Supplementary Comparison EUROMET.EM.RF-S25 EURAMET Project 819. Comparison of Electrical Field Strength Measurements above 1 GHz.

Transcription

1 EURAMET.EM.RF-S25 Final Report page 1 of 120 Supplementary Comparison EUROMET.EM.RF-S25 EURAMET Project 819 Comparison of Electrical Field Strength Measurements above 1 GHz Final Report Karel Dražil Czech Metrology Institute Radiová Praha Czech Republic February 2011

2 EURAMET.EM.RF-S25 Final Report page 2 of 120 CONTENTS 1 Introduction Participants and organisation of the comparison List of participants Comparison schedule Organisation of the comparison Unexpected incidents Travelling standard and measurement instructions Description of the standard Quantity to be measured Measurement instructions Methods of measurement Repeated measurements of the pilot institute, behaviour of the travelling standard Measurement results Results of the participating institutes Traceability of measurements Normalisation of the results Calculation of the reference value and its uncertainty Degree of equivalence Summary and conclusions References Appendices... 25

3 EURAMET.EM.RF-S25 Final Report page 3 of Introduction In the frequency range of about 1 GHz to 2.5 GHz, various physical principles of calculable electromagnetic field strength generators (e.g. small TEM cell, tapered TEM cell, rectangular waveguide, standard antenna) are currently in use in the national metrology laboratories. The objective of the project was to compare the field strength standards as realised in the participating laboratories. Both the travelling standard and technical rules were very similar to those used in the key comparison CCEM.RF-K20. The task for the participants was to calibrate a field strength meter with a small linearly polarized E-field sensor at seven frequency points in the range from 1 GHz to 2.5 GHz. 2 Participants and organisation of the comparison 2.1 List of participants List of participants (in alphabetical order) is in Table 1. Acronym Laboratory Country ČMI Český metrologický institut Czech Republic INRIM Istituto Nazionale di Ricerca Metrologica Italy LNE Laboratoire national de métrologie et d essais France METAS Federal Office of Metrology METAS Switzerland NMi-VSL *) Nederlands Meetinstituut - Van Swinden Laboratorium The Netherlands NPL National Physical Laboratory United Kingdom PTB Physikalisch - Technische Bundesanstalt Germany SP Technical Research Institute of Sweden Sweden STUK Radiation and Nuclear Safety Authority Finland UP University of Pretoria South Africa *) Nederlands Meetinstituut Van Swinden Laboratorium (NMi-VSL) is renamed as VSL, Dutch Metrology Institute (VSL) since March 2009 Table 1 List of participants 2.2 Comparison schedule Table 2 gives the date of measurement at each of the participating laboratories.

4 EURAMET.EM.RF-S25 Final Report page 4 of 120 Date Acronym Action November 2007 ČMI stability check, measurement December 2007 NMi-VSL measurement January 2008 NPL measurement February 2008 ČMI stability check February 2008 LNE measurement March 2008 STUK measurement April 2008 SP measurement May 2008 ČMI stability check June 2008 INRIM measurement July 2008 PTB measurement August 2008 METAS measurement August 2008 ČMI stability check November 2008 UP measurement January 2009 ČMI stability check Table 2 Participants measurement dates 2.3 Organisation of the comparison The project was coordinated by ČMI. The travelling standard was transported in a suitcase adapted with regard to antistatic precautions, packed in a carton box. ATA Carnets were used for the shipment to participants outside the European Union. For most of the participants, the device was sent via DHL. The transfer standard was periodically returned to the pilot laboratory to monitor its long-term stability. 2.4 Unexpected incidents Near to the end of the measurement phase of the comparison, a problem with the customs procedure was encountered. During the transport to South Africa, the ATA Carnet was lost. After some period of unsuccessful searching, the duplicate had to be issued and sent to the shipper (DHL) office at the Johannesburg International airport. A delay of several weeks was caused by the loss of the document. 3 Travelling standard and measurement instructions 3.1 Description of the standard The travelling standard involved in the comparison is a field strength meter system (Fig. 1, Fig. 2), which consists of a small linearly polarized E-field sensor with a Schottky diode detector, high resistance cable, box with measurement electronics, optical interface and control program for PC running under Windows. The electronics box contains a precision differential amplifier, analog-to-digital converter, some electro-optical modules and a battery. The control program includes three files with individual calibration data for the electronics box and sensor (TFSMeter.def, lrs002.frq, lrs002.lin). The equipment was designed by PTB and a similar model is commercially available.

5 EURAMET.EM.RF-S25 Final Report page 5 of 120 The devices were packaged in a carrying case with a size of approximately (0.45 x 0.35 x 0.15) m. The weight with the enclosure is approximately 4 kg. Fig. 1 The E-field sensor with electronics box and charging cable Fig. 2 Detail of the E-field sensor

6 EURAMET.EM.RF-S25 Final Report page 6 of 120 The length and diameter of the dipole element of the probe (see Fig. 2) is 6.5 mm and 1.2 mm, respectively. The dimensions of the probe are relatively small in comparison with the dimensions of TEM cells and waveguides used for the appropriate frequency band. As a consequence, little effects of the probe on the field distribution and only negligible errors are expected. The sensitivity of the probe to the E - field component oriented in parallel with the handle is quite small (experimentally verified), hence the influence of the E field component oriented in parallel with the septum and transversally to the TEM cell axis (increasing toward the sides of the TEM cells) can be expected to be very small as well. 3.2 Quantity to be measured Each participant should expose the probe of the circulated field strength meter to his own realisation of the standard electrical field. The value to be supplied as a measurement result is the actual field strength required to produce a field strength meter reading of 20 V/m. The measurement should be performed at frequencies 1 GHz, 1.25 GHz, 1.5 GHz, 1.75 GHz, 2 GHz, 2.25 GHz and 2.5 GHz. 3.3 Measurement instructions First of all, the offset of the amplifier is to be checked. When no field is applied or the sensor is disconnected from the electronics box, the average value of the field strength meter indication should be in the order of tenths of V/m. If this condition is not fulfilled, the file TFSMeter.def has to be modified before a new start of the program. The second number at the seventh line in the mentioned file corresponds to the offset of the amplifier and has to be changed appropriately. The goal of the exercise is to calibrate the circulated field strength meter. Each participant should expose the probe to his own electrical field and apply measurement facilities and methods as he does for the accurate transfers. Before starting of measurements, the sensor has to be adjusted with respect to the field orientation for the maximum indicated field strength. The handle of the sensor should be aligned perpendicularly to the direction of the electromagnetic wave propagation. At the given frequency points, the field strength should be adjusted for a probe reading of 20 V/m. During all measurements, the probe reading value should be kept within ±0.5 V/m of the nominal value of 20 V/m. For the comparison, the following frequencies were chosen: 1 GHz, 1.25 GHz, 1.5 GHz, 1.75 GHz, 2 GHz, 2.25 GHz, 2.5 GHz. Temperature of the probe s test volume: (23±1) C The field probe is controlled by a PC program. As the system does not measure the temperature of the probe and the frequency, these values have to be entered by the operator in designated windows.

7 EURAMET.EM.RF-S25 Final Report page 7 of Methods of measurement Basically, four types of calculable field strength realisations were used: in rectangular waveguide in small TEM cell in anechoic chamber by using of reference transmitting antenna in anechoic chamber by using of reference receiving antenna or transfer probe. Reference antennas were either calibrated by some commonly used method or characterised on the basis of dimensional measurements (horn antennas). The equipment used by each participant is summarised in Table 3. The full reports from each of the participants can be found in Appendix A. Acronym ČMI INRIM LNE METAS NMi-VSL NPL PTB SP STUK UP Calculable field strength realisation rectangular waveguide reference transmitting antenna open-ended waveguide at 1 GHz, horn antennas reference transmitting antenna horn antennas reference receiving antenna small biconical antenna small TEM cells mini TEM cell (septum distance 60 mm) up to 1.5 GHz micro TEM cell (septum distance 30 mm) reference transmitting antenna horn antennas micro TEM cell (septum distance 35 mm) reference transmitting antenna horn antennas at 1.25 GHz and higher frequencies transfer probe calibrated in micro TEM cell at 1 GHz reference transmitting antenna horn antennas rectangular waveguide reference transmitting antenna open-ended waveguide up to 1.12 GHz standard gain horn antennas at higher frequencies Table 3 Field strength realisation details 5 Repeated measurements of the pilot institute, behaviour of the travelling standard For the EURAMET.EM.RF-S25 comparison, the use of the TFS 1100 field strength meter was originally planned. As a similar system, used in CCEM.RF-K20 key comparison, suffered from instability problems, an improved version of the field strength meter was prepared and supplied by PTB. Because the exact cause of the drift of the probe sensitivity in the recent key comparison was not clear, additional measures were adopted. Within the scope of the prevention against damage of the sensitive diode probe by overvoltage, the transport suitcase was filled with conductive foam and equipped with an antistatic wrist strap, antistatic pad, grounding cables and exact instructions how to use these items. The above precautions reduce the unwanted common mode signal at the input of the precision amplifier as well. On

8 EURAMET.EM.RF-S25 Final Report page 8 of 120 the basis of preliminary tests of the travelling standard, it was decided to add a check (and adjustment, if necessary) of the amplifier offset to the measurement instructions for the comparison. The results of the stability monitoring measurements performed by the pilot laboratory can be seen in Table 4, while the corresponding offset settings are summarized in Table 5. Measurements were performed in rectangular waveguides (see Appendix A). The temperature of the waveguides was stabilised during measurements in the range of ± 0.1 C. In order to reduce the uncertainty caused by the reading resolution of the travelling standard, in contrast to standard ordinary participants measurements, the stability check was performed for the field strength meter indication of 60 V/m with modified data in the file lrs002.frq (correction factor constants multiplied by 3). All repeated measurements were performed in exactly the same configuration, with the same and identically oriented power splitter, the same power sensor and power meter. Observed differences between measured values are in any case less than a half percent. It can be concluded that the stability of the device during the comparison was very good and the procedures given in Reference [2] could be (from this view point) used for the evaluation of the comparison data. Note that the observed change of the field strength meter indication is about 0.3 % per 10 μv change of the offset voltage setting for field strength values of about 20 V/m. Frequency (GHz) E(V/m) E(V/m) E(V/m) E(V/m) E(V/m) Table 4 Stability of the travelling standard Date Offset setting (mv) Table 5 Offset settings used for the stability monitoring measurements 6 Measurement results 6.1 Results of the participating institutes The results supplied by the participants are listed in the following tables. Two laboratories (NMi-VSL and PTB) supplied two different sets of results obtained by two various field strength realisations at several frequencies (see Appendix A). In these cases, the results obtained by the methods used routinely in the laboratories were selected for the evaluation of the comparison data.

9 EURAMET.EM.RF-S25 Final Report page 9 of 120 Laboratory Field strength Standard uncertainty (V/m) (V/m) ČMI NMi-VSL NPL LNE STUK SP INRIM PTB *) METAS University of Pretoria *) Data removed on request of the participant because technical problem occured during the measurement Table 6 Results of the participating institutes at frequency 1 GHz Laboratory Field strength Standard uncertainty (V/m) (V/m) ČMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS University of Pretoria Table 7 Results of the participating institutes at frequency 1.25 GHz Laboratory Field strength Standard uncertainty (V/m) (V/m) ČMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS University of Pretoria Table 8 Results of the participating institutes at frequency 1.5 GHz

10 EURAMET.EM.RF-S25 Final Report page 10 of 120 Laboratory Field strength Standard uncertainty (V/m) (V/m) ČMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS University of Pretoria Table 9 Results of the participating institutes at frequency 1.75 GHz Laboratory Field strength Standard uncertainty (V/m) (V/m) ČMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS University of Pretoria Table 10 Results of the participating institutes at frequency 2 GHz Laboratory Field strength Standard uncertainty (V/m) (V/m) ČMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS University of Pretoria Table 11 Results of the participating institutes at frequency 2.25 GHz

11 EURAMET.EM.RF-S25 Final Report page 11 of 120 Laboratory Field strength Standard uncertainty (V/m) (V/m) ČMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS University of Pretoria Table 12 Results of the participating institutes at frequency 2.5 GHz The ambient conditions of the participants measurements and values of offset setting are reported in Table 13. Acronym Date Temperature ( C) Offset setting (mv) ČMI November 18, (pilot measurement) (participant measurement) NMi-VSL December 3 28, NPL January 15, ČMI February 2, LNE February 26, STUK March, 19 29, SP April 21 22, n/a ČMI May 10, INRIM June 16 20, PTB July 7 11, METAS August 12, ČMI August 31, UP November 26 27, ČMI January 11, Table 13 Temperature and offset setting values

