An electromagnetic interference measurement device and measurement method

Abstract

The present invention relates to a measuring device for electromagnetic interference, comprising: - an arbitrary waveform generator for generating a combination of N test signals and injecting them into a coupling network; - a coupling network for coupling the N test signals to the EUT and coupling its response and those signals generated by the EUT itself to a measuring unit; - a measuring unit for measuring the electrical signals provided by the coupling network; and - a processing unit for processing the N test signals and the measured electrical signals to obtain: - the electromagnetic signals, noise or EMI generated by the EUT; and - the Z, Y or S parameters of the EUT or any other meaningful set of parameters that can be calculated from the above parameters or from voltage and current. The present invention also relates to a method for measuring electromagnetic interference suitable for performing method steps with the device of the present invention.

Description

Technical Field

[0001] In a first aspect, the present invention relates to a measuring device for measuring and obtaining an electromagnetic signal or noise or electromagnetic interference (EMI) (hereinafter referred to as conducted emissions) generated by a device under test (EUT), and the Z or Y or S parameters of the EUT or any other meaningful set of parameters (hereinafter referred to as characterization parameters) that can be calculated from the above parameters or from voltage and current.

[0002] The measuring device can be used to design filters to attenuate the conducted emissions generated by the EUT or the matching network.

[0003] A second aspect of the present invention relates to a measuring method adapted to perform method steps with the device of the first aspect of the present invention. Background Art

[0004] In the prior art, the measurement of the conducted emissions generated by the EUT and the characterization parameters of the EUT are performed independently by two separate devices, for example, by means of a spectrum analyzer for interference signals and an impedance analyzer or network analyzer for the characterization parameters of the EUT.

[0005] In the prior art, the use of the above two separate devices is accepted as a necessary device for performing the measurement, but it exhibits several problems. First, if the EUT generates conducted emissions at the measurement port, its characterization through the characterization parameters of the EUT may be very poor (conducted emission interference measurement).

[0006] In addition, the use of those separate measuring devices provides measurements at different times and under different operating conditions, making it impossible to know which are the conducted emissions generated by the EUT at exactly the same time and the same operating conditions as the measurement of the characterization parameters, and vice versa.

[0007] This is a major drawback, which makes it very difficult and error-prone for technicians who need both types of measurement information (e.g., for modeling the EUT). Therefore, these two types of information are related by an estimate, which is always a source of error and hinders meeting the strict low error tolerance required by some regulations.

[0008] Some prior art documents are identified and briefly described below because they show relevant examples of some instruments and / or methods that can be used to measure excitation-related parameters of the EUT (e.g., impedance, S parameters, conversion loss of mixers, conversion efficiency, etc.).

[0009] US2002053899A1 describes a test set for measuring the S-parameters of an EUT (equipment under test) with more than two ports (multi-port) through a switch matrix placed between a signal generator, a receiver, and the EUT. The instrument includes signal processing to transform conventional S-parameters into mixed-mode S-parameters on one side and a time-domain representation (equivalent to reflectometer measurements) on the other side.

[0010] GB2466028A describes a high-frequency non-linear measurement system for analyzing the behavior of high-power and high-frequency amplifiers. The measurement system includes a multiplexer and a demultiplexer formed by filters, directional couplers, and splitters, improving on previous measurement systems that only used multiplexers and demultiplexers and had poor transmission and reflection characteristics at certain frequencies.

[0011] The paper "Measurement of Passive R, L, and C Components Under Nonsinusoidal Conditions: The Solution of Some Case Studies", Luigi Ferrigno et al., IEEE Transactions on Instrumentation and Measurement, vol. 57, no. 11, November 2008, pp2513 - 2521, describes a method that finds the values of passive R, L, and C components based on linear system identification and modal parameter estimation techniques when the measurement signal is non-sinusoidal.

[0012] WO2013127911A1 describes a method for characterizing the reflected wave of a frequency conversion device (such as mixers, in-phase / quadrature modulators, and demodulators, etc.) with at least two ports at a given frequency. The proposed method determines the frequency conversion factor of the EUT (with an integrated LO) by measuring the reflection coefficient using a one-port network analyzer while applying a known impedance and a filter for image rejection at another port of the EUT. The method requires an assumption of reciprocity between up-conversion and down-conversion.

[0013] US6356852B1 describes an interface that allows a two-port network analyzer to be connected to a multi-port EUT (i.e., having more than two test ports). The interface device has at least two stages of switches and is adapted to be coupled between the test ports of the EUT and the two-port network analyzer.

[0014] However, while the instruments or methods disclosed in the prior art documents cited above are only designed to measure some specific parameters of the EUT (such as S-parameters or frequency conversion), none of them are constructed to measure both the excitation-independent and excitation-dependent parameters of the EUT, let alone in a coherent and integrated manner.

[0015] Therefore, it is necessary to provide an alternative solution to cover the gaps found in the prior art by providing a measuring device that allows for the measurement of the conducted emissions generated by the EUT and its characterizing parameters. SUMMARY OF THE INVENTION

[0016] To this end, the present invention relates to a measuring device, comprising:

[0017] - An arbitrary waveform generator having N ports, where N is a natural number, which is configured and arranged to generate a combination of N test signals (linearly independent at all frequencies), one for each port, and to inject the generated N test signals into the N ports of a coupling network;

[0018] - The coupling network is configured to couple the N test signals from the arbitrary waveform generator to an EUT having M ports, where M is equal to, less than, or greater than N, and to couple the response of the EUT to these N test signals and those signals generated by the EUT itself to a measurement unit;

[0019] - The measurement unit having at least N ports is configured and arranged to measure the electrical signals provided by the coupling network; and

[0020] - A processing unit, configured and arranged to process the N test signals and the measured electrical signals to obtain:

[0021] - The conducted emissions generated by the EUT at at least some of its ports; and

[0022] - The characterizing parameters of the EUT.

