Phase Comparison of High Current Shunts up to 100 khz

Transcription

1 Accepted for publication in IEEE I&M 1 Phase Comparison of High Current Shunts up to 100 khz ian Carlo Bosco, Martin arcocz, Kåre Lind, Umberto Pogliano, ert Rietveld, Valter Tarasso, Boštjan Voljč. and Věra Nováková Zachovalová. Abstract - A comparison of the phase error of existing shunts for currents between 10 A and 100 A and frequencies from 500 Hz to 100 khz has been organized. The measurements were performed at INRIM on some shunts using the first prototype of a new type of phase comparator and a step-up method for the data processing. The results of the relative measurements have shown a good repeatability. The method for the evaluation of the reference value was based on similar shunts and suitable assumptions and derived by extrapolation. For the type of the resistance elements employed in these shunts, the derivation of the reference is not very accurate and still needs to be improved. Index Terms Measurement, Electrical variable measurement, Measurement standards, Current measurement, Phase measurement T I. INTRODUCTION HE measurement of power and power quality with high accuracy needs a lot of skill and competence in a wide area of ac measurements. So, a co-operation activity was undertaken between many national metrological institutes (NMIs) in the framework of a imera Plus project The Next eneration of Power and Energy Measurements. In this project it was recognized that the current state-of-the-art in ac power metrology was mainly involving sinewave signals and that NMIs were not adequately equipped to underpin a regulatory system for power quality. For this purpose some Manuscript received June 5, The research presented in this paper is part of the EURAMET joint research project on "Power and Energy" and has received funding from the European Community's Seventh Framework Programme, ERA-NET Plus, under rant Agreement No C. Bosco and U. Pogliano are with the Electromagnetics Division of the Istituto Nationale di Ricerca Metrologica (I.N.RI.M), Strada delle Cacce 91, 10135, Torino, Italy. M. arcocz is with the Bundesamt für Eich und Vermessungswesen (BEV), Arltgasse 35, Vienna, Austia. K. Lind is with the Justervesenet-Norwegian Metrology Service, Fetveien 99, N-2007 Kjeller, Norway.. Rietveld is with the Van Swinden Laboratorium (VSL), Thijsseweg 11, 2629 JA Delft, The Netherlands. V. Tarasso is with the Technical Research Institute of Sweden (SP), Box 857, SE , Boras, Sweden. B. Voljč. is with the Slovenian Institute of Quality and Metrology (SIQ), Trzaska 2, 1000 Lubjana, Slovenia. V. N. Zachovalová is with.český Metrologický Institut (CMI), Okruzni 31, Brno, Czech Republic. Correspondent author U. Pogliano, phone: ; fax: ; u.pogliano@inrim.it. technological objectives were defined, such as for example to accurately measure ac voltage and current waveforms that are highly distorted, contain discontinuities and are non-stationary. An essential part of ac metrology for power measurement, mainly where digitizers are employed, is scaling the voltage signals and converting current into voltage from the working levels to values convenient for the inputs of the digitizer. The transducers used for these aims must be stable and insensitive to temperature and signal level. They also need a phase error that is both low and stable within the bandwidth of interest. For current measurements the project aim is to develop shunts with different current ranges, having high stability of impedance with load current, temperature and time, starting from the technology already available. The first type of technology is the so-called cage designs, where a number of resistive elements are connected in parallel, in a cylindrical in cage-like design. This configuration has the advantages of reducing the self-inductance by connecting in parallel the resistive elements and of dissipating efficiently the heat produced. The second technology for shunt construction utilizes a coaxial structure of resistive alloys in metal foils arranged in coaxial structures which minimizes the space between the current paths, so reducing the inductance that, for these shunts, is the main source of error and phase shift. Analytical models can lead to proper design, with minimization of stray parameters and optimization of these shunts for absolute measurements of current and phase angle. However, beside models, there is always the need of an experimental verification. So, the measurement of the impedance components of the shunts and their changes with frequency is required and methods for such characterization have been undertaken in this study. In particular, while other laboratories have investigated other types of problems, at INRIM the measurements of the phase for high current shunts (over 10 A) has been considered and a phase comparator has been developed. The characterization performed and reported in this paper is related to the existing shunts, in the range form 10 A to 100 A, built by different laboratories [1], [2], [3], [4], [5], [6]. The aim of this comparison, beside the construction of the reference for phase (as for example in [7]), is to establish a connection between the analytical models and the resulting devices. Then, in the following part of the project, the measurements will be applied to the characterization of the shunts produced in the project.