12 EURAMET.EM.RF-S25 Final Report page 12 of Traceability of measurements In order to assess the independence of results, the participants were asked for details on traceability of their measurements. All participants realised their field strength measurements independently without traceability of this quantity to another laboratory. In the realisation of the electric field strength unit, RF power and sometimes attenuation are the quantities of primary importance. The major part of the participants have their power and attenuation measurements traceable to their own national primary standards. Several laboratories have their power or attenuation measurement standards traceable back to other NMIs, however with relatively small uncertainties. Therefore, measurements of all laboratories were considered as mutually independent and procedure B according to Reference [2] was used for the comparison evaluation. 6.3 Normalisation of the results As the stability of the probe was found to be very good, no elimination of the effect of drift was performed. The effect of ambient conditions (temperature) has already been eliminated by the participants measurements since the probe has been calibrated in the needed temperature range and the temperature dependence of the probe calibration factor has been taken into account by the field strength meter control program. 6.4 Calculation of the reference value and its uncertainty The comparison reference value (RV) and its uncertainty was computed by performing Monte-Carlo simulation (10 6 trials) according to Reference [2], procedure B, the median was used as an estimator. From the calculation of the RV, the results of the University of Pretoria were excluded since this institution is neither a NMI nor a designated institute listed at the BIPM website. The reference values with their uncertainties can be found in Table 14. Frequency (GHz) Reference value (V/m) Standard uncertainty (V/m) Table 14 Supplementary comparison reference values and uncertainties

13 EURAMET.EM.RF-S25 Final Report page 13 of Degree of equivalence The degree of equivalence is expressed quantitatively by two terms: 1) Deviation from the reference value, D i, D i = X i X RV. 2) The uncertainty of this deviation at the 95 % level of confidence U(D i ). The shortest coverage interval at the 95 % level of confidence is computed according to Reference [2], procedure B. For measurement results not used for RV calculation, the uncertainty is calculated according to the formula 2 U D = u + u, ( ) 2 i 2 RV i where u i and u RV are the standard uncertainties related to the measured value of i-th participant X i and the reference value X RV, respectively. The degree of equivalence between pairs of measurements is expressed quantitatively by two terms: 1) The difference of their deviations from the reference value, D ij, D ij = ( X i X j ). 2) The uncertainty of this difference at the 95 % level of confidence. The shortest coverage interval at the 95 % level of confidence is computed according to Reference [2], procedure B. No correlation between laboratories is assumed. The degrees of equivalence with respect to the reference value are shown in Table 15 to Table 21 and Fig. 3 to Fig. 9 where the limit bars show the calculated U(D i ) at each point. The degrees of equivalence between pairs of participants can be found in Table 22 to Table 28.

14 EURAMET.EM.RF-S25 Final Report page 14 of 120 Laboratory Deviation from RV Uncertainty at 95 % level of confidence (V/m) (V/m) ČMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS University of Pretoria Table 15 Deviations from comparison ref. value with uncertainties at frequency 1 GHz Inconsistent results highlighted in italics 4.00 Deviation from RV (V/m) CMI NMi-VSL NPL LNE STUK SP INRIM METAS Univ. of Pretoria Fig. 3 Deviations from comparison reference value at frequency 1 GHz Results marked by empty dots were excluded from the RV calculation

15 EURAMET.EM.RF-S25 Final Report page 15 of 120 Laboratory Deviation from RV Uncertainty at 95 % level of confidence (V/m) (V/m) ČMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS University of Pretoria Table 16 Deviations from comparison ref. value with uncertainties at frequency 1.25 GHz Inconsistent results highlighted in italics 4.00 Deviation from RV (V/m) CMI NMi-VSL NPL LNE SP INRIM PTB METAS Univ. of Pretoria Fig. 4 Deviations from comparison reference value at frequency 1.25 GHz Results marked by empty dots were excluded from the RV calculation

16 EURAMET.EM.RF-S25 Final Report page 16 of 120 Laboratory Deviation from RV Uncertainty at 95 % level of confidence (V/m) (V/m) ČMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS University of Pretoria Table 17 Deviations from comparison ref. value with uncertainties at frequency 1.5 GHz Inconsistent results highlighted in italics 4.00 Deviation from RV (V/m) CMI NMi-VSL NPL LNE SP INRIM PTB METAS Univ. of Pretoria Fig. 5 Deviations from comparison reference value at frequency 1.5 GHz Results marked by empty dots were excluded from the RV calculation

17 EURAMET.EM.RF-S25 Final Report page 17 of 120 Laboratory Deviation from RV Uncertainty at 95 % level of confidence (V/m) (V/m) ČMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS University of Pretoria Table 18 Deviations from comparison ref. value with uncertainties at frequency 1.75 GHz Inconsistent results highlighted in italics 4.00 Deviation from RV (V/m) CMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS Univ. of Pretoria Fig. 6 Deviations from comparison reference value at frequency 1.75 GHz Results marked by empty dots were excluded from the RV calculation

18 EURAMET.EM.RF-S25 Final Report page 18 of 120 Laboratory Deviation from RV Uncertainty at 95 % level of confidence (V/m) (V/m) ČMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS University of Pretoria Table 19 Deviations from comparison ref. value with uncertainties at frequency 2 GHz Inconsistent results highlighted in italics 4.00 Deviation from RV (V/m) CMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS Univ. of Pretoria Fig. 7 Deviations from comparison reference value at frequency 2 GHz Results marked by empty dots were excluded from the RV calculation

19 EURAMET.EM.RF-S25 Final Report page 19 of 120 Laboratory Deviation from RV Uncertainty at 95 % level of confidence (V/m) (V/m) ČMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS University of Pretoria Table 20 Deviations from comparison ref. value with uncertainties at frequency 2.25 GHz Inconsistent results highlighted in italics 4.00 Deviation from RV (V/m) CMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS Univ. of Pretoria Fig. 8 Deviations from comparison reference value at frequency 2.25 GHz Results marked by empty dots were excluded from the RV calculation

20 EURAMET.EM.RF-S25 Final Report page 20 of 120 Laboratory Deviation from RV Uncertainty at 95 % level of confidence (V/m) (V/m) ČMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS University of Pretoria Table 21 Deviations from comparison ref. value with uncertainties at frequency 2.5 GHz Inconsistent results highlighted in italics CMI NMi-VSL NPL LNE STUK SP Deviation from RV (V/m) INRIM PTB METAS Univ. of Pretoria Fig. 9 Deviations from comparison reference value at frequency 2.5 GHz Results marked by empty dots were excluded from the RV calculation

21 EURAMET.EM.RF-S25 Final Report page 21 of 120 Degree of equivalence Lab(j) with RV CMI NMi-VSL NPL LNE STUK SP INRIM METAS Univ. of Pretoria Lab(i) D i U(D i ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) CMI NMi-VSL NPL LNE STUK SP INRIM METAS Univ. of Pretoria Table 22 Degrees of equivalence at frequency 1 GHz Degree of equivalence Lab(j) with RV CMI NMi-VSL NPL LNE SP INRIM PTB METAS Univ. of Pretoria Lab(i) D i U(D i ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) CMI NMi-VSL NPL LNE SP INRIM PTB METAS Univ. of Pretoria Table 23 Degrees of equivalence at frequency 1.25 GHz

22 EURAMET.EM.RF-S25 Final Report page 22 of 120 Degree of equivalence Lab(j) with RV CMI NMi-VSL NPL LNE SP INRIM PTB METAS Univ. of Pretoria Lab(i) D i U(D i ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) CMI NMi-VSL NPL LNE SP INRIM PTB METAS Univ. of Pretoria Table 24 Degrees of equivalence at frequency 1.5 GHz Degree of equivalence Lab(j) with RV CMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS Univ. of Pretoria Lab(i) D i U(D i ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) CMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS Univ. of Pretoria Table 25 Degrees of equivalence at frequency 1.75 GHz

23 EURAMET.EM.RF-S25 Final Report page 23 of 120 Degree of equivalence Lab(j) with RV CMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS Univ. of Pretoria Lab(i) D i U(D i ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) CMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS Univ. of Pretoria Table 26 Degrees of equivalence at frequency 2 GHz Degree of equivalence Lab(j) with RV CMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS Univ. of Pretoria Lab(i) D i U(D i ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) CMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS Univ. of Pretoria Table 27 Degrees of equivalence at frequency 2.25 GHz

24 EURAMET.EM.RF-S25 Final Report page 24 of 120 Degree of equivalence Lab(j) with RV CMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS Univ. of Pretoria Lab(i) D i U(D i ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) D ij U(D ij ) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) (V/m) CMI NMi-VSL NPL LNE STUK SP INRIM PTB METAS Univ. of Pretoria Table 28 Degrees of equivalence at frequency 2.5 GHz

25 EURAMET.EM.RF-S25 Final Report page 25 of Summary and conclusions An international comparison of electric field strength in the frequency range (1 to 2.5) GHz has been carried out among 10 laboratories. The parameter compared was the field strength appropriate for the circulated field strength meter indication of 20 V/m. Basically, four types of calculable field strength realisations were used. The size of the uncertainties of participants results varies from 1.0 % to 6.7 % (coverage factor k=1). Some values reported by various participants obtained in anechoic chambers by using reference transmitting antennas ( standard field method ) and in TEM cell are inconsistent with the claimed uncertainties. A good agreement between measurements in waveguides and (consistent) measurements in anechoic chambers was observed. This comparison exercise results could be helpful for some participants as warning on potential systematic errors or some inadequacy in the uncertainty budget. 8 References [1] CCEM Guidelines for Planning, Organizing, Conducting and Reporting Key, Attached, Supplementary and Pilot Comparisons, March 29, 2004 [2] M. G. Cox: The evaluation of key comparison data, Metrologia, vol. 39, pp , Appendices Measurement reports supplied by participating laboratories can be found in Appendix A. Uncertainty budgets of the participants supplied in the EXCEL template can be found in Appendix B.

26 EURAMET.EM.RF-S25 Final Report page 26 of 120 Appendix A Participants measurement reports A1 ČMI

27 EURAMET.EM.RF-S25 Final Report page 27 of 120 Page 2 Description of the traveling standard: The traveling standard is a high frequency field strength meter system designed and supplied by PTB. It consists of a small linearly polarized E-field sensor, high resistance cable, box with electronics, optical interface and control program for PC. Description of the realisation of the electric field: Measurements have been performed in two rectangular waveguides R14 and R22 according to IEEE std. C The waveguides are used below the cut off frequency of all higher order modes, hence only the dominant TE 10 mode is present. On the basis of waveguide theory, the electric field in the rectangular waveguide can be calculated from the RF power P propagating in the waveguide. Equation (1) can be used for calculating the electric field E in the center of the waveguide, where a is the width and b is the height of the waveguide, c is the velocity of light in vacuum and f is the frequency. E = 2P ab 120π c 1 2af In the frequency range (1 1.7) GHz, the primary standard of CMI for calibrating small E-field sensors is consisting of a 80 cm long straight section of R14 waveguide, N-type to waveguide transition and waveguide termination. Similarly, in the frequency range ( ) GHz a 40 cm long straight section of R22 waveguide, N-type to waveguide transition and waveguide termination is used. The schematic diagram of the measurement setup (fig. 1) is identical for both frequency bands. 2 (1) 2-resistor power splitter Power sensor Signal generator Straight waveguide section N-type to waveguide adapter Probe Waveguide termination Fig.1 Schematic diagram of the measurement setup In order to improve the accuracy, errors due to reflections and losses were, under some simplifying assumptions, corrected for. Assuming the probe position in the center of the straight waveguide section and neglecting this waveguide section reflections, the correction factor regarding the standing waves is

28 EURAMET.EM.RF-S25 Final Report page 28 of 120 Page 3 k s 1+ ΓL s21 W =, (2) where Γ L is the reflection coefficient of the waveguide termination and s 21W is transmission coefficient of the straight waveguide section. The correction factor regarding the N-type to waveguide adapter reflections and losses is k p s s 21A 21W = A 21W ( s s Γ ) L, (3) where s xxa are the scattering parameters of the N-type to waveguide adapter. Thus, the final equation for electric field calculation is E 2P 120π i 21A 21W = + ΓL s21 W 2 ab A 21W c 1 2af 1 s s ( s s Γ ) L, (4) where P i is the power indicated by the power meter. Scattering parameters of the waveguide components were measured by a vector network analyzer combining TRL (waveguide) and OSL (Ntype) calibration. List of equipment: Power sensor NRV-Z51, ROHDE & SCHWARZ, ser. no /010 Power meter NRVS, ROHDE & SCHWARZ, ser. no Signal generator GT9000S, Giga-tronics, ser. no Power splitter 11667A, HEWLETT PACKARD, ser. no Waveguide R14 straight section QAV , Tesla HTT ser. no Adapter N/R14 QBV , Tesla HTT ser. no Waveguide R14 termination QGV , Tesla HTT ser. no. 001 Waveguide R22 straight section QAV /S, Tesla HTT ser. no Adapter N/R22 QBV , Tesla HTT ser. no. 002 Waveguide R22 termination QGV , Tesla HTT ser. no Measurement results: Field Strength (V/m) Frequency (GHz) Uncertainty (V/m) k=1 Table 1 The electric field strength values required for reading of 20 V/m