[0023] As described above, the EUT may have fewer than N ports (in which case some ports of the EUT will remain unused) or more than N ports, in which case some ports of the EUT will remain unmeasured. Based on this understanding, in this section, the EUT will be assumed to be an M-port device.

[0024] For a preferred embodiment of the measuring device according to the first aspect of the present invention, the arbitrary waveform generator is configured and arranged to generate the combination of the N test signals (simultaneously or sequentially) from a discrete sequence of length L, which has an autocorrelation where x * denotes the complex conjugate, [l + n] LIndicates that for n ≠ 0, the modulus outside the origin is less than or equal to cyclic shifts, and cross-correlation the modulus of which has a modulus less than or equal to modulus.

[0025] According to some embodiments, the measurement unit has N, 2N or 3N ports.

[0026] For the measuring device according to the first aspect of the present invention, in one embodiment, the arbitrary waveform generator is configured and arranged to simultaneously generate a combination of the N test signals and / or simultaneously inject the generated N test signals into the N ports of the coupling network, and wherein:

[0027] - The measurement unit is configured and arranged to simultaneously measure the electrical signals provided by the coupling network; and

[0028] - The processing unit is configured and arranged to process the N test signals and the measured electrical signals to simultaneously obtain:

[0029] - The conducted emissions generated by the EUT at at least some of its ports; and

[0030] - Characterization parameters of the EUT.

[0031] For an alternative embodiment, the arbitrary waveform generator is configured and arranged to inject the generated N test signals into the N ports of the coupling network, and wherein:

[0032] - The measurement unit is configured and arranged to sequentially measure the electrical signals provided by the coupling network;

[0033] - The processing unit is configured and arranged to process the N test signals and the measured electrical signals to sequentially obtain:

[0034] - The conducted emissions generated by the EUT at at least some of its ports; and

[0035] - Characterization parameters of the EUT.

[0036] According to the embodiment, the foregoing N test signals are single-tone signals or chirp signals or modulated signals or pulse signals or impulse signals or broadband signals that cover the frequency range to be measured.

[0037] For a preferred implementation of the embodiment where the N test signals are pulses, they form a pseudo-noise (PN) sequence signal.

[0038] According to another embodiment, the processing unit includes processing means for processing the received measured electrical signal using a correlation technique with the N injected test signals to separate data representative of the conducted emissions generated by the EUT from data representative of the characterization parameters of the EUT.

[0039] For one embodiment, the coupling network includes a line impedance stabilization network (LISN) channel configured and arranged to:

[0040] - Electrically couple the AC power supply to at least some ports of the EUT; and

[0041] - Electrolytically decouple the arbitrary waveform generator and the measurement unit from the AC power network.

[0042] According to one embodiment, the processing unit is configured to calculate a modal decomposition of the data representative of the above-mentioned measured electrical signal.

[0043] For one embodiment, the processing unit includes the EMC (electromagnetic compatibility) detector (peak, quasi-peak, and average detector) directly applied to the modal decomposition of the data representative of the aforementioned measured electrical signal.

[0044] For one embodiment, the signal generator is configured to generate and inject N test signals, the period of which is less than the switching period of the EUT to which it is connected or to be connected, to characterize the changes over time of the conducted emissions generated by the EUT and the characterization parameters of the EUT, either because the signal generator is adapted to operate only with EUTs having a known switching period that is always greater than the period provided by the signal generator, or preferably because the signal generator, in particular the period of the test signal, can be adapted to multiple different switching periods of different EUTs.

[0045] In this sense, how to directly obtain the information required for the EUT from the measurement results (b 2M and b 4M ) is disclosed in the later part of this article. For example, if the measured EUT has a switched-mode power supply at its ports, the switching period can be easily extracted from a single measurement of the conducted emissions (the first harmonic in the spectrum of these emissions provides the switching speed). Therefore, the instrument does not need to have preliminary information about the EUT (although this case is also included in other embodiments of the present invention), but in order to perform the measurement of the conducted emissions, detect the first harmonic of the emission, and then inject a PN sequence (or other types of excitation) suitable for measuring the varying impedance of the specific switched-mode power supply. This also applies to other types of switching devices, such as AC-AC, AC-DC, DC-AC, and DC-DC converters.

[0046] It should be emphasized that measurement is a complex process in which the instrument may have to interact with the EUT several times in order to fully characterize it. At each iteration, even in the case of time variations as described above, the instrument can generate different kinds of excitations (V g ) to find the characteristics of the EUT that allow S EUT and V N to be fully characterized (see the description of these parameters in the later part of this article).

[0047] According to one embodiment, the processing unit is configured and arranged to process the N test signals and the measured electrical signals, and also to design a filter to attenuate the conducted emissions generated by the EUT.

[0048] As another embodiment alternative or supplementary to the above embodiment, the processing unit is configured and arranged to process the N test signals and the measured electrical signals, and also to design a matching network for optimizing the best transfer of the conducted emissions generated by the EUT.