2 Accepted for publication in IEEE I&M 2 A. The phase comparator II. COMPARISON METHOD The measurements of phase differences between two shunts were performed by means of a phase comparator, which was specifically built for this purpose and which is a first simplified version of a model studied at INRIM for this project. AC CURRENT ENERATOR SH2 SH1 The design of the phase comparator has already been partially reported in [8], and it is going to be more extensively described in [9]. Its basic scheme, shown in Fig. 1, contains an ac current generator (built by connecting an AC calibrator to the input of a transconductance amplifier). The resulting generator is suitable for currents up to 100 A in the frequency range up to 100 khz. It supplies the inputs of the two shunts under comparison connected in series. The outputs of the two shunts are connected to a two-input phase analyzer through two active guarded transformers (ATs) employed as wideband decoupled precision voltage transmitters. The two ATs, of identical construction, with relatively low phase errors (about 350 rad at 100 khz), are made as double stage transformers with two separate screens between the primary and the secondary windings. One screen is connected to one side of the input, while the other one is connected to the reference ground of the output. A buffer supplied by rechargeable batteries drives the magnetizing winding. In the first prototype used for this comparison the phase detector was built using a commercial board put inside the computer. Two synchronized inputs of this board, operating at a sampling rate up to 1 Msamples/s with 16 bits resolution are employed for the phase determination. The phase relation of the two signals between the outputs of the active transformers and the inputs of the phase detector is maintained by connectors and balanced cables. The current generator and the digitizing board are controlled by a computer that sets the desired currents at the suitable frequencies and drives the digitizing board to acquire the samples, which are subsequently processed to derive the phase difference between the two inputs. B. The program for the determination of the phase A software program was developed for acquiring and processing the samples of the two signals at the output of the AT AT Fig. 1 Basic circuit of the comparison system. SH1 and SH2 denote the two shunts being compared. Two active guarded transformers (ATs) are employed as wideband decoupled precision transmitters. IN2 DIITAL PHASE ANALYSER IN1 shunts. Here??? the sampling rate, for the best phase resolution, is generally near the maximum (about 1 Msample/s), the samples are simultaneously acquired in a selected number, as consecutive bursts of nominally simultaneous samples from the two inputs. For each burst the phase is evaluated by an asynchronous method evaluating first the frequency throughout the zero transitions [9] and, then, the phase by the best fit on the fundamental. The mean and the standard deviation of all burst of a group of measurement are eventually computed and produced as a result of the measurement for each current and frequency. A. Data acquisition III. MEASUREMENT PROCEDURE The procedure for measurements employed for this comparison was developed after some preliminary investigations with the data obtained in experimental tests, made by means of the phase comparator. The repeatability of these results was analyzed in order to take the greatest advantage of the time employed in the process. Then, the procedure for the measurement of the phase difference between two shunts for this specific comparison was set up and standardized. 1) For each pair of shunts under comparison, they are firstly put in series, with the external screen of one of them connected to the ground potential and the other one with its screen connected to the opposite side of the current generator, as in Fig. 2a. In this connection an exact opposite current is applied to the High terminals of the two shunts and the stray capacitive currents produced by the screen voltages do not flow within the shunts, with only a relatively small effect on the phase of the shunt output (if the screen resistance is low). SH2 SH1 AT2 a) AT1 b) SH1 SH2 2) Three sets of measurement samples are taken for each current and for each frequency planned for this couple of shunts. Every set consists of 38 bursts of consecutive samples for both voltages at the outputs of the two shunts. The sample rate for all measurement frequencies is 800 ksample/s. The two series of data are acquired and stored in the computer and, then, with a AT2 AT1 Fig. 2 a and b. Connections of the shunts SH1 and SH2 with the two active guarded transformers (ATs) at the input of the phase comparator, in the two successive measurements for the compensation of the systematic errors