29 EURAMET.EM.RF-S25 Final Report page 29 of 120 Page 4 Traceability of measurements: In the field strength realisation used, RF power and length are the quantities of principal importance. Measurements of RF power are traceable to PTB through the thermistor power sensor 8478B calibration (uncertainty of 0.2 % in the frequency range of interest). Dimensional measurements are traceable to the primary standard of CMI. Uncertainty budgets: Source of uncertainty Type of uncertainty Estimated value (%) Degree of freedom Power meter accuracy B 0.50 Inf. Power splitter asymmetry B 0.52 Inf. Mismatch splitter/sensor B 0.03 Inf. Mismatch splitter/waveguide B 0.48 Inf. Waveguide dimension (a) B 0.52 Inf. Waveguide dimension (b) B 0.17 Inf. TFS meter resolution B 0.14 Inf. Sensor position - transversal B 0.09 Inf. Sensor position - longitudinal B 0.10 Inf. Sensor position - angle B 0.32 Inf. Temperature effect B 0.35 Inf. Error of VSWR correction B 0.85 Inf. Attenuation error B 0.46 Inf. Repeatability A Overall combined uncertainty 1.50 >10000 Expanded uncertainty (k=2) 3.00 Table 2. Uncertainty budget for the 1.00 GHz measurement Source of uncertainty Type of uncertainty Estimated value (%) Degree of freedom Power meter accuracy B 0.50 Inf. Power splitter asymmetry B 0.52 Inf. Mismatch splitter/sensor B 0.03 Inf. Mismatch splitter/waveguide B 0.04 Inf. Waveguide dimension (a) B 0.25 Inf. Waveguide dimension (b) B 0.26 Inf. TFS meter resolution B 0.14 Inf. Sensor position - transversal B 0.22 Inf. Sensor position - longitudinal B 0.05 Inf. Sensor position - angle B 0.32 Inf. Temperature effect B 0.35 Inf. Error of VSWR correction B 1.13 Inf. Attenuation error B 0.46 Inf. Repeatability A Overall combined uncertainty 1.57 >10000 Expanded uncertainty (k=2) 3.13 Table 3. Uncertainty budget for the 2.00 GHz measurement

30 EURAMET.EM.RF-S25 Final Report page 30 of 120 A2 NMi-VSL

31 EURAMET.EM.RF-S25 Final Report page 31 of 120 NMi-VSL report S-O&O-EL-HF 0801 Table of Contents Table of Contents Introduction Abstract Description of the generating facilities Mini TEM cell Micro TEM cell EM field generating circuitry Geometrical positioning facilities Description of the travelling standard Measurement set up Measurement conditions Actual measurements Repeatability Reproducibility Measurement programs used Traceability Measurement uncertainty Derivation of the model equation Power measurement Mismatch losses Electric field strength calculation Standing waves Form factor Temperature sensitivity Overall model equation Evaluation of parameter values and uncertainties Type-A evaluation Type-B evaluation Result Euramet project no. 819 NMi VSL contribution report nr. S-O&O-EL-HF 0801 Page 2 of 18

32 EURAMET.EM.RF-S25 Final Report page 32 of 120 NMi-VSL report S-O&O-EL-HF Introduction The purpose of the project is an international comparison of the generation of the electrical component of electromagnetic fields in the frequency range of 1 GHz up to 2,5 GHz, at a level considered as an optimum between the customary level for immunity measurements and for exposure of human beings, next to the properties of the travelling standard itself. The measuring device was sent round by the coordinator in the manner described in the project protocol. The measurements have been performed in the period of 23 November 2007 until 28 December 2007 using the EM field facilities of the NMi Van Swinden Laboratorium (NMi VSL) in Delft, The Netherlands. The results are presented in this report. 2 Abstract Presented is the result of the series of measurements performed, while exposing the Travelling Standard (TS) to an EM field within the two similar NMi VSL RF frequency EM-field facilities called mini TEM cell and micro TEM cell respectively. During Euromet 520 intercomparison the same mini TEM cell has been used. The measurements in the micro TEM can be considered as the first serious use (calibration) after the evaluation of this facility and the calculation of the uncertainty contribution. From this results it can be concluded that the contribution of the reflection coefficient of the cell is and the positioning of the probe (rotation 180 degrees) inside the cell are the predominant factors in the total measurement uncertainty value. The magnitude of the mentioned contributions however are within the range which could be expected. It also can be concluded that a systematic difference in the result occurred depending on the rotational orientation (0 degrees versus 180 degrees axial rotation) of the travelling standard. 3 Description of the generating facilities 3.1 Mini TEM cell A mini TEM cell has been used for the generation of the required EM fields from 1 GHz up to 1,5 GHz. This cell was constructed by NMi VSL using in company experiences in construction of look alike larger cells. The inner dimensions of the square mid section of the cell are nominal 120 x 120 mm. (a=b=60 mm) The tapered ends each are constructed out of one solid block of aluminium. The mid section consists of 3 separate almost identical slices, each of them constructed out of one solid block of aluminium. The cell has a 1 mm thick brass septum connected with special brass pencil shape rods to N-type connectors and is supported by foam material in the first and third part of the mid section. Euramet project no. 819 NMi VSL contribution report nr. S-O&O-EL-HF 0801 Page 3 of 18

33 EURAMET.EM.RF-S25 Final Report page 33 of 120 NMi-VSL report S-O&O-EL-HF 0801 Fig. 1 Probe in mini TEM cell 3.2 Micro TEM cell A micro TEM cell has been used for the generation of the required EM fields from 1 GHz up to 2,5 GHz. Also this cell was constructed by the NMi VSL workshop using in company experiences in construction of look alike larger cells. The inner dimensions of the square mid section of the cell are nominal 60 x 60 mm. (a=b=30 mm) The tapered ends are constructed each out of one solid block of aluminium. The mid section consists of 1 section constructed out of one solid block of aluminium. The cell has a 1 mm thick brass septum connected with special brass pencil shape rods to N-type connectors and is supported in the mid section by foam material. 3.3 EM field generating circuitry The following circuitry has been used: For the generation of the field in the cells in the frequency band from 1 GHz until 2,5 GHz the following circuitry was used: - a RF power generator was connected to a (nominal) 10 db power attenuator, which in its turn was connected to the TEM cell (port 1); - a RF power measurement device directly connected the output of the cell (port 2) terminated the transmission line. 3.4 Geometrical positioning facilities For the positioning purposes the probe was mounted on a 1 axis positioning lift. A vernier calliper was used to bring the probe axis in line with the cell. Euramet project no. 819 NMi VSL contribution report nr. S-O&O-EL-HF 0801 Page 4 of 18

34 EURAMET.EM.RF-S25 Final Report page 34 of 120 NMi-VSL report S-O&O-EL-HF Description of the travelling standard A PTB developed field strength measuring instrument was used as travelling standard. This device consists of a one directional E-field probe (dipole) operating in the frequency band to about 2,5 GHz. The probe is connected by means of an high impedance line to an electronics box in which the electrical signal is converted to an optical signal. An optical cable including an optical to electrical converting interface, is used to connected to a USB port of the PC. The measurement values can be gained using some special software installed on the PC. 5 Measurement set up The probe was installed in the centre of the higher part of the cells using an one-axis adjustable support. A vernier calliper was used to line-out the probe. In this way it was managed to install the probe within a radius of 0,5 mm from this geometrical centre. All readings were registered using the accompanying software installed on a PC. Fig. 2 Installed in micro TEM cell with probe cable upward 6 Measurement conditions During the measurements the temperature of the ambient has been continuously monitored. Also the TEM cell temperature has been registered and at the start of each measurement cycle this value the probe was corrected for its value using the probe software. Measurements have been performed at an ambient temperature of (23,0 ± 0,4) C and an ambient relative air humidity of (45 ± 10) % and a cell temperature of (23,1 ± 0,1 up to 23,7 ± 0,1) C 7 Actual measurements Measurements have been performed at all the frequencies stated in the protocol at a nominal indicated level on the TS of 20 V/m keeping the actual reading of the TS within 19,5 and 20,1 V/m. Euramet project no. 819 NMi VSL contribution report nr. S-O&O-EL-HF 0801 Page 5 of 18

35 EURAMET.EM.RF-S25 Final Report page 35 of 120 NMi-VSL report S-O&O-EL-HF Repeatability 4 cycles of measurements were performed. Each cycle consisted of at least 7 measurements as to measure the repeatability. 7.2 Reproducibility The reproducibility was tested by removing and re-installing the probe and turning it around its ( horizontal) axis between the measurement cycles and through exposing the probe in the two different facilities at all frequencies to the same field strength level. Note: The results appeared to be completely in line with each other considering the uncertainty in the measurements. Fig. 3 Installed in micro cell with probe cable upward 7.3 Measurement programs used The software program, which was supplied together with the TS in the same carriage case, was applied during the measurements. The measured temperature of the TEM cells and the signal frequency were entered in this software before each measurement cycle. Separate VSL TEM cell software was used for setting the frequency and field strength level. This software contains an automated procedure, in which the field strength and the desired series of signal frequencies are pre-programmed. The actual field strength is automatically calculated by the program from the power meter reading, using various calibration constants and parameters taken from lookup tables. The measured values are stored and printed after each run. 8 Traceability The results of the measurements are based on the measurement of the RF power measured at the end of the cell; the geometry of the main section of the cell, the alignment of the probe, the transmission line properties of the cell and power measurement head. Euramet project no. 819 NMi VSL contribution report nr. S-O&O-EL-HF 0801 Page 6 of 18

36 EURAMET.EM.RF-S25 Final Report page 36 of 120 NMi-VSL report S-O&O-EL-HF 0801 Traceability to primary and/or (inter)national standards is assumed for the geometric measurements. Calibration of the electrical devices has been performed by the HF section of the electrical standards department using a Vector Network Analyser A vernier calliper, calibrated by the NMi VSL mechanical standards department, has been used for the geometric measurements. 9 Measurement uncertainty 9.1 Derivation of the model equation P m P in P E P s Power measurement P in P P s P m = Power entering into the TEM cell (not used for the calculations) = Power passing through the TEM cell at the position of the TS = Power entering the power meter sensor = Power reading on the power meter No attenuator is used at the output of the TEM cell. P = P m k(a) k(f) M tot k(a) = Correction factor for the power meter to account for non-linearities with respect to the power level k(f) = Correction factor for the power meter to account for the frequency response M tot = Mismatch factor = P/P s Mismatch losses Mismatch causes a difference in power between a source S and a load L, which is expressed as a mismatch factor M SL. The magnitude and phase of this factor vary with the frequency and the position in the transmission line, but the extremes are given by the interval: [ M SL ] max min = 1 ± 2 where Γ S and Γ L are the reflection coefficients of the source (TEM cell) and the load (power sensor) respectively. The expectation value of the mismatch factor is 1, being the middle of the interval for M SL. Applying this to the output of the TEM cell, we can make the following observations: Γ S Γ L Euramet project no. 819 NMi VSL contribution report nr. S-O&O-EL-HF 0801 Page 7 of 18

37 EURAMET.EM.RF-S25 Final Report page 37 of 120 NMi-VSL report S-O&O-EL-HF 0801 There is no information on the relative positions of the connectors and other places where mismatch occurs, except that we know that the distances are very small compared to the wavelength, so a net overall effect Γ 2 = Γ S Γ L can be assumed between the TEM cell and the power sensor. As a worst case estimate for the uncertainty we take the extreme values: ±2 Γ 2, with a U-shaped probability, resulting in a standard uncertainty of 2 Γ 2. The reflection coefficients can be evaluated by scalar network analysis from the input of the TEM cell. From the pattern obtained over the whole frequency spectrum of interest, a worst case estimate can be made for the combined mismatch Γ 2 at the output of the TEM cell. Thus the expectation value of M tot is 1 with a standard uncertainty equal to 2 Γ 2. N.B: The relationship between return loss a and Γ is: a = -10 log 10 ( Γ 2 ) Electric field strength calculation The transverse electric field E is given by the formula: 1000 E = P Z L (1 + SW ) f ( x, y) d E = electric field strength at position (x,y), in V/m d = inner distance between septum and upper wall of TEM cell, in mm P = actual power in TEM cell, in watts Z L = 50 Ω characteristic impedance of the system SW = correction term for standing waves in the TEM cell, scalar f(x,y,) = Form factor for the field strength as a function of the position in the TEM cell, scalar x is the horizontal transferral position, y is the vertical position Standing waves Standing waves in the TEM cell are caused by various reflections, which are also the origins of the mismatch losses. Assuming that the reflection coefficients at the input and output of the TEM cell are small, only the first order reflection is considered (i.e. the reflected wave travelling to the left from the output terminal of the cell). Because the phase of the incoming and reflected waves is not known, the correction term has an expectation value 0. The probability distribution is U-shaped. The same evaluation of the combined mismatch Γ 2 is used to estimate the worst case effect of the standing waves: 1 VSWR 1 Γ u( SW ) = = 2 VSWR Form factor The VSL TEM cells that were used for these measurements have a w/b ratio of 0,83, where w is the width of the septum and b inner distance between the side walls of the cell. The form factor in the centre between the septum and the top wall has been derived from the tables in ref.1. Also the gradients (in x- and y-directions) of the field strength have been evaluated on the hand of these tables. In the centre the gradient in the x-direction is actually 0, due to symmetry. The curvature (δe/δx) is so small, that an uncertainty of 5 mm in the measurement probe position in the x-direction leads to a Euramet project no. 819 NMi VSL contribution report nr. S-O&O-EL-HF 0801 Page 8 of 18

