Maintain the farad and tie the U.S. legal farad to the international system of units, support and improve NIST’s impedance measurement services, and ensure the critically needed access of the U.S. industrial base to internationally consistent, reliable, reproducible, and traceable electrical measurements.
|
Yicheng Wang sets up an automated system for determining the frequency dependence of fused-silica capacitors. |
This project ties the U.S. legal system of electrical units to the International System of Units (SI) through the realization of the SI unit of capacitance. This work also forms the foundation of NIST’s measurement services for electrical impedance, ensuring the sound metrological basis for all impedance measurements, both nationally and internationally, and ensuring that the claims of measurement accuracy by U.S. industries are recognized and accepted worldwide. The need continues for better representation of capacitance and also for better test and calibration tools at NIST with which to verify objectively claims of improved performance specifications, to achieve consistency, and to help avoid technical trade barriers.
The primary facility for connecting the U.S. legal system of electrical units to the international system of units is the NIST calculable capacitor, with which the measurement of capacitance is effectively achieved through a measurement of length. Both the calculable capacitor and the chain of high-precision measurements that transfers the SI unit to the calibration laboratories must be maintained, improved, and compared with other national metrology laboratories to ensure measurement consistency on an international level.
AC measurements linking the calculable capacitor to the set of standards comprised by the National Farad Bank have been performed only at 1592 Hz and 1000 Hz. However, customer standards are often calibrated at other frequencies; as a result, the uncertainty provided for customer calibrations was significantly increased to account for differences in the capacitance unit due to frequency dependence. In order to better support customers’ needs in the broader frequency range from 50 Hz to 20000 Hz, the frequency dependence and dissipation factor of the Farad Bank need to be determined.
Many national laboratories are developing the capability to do ac quantum Hall resistance (QHR) measurements as a means to obtain a capacitance unit because of the difficulty in establishing a calculable capacitor measurement system. The remaining technical challenge for using ac QHR as a quantum representation of impedance is to characterize the linear frequency dependence of a QHR device. With the availability at NIST of the calculable capacitor, NIST is ideally situated to perform measurements to link the capacitance as determined with the ac QHR to NIST’s present unit of capacitance.
Development of wideband impedance measurement services requires reference standards that can be characterized over the impedance and frequency ranges of interest. NIST has developed a system to characterize commercial four-terminal-pair (4TP) capacitance standards from 1 pF to 1 nF over the frequency range from 1 kHz to 10 MHz. This system is being used to offer special tests for 4TP capacitors as well as provide reference standards for general impedance measurements using a commercial LCR meter. A bootstrapping technique using the LCR meter and an inductive voltage divider (IVD) can be used to extend the characterization from the 1 nF standard to higher-valued capacitance standards up to 10 µF. A major source of error associated with the bootstrapping technique is due to the in-phase and quadrature errors of the IVD. The newly developed straddling bridge will allow for the accurate self-calibration of single-decade, two-stage reference IVDs over the frequency range of 20 Hz to 100 kHz.
Developed special tests for the frequency dependence of capacitance standards over the entire audio frequency range based on recent advances in capacitance measurement capabilities, driven by several requests from Sandia, other national laboratories, and from industry. The value of a typical standard capacitor may vary slightly with frequency due to the effects of electrode surface films, an imperfect medium between the electrodes, and non-ideal connecting leads. Using a combination of a 1 pF “cross capacitor” having a negligible frequency dependence due to surface films, and a 10 pF nitrogen dielectric capacitor with a very small residual inductance as references, we have measured the frequency dependence of 10 pF and 100 pF reference fused-silica capacitors from 50 Hz to 20 kHz. Once these standards were characterized, we were then able to measure the frequency dependence of two reference capacitors provided by Sandia National Laboratories’ Primary Standards Lab. The capacitance uncertainties for Sandia’s standards at frequencies of 100 Hz and 400 Hz are smaller by a factor of 5 than those previously assigned. This improvement allows the Primary Standards Lab to reduce the uncertainties reported on calibrations they provide to their internal Sandia customers and to other DOE laboratories.
Developed and characterized a programmable capacitance standard to provide accurately known values of capacitance over 6 orders of magnitude below 100 pF with a resolution of 0.1 fF. The potential applications of a well-characterized programmable capacitance standard are numerous. The current work on a capacitance standard based on single electron tunneling (SET) involves an “odd value” cryogenic capacitor. The programmable capacitor will provide a crucial link to compare an SET-based capacitance standard with the SI farad represented by the Farad Bank. This comparison is not only important to confirm the accuracy of the new standard but also necessary for closing the metrology triangle via the SET-based capacitance standard. Another example that requires precise measurement of an “odd value” capacitor is a new pressure standard based on measurements of the dielectric constant of helium. The NIST Chemical Science and Technology Laboratory has proposed to use cross capacitors for accurate measurements of the dielectric constant; however, it is difficult to construct a cross capacitor with a precise value. A programmable reference capacitor can also be used to test the linearity of commercial precision capacitance bridges.
 |
Scott Shields assists with the installation of the ac QHR measurement system. |
Performed and analyzed first NIST ac QHR measurements. We have demonstrated that all dc QHR guideline properties and all dc and ac QHR values can be measured without changing sample probe lead connections at the QHR device and that ac QHR values converge to the dc value. We have also demonstrated that quadratic frequency dependence of an ac QHR device can be analytically modeled and corrected within a few parts in 109. The remaining challenge for using ac QHR as a quantum representation of impedance is to characterize and minimize the linear frequency dependence.
