noise 402 is generated from a noise source 401 in a parallel plate, is transmitted through the parallel plate, and a noise waveform is simultaneously observed as time versus voltage at a plurality of observation points 403a, 403b, and 403c. In order to simplify the explanation, there are three observation points in FIG. 4, but the number is actually the same as the number of voltage probes. The noise waveform simultaneously measured at the observation point is recorded in the memory 12 in the measurement unit 10.

  Next, the time reversal signal calculation unit 21 of the calculation unit 20 generates a time reversal signal (step ST2). FIG. 5 shows generation of a time reversal signal of a noise waveform. As illustrated, the noise waveforms 500a, 500b, and 500c are read from the memory 12 and converted into time-reversed signals 501a, 501b, and 501c by the time-reversed signal calculation unit 21. Here, the time inversion signal is obtained by rearranging measured data in reverse order in time.

Next, a time domain electromagnetic field simulation (for example, FDTD method) is executed by the calculation initial setting unit 22, the electromagnetic field calculation unit 23, and the electromagnetic field calculation end determination unit 24 (step ST3 to step ST5). However, in a normal time domain simulation, it is common to calculate the electromagnetic field at t + Δt from the electromagnetic field at time t, but in this method, the electromagnetic field at t−Δt is calculated from the electromagnetic field at time t. Formulate as follows. In other words, the calculation goes back to the past.
The minute time step Δt and the model cell size in this simulation are preferably about 1/10 of the wavelength of the main frequency component of the measurement waveform. These setting values are assumed to be set from the analysis setting / model input unit 30, but any one of the calculation initial setting unit 22, the electromagnetic field calculation unit 23, and the electromagnetic field calculation end determination unit 24 is determined from the measured waveform. An automatic calculation configuration may be used.

  FIG. 6 is an explanatory diagram schematically showing electromagnetic field analysis that goes back in time in the present embodiment. It is assumed that the calculation unit 20 includes a detailed model of a printed circuit board to be measured. Simulations dating back to the past with positions 603a, 603b, and 603c in the model corresponding to observation points 403a, 403b, and 403c in FIG. 4 as analysis signal source positions and time reversal signals 501a, 501b, and 501c as analysis signal source waveforms. Execute. The simulation is executed while going back by a minute time Δt for each step until the end condition is satisfied by the electromagnetic field calculation end determination unit 24 (step ST4 to step ST5). The termination condition in step ST5 is generally defined by time in many cases, and it is desirable that the time is longer than the signal passes through the analysis target. The end condition is assumed to be set from the analysis setting / model input unit 30, but since the analysis model information is included, the calculation initial setting unit 22, the electromagnetic wave is determined based on the size and dielectric constant of the measurement target. Either the field calculation unit 23 or the electromagnetic field calculation end determination unit 24 may automatically calculate. Moreover, the structure on condition that a peak arises, or the structure cut off manually may be sufficient.

  If the time change of the electromagnetic field due to the above simulation results is continuously observed, the waves 602a, 602b, and 602c from all the signal sources strengthen each other, and a peak occurs at a certain point 601. This is the original noise source position. This peak is detected by the peak detector 25 based on the setting from the threshold value input unit 40 (step ST6), and the peak position is displayed on the result display unit 50 (step ST7).

Examples of electromagnetic field simulations that go back in time are shown in FIGS. FIG. 7 shows the electric field intensity distribution immediately after the time reversal signal is input, and FIG. The signal source position of the simulation, that is, the observation point position in the actual measurement is a total of eight places at the four corners and the center of the four sides, and the noise source position is (x, y) = (100, 50). From FIG. 8, it can be confirmed that a peak is formed at the noise source position (x, y) = (100, 50).
With the configuration and operation described above, it is possible to realize a measuring apparatus that can determine the generation position of noise transmitted through the inner layer of the substrate from the measurement results of limited measurement points.