38 EURAMET.EM.RF-S25 Final Report page 38 of 120 NMi-VSL report S-O&O-EL-HF 0801 negligible uncertainty contribution in the form factor. The gradient in the y-direction, however, can lead to a significant uncertainty contribution Temperature sensitivity During the measurements the temperature of the TEM cell was measured within an accuracy smaller than 0,1 K, while the environmental temperature was also monitored and seen to be very close to the measured temperature of the TEM cell (within 0,3 K). The measured temperature of the TEM cell was entered into the field strength measurement program, so that corrections were automatically effected in the measuring program. Under the assumption that the TS has the same temperature as the TEM cell (within 0,1 K), the uncertainty contribution due to the temperature sensitivity of the TS can be neglected Overall model equation The final measurement result is expressed as a normalised field strength: E nor = E n /E m E m is the field strength reading from the transfer system. Included is a term E res to cater for the uncertainty due to the resolution of the field strength transfer system readout (resolution is 0,01 V/m). The resulting model equation thus becomes: E nor En = E + E m res 1000 d P m k( A) k1( f ) k2( f ) k3( f ) Mtot D ZL (1 + SW) f ( x, y) 9.2 Evaluation of parameter values and uncertainties Type-A evaluation The measurements consist of pairs of readings (power meter, TS). In each measurement run these pairs are recorded for a series of frequencies. In order to avoid the effects of possible covariance, the type-a evaluation is obtained from the spread in the E n values, taken from a number of independent runs. The reproducibility is investigated by removing and reinserting the field strength probe between the runs. The symmetry is checked by rotating the probe around its axis by 180 degrees. Since a systematic difference in the measurements was detected depending on these 2 rotational positions their results have been presented separately. The type A evaluation on the readings therefore is based on each of these two positions separately. For the uncertainty evaluation only one overall budget is presented. The approach has been conservative by taking into account for each frequency only the highest of the two type A evaluated values. The results are detailed in the tables below Type-B evaluation E m The spread in the readings is investigated by type-a evaluation as indicated above. E res The expectation value for the resolution term is 0. The resolution of the readout of the TS is 0,01 V/m, leading to a uniform distribution of half-width 0,005 V/m. Euramet project no. 819 NMi VSL contribution report nr. S-O&O-EL-HF 0801 Page 9 of 18

39 EURAMET.EM.RF-S25 Final Report page 39 of 120 NMi-VSL report S-O&O-EL-HF 0801 d P m The distance was measured with a calibrated digital depth calliper, using a small hole in the top wall of the TEM cells. The value measured was 59,50 mm (mini cell) resp. 29,70 mm (micro cell), with an uncertainty of 0,1 mm (half-width of uniform distribution). The power readings are automatically translated into field strength values E by the measurement program. The resolution has a negligible uncertainty contribution. The type-a contribution from the spread in readings is already taken care of (see above). k(a) The calibration constant for the power dependent non-linearity of the power sensor has been determined at (a frequency of 50 MHz ) very close to the levels used during the field strength measurements. The values found were 1,000 with an uncertainty (k=2, normal distribution) of 0,002 k1(f) The frequency response was taken from the calibration data of the sensor-meter combination. The minor corrections compared to the internal frequency correction of the power meter were taken into account Value: 1,0000 (1 GHz) up to 1,0013 (2,5 GHz) (due to the internal linearization of the power meter) Uncertainty, including drift and temperature sensitivity: 1,7 % (k=2, normal distribution). k2(f) The frequency dependent attenuation (S21) which was taken from the calibration data of the cell. (Value: mini cell approx 0,5 % at 1 GHz and micro cell approx 0,7 % at 1GHz to approx 1,0 % at 2,5 GHz). Uncertainty: 0,05 % (k=2, normal distribution). M tot Mismatch term: value = 1. Uncertainty, evaluated from Scalar Network Analysis: mini cell 0,01 % (1 GHz) up to 0,1 % (1,5 GHz); micro cell 0,01 % (1 GHz) up to 0,2 % (2,5 GHz) (half-width of interval, U-shaped distribution). Z L The nominal impedance is 50 Ω, which is taken as the actual value. The uncertainty has been estimated to be 0,2 Ω (half-width of uniform distribution). SW Estimated value = 0. Uncertainty, from Vector Network Analysis: Mini cell 0,010 up to 0,030 (half-width of interval, U-shaped distribution); Micro cell 0,010 (1GHz) to 0,040 (2,5 GHz). f(x,y) Calculated value 0,99. Uncertainty, from calculated y-gradient (0,00958 per mm) and estimated uncertainty of y-position of probe: 0,00958[mm -1 ] 0,3[mm] = 0,00287 (half-width of uniform distribution) Euramet project no. 819 NMi VSL contribution report nr. S-O&O-EL-HF 0801 Page 10 of 18

40 EURAMET.EM.RF-S25 Final Report page 40 of 120 NMi-VSL report S-O&O-EL-HF Result In the following tables the results of the measurements have been presented. Results in Mini TEM cell facility. Probe rotated such that probe cable lays upwards Probe rotated such that probe cable lays downwards frequency field strength meter field strength reading value ambient temperature DC offset field strength meter average reading for frequency field strength meter field strength reading value ambient temperature DC offset field strength meter average reading for 0V/m 0V/m (GHz) (V/m) (V/m) ( C) (mv) (V/m) (GHz) (V/m) (V/m) ( C) (mv) (V/m) 1 20,00 19,04 23,35 0,289 0, ,00 18,75 23,50 0,289 0,50 20,00 19,08 23,48 0,289 0,50 20,00 18,62 23,34 0,300 0,20 20,00 18,71 23,33 0,300 0,20 20,00 18,60 23,42 0,300 0,20 20,00 18,82 23,48 0,300 0,20 20,00 18,62 23,44 0,300 0,20 20,00 18,82 23,52 0,300 0,20 20,00 18,64 23,45 0,300 0,20 20,00 18,82 23,45 0,300 0,30 20,00 18,62 23,45 0,300 0,20 20,00 18,82 23,42 0,300 0,10 20,00 18,67 23,46 0,300 0,20 20,00 18,84 23,39 0,300 0,10 1,25 20,00 19,38 23,35 0,289 0,50 1,25 20,00 18,91 23,50 0,289 0,50 20,00 19,42 23,48 0,289 0,50 20,00 18,85 23,34 0,300 0,20 20,00 19,11 23,33 0,300 0,20 20,00 18,85 23,43 0,300 0,20 20,00 19,20 23,47 0,300 0,20 20,00 18,85 23,45 0,300 0,20 20,00 19,20 23,50 0,300 0,20 20,00 18,85 23,45 0,300 0,20 20,02 19,22 23,44 0,300 0,30 20,00 18,85 23,44 0,300 0,20 20,00 19,20 23,40 0,300 0,10 20,00 18,85 23,44 0,300 0,20 20,00 19,22 23,41 0,300 0,10 Euramet project no. 819 NMi VSL contribution report nr. S-O&O-EL-HF 0801 Page 11 of 18

41 EURAMET.EM.RF-S25 Final Report page 41 of 120 NMi-VSL report S-O&O-EL-HF 0801 Results in Mini TEM cell facility. Probe rotated such that probe cable lays downwards Probe rotated such that probe cable lays upwards 1,5 20,00 19,49 23,36 0,289 0,50 1,5 20,00 19,18 23,50 0,289 0,50 20,00 19,53 23,48 0,289 0,50 20,00 19,02 23,34 0,300 0,20 20,00 19,16 23,33 0,300 0,20 20,00 19,09 23,43 0,300 0,20 20,00 19,36 23,46 0,300 0,20 20,00 19,11 23,45 0,300 0,20 20,00 19,33 23,51 0,300 0,20 20,00 19,09 23,45 0,300 0,20 20,00 19,36 23,44 0,300 0,30 20,00 19,13 23,45 0,300 0,20 20,00 19,33 23,40 0,300 0,10 20,00 19,11 23,45 0,300 0,20 20,00 19,36 23,41 0,300 0,10 Overall result in Mini TEM cell. Field strength values with standard uncertainty (k=1) for field strength meter reading of 20 V/m frequency (GHz) field strength (V/m) frequency (GHz) field strength (V/m) 1,00 18,87 ± 0,19 1,00 18,65 ± 0,19 1,25 19,24 0,42 1,25 18,86 0,42 1,50 19,36 0,43 1,50 19,11 0,43 1,75 1,75 2,00 2,00 2,25 2,25 2,50 2,50 Probe rotated such that probe cable lays upwards Probe rotated such that probe cable lays downwards Euramet project no. 819 NMi VSL contribution report nr. S-O&O-EL-HF 0801 Page 12 of 18

42 EURAMET.EM.RF-S25 Final Report page 42 of 120 NMi-VSL report S-O&O-EL-HF 0801 Results in Micro TEM cell facility. Probe rotated such that probe cable lays upwards Probe rotated such that probe cable lays downwards frequency field strength meter field strength reading value ambient temperature DC offset field strength meter average reading for frequency field strength meter field strength reading value ambient temperature DC offset field strength meter average reading for 0V/m 0V/m (GHz) (V/m) (V/m) ( C) (mv) (V/m) (GHz) (V/m) (V/m) ( C) (mv) (V/m) 1 20,00 18,65 23,19 0,310 0,1 1 20,00 18,25 23,31 0,310 0,0 20,00 18,69 23,17 0,310 0,1 20,00 18,21 23,29 0,300 0,2 20,00 18,72 23,16 0,310 0,1 20,00 18,23 23,26 0,300 0,2 20,00 18,67 23,15 0,310 0,1 20,00 18,19 23,24 0,300 0,2 20,00 18,72 23,13 0,310 0,1 20,00 18,23 23,14 0,300 0,2 20,00 18,72 23,14 0,310 0,1 20,00 18,19 23,13 0,300 0,2 20,00 18,69 23,11 0,310 0,1 20,00 18,19 23,11 0,300 0,2 20,00 18,72 23,10 0,310 0,0 20,00 18,19 23,11 0,300 0,2 20,00 18,72 23,10 0,310 0,0 20,00 18,16 23,09 0,300 0,2 20,00 18,74 23,10 0,310 0,0 20,00 18,19 23,07 0,300 0,2 1,25 20,00 18,78 23,18 0,310 0,1 1,25 20,00 18,18 23,31 0,310 0,0 20,00 18,75 23,17 0,310 0,1 20,00 18,14 23,29 0,300 0,2 20,00 18,78 23,16 0,310 0,1 20,00 18,14 23,27 0,300 0,2 20,00 18,78 23,16 0,310 0,1 20,00 18,14 23,24 0,300 0,2 20,00 18,82 23,14 0,310 0,1 20,00 18,14 23,14 0,300 0,2 20,00 18,82 23,14 0,310 0,1 20,00 18,10 23,13 0,300 0,2 20,00 18,78 23,12 0,310 0,1 19,98 18,12 23,12 0,300 0,2 20,00 18,80 23,11 0,310 0,0 20,00 18,10 23,10 0,300 0,2 20,00 18,84 23,09 0,310 0,0 20,00 18,08 23,09 0,300 0,2 20,00 18,80 23,09 0,310 0,0 20,00 18,12 23,07 0,300 0,2 Euramet project no. 819 NMi VSL contribution report nr. S-O&O-EL-HF 0801 Page 13 of 18

43 EURAMET.EM.RF-S25 Final Report page 43 of 120 NMi-VSL report S-O&O-EL-HF 0801 Results in Micro TEM cell facility. Probe rotated such that probe cable lays upwards Probe rotated such that probe cable lays downwards frequency field strength meter field strength reading value ambient temperature DC offset field strength meter average reading for frequency field strength meter field strength reading value ambient temperature DC offset field strength meter average reading for 0V/m 0V/m (GHz) (V/m) (V/m) ( C) (mv) (V/m) (GHz) (V/m) (V/m) ( C) (mv) (V/m) 1,5 20,00 19,00 23,66 0,310 0,1 1,5 20,00 18,50 23,15 0,310 0,0 20,00 19,26 23,34 0,310 0,1 20,00 18,59 23,52 0,300 0,2 20,00 19,13 23,18 0,310 0,1 20,00 18,63 23,30 0,300 0,2 20,00 19,15 23,17 0,310 0,1 20,00 18,61 23,29 0,300 0,2 20,00 19,15 23,16 0,310 0,1 20,00 18,61 23,27 0,300 0,2 20,00 19,15 23,16 0,310 0,1 20,00 18,61 23,24 0,300 0,2 20,00 19,20 23,13 0,310 0,1 20,00 18,61 23,14 0,300 0,2 20,00 19,20 23,13 0,310 0,0 20,00 18,59 23,12 0,300 0,2 20,00 19,15 23,11 0,310 0,0 20,00 18,59 23,12 0,300 0,2 20,00 19,17 23,11 0,310 0,0 20,00 18,59 23,10 0,300 0,2 1,75 20,00 18,52 23,18 0,310 0,1 1,75 20,00 17,96 23,30 0,310 0,0 20,00 18,52 23,17 0,310 0,1 20,00 17,93 23,28 0,300 0,2 20,00 18,54 23,16 0,310 0,1 20,00 17,96 23,26 0,300 0,2 20,00 18,54 23,15 0,310 0,1 20,00 17,93 23,24 0,300 0,2 20,00 18,56 23,12 0,310 0,1 20,00 17,96 23,14 0,300 0,2 20,00 18,56 23,13 0,310 0,1 20,00 17,89 23,12 0,300 0,2 20,00 18,54 23,11 0,310 0,1 20,00 17,89 23,10 0,300 0,2 20,00 18,59 23,11 0,310 0,0 20,00 17,93 23,10 0,300 0,2 20,00 18,56 23,09 0,310 0,0 20,00 17,89 23,08 0,300 0,2 20,00 18,56 23,08 0,310 0,0 20,00 17,91 23,06 0,300 0,2 Euramet project no. 819 NMi VSL contribution report nr. S-O&O-EL-HF 0801 Page 14 of 18