3 Accepted for publication in IEEE I&M 3 proper software, the controller of the phase comparator evaluates the values of the phase differences for every burst [9]. 3) The same shunts are then put in series in the reversed order, see Fig. 2b, and the same inversion is also applied in the connections of the external screens of the ATs. In this way the voltages applied to the guards of the ATs are the same for both measurements and the phase shifts of each of them does not change. Finally, three other sets of measurements are taken with the same methodology as before. After the measurement, the phase differences for each burst are stored in a file and a first processing is performed by an automatic program assisted by the operator. During this processing the operator inspects the graphs of the measurement results for all bursts to see the consistency of the data and the possible presence of outlayers. If any of them are found, they are removed from the results. This was found necessary especially for higher frequencies where, for the noise, sometimes the program for the identification of the frequency fails and, consequently, produces a wrong value of the phase. This first processing evaluates the mean values for each measurement set and for the three sets together and the related standard deviations. B. Compensation of systematic errors Three sets of measurements, with an inversion in the series of the shunts, have been introduced in order to cancel some of the systematic errors in the comparator and the non negligible phase errors of the ATs. In the measurement process the data are taken with the shunts put in series in a given order and, then, in the reversed order. In this way, the phase differences of the signals measured by in each position ( A and B ) are related to the phase of the two shunts ( SH1 and SH2 ) by the relations [8]: A B S H2 S H1, (1) 2 while the quantity: A B (2) 2 is the phase difference between the two channels and it is specific for the phase comparator. Its stability for a given frequency indicates the good functioning of the measurement system. Its value depends on the frequency, it is approximately linear with frequency and in the worst conditions at 50 A and 100 khz is about 1.5 mrad with a spread of variations, when the measurement works correctly, contained in 130 rad (standard deviation). Eventually, the phase difference of the two-shunt comparison is computed by relation (1) and its standard deviation is obtained by the square root of the quadratic composition of the components. IV. DATA PROCESSIN AND UNCERTAINTY EVALUATION A. Comparison between the shunts at the same current The differences of the phases between two shunts and the standard deviations resulting from the measurements performed by the procedure in the previous chapter are the starting point for processing the results and deriving the uncertainty budget. Additional components of uncertainty (type B) for each measurement link has been derived from specific tests performed on the comparator [9]. The possible effect of the Tee was tested for some measurements at 20 A by using two very different models: one built by about 50 cm of coaxial cable connected with a brass box and two N connector, the other one, for high currents, built by a quasi-coaxial structure (internal part in copper and external part in aluminum) with LC connectors (for connection proper LC-N adapters have been used). The differences in the results were found in the normal spread of the results and a specific uncertainty component for the Tee was not taken into account in the budget. The data processing on the whole measurement set at the same current is aimed at consolidating the results to satisfy the additional requirement introduced by the coherence principle, which demands that, in all closed loops, the sum of the phase differences is zero. These additional constraints allow to adjust the resulting values by means of a suitable optimization performed by a best fit which also can estimate the uncertainties as a function of the frequency. For this purpose after the collection of measurement data between all shunts, the data were organized in separate groups one group for each current value and for every group a diagram of the comparison links was derived from the measurement data. For each frequency, the values are adjusted by a best fit to obtain a null sum of the residuals in every closed loop of these diagrams, minimizing the quadratic sum of the displacements. The evaluation of the uncertainty is made a priori from the means of the variance divided by the number of degrees of freedom in the comparison [10]. A successive evaluation can be made a posteriori from the quadratic mean of the displacements. The maximum of these two values, multiplied by a proper coverage factor, is taken as the expanded uncertainty for the phase difference of one shunt relative to others of the same group. Specifically, for all the four measurement groups considered, the a-priory estimated uncertainties were greater than the a-posteriori ones, which indicate that main uncertainty components taken into account, were not underestimated. B. Linking the phase of the shunts at different current To link the phase differences of the shunts with different nominal current, a typical step-up procedure, widely used in ac-dc transfer difference measurement, has been utilized. In a group of measurements at the same current the shunts that are considered are generally those that have the same nominal current. However some additional shunts with higher nominal current are also introduced into the group. These shunts are also associated with a measurement group of higher current. This allows us to link the values of the phase differences,