In the above description, a single noise source position measuring device is used. However, even if the function is divided into a plurality of devices and a noise source position measuring device including a plurality of devices is configured, an equivalent function can be realized. For example, a single sensor may be combined and a simulation may be performed based on each output signal. In such a case, since an existing measuring instrument can be used, there is an advantage that the apparatus cost can be reduced.
In addition, the configuration using the electromagnetic field analysis that goes back in time has been explained. However, under conditions where the loss is sufficiently small and there is no irreversible structure in the model, the electromagnetic wave that advances the normal time instead of the electromagnetic field analysis that goes back in time is used. Equivalent results can be obtained using field analysis. Also in this case, since the existing software can be used, there is an advantage that the cost can be reduced.

  Further, the arithmetic unit 20 may use the physical structure and electrical characteristics of the measurement target such as a printed circuit board as a model for electromagnetic field analysis from CAD data and additional information. That is, the physical structure of the substrate such as the substrate outer shape and the shape and size of the conductor pattern is acquired as CAD data, and the electrical characteristics of the substrate such as the dielectric constant of the substrate dielectric, the dielectric loss tangent, and the conductivity of the conductor are set as additional information. Therefore, it can be used as a model for electromagnetic field analysis.

  In the above embodiment, three observation points (signal source positions) are used as shown in FIGS. 4 to 6. However, if there is no practical problem, two observation points may be used. In this case, there are two points where the waves from the signal source in FIG. 6 strengthen each other, but it is practical enough if these two points are close to each other or if the noise source position is narrowed down to two points. In some cases, it is possible to have two observation points.

In addition, when the measurement space is a closed space having a boundary, a multipath passing through the boundary is generated. FIG. 9 is an explanatory diagram showing the occurrence of multipath due to such a boundary.
As shown in FIG. 9, there are two actual signal sources: a signal source 901a which is a first signal source and a signal source 901b which is a second signal source. However, since the multipath of the signal source 901b is generated by the boundary 902, and this operates as the virtual wave source 901c outside the boundary 902, the virtual wave source 901c is the third signal source, and the noise is the same as in FIG. Location 903 can be identified. Therefore, when multipath by the boundary 902 can be used, even if there are two observation points, it can be regarded as three signal source positions including the position of the virtual wave source 901c. However, the noise position 903 can be identified. That is, a virtual location such as the virtual wave source 901c can be included as a location for acquiring a time change of the noise waveform to be measured.

  Furthermore, when multiple virtual wave sources by multipath can be obtained, or when it is sufficient to narrow down the above-mentioned noise source position to two points, even if only one actual observation point is included, a virtual part is included. If the time change of the noise waveform at a plurality of measurement points can be acquired as the same time region, the noise source position can be estimated.

  As described above, according to the noise source position estimation apparatus of the first embodiment, in the noise source position estimation apparatus that estimates the generation source of the noise generated in the measurement object, the noise of the measurement object in a plurality of measurement locations Calculates the electromagnetic field analysis when measuring the time change of the waveform in the same time domain, the signal source position for multiple measurement locations, and the excitation waveform of the signal source for the waveform that changes over time. Thus, the noise position can be measured regardless of the shape of the object to be measured, because the operation unit detects the peak of the electromagnetic field intensity and determines the position where the peak is detected as the noise source.

  In addition, according to the noise source position estimation apparatus of the first embodiment, since the calculation unit inverts the noise waveform acquired by the measurement unit in time and performs electromagnetic field analysis going back in time, it is limited. The noise generation position can be determined from the measurement result of the measurement point.

  In addition, according to the noise source position estimation apparatus of the first embodiment, since there are three or more measurement locations, the noise generation location can be determined for any measurement target. .

  Further, according to the noise source position estimation apparatus of the first embodiment, the calculation unit obtains the electromagnetic analysis time-out time from the size of the measurement object and the dielectric constant, and therefore inputs the time-out time from the outside. Electromagnetic field analysis can be performed without any problem.