44 EURAMET.EM.RF-S25 Final Report page 44 of 120 NMi-VSL report S-O&O-EL-HF 0801 Results in Micro TEM cell facility. Probe rotated such that probe cable lays upwards Probe rotated such that probe cable lays downwards frequency field strength meter field strength reading value ambient temperature DC offset field strength meter average reading for frequency field strength meter field strength reading value ambient temperature DC offset field strength meter average reading for 0V/m 0V/m (GHz) (V/m) (V/m) ( C) (mv) (V/m) (GHz) (V/m) (V/m) ( C) (mv) (V/m) 2 20,00 18,02 23,18 0,310 0,1 2 20,00 17,59 23,30 0,310 0,0 20,00 18,04 23,16 0,310 0,1 20,00 17,53 23,28 0,300 0,2 20,00 18,04 23,16 0,310 0,1 20,00 17,57 23,25 0,300 0,2 20,00 18,06 23,15 0,310 0,1 20,00 17,55 23,23 0,300 0,2 20,00 18,09 23,13 0,310 0,1 20,00 17,57 23,14 0,300 0,2 20,00 18,09 23,14 0,310 0,1 20,00 17,53 23,13 0,300 0,2 20,00 18,04 23,11 0,310 0,1 20,00 17,55 23,11 0,300 0,2 20,00 18,09 23,11 0,310 0,0 20,00 17,53 23,09 0,300 0,2 20,00 18,06 23,09 0,310 0,0 20,00 17,51 23,08 0,300 0,2 20,00 18,06 23,08 0,310 0,0 20,00 17,53 23,07 0,300 0,2 2,25 20,00 18,40 23,19 0,310 0,1 2,25 20,03 17,93 23,31 0,310 0,0 20,00 18,42 23,16 0,310 0,1 20,00 17,88 23,29 0,300 0,2 20,00 18,42 23,16 0,310 0,1 19,98 17,88 23,25 0,300 0,2 20,00 18,44 23,15 0,310 0,1 20,00 17,88 23,23 0,300 0,2 19,97 18,46 23,13 0,310 0,1 20,00 17,86 23,13 0,300 0,2 20,00 18,44 23,14 0,310 0,1 20,00 17,84 23,12 0,300 0,2 20,00 18,42 23,12 0,310 0,1 20,00 17,86 23,11 0,300 0,2 20,00 18,42 23,10 0,310 0,0 20,00 17,84 23,10 0,300 0,2 20,00 18,46 23,08 0,310 0,0 20,00 17,84 23,08 0,300 0,2 20,00 18,46 23,09 0,310 0,0 20,00 17,86 23,06 0,300 0,2 Euramet project no. 819 NMi VSL contribution report nr. S-O&O-EL-HF 0801 Page 15 of 18

45 EURAMET.EM.RF-S25 Final Report page 45 of 120 NMi-VSL report S-O&O-EL-HF 0801 Results in Micro TEM cell facility. Probe rotated such that probe cable lays upwards Probe rotated such that probe cable lays downwards frequency field strength meter field strength reading value ambient temperature DC offset field strength meter average reading for 0V/m frequency field strength meter field strength reading value ambient temperature DC offset field strength meter average reading for 0V/m (GHz) (V/m) (V/m) ( C) (mv) (V/m) (GHz) (V/m) (V/m) ( C) (mv) (V/m) 2,5 20,00 18,34 23,18 0,310 0,1 2,5 20,00 17,79 23,30 0,310 0,0 20,00 18,36 23,17 0,310 0,1 20,00 17,75 23,29 0,300 0,2 20,00 18,38 23,17 0,310 0,1 20,00 17,75 23,25 0,300 0,2 20,00 18,34 23,14 0,310 0,1 20,00 17,77 23,22 0,300 0,2 20,00 18,38 23,14 0,310 0,1 20,00 17,79 23,12 0,300 0,2 20,00 18,38 23,13 0,310 0,1 19,80 17,74 23,10 0,300 0,2 19,98 18,36 23,12 0,310 0,1 20,00 17,75 23,11 0,300 0,2 20,00 18,36 23,11 0,310 0,0 20,00 17,73 23,09 0,300 0,2 20,00 18,40 23,09 0,310 0,0 20,00 17,73 23,08 0,300 0,2 20,00 18,38 23,09 0,310 0,0 20,00 17,73 23,07 0,300 0,2 Euramet project no. 819 NMi VSL contribution report nr. S-O&O-EL-HF 0801 Page 16 of 18

46 EURAMET.EM.RF-S25 Final Report page 46 of 120 NMi-VSL report S-O&O-EL-HF 0801 Overall result in Micro TEM cell. Field strength values with standard uncertainty (k=1) for field strength meter reading of 20 V/m frequency (GHz) field strength (V/m) frequency (GHz) field strength (V/m) 1,00 18,70 ± 0,17 1,00 18,20 ± 0,17 1,25 18,79 0,23 1,25 18,12 0,23 1,50 19,16 0,30 1,50 18,59 0,30 1,75 18,55 0,42 1,75 17,93 0,42 2,00 18,06 0,43 2,00 17,55 0,43 2,25 18,44 0,46 2,25 17,87 0,46 2,50 18,37 0,53 2,50 17,76 0,53 Rotated such that probe cable lays upwards Rotated such that probe cable lays downwards Euramet project no. 819 NMi VSL contribution report nr. S-O&O-EL-HF 0801 Page 17 of 18

47 EURAMET.EM.RF-S25 Final Report page 47 of 120 NMi-VSL report S-O&O-EL-HF 0801 Example of measurement uncetainty calculation table (Micro TEM cell at 1 GHz) Contributions: parameter unit value U i dist u(x i ) C i [C i ] u i (y) (V/m) u(x i ) 2 Field strength reading E m V/m 20,000 none Resolution UUT E res V/m 0,000 0,01 uniform 0, ,0029 0, Distance Septum-wall d mm 29,70 0,10 uniform 0, , V.m -1.mm -1 0,0366 0, Power reading P m W 0,00632 none Non-linearity of power sensor k(a) 1,000 0,002 normal 2s 0,001 9, V.m -1 0,0094 0, Frequency response of power sensor k1(f) 1,000 0,017 normal 2s 0,0085 9, V.m -1 0,0800 0, Inverse frequency dependant attenuation of the cell k2(f) 1,008 0,005 normal 2s 0,0025 9, V.m -1 0,0233 0, Mismatch losses M tot 1,000 0,000 U-shaped 0,0001 9, V.m -1 0,0009 0, TEM Cell impedance Z L Ω 50,00 0,20 uniform 0, , V.m -1. Ω -1 0,0217 0, Standing waves SW 0,000 0,010 U-shaped 0, ,818 V.m -1 0,1331 0, Form factor f(x,y) 0,990 0,00570 uniform 0, , V.m -1 0,0626 0, Calculated field strenght (intermediate result) E V/m 18,818 Type-A: independent repeat measurements E n V/m 18,818 0,01 normal 1s 0,01 1 0,0080 0, RESULT u(y) k U(y) Normalised field strength (result) E n V/m 18,818 0, ,00 0,349 V/m 0, Relative U(y)/E 0,019 rel Euramet project no. 819 NMi VSL contribution report nr. S-O&O-EL-HF 0801 Page 18 of 18

48 EURAMET.EM.RF-S25 Final Report page 48 of 120 A3 NPL

49 EURAMET.EM.RF-S25 Final Report RESTRICTED page 49 of 120 NPL Report MS001 NPL Contribution to EUROMET Comparison 819: Comparison of Electrical Field Strength Measurements above 1 GHz E R GOODALL Measurement Services Team National Physical Laboratory Teddington Middlesex United Kingdom, TW11 0LW ABSTRACT This report describes the measurement techniques used and results obtained for the calibration of an electric field strength probe under EUROMET comparison 819. The transfer probe was calibrated at 20 V/m at frequencies between 1 GHz and 2.5 GHz using three horns in an anechoic chamber. Uncertainty budgets are given for the three horns used.

50 EURAMET.EM.RF-S25 Final Report RESTRICTED page 50 of 120 NPL Report MS001 Crown copyright 2008 Reproduced with the permission of the Controller of HMSO and Queen's Printer for Scotland National Physical Laboratory Hampton Road, Teddington, Middlesex, TW11 0LW This report is Restricted - Commercial and must not be exposed to casual examination. It is not for general distribution and should not be cited as a reference other than in accordance with the contract. Approved on behalf of the Managing Director, NPL By Seton Bennett Director, Measurement Services

51 EURAMET.EM.RF-S25 Final Report page 51 of 120 RESTRICTED COMMERCIAL NPL Report MS001 CONTENTS Page 1. INTRODUCTION COMPARISON TRANSFER STANDARD MEASUREMENT SCHEDULE MEASUREMENTS MEASUREMENT RESULTS UNCERTAINTIES TABLES Table 1: The applied electric field strength required to produce an indicated reading of 20.0 V/m...5 Table 2: Uncertainty budget for the 1.00 GHz Measurement Table 3: Uncertainty budget for the 1.25 to 2.25 GHz Measurements Table 4: Uncertainty budget for the 2.50 GHz Measurement.....8

52 EURAMET.EM.RF-S25 Final Report page 52 of 120 RESTRICTED COMMERCIAL NPL Report MS INTRODUCTION This report describes the measurements carried out at NPL as part of EURAMET comparison 819: Comparison of Electrical Field Strength Measurements above 1 GHz. The pilot laboratory for the comparison is the Czech Metrology Institute (Czech Republic) with PTB (Germany) providing the transfer standard. The list of participating organisations is: CMI INRIM LNE METAS NMi NPL PTB SP STUK UP (Český metrologický institut), Czech Republic (Istituto Nazionale di Ricerca Metrologica), Italy (Laboratoire National de Métrologie et d essais), France (Swiss Federal Office of Metrology and Accreditation), Switzerland (Nederlands Meetinstituut-Van Swinden Laboratorium), Netherlands (National Physical Laboratory), United Kingdom (Physikalisch Technische Bundesanstalt), Germany (SP Technical Research Institute of Sweden), Sweden (Radiation and Nuclear Safety Authority), Finland (University of Pretoria), South Africa The measurements described in this report were carried out during the period January 2008 using the facilities and equipment used by NPL to provide a field strength calibration service to industry. Three separate horn antennas were used to cover the frequency range in an anechoic chamber. 2. COMPARISON TRANSFER STANDARD The transfer field strength sensor and metering unit, supplied by PTB, includes optical fibre connectors with PC interface and software to compensate for temperature variations and the frequency related characteristics of the probe. The active part of the sensor was in the form of a dipole with a Schottky detector, which was connected to the metering unit by high resistance leads. The sensor was mounted at the end of a thin walled plastic tube, which was connected to a plastic holder. The metering unit contained electronics and a battery, which amplifies and filters the analogue dc voltage, before converting it into a digital bit sequence. This is then sent to the computer via a fibre optical cable. The unit also has connection sockets for an external power supply/charger. During all the measurements reported here, the internal battery powered the metering unit. The measurement software provided a readout of the measured field strength and allowed the ambient temperature and measurement frequency to be inputted by the operator to apply software correction. A Perspex bracket and Tufnol tube was used by NPL to support the sensor. The flexible, high resistance leads passed along the Perspex bracket and Tufnol tube to the metering unit. The Tufnol tube could be readily rotated in a purpose built support mechanism. This arrangement was used for all measurements. 1

53 EURAMET.EM.RF-S25 Final Report page 53 of 120 RESTRICTED COMMERCIAL NPL Report MS MEASUREMENT SCHEDULE The measurement protocol required all measurements to be made at an indicated field strength of 20.0 V/m ± 0.5 V/m and at a temperature between 22 C and 24 C. For all measurements made, the applied field was adjusted so that the sensor registered 20.0 V/m. The ambient temperature was maintained between 23.6 C and 24.0 C for all measurements. The measurement frequencies were: 1.00 GHz, 1.25 GHz, 1.50 GHz, 1.75 GHz, 2.00 GHz, 2.25 GHz & 2.50 GHz. 4. MEASUREMENTS Three horn antennas in an anechoic chamber were used to generate a calculable electromagnetic field in which the transfer standard was placed. The general setup consists of a synthesised signal generator feeding a solid-state amplifier, which is connect via a low pass filter to the input of a directional coupler. A calibrated power meter on the side arm of the coupler monitors the power, P incident on an antenna connected to the output of the coupler. Knowing the gain, G of the antenna at the distance, R at which the calibration is performed, it is then possible to calculate the electric field strength, E. E = PGZ 4πR 0 2 [V/m] The power P is determined from the power indicated by the power meter attached to the side arm of the coupler. The precise relationship between the power accepted by the antenna and the power indicated on the power meter is given below for the case of: a) Coaxial coupler systems; 1.00 GHz to 2.25 GHz b) WG10 Waveguide coupler system; 2.5 GHz a) Coaxial coupler systems; 1.00 GHz to 2.25 GHz The power sensors and couplers are calibrated separately. The method of the coupler calibration is set out in an internal NPL procedure (QPCETM-B-392) and this technique allows any combination of coaxial coupler, power sensor and coaxial antenna to be used. The power available to a matched load on the side arm of the coaxial coupler, P cs is related to the power meter reading, P r by Pr Pcs = PMF 1- Γp Γ Where Γ p and Γ cs are the reflection coefficients of the power sensor and coupler sidearm respectively and the PMF is the Power Meter Calibration Factor. 2 cs 2