4 Accepted for publication in IEEE I&M 4 between measurements made at different current. For the determination of the relative value of the phase difference, one of the shunts is taken as a preliminary reference and the optimization inside the group of the measurements a step-up and a step-down is performed. For each step a specific shunt is assumed to have the same phase difference at two currents with an additional uncertainty that can be evaluated by the relative changes of the shunts measured at the two currents. between the two groups in the step-up. The same happens also for the shunt V1240A2, which is used only for the determination of the step-up uncertainty. An analogous procedure was applied for linking the 50 A group of measurement, which is shown in Fig. 5 and the one of 100 A shown in Fig. 6. In this case, unfortunately, the number of shunts involved was only three, resulting in a smaller V. COMPARISON RESULTS A. Determination of the relative phase differences The first step for measuring the phases of the shunts was the determination of the relative phases. The shunts sent by the NMIs and those of INRIM (some of them acquired as commercial products) were considered in the investigation. Four groups of measurements were organized, respectively, at the currents of 10 A, 20 A, 50 A and 100 A. The frequencies Fig. 4 Comparisons the phase difference between the shunt at 20 A Fig. 3 Comparisons the phase difference between the shunt at 10 A the boxes represent the shunts under comparisons, while the arrows the measurement links. Fig. 5 Comparisons the phase difference between the shunt at 50 A Fig. 6 Comparisons the phase difference between the shunt at 100 A considered were: 500 Hz, 1 khz, 2 khz, 5 khz, 10 khz, 20 khz, 50 khz and 100 khz. The structure of the measurements at 10 A is shown in Fig. 3. One of the shunts (denoted by the label V1620A3) was taken as the relative reference. The same shunt, having a nominal current of 20 A, is also in the graphs shown in Fig. 4 of the measurements at the current of 20 A and is the link reduction of the uncertainties due to the optimization. By applying in sequence the concept illustrated in chapter III and IV, the value of the relevant uncertainty can be computed. Table I shows these relative phase differences of each shunt (if a shunt is used to as a step-up element the same value is reported for both currents). In the second column the type of shunt is indicated (MF = metal foil, CL = cage-like, O = other). In the last lines the relative uncertainty at each current value is reported. B. Reference value The phases evaluated in the previous section are relative to the phase of the shunt V1620A3 at 10 A, which is taken as a

5 Accepted for publication in IEEE I&M 5 reference. In order to evaluate the absolute phases of all shunts linked in the measurement it is necessary to compute with some method the absolute phase of this shunt. The value was evaluated firstly using a step-down procedure, like that shown in the previous section, for the steps 5 A, 2 A, 1 A and 0.5 A and deriving the relative values of the relative phases of the shunts involved. TABLE I RELATIVE PHASE OF THE SHUNTS UNDER COMPARISON, RELATIVE TO THE SHUNT V16A20A3 AT 10 A VALUE IN mrad, UNCERTAINTIES COVERAE FACTOR (k=2) Type I 500 Hz 1 khz 2 khz 5 khz 10 khz 20 khz 50 khz 100 khz B7A10A MF CMI10A1 CL JVNT10A CL MU10ASIQ CL TDA20A O A40A20A O A40B50A CL HT20A O TDA20A O V1240A2 MF V1280A2 MF A40B50A CL CS1D0701 CL CS2D0702 CL V1280A2 MF CS1D0701 CL CS2D0702 CL V1280A2 MF U U U U A determination of the absolute value of the reference was accomplished by considering that at INRIM there were three types of shunts of identical structure having nominal currents of 2 A, 1 A and 0.5 A (a couple of each of them), and The phase of the reference shunt V1620A3 as a function of the frequency is shown in Table II with the expanded uncertainties. From these values it is possible to derive, by composition with the data in Table I, all the absolute phases and the uncertainties for all the shunt taking part in the comparison. TABLE II PHASE OF THE REFERENCE SHUNT IN mrad, UNCERTAINTIES COVERAE FACTOR (k=2) I 500 Hz 1 khz 2 khz 5 khz 10 khz 20 khz 50 khz 100 khz V1620A U C. Summary of the results From the results of the comparison, much information about the phases of the different shunts could be derived. At 10 A, which was the best measurement point, given the large number of shunts and measurements between them, the expanded uncertainty of the phase differences