Fully tested a straddling bridge. The bridge has been used to self-calibrate two IVD — a low-frequency (20 Hz to 20 kHz), single-decade IVD and a high-frequency (1 kHz to 1 MHz) IVD. Test results indicate that the low-frequency IVD exhibits in-phase and quadrature errors of less than ±5 parts in 108 of full-scale ratio at 1 kHz, increasing to ±1 part in 106 at 20 kHz. In addition, a method to compensate for the internal admittance loading errors of the low-frequency IVD was verified by calibrating the IVD with the straddling bridge both with and without compensating networks installed. The measurements indicate that the compensation scheme reduces the IVD’s in-phase and quadrature errors by more than an order of magnitude at 20 kHz.
Designed and implemented a system to characterize four-terminal-pair capacitance standards of values from 1 pF to 100 mF for use as primary reference standards for measuring general impedances (inductors, capacitors, and resistors) at frequencies from 20 Hz to 100 kHz. New commercial capacitance bridges and LCR meters are becoming more and more accurate. To satisfy the need for this demanding calibration service by U.S. industry, we have to modernize the NIST calibration facilities. The new calibration system has been designed with particular emphasis on automation. An uncertainty analysis is underway.
Completed the key international comparison for inductive voltage dividers (CCEM-K9). The existing NIST IVD calibration system was used with the new straddling bridge to determine the in-phase and quadrature errors of the precision IVD used for a CCEM sponsored international comparison. Results indicate that agreement between the two systems is better than 3 parts in 108 for in-phase error components, and better than 5 parts in 108 for quadrature error components. The results of the worldwide comparison will be available in 2005.
Acted as pilot laboratory for SIM capacitance key comparison SIM-EM.K4 and supplemental comparisons SIM-EM.S4 and SIM-EM.S3, completed protocol documentation and first round of measurements. The CCEM K4 key comparison for capacitance standards was completed in 2000 with NIST serving as the pilot laboratory. As the only country with a working calculable capacitor in the Americas, NIST has the obligation to provide the link between SIM and CCEM for capacitance measurements. In partial fulfillment of this obligation, NIST is serving as the pilot lab for an international comparison of capacitance measurements between many of the National Metrology Labs in North, Central, and South America.
Completed moving the Farad Bank, the quad bridge bank, the capacitance calibration lab, and ac QHR lab into the new Advanced Measurement Laboratory (AML) without causing major disruption of impedance dissemination. We have checked the Farad Bank against the calculable capacitor after the move. Initial results indicate that the Bank value changed by less than 2 parts in 108. We will to continue monitoring the Farad Bank for six more months before moving the calculable capacitor to its new environmentally controlled room in the AML.
206 tests were performed on 123 artifacts for 85 customers for impedance standards and inductive voltage dividers, providing income to the division of approximately $209,763 (October 1, 2003 to September 30, 2004).
Yicheng Wang collaborated with the Physical and Chemical Properties Division in a competence project to develop an atomic standard of pressure based on capacitance metrology.
Y. Wang and L. Lee, “A digitally programmable capacitance standard,” Rev. Sci. Instrum. 75, pp. 1158 (2004).
Y. Wang and S. Shields, “Improved capacitance measurements with respect to a 1 pF cross capacitor from 200 Hz to 2000 Hz,” Digest, Conference on Precision Electromagnetic Measurements, (London, England, 27 June – 2 July, 2004), pp. 489-490.
J.W. Schmidt, M.R. Moldover, and Y. Wang, “Measured permittivity ratio of Argon/Helium using cross capacitors,” Digest of Conference on Precision Electromagnetic Measurements (Conference on Precision Electromagnetic Measurements 2004), (London, England, 27 June – 2 July, 2004),
p. 491.
S. Avramov-Zamurovic, B. Waltrip, K. Stricklett, and A. Koffman, “A balancing algorithm for systems with correlated injections,” Proc., of the IEEE Instrumentation and Measurement Technology Conference (IMTC/2004), 18–20 May, 2004, Como, Italy.
Y. Wang, “Frequency dependence of capacitance standards,” Rev. Sci. Instrum. 74, pp. 4212 (2003).
N. M. Zimmerman, M. El-Sabbagh, and Y. Wang, “Improved Cryogenic Capacitor for the ECCS: Larger Value (10 pF) and SI measurement by tuning the calculable capacitor,” IEEE Trans. Instrum. Meas. 52, pp. 608 (2003).
A.-M. Jeffrey and A. D. Koffman, “Improved 1 kHz capacitance calibration uncertainty,” IEEE Transactions on Instrumentation and Measurement, 52, (4), pp. 1284-1288 (August 2003).