[0049] The present invention also relates in a second aspect to a measurement method, comprising:

[0050] a) generating test signals and injecting the test signals into at least some of the ports of the EUT;

[0051] b) after the test signals have been injected into the at least some of the ports of the EUT, receiving (i.e., measuring) electrical signals from the at least some of the ports,

[0052] c) measuring (i.e., processing / calculating) simultaneously or sequentially in the received electrical signals:

[0053] - the conducted emissions generated by the EUT at at least some of its ports; and

[0054] - the characterization parameters of the EUT.

[0055] Preferably, the method of the second aspect of the present invention includes using the measurement device of the first aspect of the present invention to perform the method steps, wherein:

[0056] - step a) includes generating a combination of the N test signals as the test signals by the arbitrary waveform generator, at least one for each port, and injecting them into the ports of the EUT simultaneously or sequentially through the coupling network;

[0057] - step b) includes receiving the electrical signals through the coupling network, the electrical signals including the response of the EUT to the N test signals and the signals generated by the EUT itself; and

[0058] step c) includes:

[0059] - Measure the electrical signals provided by the coupling network simultaneously or sequentially with the measurement unit; and

[0060] - Process the N test signals and the measured electrical signals with the processing unit to simultaneously or sequentially obtain:

[0061] - Conducted emissions generated by the EUT at at least some of its ports; and

[0062] - Characteristic parameters of the EUT.

[0063] For another embodiment, the method of the second aspect of the present invention includes using the measuring device of the first aspect of the present invention to perform method steps, wherein:

[0064] - Step a) includes generating the combination of the N test signals as the test signals by the arbitrary waveform generator, and injecting the combination of the N test signals into at least some ports of the EUT simultaneously or sequentially through the coupling network;

[0065] - Step b) includes measuring the electrical signals provided by the coupling network simultaneously or sequentially with the measurement unit, including the response of the EUT to the N test signals and the signals generated by the EUT itself; and

[0066] Step c) includes:

[0067] - Processing the N test signals and the measured electrical signals with the processing unit to simultaneously or sequentially calculate:

[0068] - Electromagnetic signals or noise or electromagnetic interference (EMI) generated by the EUT at at least some of its ports; and

[0069] - Z or Y or S parameters of the EUT or any other meaningful set of parameters that can be calculated from the above parameters or from voltage and current.

[0070] According to an embodiment of the method of the second aspect of the present invention, the method includes:

[0071] - Constructing a circuit model or modal model of the EUT; and

[0072] - Designing an optimal filter or matching network and / or its components by predicting the level of conducted emissions generated by the EUT when the constructed circuit and / or modal model is virtually connected to a filter or matching network component and simulating its operation.

[0073] For an embodiment of the method according to the second aspect of the present invention, the step of designing the optimal filter further comprises performing an optimization process in order to reduce the number of filter components to be virtually connected to and combined with the constructed circuit model and modal model and simulated therewith.

[0074] According to an implementation manner of this embodiment, the optimization algorithm includes at least one of the following algorithms or a combination thereof: genetic algorithm, gradient algorithm, conjugate gradient algorithm, and Broyden-Fletcher-Goldfarb-Shanno algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Hereinafter, some preferred embodiments of the present invention will be described with reference to the accompanying drawings. They are provided for illustrative purposes only and do not limit the scope of the present invention.

[0076] Figure 1 is an EUT model that can be generated by an instrument using the measuring device according to the first aspect of the present invention, which includes: a) an N-port EUT; b) a Thevenin equivalent model; c) a Norton equivalent model.

[0077] Figure 2 is a block diagram of the measuring instrument / device according to the first aspect of the present invention electrically coupled to the EUT and the power line network according to an embodiment.

[0078] Figure 3 is a block diagram of a model of the measuring instrument according to the first aspect of the present invention according to an embodiment.

[0079] Figure 4 is for analyzing Figure 3 the definition of the normalized wave in the frequency domain of the model.

[0080] Figure 5 is a two-port EUT model: the S-parameter matrix S EUT characterizes the internal impedance of the EUT, and the two voltage sources (V n1 and V n2 ) characterize the conducted emissions generated by the EUT, for the embodiments used in the examples of methods A and B described below.

[0081] Figure 6 shows the S-parameters of the EUT for the EUT model used in Figure 5 the examples of methods A and B.

[0082] Figure 7 shows for Figure 5 the EUT model used in the examples of methods A and B, the spectral amplitudes of the two noise voltage sources. For comparison purposes, the CISPR limit for type A devices is plotted.

[0083] Figure 8 shows for the example of method A, at Vg1 and V g2 The signal (1)b measured in the case of turn-off 2M0 and (2)b 4M0 .

[0084] Figure 9 Shows the same example for method A, the signals (1)b g1 measured when V is on and V g2 is off, and (2)b 2M1 and (2)b 4M1 .

[0085] Figure 10 Shows for Figure 8 and Figure 9 the same example of method A as shown, the signals (1)b g1 measured when V is off and V g2 is on, and (2)b 2M2 and (2)b 4M2 .

[0086] Figure 11 Shows for Figure 8 , 9 and 10 the comparison of the four original and estimated S-parameters (S EUT ) of the EUT for the same example of method A.

[0087] Figure 12 Shows Figures 8 to 11 the comparison of the original voltage noise source and the estimated voltage noise sources V n1 and V n2 for the same example of method A.

