Microsoft Word - 24-145-1-GA.doc ACTA IMEKO  July 2012, Volume 1, Number 1, 65‐69  www.imeko.org    ACTA IMEKO | www.imeko.org  July 2012 | Volume 1 | Number 1 | 65  New generation of AC‐DC current transfer standards at  Inmetro  M. Klonz 1 , R. Afonso 2 , R.M. Souza 2 , R.P. Landim 2   1  Retired from Physikalisch‐Technische Bundesanstalt, Bundesallee 100, D‐38116 Braunschweig, Germany  2  Instituto Nacional de Metrologia, Qualidade e Tecnologia, Av. Nossa Senhora das Graças, 50 – Xerém, RJ25250020 Duque de Caxias, Brazil      Keywords: ac‐dc difference; PMJTC; comparison  Citation: M. Klonz, R. Afonso, R.M. Souza, R.P. Landim, New generation of AC‐DC current transfer standards at Inmetro, Acta IMEKO, vol. 1, no. 1, article 13,  July 2012, identifier: IMEKO‐ACTA‐01(2012)‐01‐13  Editor: Pedro Ramos, Instituto de Telecomunicações and Instituto Superior Técnico/Universidade Técnica de Lisboa, Portugal  Received January 10 th , 2012; In final form May 11 st , 2012; Published July 2012  Copyright: © 2012 IMEKO. This is an open‐access article distributed under the terms of the Creative Commons Attribution 3.0 License, which permits  unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited  Funding: Information not available  Corresponding author: R. M. Souza, e‐mail: rmsouza@inmetro.gov.br    1. INTRODUCTION  Inmetro, the Brazilian National Metrology Institute, is responsible for developing new calibration set-ups and standards that will improve the capacity of providing a higher quality calibration service, specially to the Brazilian accredited laboratories, which are responsible for providing calibration services to all other laboratories and industries in Brazil. The Inmetro started to invest in PMJTCs (Planar Multijunction Thermal Converters), to replace SJTCs (Single Junction Thermal Converters) as primary standard at 10 mA current level. For higher currents the new TCCs (Thermal Current Converters) are built from high quality coaxial shunts, model A40B, manufactured by Fluke in parallel to PMJTCs. This Fluke design of shunts with small current level effect on the ac-dc current transfer difference follows the design by several authors in different national institutes [1, 2, 3, 4, 5]. Thermal converters are capable of comparing the joule heating between ac and dc modes at 0.1 µV/V level, and are widely employed as the ac-dc current transfer standards in most of the national metrology institutes. The existing ac-dc current transfer standards of Inmetro are made of SJTCs which have one thermocouple at the midpoint of the heater and are enclosed in an evacuated glass bulb to improve its sensitivity [6]. The fundamental limitations of the performance of an SJTC are thermoelectric errors (Thomson and Peltier effects) in the heater due to the rather large temperature gradient along the heater (about 200 ºC), level dependence of the ac-dc difference, small output voltage and therefore small dynamic range. Moreover such an SJTC based ac-dc current transfer system needs to be recalibrated against higher level standards at least every 5 years to obtain small uncertainties. To reduce these thermoelectric errors, the MJTC uses as many as two hundred thermocouples spaced along a much longer heater wire [6], which results in a larger output voltage and negligible temperature gradients along the heater. However, the MJTC fabrication process is complicated and expensive. The design of PMJTCs is suitable to mass production without degradation of the performance of the MJTC. PMJTCs provide long-term stability together with high sensitivity and high dynamic range. They are well known for very small ac–dc current transfer differences at audio frequencies [7, 8]. Moreover the shunts used in the former ac-dc current transfer standards (model A40 shunts made by Fluke) were replaced by high quality coaxial shunts (model A40B manufactured by Fluke). In order