was less than 6 rad up to 10 khz and about 40 rad at 100 khz. The uncertainty increases respectively to about 30 rad and 250 rad at the same frequencies for the shunts measured at 50 A. In the absolute determination of phase, the uncertainty in all shunts is dominated by the uncertainty of the reference. In fact, by reasonable assumptions, the expanded uncertainty of the reference has been evaluated to be less than 60 rad up to 10 khz and 600 rad to 100 khz. The phase of all studied shunts built by NMIs, in the 10 A to 100 A range was generally linear as a function of frequency (a slight nonlinearity was observed over 50 khz in MF type of shunts) and very stable, even thought up to now the measurement were repeated only within less than one year. For all the shunts, apart from a 20 A preliminary prototype from INRIM (TDA20), the spread of the results was contained within 60 rad/khz. Fig. 7 Internal resistive structure of the 2 A, 1 A, 0.5 A shunts employed for deriving the absolute reference for phase resistance R respectively of 0.5, 1, 2. From the identical internal construction, shown if Fig. 7, the same value of the inductance L was assumed, even if the contribution of the inductance of the resistive components is not negligible, and can be different for each one of them. Then, the reference value was computed by assuming that, for every pair of shunts of this type and with different resistances, the absolute of phase error values can be obtained from the relative ones, considering their time constants proportional to 1/R. So, the differences of the time constants can be assumed from the measurements, while their ratio is given by the inverse proportionality of the resistances. As there are several shunts different available, possible extrapolations are possible and their spread was considered as an indication of the accuracy. VI. CONCLUSIONS The phase values for a set of 10 A to 100 A shunts have been measured up to 100 khz and the values obtained seem to be stable in time and linear as a function of the frequency. This comparison and the other tests were performed for characterizing the shunts already built. The data collected are useful, within the project, for identifying the models that are more suitable for the different current ranges and for improving the design of the shunts. In fact, in the design process, it is necessary to improve the models by identifying the differences between their previsions and the experimental results. Furthermore, the results will be useful, with an improved phase comparator and a new determination of the reference, to characterize the prototype of the shunt built in the project. At the moment, the accuracy level of the determination of the reference is not satisfactory. However, by using the described method of the evaluation by extrapolation, with suitable metal film shunts (one of which has been already

6 Accepted for publication in IEEE I&M 6 built), it will be possible to widely improve the accuracy of this determination, because in this case the inductance of each element is almost equal. REFERENCES [1] M. arcocz, P. Scheibenreiter, W. Waldmann and. Heine: Expanding the measurement capability for AC-DC Current Transfer at BEV, CPEM 2004 Digest, pp [2] K. Lind, T. Sorsdal, H. Slinde, Design, Modeling, and Verification of High-Performance AC DC Current Shunts From Inexpensive Components, IEEE Trans. on Instr. and Meas., vol. 57, no. 1, pp , [3] K. -E. Rydler, V. Tarasso, Extending ac-dc current transfer measurement to 100 A, 100 khz, CPEM 2008 Digest, pp [4] V. N. Zachovalova, AC-DC current transfer difference in CMI, CPEM 2008 Digest, pp [5] B. Voljc, M. Lindic, R. Lapuh, Direct Measurement of AC Current by Measuring the Voltage Drop on the Coaxial Current Shunt, IEEE Trans. on Instr. and Meas., vol. 58, no. 4, pp , [6] U. Pogliano,. C. Bosco, D. Serazio, Coaxial Shunts as AC DC Transfer Standards of Current, IEEE Trans. on Instr. and Meas., vol. 58, no. 4, pp , [7] I. Budovsky, "Measurement of Phase Angle Errors of Precision Current Shunts in the Frequency Range From 40 Hz to 200 khz," in IEEE Trans. on Instr. and Meas., vol 56, no 2, pp , [8] U. Pogliano, D. Serazio, B. Trinchera, Wideband Phase Comparator for High Current Shunts, CPEM 2010 Digest, Deajeon, Korea, June [9] U. Pogliano, D. Serazio, B. Trinchera, Wideband Phase Comparator for High Current Shunts, submitted to IEEE Trans. on Instr. and Meas. [10] O. Cardfeld, The Use of an Electric Network Analogue in Least Square Method Evaluations of Measurements, Metrologia, vol. 5, no. 2, ian Carlo Bosco was born in Turin, Italy, in He received the high school degree in electrotechnics in He joined the Electrical Metrology Department of the Istituto Elettrotecnico Nazionale "alileo Ferraris," now Istituto Nazionale di Ricerca Metrologica (I.N.RI.M.), Turin, in 1967, where he has been involved in low-frequency measurements devoted particularly to establish and maintain the standards of inductance, capacitance and ac-dc transfer. At present he is responsible for the