  Further, according to the noise source position estimation apparatus of the first embodiment, the calculation unit obtains the space cell size and the time step size of the electromagnetic field analysis model from the measurement waveform. Electromagnetic field analysis can be performed without inputting the time step size.

  In addition, according to the noise source position estimation apparatus of the first embodiment, the calculation unit holds the physical structure and electrical characteristics of the measurement target as a model for electromagnetic field analysis from CAD data and additional information. An electromagnetic field analysis model can be obtained.

  In addition, according to the noise source position estimation program of the first embodiment, using the measurement unit that acquires the time change of the noise waveform of the measurement target at a plurality of measurement locations in the same time domain, the noise generated in the measurement target is detected. The computer that implements the noise source position estimation device that estimates the source is used to calculate the electromagnetic field analysis when multiple measurement locations are signal source positions and the time-varying waveform is the excitation waveform of the signal source. On the other hand, the peak of the electromagnetic field strength is detected, and the position where the peak is detected is operated as a calculation unit that determines the noise source. Therefore, the noise source can measure the noise position regardless of the shape of the measurement target. The position estimation device can be realized on a computer.

  In the present invention, any constituent element of the embodiment can be modified or any constituent element of the embodiment can be omitted within the scope of the invention.

  DESCRIPTION OF SYMBOLS 10 Measurement part, 11a-11n Sensor, 12 Memory, 20 Calculation part, 21 Time inversion signal calculation part, 22 Calculation initial setting part, 23 Electromagnetic field calculation part, 24 Electromagnetic field calculation end determination part, 25 Peak detection part, 30 Analysis Setting / model input unit, 40 threshold value input unit, 50 result display unit, 200 apparatus main body, 201a, 201b flat plate, 202a, 202b, 202c, 202d voltage probe, 203a, 203b, 203c, 203d cable, 401 noise source, 402 Noise, 403a, 403b, 403c Observation point, 500a, 500b, 500c Noise waveform, 501a, 501b, 501c Time reversal signal, 601 Single point (peak), 602a, 602b, 602c Wave, 603a, 603b, 603c Position in model 901a, 901b Issue source, 901c virtual wave source, 902 boundary 903 noise position.

Claims (7)

In the noise source position estimation device that estimates the source of noise generated in the measurement target,
A measurement unit that obtains a time change of the noise waveform of the measurement object in a plurality of measurement locations in the same time domain;
The electromagnetic field analysis is calculated when the plurality of measurement points are signal source positions and the time-varying waveform is the excitation waveform of the signal source, and the peak of the electromagnetic field intensity is detected with respect to the calculation result. A noise source position estimation apparatus comprising: an arithmetic unit that determines a detected position as a noise source.   The noise source position estimation apparatus according to claim 1, wherein the calculation unit reverses the noise waveform acquired by the measurement unit with respect to time and performs electromagnetic field analysis that goes back in time.   The noise source position estimation apparatus according to claim 1, wherein the plurality of measurement points are three or more.   4. The noise source position estimation according to claim 1, wherein the calculation unit obtains the electromagnetic analysis time-out period from a size and a dielectric constant of the measurement target. 5. apparatus.   5. The noise source according to claim 1, wherein the calculation unit obtains a space cell size and a time step size of the electromagnetic field analysis model from the measurement waveform. 6. Position estimation device.   The said calculating part hold | maintains the physical structure and electrical property of the said measuring object as a model of the said electromagnetic field analysis from CAD data and additional information, The any one of Claims 1-5 characterized by the above-mentioned. Noise source position estimation apparatus. Using a measurement unit that acquires the time change of the noise waveform to be measured at multiple measurement locations in the same time domain,
A computer that realizes a noise source position estimation device that estimates a generation source of noise generated in the measurement target,
The electromagnetic field analysis is calculated when the plurality of measurement points are signal source positions and the time-varying waveform is the excitation waveform of the signal source, and the peak of the electromagnetic field intensity is detected with respect to the calculation result. A noise source position estimation program that operates as a calculation unit that determines a detected position as a noise source.

Images