to validate the new system, an unofficial comparison was made between PTB (Physikalisch- Technische Bundesanstalt, Germany) [9] and Inmetro standards. The This  paper  describes  the  new  primary  standard  for  the  ac‐dc  current  transfer  at  Inmetro,  based  on  PMJTCs  and  the  new  shunts  manufactured by Fluke for rated currents from 10 mA up to 20 A. The build‐up of the ac‐dc current scale is described together with the  uncertainty budgets which result in final uncertainties at 5 A of 6 µA/A to 12 µA/A in the frequency range from 10 Hz to 100 kHz. The  recalibration of the standards after one year showed very small differences which are included in the uncertainty budget.    ACTA IMEKO | www.imeko.org  July 2012 | Volume 1 | Number 1 | 66  measurements were performed at 10 mA and 5 A, in the whole frequency range from 10 Hz to 100 kHz. 2. CALIBRATION SET‐UP  The basic standard for ac-dc current transfer is the 10 mA PMJTC providing traceability to PTB. All standards for higher currents contain a shunt associated with a dedicated PMJTC which measures the voltage across the shunt. To build-up the current scale from 10 mA to 20 A, the different current ranges have to be calibrated against each other. In this step-up method starting from 10 mA the next higher current standard for 20 mA is calibrated at the current of 10 mA. Under the assumption that it does not change its ac-dc current transfer differences, it is used then at 20 mA. This procedure continues step by step for all frequencies from 10 Hz to 100 kHz and currents up to 20 A. The calibration set-up used is shown in detail in Figure 1. Two separate calibrators deliver ac and dc voltages. An ac-dc switch connects the ac and dc voltages to the transconductance amplifier which converts the voltage to the necessary current. Both ac-dc current transfer standards are connected in series and therefore get the same current for this comparison. The two nanovoltmeters Keithley 182 measure the output voltage of the PMJTCs. The nanovoltmeters are modified because their input amplifiers should be driven at the potential of the ac-dc transfer standards. The basic design of the measurement set-up showing only PMJTCs for 10 mA is given in Figure 1. For higher currents, coaxial shunts, model A40B, manufactured by Fluke, are associated to them. The different earth connections are chosen in a specific way to avoid any earth loops which may change the measured values in an unknown way. A coaxial choke (CC) has been introduced to suppress earth currents causing common mode voltages at the input of the transconductance amplifier. The introduction of potential driven guards (Figure 2) in the comparison circuit of the two ac-dc transfer standards avoids systematic changes of the ac-dc transfer difference of the standard, which is at the higher potential in the series connection of the two standards, especially at higher frequencies [10, 11]. This is necessary because the standard is calibrated at low potential and used at high potential. Photos in Figure 3 and Figure 4 show the calibration set-up. With this calibration set-up, standards for all current ranges have been built-up with a standard deviation of the measurement smaller than 1 µA/A. In Figure 5 the different steps in the step-up procedure are shown. All PMJTCs called 90 have heater resistances of 90 Ω, whereas the 400 has a heater resistance of 400 Ω and the 900 has a 900 Ω one. The second name is the shunt for the different currents. In Figure 6 some thermal converters are shown. The first one from the left is a 10 mA PMJTC; the second one is a boxed                Clarke Hess ‐ 8100   Hi Lo chassis 20 A 20 mA a 2 A    DC Calibrator         Gnd     OUTPUT   SENSE   Grd     Hi     Lo     AC Calibrator       Gnd   OUTPUT   SENSE   Grd   Hi   Lo     External Guard OFF   External Guard OFF      K‐182 PMJTC      K‐182   PMJTC CC   AC‐DC Switch   Transcoductance  Amplifier Figure 1. Measurement set‐up for ac–dc current transfer difference  measurements.  