[0088] Figure 13 is the first sample of the MLS sequence of the 32767 chip used in the example of method B.

[0089] Figure 14 is the comparison result of b 2M,1 and b 2M,2 between the matrix obtained by the previous method and the matrix obtained by the PN sequence.

[0090] Figure 15 is the comparison between the actual S-parameters and the estimated S-parameters of the EUT.

[0091] Figure 16 is the comparison between the interference sources of the EUT and the estimated interference sources (V n1 and V n2 ).

[0092] Figure 17Four different examples of the coupling network of the measuring device according to the first aspect of the present invention are shown, for different embodiments, in particular: a) voltage follower; b) galvanometer; c) transformer; d) directional coupler. Detailed Description

[0093] In this section, some working embodiments of the measuring device according to the first aspect of the present invention and the different signals involved in its operation will be described with reference to the accompanying drawings.

[0094] The following description relates to embodiments of the device / method of the present invention for performing sequential measurements of conducted emissions and impedance (Method A) and simultaneous measurements thereof (Method B).

[0095] Measurement step for performing sequential or simultaneous measurements of conducted emissions and impedance:

[0096] The embodiments described above for the measuring device according to the first aspect of the present invention allow the calculation of the conducted emissions and characterization parameters of the EUT. These can be combined to obtain a general equivalent Thévenin / Norton model of the EUT. Herein, by a general Thévenin / Norton equivalent, one can understand the characterization of any EUT of the form y = Ax + y0 ( Figure 1 .a), where y is a column vector of electrical values as a response, x is a column vector of electrical values as an excitation, the matrix A encompasses the response of the device to these excitations (including its characterization parameters), and y0 includes the effect of conducted emissions. The electrical values can be any combination of voltage and current, and x, y, and y0 can be understood as including the time or frequency characterization of the electrical values (if time-domain values are used, Ax must be understood as a matrix convolution). Thus, the matrix A can be, among others, any of the commonly used parameters (scattering (S) parameters, impedance parameters (Z), admittance parameters (Y), chain scattering or chain transfer (T) parameters, hybrid (H) parameters, chain (ABCD) parameters, etc.), and the column vector y0 can represent a series voltage source, a parallel current source, a wave source, etc. In the graphical representation of these parameters, elements from different representations are usually mixed, as Figure 1 .b shows, where a series voltage source naturally fits the EUT characterization using the Z-parameter matrix, or in Figure 1 .c, where a parallel current source naturally fits the characterization using the Y-parameter matrix. Since most matrix characterizations can be converted into each other, for example Figure 1 .b or Figure 1 .c, those mixed graphical representations are also possible. To measure the equivalent Thévenin / Norton model of the EUT, the instrument measures the conducted emissions and characterization parameters of any EUT connected to it simultaneously or sequentially, and uses this information to construct its equivalent model.

[0097] According to an embodiment of the measuring device according to the first aspect of the present invention, in Figure 2The block diagram of an instrument that can perform these measurements is shown. The EUT (Equipment Under Test) is designed to be measured at N ports (if the EUT has fewer than N ports, some of the EUT's ports will remain unused; if the EUT has more than N ports, some of the EUT's ports will remain unmeasured; based on this understanding, hereinafter, for the embodiments described herein, the EUT will be assumed to be an N-port device). The instrument includes an N-port arbitrary waveform generator; a kxN-port measurement unit, where k is typically 1, 2, or 3; and N coupling networks that inject signals dependent on the signals generated by the arbitrary waveform generator into the ports of the EUT and inject signals dependent on the EUT's response to the above excitation into the kxN ports of the measurement unit. The processing unit will perform most of the calculations detailed below. In Figure 2 In the specific embodiment shown, the coupling network includes the circuitry of a typical LISN channel because it is designed to characterize the mains or power supply terminals of the EUT. In different embodiments of the present invention designed to characterize other types of terminals, there will be no coupling network connected to the mains through an LISN channel.

[0098] The arbitrary waveform generator and the measurement unit can operate in a baseband configuration or include mixers, upconverters, downconverters, etc. The measurement unit contains kxN signal measurement devices, which can be physical or equivalent (a multiplexing mode can be used if needed).

[0099] The processing unit can be embedded in the physical instrument or hosted on an external PC or in the cloud.

[0100] The coupling network can be in various configurations, none of which refer to a switch matrix. For example, using power dividers and directional couplers, impedance bridges, circulators, voltage or current probes, etc. This definition means that in such a coupling network, all ports are always interconnected (contrary to what may happen in a switch matrix with more inputs than outputs (or vice versa), where only those ports in the switch positions are interconnected).

[0101] To demonstrate the feasibility of the instrument, it can be as Figure 3It is modeled as shown and described below. In this analysis, a four-port coupled network is assumed, although a three-port coupled network is sufficient for the analysis performed. In a practical implementation, the fourth port can be used, for example, with an appropriate coupled network to sense the level of the signal injected by an arbitrary waveform generator. In the following analysis, the coupled network is very general. The analysis is performed in the frequency domain. Since any signal admits a time-domain or frequency-domain characterization, the analysis performed is general. For the purposes of the analysis, normalized waves are used, but it should be understood that the instrument can measure other kinds of electrical signals (which can be represented as combinations of normalized waves). For the following analysis, port 1 of the coupled network has a reference impedance equal to the internal impedance of the corresponding arbitrary waveform generator port. Ports 2 and 4 have reference impedances equal to the input impedances of the corresponding measurement unit ports (assuming a value of k = 2, although a similar analysis can be performed for other values of k).