calibration activity in electrical metrology. Martin arcocz was born in Vienna, Austria, in He received the Ing. degree in electrical engineering from the Höhere Technische Bundes- Lehr- und Versuchsanstalt Vienna in He joined the Bundesamt für Eich- und Vermessungswesen (BEV), Vienna, in 1977 and is working in the field of ac-dc transfer and ac quantities. Since 2004 he is head of the laboratory Electrical Quantities at BEV. Kåre Lind was born in Tromsø, Norway, in He received the degree sivilingeniør in electrical engineering from the Norwegian Institute of Technology, Trondheim, in He joined the Norwegian Telecom s Research and Development department (NT R&D), where he worked on the development of calibration, test, and measurement methods for telecom equipment. In 1986, he became the Head of the calibration and measurement methods group at NT R&D. The group was granted accreditation as an electrical calibration laboratory in In addition, he was engaged in microwave radio propagation studies. In 1998, he joined Justervesenet, the Norwegian metrology and accreditation service, Kjeller, Norway, where he currently works primarily with time and frequency and developing methods for electrical calibration. Umberto Pogliano was born in Leinì, Turin, Italy, in He received the Dr. Ing. degree in Electronic Engineering in 1975 and the Ph.D degree in metrology in 1987, both from the Politecnico di Torino. In 1977 he joined the Electrical Metrology Department of the Istituto Elettrotecnico Nazionale "alileo Ferraris," now Istituto Nazionale di Ricerca Metrologica (I.N.RI.M.), Turin. His research activity has been focused on the modelling of standards, on the development of systems and procedures in the field of precision dc and ac low-frequency measurements. At present his main interests are in the ac-dc transfer standard, in the ac voltage, current and power measurements, and in the generation, acquisition and reconstruction of electrical signals. From 2002 to 2007 he was responsible for the Electrical Metrology Department and now is he responsible for the research group The metrology of voltage, current and power in variable regime and in high frequency. Dr. Pogliano is an Italian delegate of CCEM and TC4. ert Rietveld was born in The Netherlands in He received the M.Sc. (cum laude) and Ph.D. degrees in low temperature and solid state physics from the Delft University of Technology, Delft, The Netherlands, in 1988 and 1993, respectively. In 1993, he joined VSL (Van Swinden Laboratorium), Delft, where he is a senior scientist in the DC/LF group of the Research & Development Department. He is involved in the development of power measurement systems and electrical quantum standards, especially the quantum Hall resistance standard. Other scientific work concerns the measurement of very small electrical currents and evaluation of selfcalibrating instruments. In addition he has worked as a program manager, coordinating the scientific work of all technological areas within VSL. Dr. Rietveld is a member of the Consultative Committee for Electricity and Magnetism (CCEM) of the International Bureau of Weights and Measures (BIPM), the contact person for VSL in the technical committee of electricity and magnetism (TCEM) of the European Association of National Metrology Institutes (EURAMET), and a member of several CCEM and EURAMET working groups. Valter Tarasso was born is Sweden in He received the M.S. degree in electrical engineering from Chalmers University of Technology, othenburg, Sweden, in He joined the Electrical Metrology Laboratory at SP Technical Research Institute of Sweden in He is involved in calibration and development of measuring methods for electrical ac low frequency quantities and ac-dc transfer of voltage and current. Internationally he is actively participating in EURAMET SC-LF group.

7 Accepted for publication in IEEE I&M 7 Boštjan Voljč was born in Ljubljana, Slovenia, in He received the B.Sc. degree in electrical engineering from the University of Ljubljana in In 2001, he joined the Slovenian Institute of Quality and Metrology (SIQ), Ljubljana, where he is currently a Senior Calibration Engineer. He is in charge of maintaining the national standard of electric current and is also a speaker at metrology related seminars. He has recently been involved in local and international research projects. His research interests include precision ac voltage and ac current measurements. Mr. Voljč is a member of the EURAMET Technical Committee for Electricity and Magnetism, of the Subcommittee for Low Frequency and Subcommittee for Power and Energy. Vera Novakova Zachovalova was born in Znojmo, Czech Republic, in She received the Ing. degree in electrical engineering from Brno University of Technology in In 2005 she joined the Department of DC and LF electrical quantities of the Czech Metrology Institute (CMI). She is responsible for AC-DC transfer difference and AC voltage and current traceability. She is involved in several national and international research projects.