54 EURAMET.EM.RF-S25 Final Report page 54 of 120 RESTRICTED COMMERCIAL NPL Report MS001 Note that the term matched load means a load of identical impedance to that of the characteristic impedance of an ideal transmission line standard. This is sometimes referred to as the normalising impedance and is 50 ohms for coaxial transmission lines. The coupler ratio, CR, is measured using a calibrated Vector Network Analyser and gives the ratio of the powers available to a matched load at the output port to that of the side-arm port. Power available to a matched load at output port CR = = P Power available to a matched load on side- arm port P Note that CR is not the conventional coupling factor, which relates the side-arm power to the input power to the coupler. The power available to a matched load at the output port of the main-arm of the coupler, P cm, can therefore be expressed in terms of the reading on the power meter by: cm cs P cm CR Pr 1- ΓpΓ = PMF The power accepted by the antenna is given by: 2 ( 1- Γh ) Pcm P = 1- Γh Γ t 2 g cs 2 where P cm Γ h Γ g is the power available to a matched load at the output of the coupler. is the reflection coefficient of the horn antenna is the effective source reflection coefficient of the coupler main arm. The power accepted by the antenna can now be re-written in terms of the above quantities CR Pr 1- Γp Γcs P = PMF 1 Γh Γ t 2 g 2 2 ( 1- Γh ) Hence, in terms of the power meter reading P r, the power flux density can be written as: 2 CR Pr 1- Γp Γc s PFD= PMF 1- Γ Γ h 2 ( 1- Γh ) 2 g 4 G π R 2 2 [W/m ] 2 In this expression, G is the true gain of the antenna. The product of ( 1- Γt ) G in the above equation is the apparent gain. The remaining mismatch factors in the numerator and denominator are not evaluated, but treated as an uncertainty. 3

55 EURAMET.EM.RF-S25 Final Report page 55 of 120 RESTRICTED COMMERCIAL NPL Report MS001 b) WG10 Waveguide coupler system; 2.5 GHz Each of the couplers has been calibrated by connecting a calibrated coaxial power sensor to its output, using a calibrated waveguide to coaxial transformer, and noting the side arm and output powers at a number of frequencies across the operating band. This technique characterises the directional coupler and power sensor combination, along with any attenuators, which may be included in the circuit to optimise dynamic range, without having to calibrate each component independently. By this method, one obtains a series of calibration factors at spot frequencies across the band, which relates the output power to the indicated side arm power. Linear interpolation is used to obtain data at frequencies between the calibrated points. The coupler ratio, CR, is the ratio of the power available to a matched load to the power indicated on the side arm power meter: CR = Power available to a matched load Power indicated on side arm meter The power accepted by the antenna is given by: p P = m gh 2 ( 1 Γh ) 1 Γ Γ where P m is the power available to a matched load and Γ h and Γ g are the reflection coefficients of the horn and effective source respectively. The coupler ratio can be written in terms of the above quantities as: P CR = m where Pc S is the indicated side arm power meter reading. Pcs Hence, in terms of the indicated side arm power meter reading, Pc S, the power flux density can be rewritten as: 2 CR Pcs G( 1- Γh ) PFD = 2 2 4π R 1-Γh Γg 2 In this expression, G is the true gain of the antenna. The product of ( 1- Γt ) h g 2 G in the above equation is the apparent gain. The remaining mismatch factor in the denominator is not evaluated, but treated as an uncertainty. For both coupler systems, the distance R is measured from the aperture of the horn antenna to the centre of the probe sensing element(s). The quasi plane wave field is set up inside an open ended electromagnetic anechoic chamber to approximate free space conditions and thereby minimise errors due to reflections from the surroundings. The anechoic chamber is lined with 36" radiation absorbing material (RAM). 4

56 EURAMET.EM.RF-S25 Final Report page 56 of 120 RESTRICTED COMMERCIAL NPL Report MS MEASUREMENT RESULTS The values in table 1 are the applied electric field strength required to produce an indicated reading of 20.0 V/m. Table 1: The applied electric field strength required to produce an indicated reading of 20.0 V/m Frequency /(GHz) Applied Field Strength /(V/m) ± ± ± ± ± ± ± Note: the uncertainties quoted in the above table are for a coverage factor k=1. At each frequency point, the applied field strength was measured five times. The purpose being to ensure that the contributions in the uncertainty budget relating to repeatability (which are derived from historical measurements) are appropriate for these measurements. The standard deviation of the five measurements was well within the expected range and so the uncertainty contributions in question are considered adequate. 6. UNCERTAINTIES The uncertainty budgets for the three horns used for the measurements are given in tables 2, 3 & 4 and were compiled in terms of Power Flux Density (PFD) as most of the uncertainty data is readily available in terms of uncertainties in power measurements. To convert the combined uncertainty to give the uncertainty in the electric field strength one must divide the values by 2. The combined uncertainties for k=1 are as follows: ± 2.71 % for the frequency of 1.00 GHz ± 1.99 % for frequencies between 1.25 GHz and 2.25 GHz ± 1.76 % for the frequency of 2.50 GHz 5

57 EURAMET.EM.RF-S25 Final Report page 57 of 120 RESTRICTED COMMERCIAL NPL Report MS001 Table 2: Uncertainty budget for the 1.00 GHz Measurement Symbol Source of Uncertainty ±Value / (%) Probability Distribution Divisor C i ±u i / (%) υ i or υ eff C pr Coupler power ratio 2.30 Normal Inf P sa Power sensor accuracy 1.30 Normal Inf P a Power meter accuracy 1.00 Rectangular Inf Pr a Power meter reference 0.90 Rectangular Inf Pr d Power reference drift 0.75 Rectangular Inf Z s Power meter zero setting 0.75 Rectangular Inf C r Connector repeatability 0.30 Normal P l Power sensor linearity 1.20 Rectangular Inf M h Coupler/horn mismatch 1.30 U - shaped Inf M ps Coupler/Psensor Mismatch 0.70 U - shaped Inf H g Horn gain 9.20 Normal Inf D s Distance encoder 0.13 Rectangular Inf H rr Horn/radome reference 0.14 Rectangular Inf H pa Horn/probe alignment 0.37 Normal Inf R c Reflections in chamber 1.00 U - shaped Inf R hp Reflections horn/probe 0.50 U - shaped Inf S r Meter Resolution 0.50 Rectangular Inf D nd Device noise and drift 1.00 Normal Inf Z d Zero drift 1.00 Rectangular Inf T e Temperature effects 1.10 Rectangular Inf C rd Coupler ratio drift 1.00 Normal Inf u c Combined uncertainty normal 5.42 > u s Standard uncertainty normal (k=1) 5.42 Note: Uncertainties in the above table are for Power Flux Density. Halve the values for uncertainties in electric field strength 6

58 EURAMET.EM.RF-S25 Final Report page 58 of 120 RESTRICTED COMMERCIAL NPL Report MS001 Table 3: Uncertainty budget for the 1.25 GHz to 2.25 GHz Measurements Symbol Source of Uncertainty ±Value / (%) Probability Distribution Divisor C i ±u i / (%) υ i or υ eff C pr Coupler power ratio 2.30 Normal Inf P Sa Power sensor accuracy 1.30 Normal Inf P a Power meter accuracy 1.00 Rectangular Inf P ra Power meter reference 0.90 Rectangular Inf P rd Power reference drift 0.75 Rectangular Inf Z s Power meter zero setting 0.75 Rectangular Inf C r Connector repeatability 0.30 Normal P l Power sensor linearity 1.20 Rectangular Inf M h Coupler/horn mismatch 2.60 U - shaped Inf M ps Coupler/Psensor Mismatch 0.70 U - shaped Inf H g Horn gain 4.50 Normal Inf D s Distance encoder 0.13 Rectangular Inf H rr Horn/radome reference 0.14 Rectangular Inf H pa Horn/probe alignment 0.37 Normal Inf R c Reflections in chamber 1.00 U - shaped Inf R hp Reflections horn/probe 0.50 U - shaped Inf S r Meter Resolution 0.50 Rectangular Inf D nd Device noise and drift 1.00 Normal Inf Z d Zero drift 1.00 Rectangular Inf T e Temperature effects 1.10 Rectangular Inf C rd Coupler ratio drift 1.00 Normal Inf u c Combined uncertainty normal 3.98 > u s Standard uncertainty normal (k=1) 3.98 Note: Uncertainties in the above table are for Power Flux Density. Halve the values for uncertainties in electric field strength 7

59 EURAMET.EM.RF-S25 Final Report page 59 of 120 RESTRICTED COMMERCIAL NPL Report MS001 Table 4: Uncertainty budget for the 2.50 GHz Measurement Symbol Source of Uncertainty ±Value / (%) Probability Distribution Divisor C i ±u i / (%) υ i or υ eff C pr Coupler power ratio 4.10 Normal >10000 P a Power meter accuracy 1.00 Rectangular >10000 P ra Power meter reference 0.90 Rectangular >10000 P rd Power reference drift 0.75 Rectangular >10000 Z s P. meter zero setting 0.75 Rectangular >10000 C r Connector repeatability 0.30 Normal P l Power sensor linearity 1.20 Rectangular >10000 M Coupler/horn mismatch 1.44 U - shaped >10000 H g Horn gain 2.20 Normal >10000 D s Distance encoder 0.29 Rectangular >10000 H rr Horn/radome reference 0.14 Rectangular >10000 H pa Horn/probe alignment 0.37 Normal >10000 R c Reflections in chamber 1.00 U - shaped >10000 R hp Reflections horn/probe 2.00 U - shaped >10000 S r Meter Resolution 0.50 Rectangular >10000 D nd Device noise and drift 1.00 Normal >10000 Z d Zero drift 1.00 Rectangular >10000 H i Horn gain interpolation 1.90 Normal >10000 T e Temperature effects 1.10 Rectangular >10000 C rd Coupler ratio drift 1.00 Normal >10000 u c Combined Uncertainty Normal 3.51 > u s Standard uncertainty Normal (k=1) 3.51 Note: Uncertainties in the above table are for Power Flux Density. Halve the values for uncertainties in electric field strength 8

60 EURAMET.EM.RF-S25 Final Report page 60 of 120 A4 LNE

61 EURAMET.EM.RF-S25 Final Report page 61 of 120 A5 STUK

62 EURAMET.EM.RF-S25 Final Report page 62 of 120 MEASUREMENT REPORT 2 1 General information STUK has two types of realisations for the electric field strength above 1 GHz. Probes can be calibrated in anechoic chamber in front of an antenna or in rectangular waveguide. The latter method is more accurate but only suitable for probes with diameter less than 10 mm due to the size of the holes in the waveguide walls. The transfer standard of this comparison fulfils this requirement, hence the waveguide method was selected for the measurements. Two waveguide realisations have been implemented based on WR975 and WR430 standard waveguides and waveguide components. The frequencies 1.25 and 1.5 GHz are not covered by these, hence they were not measured in STUK. The detailed descriptions of the setups are presented in the following sections.

63 EURAMET.EM.RF-S25 Final Report page 63 of 120 MEASUREMENT REPORT 3 2 Description of the realisations of the electric field in STUK 2.1 Basic principle The electric field in the rectangular waveguide can be calculated from the RF power propagating in the waveguide based on the basic theory of waveguides. The waveguides are used below the cut off frequency of all higher order modes, and hence only the TE 10 is propagating. Thus, the equations 1 and 2 can be used for calculating the electric field in the centre line of the wave guide. The theory of the rectangular wave guides is discussed e.g. in the reference [1]. E rms 2Pp ZTE10 =, (1) ab Z TE 10 = η 0 λ 1 2a 2, (2) where E rms is the rms (root mean square) value of electric field strength in the centre line of the waveguide, P p is the RF power propagating in the waveguide, Z TE10 is the wave impedance for TE 10 mode, a is the width and b is the height of the waveguide GHz setup The primary calibration system of STUK for calibrating small E-field probes in the frequency range GHz is consisting of a 1.5 m long straight section of WR975 waveguide (width 248 mm, height 124 mm), N-type-to-waveguide transition and a slid-

64 EURAMET.EM.RF-S25 Final Report page 64 of 120 MEASUREMENT REPORT 4 ing waveguide termination as illustrated in the figure 1. The components are manufactured by Mega Industries Ltd., Gorham, ME, USA. The transfer standard was positioned with a jig that holds the probe handle perpendicular to the wall of the waveguide at the centre line of the side b. The positioning in the direction of the side a i.e. the depth to which the probe is inserted, is measured with a slide calliper. The probe should be positioned at the centre line of the waveguide, i.e. at the maximum of the electric field. The positioning is ascertained by slightly moving the probe up and down from the measured position to see that the meter reading decreases to both directions. Finally, the alignment of the dipole arms with the direction of the E- field was set by rotating the transfer standard around the axis of the probe handle to find the position giving the maximum E-field reading. The mismatch of the termination is compensated by finding the adjacent positions of the sliding termination giving the maximum and minimum E-field meter readings. The actual calibrations were made with the termination positioned in the middle of these two positions. The separation of the sliding termination giving the maximum and minimum should be quarter wavelengths in waveguide (λ g ). This was ascertained by calculating the λ g (equations 3 and 4) and comparing it to the measured positions. λ λ g =, (3) 2 f 1 c f c f c =, (4) 2a where λ is the corresponding free space wavelength, f c is the cut off frequency of the TE 10 mode, f is the frequency and c the speed of light in vacuum.