[0102] Figure 2 The N signal generators of the block diagram of can be characterized, without loss of generality, by their open-circuit voltages V gi and internal impedances and the k·N measurement unit ports (in Figure 3 , 2N) can be characterized, without loss of generality, by their input impedances and . The arbitrary waveform generators and measurement unit devices are coupled to the EUT ports by means of N coupled networks, which, without loss of generality, are again characterized by their S-parameter matrices, .

[0103] Without loss of generality, the following analysis is performed under the assumption of two measurement unit ports for each EUT port. The analysis can also be performed for any number of measurement unit devices per EUT port (k). In particular, the case of k = 1 can be easily considered by setting the relevant S-parameters of the relevant matrix equal to 0.

[0104] As Figure 4 shown, the following analysis has been performed using a very general definition of normalized waves (and thus, S-parameters). The parameters k and Z0 used in the definition of the normalized waves are indicated below the ports. It can be seen that, for simplicity of calculation, the values of the different parameters Z0 used at ports 1, 2, and 4 of the N coupled networks are equal to the internal impedance of the arbitrary waveform generator and the input impedance of the measurement unit ports. The values of k and Z0 at its port 3 are set to accommodate the desired wave definition at the port of the EUT.

[0105] Let it be the following column vectors:

[0106]

[0107] where = 1, ..., 4. If the S-parameter matrices of the EUT and the coupling network are

[0108] and

[0109] where i = 1, ..., N, making it a diagonal matrix

[0110]

[0111] where i = 1, ..., 4, j = 1, ..., 4. Finally, making it a diagonal matrix

[0112]

[0113] where j = 1, ..., 4

[0114] Then,

[0115] b = S EUT a

[0116] a 1M = K 1M V g

[0117] a 3M = K 3M V n + b

[0118] b 3M = K 3M V n + a

[0119] b 1M = S 11M a 1M + S 13M a 3M

[0120] b 3M = S 31M a 1M + S 33M a 3M

[0121] b 2M = S 21M a 1M + S 23 a 3M

[0122] b 4M = S 41M a 1M + S 43M a3M .

[0123] From these equations, the following is followed:

[0124] b 3M =(I N S 33M S EUT ) -1 S 31M a 1M +(I N -S 33M S EUT ) -1 S 33M (I N -S EUT )K 3M V n

[0125] a 3M =S EUT (I N -S 33M S EUT ) -1 S 31M a 1M +(I N +S EUT (I N -S 33 S EUT ) -1 S 33M )(I N -S EUT )K 3M V n .

[0126] Based on these, all other waves (and thus, voltages and currents) at all ports of the circuit of Figure 3 can be easily calculated.

[0127] Based on these equations, several measurement strategies (time domain, frequency domain, mixed domain, or spread spectrum) can be envisioned.

[0128] For example, two very basic methods that can be enriched in several stages will be described as follows.

[0129] Method A :

[0130] Assume the EUT emits a fixed interference. First, when V g =0 (a 1M =0), measure the influence of V n to obtain

[0131] b 2M0 =S 23a 3M = S 23 (I N + S EUT (I N - S 33 S EUT ) -1 S 33M )(I N - S EUT )K 3M V n

[0132] b 4M0 = S 43M a 3M = S 43M (I N + S EUT (I N - S 33M S EUT ) -1 S 33M )(I N - S EUT )K 3M V n

[0133] If then measured appropriately timed (synchronized with interference or with the 50-Hz mains signal,...) with V g ≠ 0 (a 1M ≠ 0), the following waves are measured,

[0134] b 2M = S 21M a 1M + S 23M a 3M

[0135] = (S 21M + S 23M S EUT (I N - S 33M S EUT ) -1 S 31M )a 1M + b 2M0

[0136] b 4M = S 41M a 1M + S 43M a 3M

[0137] = (S 41M + S 43M S EUT (I N - S33M S EUT ) -1 S 31M )a 1M +b 4M0

[0138] If N linearly independent (at all frequencies) (column) vectors a 1M，k , k = 1, ..., N are generated and their responses are measured, the following excitation and response matrices (formed by column vectors) can be constructed,

[0139] A = [a 1M，1 … a 1M，N

[0140] B2 = [b 2M，1 -b 2M0 … b 2M，N -b 2M0

[0141] B4 = [b 4M，1 -b 4M0 … b 4M，N -b 4M0 ,

[0142] There is:

[0143] B2 = (S 21M +S 23M S EUT (I N -S 33M S EUT ) -1 S 31M )A

[0144] B4 = (S 41M +S 43 S EUT (I N -S 33M S EUT ) -1 S 31M )A.

[0145] Since A is invertible, S can be calculated from either expression EUT . For example,

[0146]

[0147] Once S EUT is known, V n can be easily calculated.

[0148] Example A: Consider using Figure 5 ​​For the case of a two-port EUT characterized by modeling (choose one, although any of the other characterizations above can be used), it consists of an S-parameter network (S EUT ) that characterizes the internal impedance of the EUT (its characterization parameter) and two noise voltage sources (V n1 and V n2 ) that characterize the interference generated by the EUT. Port 1 of the EUT is the line-ground port and port 2 is the neutral-ground port. The S-parameters used in this example (based on actual measurements) are shown in Figure 6 , and the amplitudes of the two noise voltage sources are shown in Figure 7 . Some of the CISPR limits are also drawn only for comparison purposes.

