Address:
|
100 Bureau Drive, Stop 8171
Gaithersburg, MD, 20899-8171, United States
|
Abstract:
|
Key and supplementary comparisons are of interest to the metrology community because they provide the means of demonstrating the degree of equivalence between the measurement capabilities of national metrology institutes. In recent years, NIST has been involved in both key and supplemental comparisons for DC resistance in the range 1 ohm to 1 Gohm. While participating in and piloting several of these comparisons, much has been learned, and as a result, both the comparison process and our view of this process has evolved. Topics that will be addressed here include development of protocols, characterization of standards, uncertainty analysis approaches, and the present status of several comparisons in the area of DC resistance. Over the past thirty years there have been significant events which have increased the need and changed the fundamentals of key, supplemental, and inter laboratory comparisons in the metrology community. Events such as the discover of the quantum Hall effect in 1980 and the subsequent redefinition of the ohm in 1990 were the motivation for resistance comparisons to satisfy mostly scientific and technical needs. More recent comparisons that have taken place following the Mutual Recognition Arrangement (MRA) of 1999 have been driven by a new focus on international trade. As the comparison process has evolved, a robust written protocol has become necessary to clearly define expectations, uncertainty budget templates, data collection procedures, handling of standards, reporting of results, statistical analysis methods, and definition of reference values. Resistance standards used in higher-level comparisons need to be well characterized for parameters such as drift rate, temperature coefficient, pressure coefficient, and voltage coefficient. Establishment of these parameters allows for data to be corrected when it is not possible for all participants to measure the standards under the same test conditions. Multiple standards are used at each level to provide redundancy and increase the statistical significance of the results. For the CCEM-K2 comparison at 10 Mohm and 1 Gohm, appropriate standards were not available at the time so NIST developed rugged transport standards. These standards are presently being used in two regional comparisons. Statistical experts have been engaged to determine approaches to the uncertainty analysis that are appropriate for analyzing the data collected during comparisons. Experience has shown us that it is essential to consult with statistical experts at the earliest stages of designing a comparison. For the resistance comparison now being piloted in SIM by NIST, statisticians were consulted and have developed appropriate analysis for specific aspects such as multiple loops and linking the results of this regional comparison back to results from previously completed key comparisons. Since 1990, NIST has participated in key comparisons in the area of DC resistance at the nominal values of 1 ohm, 100 ohm, 10 kohm, 10 Mohm, and 1 Gohm. Key comparisons CCEM-K1 (1 ohm and 10 kohm) and CCEM-K2 (10 Mohm and 1 Gohm) have been completed and the results are included in the key comparison database (KCDB). Key comparison CCEM-K10 (100 ohm) has been completed and the report is in the Draft A status. Presently NIST is supporting several regional comparisons of DC resistance by providing standards for use in EUROMET.EM-K2 (10 Mohm and 1 Gohm) and piloting comparisons SIM.EM-K1 (1 ohm), SIM.EM-K2 (1 Gohm), and SIM.EM-S6 (1 Mohm). Other current activities are to provide opening and closing measurements for a NCSLI sponsored inter laboratory comparison of high resistance at 1 Gohm.
|