65 EURAMET.EM.RF-S25 Final Report page 65 of 120 MEASUREMENT REPORT 5 Measurement PC (Dell Latitude D610) Bi-directional coupler Signal generator N-type coaxial cable RF-power meter (input power) RF-power meter (reflected power) Calibration point USB cable Opto/USB interface box side a side b Fiberoptic cable Electronics box cables Positioner Transfer standard N-type-to-WR975 adapter WR975 waveguide (1.5 m) WR975 sliding termination Figure 1. Realisation of E-field at 1 GHz. The measurement hole in the side a was not used in this comparison and it was plugged during the measurements GHz setup The primary calibration system of STUK for calibrating small E-field probes in the frequency range GHz is consisting of a 0.6 m long straight section of WR430 waveguide (width 109 mm, height 54.6 mm), N-type-to-waveguide transition and a waveguide termination as illustrated in the figure 2. The transition and termination are

66 EURAMET.EM.RF-S25 Final Report page 66 of 120 MEASUREMENT REPORT 6 manufactured by Arra Inc., Bay Shore, NY, USA, and the straight section by Spinner GmbH, München, Germany. The positioning procedure is identical to the procedure at 1 GHz setup (see the previous section). Two holes in the side a (not used in this comparison) and additional hole in the side b are plugged while one hole in the side b is used to put the probe into the waveguide. The separation of the two calibration points is quarter wavelength at 1800 MHz and they can be used to compensate the effect of mismatch of the termination MHz was not measured in this comparison, hence the possible mismatch causes additional uncertainty to the calibration. The 10 repetition measurements were made randomly at both calibration points. Any systematic difference of electric field between the two points was not observed.

67 EURAMET.EM.RF-S25 Final Report page 67 of 120 MEASUREMENT REPORT 7 Measurement PC (Dell Latitude D610) Signal generator N-type coaxial cable Bi-directional coupler USB cable RF-power meter (input power) RF-power meter (reflected power) Calibration points Opto/USB interface box 1 2 Fiberoptic cable side a side b Electronics box cables Positioner Transfer standard N-type-to-WR430 adapter WR430 waveguide (0.6 m) WR430 termination Figure 2, Realisation of E-field for 1.75, 2, 2.25, and 2.5 GHz. The measurement holes in the side a were not used in this comparison and they were plugged during the measurements as well as the additional hole in the side b.

68 EURAMET.EM.RF-S25 Final Report page 68 of 120 MEASUREMENT REPORT 8 3 Description of the used measuring equipment 3.1 RF-power, input The input power was measured with Agilent E4416A meter supplied with Agilent 8482A probe (Agilent, Inc., Santa Clara, CA, USA). The probe is calibrated at 1.Jan 2005 by Agilent Technologies, Bayan Lepas Free Industrial Zone, Penang, Malaysia, and the meter at 9.Jan 2002 by Agilent Technologies EPSG Queensferry, South Queensferry, West Lothian, Scotland. 3.2 RF-power, reflected The reflected power was measured with HP 435A meter supplied with HP 8481A Probe (Hewlett & Packard Inc., Santa Clara, CA, USA). The setup was checked in house against Agilent E4416A and 8482A probe. 3.3 Directional coupler The RF power meters were connected to bi-directional coupler Narda 8022 (Narda Microwave - East, Hauppage, NY, USA). The port matching used in the uncertainty budget is given by the manufacturer. The coupling of the side arms of the bi-directional coupler was measured with vector network analyser (VNA) HP 8752C (Hewlett-Packard, Santa Clara, CA, USA). The VNA is calibrated at 20.Sept 2007 by Agilent Technologies, Customer Service Center, Espoo, Finland. The coupling was also checked with abovementioned RF-power meters.

69 EURAMET.EM.RF-S25 Final Report page 69 of 120 MEASUREMENT REPORT Signal source The abovementioned VNA (HP 8752C) was used also as a signal source at all frequencies. The output power was adequate without external amplifiers. 3.5 Instrument for positioning The positioning of the probe was made with a slide calliper, which is checked in house against cauge blocks (Opus Metrology Ltd., Northamptonshire, England) calibrated by the manufacturer at 10 th April Ambient temperature The ambient temperature was monitored during the measurements with precision glass capillar thermometer (Amarell GmbH & Co., Kreuzwertheim, Germany), calibrated by the manufacturer at 10th Oct 2001.

70 EURAMET.EM.RF-S25 Final Report page 70 of 120 MEASUREMENT REPORT 10 4 Uncertainty of the measurements The estimated uncertainty components of the E-field realisations of STUK are presented in the Tables 2 and 3. The RF power measurement uncertainty consists of the calibration uncertainty and mismatch of the RF sensor, uncertainty of RF meter including the reference oscillator and the measurement uncertainty of the coupling of the bi-directional coupler. The weighting factor of the RF power uncertainty is 0.5 since the E-field in the waveguide is directly proportional to the square root of the RF power. The uncertainty values are given in calibration certificates and device manuals. The details of the uncertainty estimate are based on the procedures given in reference [2]. The dimensions of the waveguides have been estimated to be within 1 % of the given values of the WR430 and within 2 % of the WR975 waveguide. The E-field is the voltage between the centre lines of the a-sides of the waveguide divided by the height, and hence the error in the dimension a causes error to the E-field directly. The spatial positioning uncertainty was estimated to be ±1 mm in the WR430 and ±2 mm in the WR975 waveguide. This results very small error in the E-field, since the derivative of the field distribution is zero at the centre line of the waveguide. However, together with the rotational positioning, the error was estimated to be 1 %. The standing wave related uncertainty is negligible in the 1 GHz setup because the effect of the standing wave can be compensated by finding the minima and maxima of the transfer standard reading by adjusting the sliding termination. In the GHz setup, the mismatch of the termination can not be compensated. The estimated uncertainty is based on the minimum matching over the whole frequency band measured by the manufacturer of the component (Arra Inc.).

71 EURAMET.EM.RF-S25 Final Report page 71 of 120 MEASUREMENT REPORT 11 Table 2. The estimated uncertainty of the calibration standard at 1.00 GHz Error source Weighting factor Standard uncertainty (± %) Power measurement (details) Calibration factor of power sensor Mismatch of the sensor and bidirectional coupler Power meter absolute accuracy Power reference calibration Power reference accuracy Power reference port mismatch Measured coupling of the bidirectional coupler (VNA) Power measurement (combined) Dimensions of the waveguide Positioning Standing wave 1 0 Combined standard uncertainty 2.5 Expanded uncertainty (95 % confidence level) (K=2) 4.9

72 EURAMET.EM.RF-S25 Final Report page 72 of 120 MEASUREMENT REPORT 12 Table 3. The estimated uncertainty of the calibration standard at 1.75, 2.00, 2.25 and 2.50 GHz Error source Weighting factor Standard uncertainty (± %) Power measurement (details) Calibration factor of power sensor Mismatch of the sensor and bidirectional coupler Power meter absolute accuracy Power reference calibration Power reference accuracy Power reference port mismatch Measured coupling of the bidirectional coupler (VNA) Power measurement (combined) Dimensions of the waveguide 1 1 Positioning 1 1 Standing wave Combined standard uncertainty 2.1 Expanded uncertainty (95 % confidence level) (K=2) 4.1

73 EURAMET.EM.RF-S25 Final Report page 73 of 120 MEASUREMENT REPORT 13 5 Measurement results The electric field strengths for the transfer standard reading of 20 V/m are presented in the Table 1. Frequency [GHz] Electric field [V/m] Standard deviation of 10 measurements [%] References [1] S. Ramo, J.R. Whinnery, T. van Duzer, Fields and Waves in Communications Electronics, John Wiley and Sons, Inc., New York, USA, [2] Agilent, Fundamentals of RF and Microwave Power Measurements, Application Note 64-1C, Agilent Technologies, Santa Clara, CA, USA, 2001

74 EURAMET.EM.RF-S25 Final Report page 74 of 120 A6 SP

75 EURAMET.EM.RF-S25 Final Report page 75 of 120 Measurement report: EURAMET Project 819 Technical Protocol: Comparison of Electrical Field strength Measurements above 1 GHz SP Sveriges Tekniska Forskningsinstitut SP Technical Research Institute of Sweden Borås, June 2008 Lars Fast Technical officer Kristian Karlsson Technical officer Jan Carlsson Deputy Technical Manager 2

76 EURAMET.EM.RF-S25 Final Report page 76 of 120 Summary Introduction References Environment Methods used by SP Method described in ref [2] Method described in ref [3] Equipment used in ref [2] and ref [3] Uncertainty ref [2] Uncertainty for ref [3] Results

77 EURAMET.EM.RF-S25 Final Report page 77 of 120 Summary In this report we summarise our results from the measurements done at SP Technical Research Institute of Sweden, during spring The measurements are a part of a Round Robin test within the Euramet organisation. The project is called EURAMET Project 819 Technical Protocol: Comparisin of Electrical Field Strength Measurements above 1 GHz. The objective with the Round Robin test is to evaluate a transfer probe and to compare its function with conventional techniques for obtaining a specific electric field at a given point. 4

78 EURAMET.EM.RF-S25 Final Report page 78 of Introduction The electrical field in a defined point is measured using a transfer probe according to ref [1]. The electrical field in this specific point is also measured according to the methods presented in ref [2] and ref [3]. The two different values for this point in space are compared for a number of frequencies between 1 GHz and 2.5 GHz. 2 References [1] EURAMET Project 819 Technical Protocol: Comparison of Electrical Field Strength Mewasurements above 1 GHz [2] SP-Metod 2896, Version 1.1, Kalibrering av elektromagnetisk fältstyrka med µtem-cell, (In Swedish) [3] SP-Metod 2781, Version 1.1, Kalibrering av elektromagnetisk fältstyrka med hornantenn, (In Swedish) [4] Calibration certificate Power meter and Power sensor NRVS, SP and SP Environment The relative humidity was 22 % during the measurement and the temperature was 23 C. The measurements were preformed in the anechoic chamber HERTZ in house number 15 at SP Technical Research Institute of Sweden, Borås during the 21 and 22 of April Methods used by SP We use two different methods to determine the electrical field; they are described in ref [2] and ref [3], respectively. In ref [2] we use the technique of calibrating a transfer probe in the frequency interval 50 MHz to 1 GHz inside a µtem-cell where the field is calculated from a simple formula. The input parameters are the geometry, the input power together with the power loss of the µtem-cell and the impedance of the cell. This transfer probe is used to measure the field strength in a point in the far field from a transmitting antenna. The power used for feeding the antenna is associated with the field detected by the transfer probe. The power is measured indirectly with a power meter through a directional coupler. Since the last quantity is not used other than as a reference level the absolute value is not of interest. In ref [3] we generate a field at a given point in space in the far field region in front of a standard gain horn antenna. The electrical field in this point is related to the power fed to the horn antenna and can be calculated with the knowledge of the antenna gain the distance to the point and the input power. The input power of the antenna is measured indirectly by using a directional coupler with a known coupling factor and a power meter. 5

79 EURAMET.EM.RF-S25 Final Report page 79 of Method described in ref [2] The technique of establishing a known field inside a µtem-cell and then calibrating a transfer probe using this field is described in ref [2]. The relation between the power and the electric field is given by the expression: where: P TEM ( E d ) = [1] Z L 2 P TEM E d Z L = power in the µtem-cell = filed strength = distance between the wall and the centre conductor inside the cell = characteristic impedance of the µtem-cell After that the electrical field has been established and the transfer probe has been calibrated this probe can be used in the anechoic chamber to calibrate other e-field probes. Once the transfer probe is calibrated, one can generate a field in an anechoic chamber with an appropriate antenna and measure the field in a specific point with the transfer probe. After that one can calibrate another field probe in this point. It is important to place both field probes in the same point to obtain good agreement. 4.2 Method described in ref [3] The technique of establishing a known electrical field in a define point inside an anechoic chamber with the used of standard gain horns is described in ref [3]. The relation between the input power of the standard gain horn and electrical field in the far field region of the horn is given by following relation: where: E 1 P 30 FWD R P = 2 G P [2] 1 E = Electric field P FWD = Forward Power P 1 = Forward power directional coupler P 2 = Forward power measured associated with P 1 G = Gain of the standard gain horn R = Distance between the antenna and the e-field probe 6