[0149] Each of the coupling networks considered for the instrument has a LISN channel of CISPR - 16 50 Ω / / 50 μH, a limiter attenuator, and a directional coupler.

[0150] The measurement steps for this case are:

[0151] a. The EUT with N ports (N being any number) is connected to the instrument and turned on.

[0152] b. The instrument measures the signals at ports 2 and 4 of the coupling network, where the two generators V g1 and V g2 are turned off. The signals measured are defined as b 2M0 and b 4M0 . Figure 8 shows this measurement.

[0153] c. The instrument measures the signals at ports 2 and 4, where the first generator is turned on (in this case V g1 ), and the second is turned off (in this case V g2 ) (an easy scheme for generating linearly independent signals). The signal used in this particular example is a signal with a flat spectrum from 9 kHz to 30 MHz, although many other signals can alternatively be used. The injected signal interacts with the EUT and is partially re-emitted together with the EUT emissions generated by V n1 and V n2 . These signals reach the measurement unit through ports 2 and 4 of the coupling network to be measured therein. Figure 9 shows the signals measured in this case.

[0154] d. In the fourth step, the instrument measures the signals in ports 2 and 4 of the coupling network, where the second generator is turned on (in this case V g2 ), and the first is turned off (in this case V g1 ), and a signal similar to the above spectrum, asFigure 10 as shown

[0155] e. Using the measurements completed in the previous step, the EUT, S EUT The S-parameters of can be calculated using Equation 1. In this example, four S-parameters are recovered, as Figure 11 shown

[0156] f. To measure the S-parameters of the EUT, the noise voltage sources can be obtained using the following formula:

[0157]

[0158] Applying this equation to our example, two voltage noise sources are perfectly recovered, as Figure 12 shown

[0159] After these five steps, all the information for constructing the Thevenin equivalent model of the EUT has been obtained.

[0160] Method B :

[0161] Now, if the excitation is a spread-spectrum signal and the signal generator generates highly uncorrelated sequences, all the above measurements can be performed simultaneously. The system will be excited simultaneously by N pseudo-noise (PN) sequences and the responses of the recorded EUT. Therefore, by performing N·N correlations on all the responses to all the PN sequences, the response column vectors as described above will be recovered, one for each excitation sequence (although this measurement and the associated correlation are in the time domain and, as mentioned before, for analysis purposes, they are characterized by their frequency-domain counterparts):

[0162] b 2M =(S 21M +S 23M S EUT (I N -S 33M S EUT ) -1 S 31M )a 1M +b 2M0

[0163] b 4M =(S 41M +S 43M S EUT (I N -S 33M S EUT ) -1 S 31M )a 1M +b 4M0

[0164] b 2M0 =S 23M a3M = S 23M (I N + S EUT (I N - S 33M S EUT ) -1 S 33M )(I N - S EUT )K 3M V n

[0165] b 4M0 = S 43M a 3M = S 43M (I N + S EUT (I N - S 33M S EUT ) -1 S 33M )(I N - S EUT )K 3M V n 。

[0166] In this case, due to the spreading effect on the correlation of signals other than the excitation PN sequence, the terms b 2M0 and b 4M0 will have low values and can generally be ignored.

[0167] Then, the matrices

[0168] A = [a 1M，1 … a 1M，N

[0169] B2 = [b 2M，1 -b 2M0 … b 2M，N -b 2M0 ≈[b 2M，1 … b 2M，N

[0170] B4 = [b 4M，1 -b 4M0 … b 4M，N -b 4M0 ≈[b 4M，1 … b 4M，N ,

[0171] can be constructed (matrix A is also constructed by appropriately recording the N·N correlation of the input PN sequence and is basically a diagonal matrix at each measurement frequency), and the S-parameter matrix of the EUT can be obtained as follows ​​

[0172]

[0173] Once S is known EUT , the interference vectors b 2M0 and b 4M0 can be recovered from the following

[0174] b 2M0 = b 2M - (S 21 + S 23M S EUT (I N - S 33M S EUT ) -1 S 31M )a 1M

[0175] b 4M0 = b 4M - (S 41M + S 43M S EUT (I N - S 33 S EUT ) -1 S 31M )a 1M ,

[0176] This time, calculations are performed directly using the PN excitation and their responses. From b 2M0 and b 4M0 , the interference vectors

[0177]

[0178] can be obtained.

[0179] This measurement scheme is provided only as an example to demonstrate that all the parameters of the (generalized) Thevenin equivalent can be measured simultaneously. As in the case of the more conventional measurement schemes described above, other measurement steps can be performed to obtain the same result. For example, the interference level can be recovered first, then the S-parameters of the circuit, or the generator can generate a superposition of PN sequences to achieve code diversity in the measurement, or the measurement of the interference and S-parameters can be performed sequentially, etc. As previously mentioned, algorithms and techniques (interpolation, multiple measurements,...) that improve the numerical accuracy of the results can be used to enrich this basic measurement scheme.

[0180] Example B: Consider the case of a DUT characterized using the model of Figure 5 with the S-parameter network (S Figure 6 ) as shown in EUT and two noise voltage sources (V Figure 7 ) as shown in n1 and Vn2 )。

[0181] Similarly, for all the coupling networks considered for the instrument, there are CISPR-16 50Ω / / 50μH LISN channels, attenuators (transient limiters), and directional couplers.