Figure 2. ac‐dc current transfer with potential driven guards (PMJTCs with  shunts for current ranges above 10 mA).  Figure 3. ac‐dc current transfer set‐up.  Figure 4. Connection of the ac‐dc current transfer standards in series.    ACTA IMEKO | www.imeko.org  July 2012 | Volume 1 | Number 1 | 67  50 mA shunt connected to a 90  PMJTC and the others are PMJTCs connected to coaxial shunts for currents up to 100 mA and for higher currents up to 20 A. 3. UNCERTAINTY ANALYSIS  The model equation is step tep -1 CA C com. mode Lev LF diff. step-ups i s i                (1) with  step i – 1: Transfer difference of standard at the step i – 1. CA: Contribution of the mean of repeated twelve measurements. C: Contribution of the measurement set-up. com. mode: Transfer difference from the common mode effect in the transconductance amplifier determined from the difference of measurements with and without the choke CC in Figure 1. Lev: Transfer difference due to level dependence of shunts estimated from the design of the shunts. LF: Transfer difference due to low frequency behavior of PMJTC. diff.step-ups: Correction with the difference of different step-up measurements performed in the same measurement set-up. The sum of the variances of the different contributions results in the variance of the result:                 2 2 2 2 step tep -1 CA C 2 2 2 2 com. mode Lev LF diff. step-ups u u u u +u u u u i s i               (2) where u2(x) represents the variance of x. The uncertainty budgets of the different current steps are given in Tables 1 to 4. 4. COMPARISON RESULTS  An unofficial interlaboratory comparison of ac-dc current transfer standards between PTB and Inmetro was performed with a travelling standard for 10 mA and one for 5 A. The current points chosen were 10 mA and 5 A, in the frequency range from 10 Hz and 100 kHz. In Inmetro each current point was measured against Inmetro standards in twelve cycles at all frequencies, and the mean was calculated using the results of the sequence which gives the ac-dc current transfer difference, represented by δInmetro. Tables 5 and 6 show the results obtained. PTB uses a similar calibration set-up and similar standards. The shunts are manufactured by the Norwegian Metrology Institute Justervesenet with a different design [12]. The results obtained in PTB are represented by δPTB. The associated expanded uncertainties are given by U. The measured differences Inmetro-PTB between both institutes were small compared to the given uncertainties which is also represented by the small En-values. This is a very satisfying result of the comparison. 5. CONCLUSIONS  In the first step a new step-up procedure was developed to build-up the ac-dc current transfer standards for 10 mA up to 20 A and to perform an uncertainty analysis for the whole build-up. The second step towards the introduction of the new ac-dc current transfer system of Inmetro was performing a comparison between Inmetro and PTB, which proved that Inmetro’s new calibration set-up and standards work as expected.     3 mA  1 mA  900‐2 400‐3  400‐3  900‐2 90‐11  90‐12 5 mA  90‐13 + Sh20mA  20 mA  90‐13 + Sh20mA  90‐14 + Sh50mA  90‐11  90‐12 10 mA  50 mA  90‐14 + Sh50mA  90‐15 + Sh100mA  100 mA  90‐15 + Sh100mA  90‐16 + Sh200mA  200 mA  90‐16 + Sh200mA  90‐17 + Sh500mA  500 mA  90‐17 + Sh500mA  90‐18 + Sh1A 1 A  90‐18 + Sh1A  90‐19 + Sh2A 2 A  90‐19 + Sh2A  90‐20 + Sh5A  5 A  90‐20 + Sh5A  90‐21 + Sh10A  90‐21 + Sh10A  90‐22 + Sh20A  10 A  20 A  90‐22 + Sh20A  Figure 5. Schematics of the step‐up procedure.  Figure 6. ac‐dc current transfer standards.    ACTA IMEKO | www.imeko.org  July 2012 | Volume 1 | Number 1 | 68  Table 1. Uncertainty analysis for the step‐up at 10 mA.  