80 EURAMET.EM.RF-S25 Final Report page 80 of Equipment used in ref [2] and ref [3] Signal generator R&S SME06 SP Signal generator R&S SMR40 SP RF Amplifier AR 100W1000M1 SP RF Amplifier AR 500A100A SP RF Amplifier AR 10W1000 SP RF Amplifier AR100S1G4 (0,8-4,2GHz) SP RF Amplifier AR 20T4G18 (4,2-18GHz) SP Directional Coupler AR DC 3001 SP Directional Coupler AR DC Directional Coupler AR DC Power meter NVRS SP Power sensor NVRS SP Antenna CHASE Bilog CBL6121A SP Thermometer/RH Testo 615 SP µtem cell Schaffner SP Transfer probe Schaffner SP Measure Leica SP Standard gain horn Horn SP Horn SP Software: HF Kalibrering Latest available version Pyramidal Horn Antenna Version: Uncertainty ref [2] The uncertainty budget is based on the following relation, where: EK = k ET R G PK PT (3) k = Transfer probe calibration factor, A2 ref [2] E T = Field strength transfer probe, relative measure, A3 ref [2] P K = Forward power from directional coupler, relative measure - ref [4] P T = Forward power from directional coupler, relative measure - ref [4] G = Rate between 1) gain when the transfer probe is measured and 2) gain when the calibration object is measured, A4 ref [2] R = Ratio between the distances between the calibration object and the antenna and the transfer probe and the antenna, A4 ref [2] The equation [3] gives the following uncertainty for a calibration using the transfer probe: 7

81 EURAMET.EM.RF-S25 Final Report page 81 of 120 Factor Example Standard uncertainty Distribution Sensitivity Contribution to standard values coefficient uncertainty k Normal E T 10 V/m Normal P K 10.8 W Normal ½ P T 10.0 W Normal ½ G Normal ½ R Rectangular E K 10.0 V/m Table 1 Uncertainty budget for Transfer probe calibration. The uncertainty of a measurement with coverage factor k=1 is 3 % 4.5 Uncertainty for ref [3] The uncertainty budget is based on the following relation, E 1 P 30 FWD R P = 2 G P (4) where: 30 = Constant, no contribution to the uncertainty budget P FWD = Output power from the amplifier, absolute power 1 P 1 P 2 G R = Forward power from directional coupler, relative power = power after directional coupler and cable, relative power = Reference antenna gain = distance between reference antenna och calibration object Equation [4] gives the following uncertainty for a calibration using horn antennas: (In the table the σ Eff power meters frequency dependent uncertainty, and the uncertainty δ(g) of the positioning, in height and sideways, is frequency dependent, see table 3.) Factor Example Standard uncertainty Distribution Sensitivity Contribution to standard value coefficient uncertainty P FWD 1.00 mw σ Eff,abs Normal ½ σ Eff,abs /2 P mW σeff,rel Normal -½ σ Eff,rel /2 P mw σ Eff,rel Normal ½ σ Eff,rel /2 G 8.82 δ(g) Rectangular ½ δ(g)/(2 3) R 2.00 m Rectangular E V/m Se table 5 Table 2. Uncertainty budget for horn antenna calibration. 8

82 EURAMET.EM.RF-S25 Final Report page 82 of 120 The frequency dependent uncertainties are: Frequency σ Eff,rel σ Eff,abs δ(g) δ(e) [GHz] (from ref [4]) (from ref [4]) (from A2 ref [3]) [%] Table 3. Frequency dependent uncertainty An example is presented a probe calibration for electrical field at 5GHz. Factor Estimated value Standard uncertainty Distribution Sensitivity coefficient Contribution to standard uncertainty P FWD mw Normal ½ P Normal -½ mw P mw Normal ½ G Rectangular ½ R 3.00 m Rectangular E V/m Table 4. Uncertainty budget for a horn antenna calibration at 5 GHz. The total uncertainties for generated electrical fields using horn antennas are: Frequency Total uncertainty [GHz] [%] Table 5. Uncertainty for a horn antenna calibration at different frequencies. 5 Results We measured the electrical field with probe having zero fields present in the chamber and obtained the following results. Frequency Probe offset [GHz] E [V/m] Table 6. Offset for the test object, zero field readings. 9

83 EURAMET.EM.RF-S25 Final Report page 83 of 120 The calibration of the probe was performed by using the method described in ref [2] for 1.0 GHz and method described in ref [3] for the frequencies 1.25, 1.50, 1.75, 2.00, 2.25 and 2.50 GHz. The results are presented in the table below. Frequency Probe Reading Reference field Calibration factor Uncertainty [GHz] E [V/m] E [V/m] - [%] Table 7. Results of the calibration. 10

84 EURAMET.EM.RF-S25 Final Report page 84 of 120 A7 INRIM

85 EURAMET.EM.RF-S25 Final Report page 85 of 120 Method of calculation of the field-strength The electric field-strength value E was calculated by means of the well known formula: E = 30 PG d where: P is the net power delivered to the radiating antenna, measured as the difference between the incident and reflected power at the antenna input; G is the numerical gain of the standard antenna, evaluated for the relevant distance; d is the separation distance between the antenna aperture and the field probe. Measurement results The measurements results and associated uncertainties are shown in the enclosed annex (file INRIM_E819.xls ). Starting from the above-mentioned equation the different uncertainty contributions were evaluated in terms of relative values and reported in details in the table uncertainty budget of the spreadsheet. For convenience the final results are also reported in the following table: Frequency (MHz) Field strength (V/m) Uncertainty (k=1) (V/m) 1,00 19,5 1,3 1,25 20,6 1,4 1,50 19,9 1,3 1,75 18,4 1,2 2,00 19,6 1,3 2,25 19,8 1,3 2,5 19,7 1,3 Note: The stability of the sensor readings during all the performed measurements, for each measuring frequency, was better than or equal to 0,1 V/m, i.e. to the resolution of the meter. Consequently, in the table raw data of the attached spreadsheet, only one row of data is reported for each frequency value. The corresponding uncertainty contribution is shown as Probe readings in the uncertainty budget. Torino, 7 October Michele Borsero INRIM Torino (Italy) 2

86 EURAMET.EM.RF-S25 Final Report page 86 of 120 A8 PTB

87 EURAMET.EM.RF-S25 Final Report page 87 of 120 Introduction Measurements on the comparison artefact have been performed in two different field generators which are a µtem cell and standard gain horn antennas. The µtem cell represents the German standard of the EM field strength between 1 MHz and 1.1 GHz and is routinely used for transfer sensor calibration in this frequency range. However, it turned out to be a reliable field generator in the frequency up to 2 GHz excluding the first higher-order mode at about 1.4 GHz. The standard gain horn antennas are routinely used for field sensor calibration above 1.1 GHz at PTB. All measurements have been carried out in a temperature range between 21 C and 23 C five times only at the distinct frequencies due to restricted resources. The field generator has been set to generate field strength values close to 20 V/m. The results have then been linearly scaled to the field strength that produces a field strength reading of exactly 20 V/m. The measurement value for each distinct frequency represents the mean of these measurements. A DC offset correction has not been applied for simplicity since its effect on the calibration result can be neglected. The uncertainty budget contains a contribution from the reference standards (type B) and a contribution from the experimental standard uncertainty (type A). Due to only five consecutive measurements the reduced overall degree of freedom requires a factor of k=2.87 for an expanded measurement uncertainty but does not affect the assigned standard uncertainty (coverage factor k=1). Contributions from thermal effects are included in the experimental standard uncertainty (type A) for the measurement conditions specified in this report. Positional effects and effects of deviant thermal conditions are not within the scope of this comparison but are within the responsibility of the user of a calibrated field sensor. 2

88 EURAMET.EM.RF-S25 Final Report page 88 of 120 Description of realization of electrical field µtem Cell The calibration setup is shown in the following figure: RF synthesizer attenuator µtem attenuator thermal RF and amplifier -10 db cell -10 db power meter RF source: attenuator: power meter: µtem cell: synthesizer and power amplifier -10 db each with best possible match to µtem cell RMS measurement with thermal head septum height 35 mm, frequency up to 2 GHz The calibration field strength in the µtem cell will be calculated from the feed-through power P, the characteristic line impedance Z L and the septum distance d of the cell according to the equation E = P Z Since the feed-through power will be measured behind an attenuator, only, and the power meter cannot be assumed to be ideal, the feed-through power P has to be calculated according to d L. ( ) ( ) P = A f k P k( f ) P DC mess where A(f) is the frequency dependent attenuation of the attenuator at the output. The measurement value P mess from the power meter will be multiplied with two correction factors of which k(f) represents the frequency dependent power efficiency (normalized to the DC value k(0)=1) and of which k DC (P) corrects for a linearity error that depends on the measured power. Since the RF power meter allows for feeding in DC voltages, these two factors can be determined independently. With these additional factors the above equation becomes 3

89 EURAMET.EM.RF-S25 Final Report page 89 of 120 E = ( ) ( ) Z A f k P k( f ) P L DC mess d and represents the fundament to set the calibration field strength in the standard measurement facility. The calibration field strength determined for the displayed field strength values are RMS values. The representation of the (empty) calibration field is realized with a relative expanded uncertainty of 2% (k=2). Horn antennas The effective value of power flux density is calculated from the power measured at a directional coupler and antenna gain, according to the following equation: S = P m D G 4 π d 2 where S = power flux density, P m = measured power, D = coupling attenuation factor of directional coupler, G = linear antenna gain d = distance antenna / EUT The electric field strength is calculated from the power flux density according to the following equation: E = S Z 0 where S = power flux density Z 0 = free space impedance The representation of the (empty) calibration field is realized with a relative expanded uncertainty of 7% (k=2). 4

90 EURAMET.EM.RF-S25 Final Report page 90 of 120 Final Results µtem Cell field strength values with standard uncertainty (k=1) for field strength meter reading of 20 V/m frequency (GHz) field strength (V/m) 1,00 20,87 ± 0,20 1,25 20,91 ± 0,20 1,50 20,88 ± 0,20 1,75 20,9 ± 0,20 2,00 20,9 ± 0,20 2,25 2,50 Horn antennas field strength values with standard uncertainty (k=1) for field strength meter reading of 20 V/m frequency (GHz) field strength (V/m) 1,00 1,25 20,41 ± 0,70 1,50 20,17 ± 0,70 1,75 20,37 ± 0,70 2,00 18,87 ± 0,71 2,25 20,36 ± 0,71 2,50 19,6 ± 0,71 5

91 EURAMET.EM.RF-S25 Final Report page 91 of 120 A9 METAS

92 EURAMET.EM.RF-S25 Final Report page 92 of 120 of the transmit axis is 0.9 m. Its polarisation is vertical. In order to reduce reflections, microwave absorbers are used below and behind the experimental setup. The computer controlled application allows controlling the stability of the whole system. Moreover, in order to increase the accuracy, the calibration is performed using two different systems of calibrated reference antennas and two different power meters. 2 Measuring instruments Device Manufacturer Type Inventory Für oberhalb 1 GHz EMC-chamber ETS Microwave absorbers TDK IS 60 - (9 pieces for the wall) Microwave absorbers ETS EMC CL24 - (6 pieces for the floor) Transmit antenna Schwarzbeck BBHA 9120 E 5130 RF-generator Rohde & Schwarz SMF 100A 6331 Directional coupler 2-18 Agilent 773D 6343 GHz HF-amplifier 800MHz bis 3 Amplifier Research 25S1G GHz Power meter Rohde & Schwarz NRVD 2001 Power meter head Rohde & Schwarz NRV-Z Power meter head Rohde & Schwarz NRP-Z Power meter head Rohde & Schwarz NRP-Z Reference antenna 1 Schwarzbeck SBA Reference antenna 2 Schwarzbeck SBA Calculation method for electric field strength The equation for electric field strength calculation is E = AF P Antenna Z L where : P Antenna the power at the receive antenna. Z L the reference impedance of 50 Ohms (conventional value without uncertainty). AF the antenna factor of the receive antenna. 4 Results The results are the field strength in V/m which is required to have a reading of 20 V/m of the transfer standard. frequency field strength Realtive uncertainty (k=1) (GHz) (V/m) ± 3.0% % % % % % % 2/3

93 EURAMET.EM.RF-S25 Final Report page 93 of 120 Federal Department of Justice and Police FDJP Federal Office of Metrology METAS 5 Uncertainty budget All frequencies except 1.75 GHz Source Distribution Type value value kvalue standard uncertainty sensitivity factor (%) () (%) () (%) Antenna Factor Calibration Certificate B - normal 0.5 db 5.70% % % Power Calibration Certificate B - normal % % % Mismatch uncertainty Calculation B U-shape 0.01 (power meter)* 0.06% % % (0.06) antenna Stability of power measurements Estimation B - normal 0.01 db 0.12% % % Homogeneity Measurements A 0.54% 0.54% % % Reading uncertainty From the digitization B- rectangular 0.1 / 20 V/m 0.50% % % Total Uncertainty (k=1) 2.92% 1.75 GHz Source Distribution Type value value kvalue standard uncertainty sensitivity factor (%) () (%) () (%) Antenna Factor Calibration Certificate B - normal 0.60 db 6.90% % % Power Calibration Certificate B - normal % % % Mismatch uncertainty Calculation B U-shape 0.01 (power meter)* 0.06% % % (0.06) antenna Stability of power measurements Estimation B - normal 0.01 db 0.12% % % Homogeneity Measurements A 0.54% 0.54% % % Reading uncertainty From the digitization B- rectangular 0.1 / 20 V/m 0.50% % % Total Uncertainty (k=1) 3.51% uncertainty uncertainty 3/3

94 EURAMET.EM.RF-S25 Final Report page 94 of 120 A10 University of Pretoria