[0182] The measurement steps for this case are as follows:

[0183] a. The EUT with 2 ports is connected to the instrument and turned on.

[0184] b. The instrument generates a PN sequence. In this example, a single maximum length (MLS) sequence of 32767 chips is used simultaneously in both ports, but with a time shift of 16384 samples (to avoid overlapping interference). The measurement period is 16384 samples. Figure 13 The first 200 samples of it are shown.

[0185] c. Measure a 1M , b 2M and b 4M . These time-domain signals are correlated with the PN sequence. Figure 14 Shows the result of b 2M,1 and b 2M,2 column vectors obtained after performing the cross-correlation with the above PN sequence. A median filter has been used to smooth the influence of interference.

[0186] d. Estimation of the S parameters of the EUT using Equation 1. Figure 15 Shows the comparison between the actual S parameters and the estimated S parameters of the EUT. A median filter has been used to smooth the influence of interference.

[0187] e. Estimation of the interference source. Figure 16 Shows the comparison between the interference sources (V n1 and V n2 ) of the EUT and the estimated interference sources (V n1 and V n2 ).

[0188] The above two methods are presented only as non-limiting examples of possible measurement strategies. The present invention includes any measurement strategy that at least includes the generation and injection of the N test signals described in the previous part of this article, and at least any measurement strategy with the above autocorrelation R XX and cross-correlation R XX .

[0189] Taking into account the definition of the term coupling network given in the previous part of this article, and considering Figure 3 the same port numbers shown in Figure 17 Some examples of coupling networks are shown.

[0190] Specifically, Figure 17 .a) shows a coupling network for a measurement unit with a single port (k = 1). This coupling network consists of a voltage follower (a circuit whose output voltage follows the input voltage at all times).

[0191] Figure 17 .b) shows a coupling network for a measurement unit with two ports (k = 2). In this case, the coupling network consists of two voltage followers and small-value resistors. It allows the measurement of the voltage at the port of the EUT (port 4) and the current from the voltage drop across the resistor ends.

[0192] Figure 17 .c) shows an example of a coupling network using a transformer.

[0193] Finally, Figure 17 .d) shows an example of a coupling network using only a directional coupler. In this case, a part of the signal generated in an arbitrary waveform generator reaches the measurement unit (port 2), and a part reaches the EUT (port 3). On the other hand, the reflected signal in the EUT or its conducted emission enters via port 3 and reaches the measurement unit via port 4.

[0194] The measuring device according to the first aspect of the present invention is more complex and complete than the measuring devices known in the prior art, and its performance cannot be achieved by any of them. It not only increases the possibility of simultaneously (or sequentially) measuring Z or Y or S parameters or any other meaningful parameter set that can be calculated from the above parameters or from voltage and current, as well as the electromagnetic signals or noise or electromagnetic interference (or its conducted emission) generated by the EUT, but also, for some embodiments, constructs a Thevenin or Norton equivalent model, and as a last resort, finds the best power line filter for reducing conducted emissions. The device is designed to speed up the design and implementation of electronic EUTs, reduce their design costs, optimize their implementation, and speed up their time to market.

[0195] Those skilled in the art can introduce changes and modifications in the described embodiments without departing from the scope of the present invention, as defined in the appended claims. For example, replacing the LISN inside the above coupling network with one or more LISNs outside it.

Claims

1. A measuring device, comprising: - An arbitrary waveform generator having N ports, where N is a natural number, configured and arranged to generate a combination of N test signals, one test signal per port, and inject the N generated test signals into N ports of a coupling network; - The coupling network is configured to couple the N test signals from the arbitrary waveform generator to a device under test EUT having M ports, where M is equal to, less than, or greater than N, and couple the response of the EUT to these N test signals and those signals generated by the EUT itself to a measurement unit; - The measurement unit having at least N ports, configured and arranged to measure the electrical signals provided by the coupling network; And - A processing unit configured and arranged to process the N test signals and the measured electrical signals to obtain: - Electromagnetic signals or noise or electromagnetic interference EMI generated by the EUT at at least some of its ports; and - of the EUT or or parameters or any other meaningful set of parameters that can be calculated from the above parameters or from voltage and current.

2. The measuring device according to claim 1, wherein the arbitrary waveform generator is configured and arranged to generate the combination of the N test signals from a discrete sequence of length L, which has an autocorrelation ( ) in which denotes the complex conjugate, and denotes a cyclic shift for whose modulus outside the origin is less than or equal to 1 , and the modulus of the cross-correlation ( ) has a modulus less than or equal to 1 .

3. The measuring device according to claim 1 or 2, wherein the measurement unit has N, 2N, or 3N ports.

4. The measuring device according to claim 1 or 2, wherein the arbitrary waveform generator is configured and arranged to simultaneously generate the combination of the N test signals and / or simultaneously inject the N generated test signals into the N ports of the coupling network, and wherein: - The measurement unit is configured and arranged to simultaneously measure the electrical signals provided by the coupling network; And - The processing unit is configured and arranged to process the N test signals and the measured electrical signals to simultaneously obtain: - Electromagnetic signals or noise or electromagnetic interference EMI generated by the EUT at at least some of its ports; and - of the EUT or or parameters or any other meaningful set of parameters that can be calculated from the above parameters or from voltage and current 5. The measuring device according to claim 1 or 2, wherein the arbitrary waveform generator is configured and arranged to at least sequentially inject the N generated test signals into the N ports of the coupling network, and wherein: - The measurement unit is configured and arranged to sequentially measure the electrical signals provided by the coupling network; And - The processing unit is configured and arranged to process the N test signals and the measured electrical signals to sequentially obtain: - Electromagnetic signals or noise or electromagnetic interference EMI generated by the EUT at at least some of its ports; and - of the EUT or or parameters or any other meaningful set of parameters that can be calculated from the above parameters or from voltage and current 6. The measuring device according to claim 1 or 2, wherein the above N test signals are single-tone signals or chirp signals or modulated signals or pulses or impulses or broadband signals covering the frequency range to be measured.