Influencing quantity  Measurement uncertainty in µA/A at the frequency in kHz  0.01  0.02  0.03  0.04  0.055  0.12  0.5  1  5  10  20  50  70  100  u( 10 mA)  1.5  1.5  1.5  1.5  1.5  1.5  1.5  1.5  1.5  1.5  1.5  1.5  1.5  1.5  u(CA)  0.5  0.4  0.3  0.5  0.6  0.3  0.3  0.3  0.3  0.3  0.3  0.3  0.4  0.6  u(C)   0.1  0.1  0.1  0.1  0.1  0.1  0.1  0.1  0.1  0.1  0.1  0.1  0.1  0.1  u(com. mode)  1  1  1  1  1  1  0  0  0  0  0  0  0  0  u(Lev)  0  0  0  0  0  0  0  0  0  0  0  0  0  0  u(diff. step‐ups)  0.2  0.2  0.0  0.3  0.1  0.0  0.1  0.2  0.1  0.1  0.1  0.1  0.1  0.2  u(LF)  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  u(mA)  1.9  1.9  1.8  1.9  1.9  1.8  1.5  1.5  1.5  1.5  1.5  1.5  1.6  1.6  U(mA) k = 2  3.8  3.8  3.6  3.8  3.8  3.6  3.0  3.0  3.0  3.0  3.0  3.0  3.2  3.2                                Table 2. Uncertainty analysis for the step‐up at 100 mA.  Influencing quantity  Measurement uncertainty in µA/A at the frequency in kHz  0.01  0.02  0.03  0.04  0.055  0.12  0.5  1  5  10  20  50  70  100  u( 50 mA)  3.0  2.6  2.5  2.5  2.5  2.4  1.7  1.7  1.7  1.7  1.7  1.7  1.8  1.9  u(CA)  0.5  0.3  0.5  0.4  0.3  0.5  0.4  0.4  0.4  0.2  0.4  0.4  0.4  0.5  u(C)   0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  u(com. mode)  1.0  1.0  1.0  1.0  1.0  1.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  u(Lev)  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.3  0.4  u(diff. step‐ups)  0.1  0.1  0.1  0.1  0.3  0.2  0.1  0.1  0.2  0.4  0.2  0.1  0.1  0.1  u(LF)  0.3  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  u(mA)  3.2  2.9  2.8  2.7  2.8  2.7  1.8  1.8  1.8  1.8  1.8  1.8  1.9  2.0  U(mA) k = 2  6.4  5.8  5.6  5.4  5.6  5.4  3.6  3.6  3.6  3.6  3.6  3.6  3.8  4.0                                Table 3. Uncertainty analysis for the step‐up at 1 A.  Influencing quantity  Measurement uncertainty in µA/A at the frequency in kHz  0.01  0.02  0.03  0.04  0.055  0.12  0.5  1  5  10  20  50  70  100  u( 500 mA)  3.8  3.5  3.4  3.3  3.4  3.4  2.1  2.0  2.1  2.1  2.1  2.1  2.1  2.5  u(CA)  0.4  0.3  0.3  0.3  0.3  0.4  0.3  0.3  0.2  0.2  0.2  0.2  0.3  0.3  u(C)   0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  u(com. mode)  1.0  1.0  1.0  1.0  1.0  1.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  u(Lev)  1.0  1.0  1.0  1.0  1.0  1.0  1.0  1.0  1.0  1.0  1.0  1.0  1.0  1.5  u(diff. step‐ups)  0.1  0.1  0.1  0.2  0.2  0.1  0.4  0.1  0.2  0.3  0.1  0.3  0.1  0.4  u(LF)  0.4  0.1  0.1  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  u(A)  4.1  3.8  3.7  3.6  3.7  3.7  2.4  2.3  2.3  2.3  2.3  2.3  2.4  2.9  U(A) k = 2  8.2  7.6  7.4  7.2  7.4  7.4  4.8  4.6  4.6  4.6  4.6  4.6  4.8  5.8    Table 4. Uncertainty analysis for the step‐up at 5 A.  Influencing quantity  Measurement uncertainty in µA/A at the frequency in kHz  0.01  0.02  0.03  0.04  0.055  0.12  0.5  1  5  10  20  50  70  100  u(  2 A)  4.4  4.1  4.0  3.9  4.0  4.0  2.7  2.6  2.5  2.8  3.1  3.8  3.9  4.6  u(CA)  0.3  0.4  0.3  0.2  0.4  0.4  0.3  0.5  0.4  0.2  0.2  0.4  0.6  0.4  u(C)   0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  0.2  u(com. mode)  1.0  1.0  1.0  1.0  1.0  1.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  u(Lev)  1.0  1.0  1.0  1.0  1.0  1.0  1.0  1.0  1.0  1.5  2.0  3.0  3.0  3.5  u(diff. step‐ups)  0.1  0.2  0.3  0.0  0.1  0.2  0.3  0.1  0.3  0.3  0.2  0.6  0.5  0.7  u(LF)  0.4  0.3  0.1  0.1  0.1  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  u(A)  4.6  4.4  4.3  4.2  4.3  4.3  2.9  2.8  2.8  3.2  3.7  4.9  5.0  5.9  U(A) k = 2  9.2  8.8  8.6  8.4  8.6  8.6  5.8  5.6  5.6  6.4  7.4  9.8  10.0  11.8    ACTA IMEKO | www.imeko.org  July 2012 | Volume 1 | Number 1 | 69  The results of both institutes’ standards agreed within 2 µA/A for 10 mA and 5 µA/A for 5 A between 10 Hz and 100 kHz. That means that this new system works reliably and is ready for use in the next international comparison of ac-dc current transfer standards (SIM.EM-K12) [13]. REFERENCES  [1] I. Budovsky, “Measurement of Phase Angle Errors of Precision Current Shunts in the Frequency Range from 40 Hz to 200 kHz”, IEEE Trans. Instrum. Meas., vol. 56, no. 2, April 2007, pp. 284-288. [2] M. Garcocz, P. Scheibenreiter, W. Waldmann, G. Heine, “Expanding the measurement capabilities for ac-dc current transfer at BEV”, 2004 Conf. on Precision Electromagnetic Measurements Digest, CPEM, June 27 to July 2, 2004, pp. 461-462. [3] P.S.Filipski and M. Boeker, “AC-DC current shunts and system for extended current and frequency ranges”, in Proc. IMTC, Ottawa, ON, Canada, May 17-19, 2005, pp. 991-995. [4] P. S. Filipski and M. Boeker, “Ac-dc current transfer standards and calibrations at NRC”, Simposio de Metrologia 2006, 25 al 27 Otubre 2006. [5] R. M. Souza, R. V. F. Ventura, F. A. Silveira, “Reevaluation of ac-dc current transfer system based on single junction thermal converters at Inmetro”, XVIII IMEKO TC4 Symposium and IX Semetro / Metrologia 2011: TC 4 IMEKO and IX Semetro, Natal, Brazil, 2011. [6] M. Klonz, “AC-DC Transfer Difference of the PTB Multijunction Thermal Converter in the Frequency Range from 10 Hz to 100 kHz”, IEEE Trans. Instrum. Meas., vol. 36, no. 2, June 1987, pp. 320-329. [7] M. Klonz, T. Weimann, T., “Accurate Thin Film Multijunction Thermal Converter on a Silicon Chip”, IEEE Trans. Instrum. Meas. IM-38, no. 2, April 1989, pp. 335-336, [8] H. Laiz, M. Klonz, E. Kessler, T. Spiegel, “New Thin-Film Multijunction Thermal Converters with Negligible Low Frequency AC-DC Transfer Differences”, IEEE Trans. Instrum. Meas., vol. 50, no. 2, April 2001, pp.333-337. [9] M. Klonz, H. Laiz, T. Spiegel, P. Bittel, “AC-DC Current Transfer Step-up and Step-down Calibration and Uncertainty Calculation”, IEEE Trans. Instrum. Meas., vol. 51, no. 5, 2002, pp. 1027-1034. [10] K. E. Rydler, “High precision automated measuring system for ac-dc current transfer standards”, IEEE Trans. Instrum. Meas., vol. 42, no.1, April 1993, pp. 608-611. [11] T. Funck, and M. Klonz, “Improved AC-DC current transfer step-up with new current shunts and potential driven guarding”, IEEE Trans. Instrum. Meas. vol. 56, April 2007, pp. 361-364. [12] K. Lind, T. Sorsdal, and H. Slinde, “Design, modeling, and verification of high-performance AC-DC current shunts from inexpensive components”, IEEE Trans. Instrum. Meas., vol. 57, no.1, Jan. 2008, pp. 176-181. [13] L. D. Lillo, “SIM.EM-K12 Comparison Protocol – AC-DC Current Transfer Difference”, 2010. Table 5.  Result of the comparison between Inmetro and PTB at 10 mA. Institutes  AC‐DC current transfer differences together with their uncertainties in µA/A at the frequencies in kHz  0.01  0.02  0.03  0.04  0.055  0.12  0.5  1  5  10  20  50  70  100  Inmetro  5.5  1.3  1.0  0.7  0.3  ‐0.1  ‐0.5  0.1  0.4  0.6  2.2  17.6  34.8  70.9  U Inmetro  4  4  4  4  4  4  3  3  3  3  3  3  3  4  PTB  5.3  1.8  0.7  0.7  ‐0.3  ‐0.4  ‐0.1  ‐0.2  0.5  1.5  3.2  18.9  36.4  72.2  U PTB  3  3  3  3  3  3  3  3  3  3  3  3  3  3  Difference Inmetro ‐ PTB  0.2  ‐0.5  0.3  0.0  0.5  0.3  ‐0.4  0.3  ‐0.1  ‐0.9  ‐1.0  ‐1.3  ‐1.6  ‐1.3  En  0.0  ‐0.1  0.1  0.0  0.1  0.1  ‐0.1  0.1  0.0  ‐0.2  ‐0.2  ‐0.3  ‐0.4  ‐0.3  Table 6.  Result of the comparison between Inmetro and PTB at 5 A.  Institutes  AC‐DC current transfer differences together with their uncertainties in µA/A at the frequencies in kHz  0.01  0.02  0.03  0.04  0.055  0.12  0.5  1  5  10  20  50  70  100  Inmetro  0.0  0.4  0.2  0.1  0.0  ‐0.3  0.0  0.1  2.7  11.6  13.2  ‐88.1  ‐187.0  ‐353.9  U Inmetro  9  9  9  9  9  9  6  6  6  7  8  10  10  12  PTB  ‐0.9  ‐0.2  0.4  0.2  ‐0.4  0.5  0.5  0.4  3.0  10.6  10.5  ‐89.9  ‐187.2  ‐349.4  U PTB  5  4  4  4  4  4  4  4  4  5  7  9  10  11  Difference Inmetro ‐ PTB  0.9  0.6  ‐0.2  0.0  0.4  ‐0.8  ‐0.5  ‐0.3  ‐0.3  1.0  2.7  1.9  0.2  ‐4.5  En  0.1  0.1  0.0  0.0  0.0  ‐0.1  ‐0.1  0.0  0.0  0.1  0.2  0.1  0.0  ‐0.3