7. The measuring device according to claim 1 or 2, wherein the processing unit comprises processing means for processing the received measured electrical signals using a correlation technique of the N injected test signals to separate data representative of the electromagnetic signals or noise or electromagnetic interference EMI generated by the EUT from data representative of the or or parameters or any other meaningful set of parameters that can be calculated from the above parameters or from voltage and current.

8. The measuring device according to claim 1 or 2, wherein the coupling network includes a line impedance stabilization network LISN channel configured and arranged to: - Electrically couple an AC power supply to the ports of the EUT; and - Decouple the arbitrary waveform generator and the measurement unit from the AC power network.

9. The measuring device according to claim 1 or 2, wherein the processing unit is configured to calculate a modal decomposition of data representative of the measured electrical signals.

10. The measuring device according to claim 1 or 2, wherein the processing unit comprises an EMC detector directly applied to the modal decomposition data of the electrical signal representative of the above-mentioned measurement.

11. The measuring device according to claim 1 or 2, wherein the waveform generator is configured to generate and inject N test signals having a period less than the switching period of the EUT to characterize the electromagnetic signals or noise or electromagnetic interference EMI generated by the EUT and the or or variation over time of the parameters or any other meaningful set of parameters that can be calculated from the above parameters or from voltage and current.

12. The measuring device according to claim 1 or 2, wherein the processing unit is configured and arranged to process the N test signals and the measured electrical signal, and is further configured to design: - a filter for attenuating electromagnetic signals or noise or electromagnetic interference EMI generated by the EUT; and / or - a matching network for optimal transfer of electromagnetic signals generated by the EUT.

13. A measurement method, comprising: Step a) generating test signals and injecting the test signals into at least some of the ports of the EUT; Step b) after the test signals have been injected into at least some of the ports of the EUT, receiving electrical signals from the at least some of the ports, Step c) measuring simultaneously or sequentially in the received electrical signals: - electromagnetic signals or noise or electromagnetic interference EMI generated by the EUT at at least some of its ports; and - of the EUT or or parameters or any other meaningful set of parameters that can be calculated from the above parameters or from voltage and current.

14. The measurement method according to claim 13, comprising performing the method steps using the measuring device according to any one of claims 1 to 12, wherein: - Step a) includes generating the combination of the N test signals as the test signals by the arbitrary waveform generator, and injecting the combination of the N test signals into at least some of the ports of the EUT simultaneously or sequentially through the coupling network; - Step b) includes receiving the electrical signals through the coupling network, the electrical signals including the response of the EUT to the N test signals and the signals generated by the EUT itself; and - Step c) includes: - measuring the electrical signals provided by the coupling network simultaneously or sequentially with the measuring unit; and - processing the N test signals and the measured electrical signal with the processing unit to simultaneously or sequentially obtain: - electromagnetic signals or noise or electromagnetic interference EMI generated by the EUT at at least some of its ports; and - of the EUT or or parameters or any other meaningful set of parameters that can be calculated from the above parameters or from voltage and current.

15. The measurement method according to claim 13, wherein: - in Step b), the receiving step is a measurement step for measuring the electrical signals from at least some of the ports after the test signals have been injected into at least some of the ports of the EUT; and - in Step c), the simultaneously or sequentially measuring step refers to a processing / calculation step for processing / calculating simultaneously or sequentially from the measured electrical signals: - electromagnetic signals or noise or electromagnetic interference EMI generated by the EUT at at least some of its ports; and - of the EUT or or parameters or any other meaningful set of parameters that can be calculated from the above parameters or from voltage and current.

16. The measurement method according to claim 15, comprising performing the method steps using the measuring device according to any one of claims 1 to 12, wherein: - Step a) includes generating the combination of the N test signals as the test signal by the arbitrary waveform generator and injecting the combination of the N test signals into at least some of the ports of the EUT simultaneously or sequentially through the coupling network; - Step b) includes measuring the electrical signals provided by the coupling network simultaneously or sequentially with the measurement unit, including the response of the EUT to the N test signals and the signals generated by the EUT itself; And - Step c) includes: - Processing the N test signals and the measured electrical signals by the processing unit to calculate simultaneously or sequentially: - Electromagnetic signals or noise or electromagnetic interference EMI generated by the EUT at at least some of its ports; and - of the EUT or or parameters or any other meaningful set of parameters that can be calculated from the above parameters or from voltage and current.

17. The measurement method according to any one of claims 13 to 16, comprising: - Constructing a circuit model or a modal model of the EUT; And - Designing an optimal power filter or matching network and / or its components by predicting the level of electromagnetic signals or noise or EMI generated by the EUT when the constructed circuit model and / or modal model is virtually connected to an electronic filter or a matching network component and its operation is simulated.

Images