Search

Behavior models of electronic components are critical in many metrological and measurement areas, e.g., time-domain simulations for analog-integrated circuit design and predistortion of the amplifier to reduce the distortion. This paper describes the use of pulse-driven ac Josephson voltage standard (ACJVS) to synthesize quantum accurate waveforms with ultralow distortion and noise floor to model the amplifier nonlinearity, which is a promising way to increase the modeling accuracy. Four Volterra series models including the memory polynomial model and three submodels of the dynamic deviation reduction model are analyzed and compared in detail. Normalized mean square error is calculated to evaluate the modeling accuracy. The difference of the measured and predicted outputs of the amplifier are compared in both time and frequency domains. The result shows that high-performance models can be achieved, which demonstrates the feasibility of modeling nonlinear amplifier behavior with the ACJVS.

After continuous investigation on the electrical compensation method for the end effect error in the new vertical calculable cross-capacitor (VCCC), the latest optimal hollow active auxiliary electrode (OHAAE) has been developed and successfully substituted for the current geometrical nosepiece at one end of the movable guard electrode at the National Institute of Metrology, China (NIM). The critical electrical characteristics of the latest OHAAE are investigated by the model tests and eventually implemented with the least mechanical modifications to the new VCCC at NIM. The standard uncertainty of the reproduction of 1 pF capacitance unit from the new VCCC by adopting the latest OHAAE is decreased to $10.5 \times 10^{-9}$ in comparison with that of $20.0 \times 10^{-9}$ by the geometrical nosepiece.

This paper describes an absolute method for calibrating commercial 1000:1 dc current transducers (DCCTs). By adding two auxiliary windings to produce 10:1 and 100:1 ratios, we use a three-step build-up method to establish the 1000:1 current ratio standard. First, we determine the 10:1 ratios of 11 DCCTs through permutation; second, we set up a two-stage composite DCCT to produce a 100:1 ratio; after switching auxiliary windings in the two-stage setup, we then derive the 1000:1 current ratio. The experimental results show that the requirements for the current source and the current meter are not stringent; their Type-B contributions to the overall uncertainties are very small. Type-A uncertainties which include factors such as the experimental environment and possible drifts of the output offsets of the DCCTs are experimentally determined. Repeated tests show that the current ratio standard has an expanded uncertainty of $0.26 \times 10^{-6}$ for the 1000:1 ratio.

To compensate the end effect error in a vertical calculable cross capacitor with a fixed and a movable guard electrode, two different methods, a nosepiece and a modified active auxiliary electrode, are usually adopted at National Measurement Institutes across the world. However, the latter electrical approach can be more difficult when assembling the mirrors in the resonant cavity of the Fabry–Perot laser interferometer. This paper describes a novel hollow active auxiliary electrode that can satisfy the optical requirements. The model test experimental results show that it has the same or better compensation effectiveness for the end effect error in comparison with the modified active auxiliary electrode.

A new type of calculable capacitor was initially built at the National Institute of Metrology, China, in cooperation with the National Measurement Institute, Australia. The capacitor realized the SI unit of capacitance at 1 pF, with an uncertainty of 20 parts in $10^{9}$ . A simplified operational approach was used that avoids the complex substitution measuring process with small-value capacitors. The details of novel aspects of the installation and adjustment processes are presented, together with the experimental results and an initial uncertainty estimation.

The precise dc-based measurement of the mutual inductance is one of the key issues in the Joule Balance method used to determine Planck's constant h. Here, a new compensation method with a standard square wave is provided. The dc value of the mutual inductance is obtained by measuring the integral area of the induced voltage. To increase the accuracy of this integral method, we just accurately measure the residual area observed between the induced voltage of the mutual inductor and a synchronously generated standard square wave. Experimental results indicate that this approach is consistent with the previous frequency extrapolation method within 1 ppm. The measurement uncertainty is estimated to be about 0.22 μH/H for the 340-mH mutual inductor at room temperature. A better result of 0.1 ppm could be obtained when the superconducting coil is used.

The precise measurement of Planck constant h to establish a kind of quantum mass standard is an important but very difficult work. The Joule Balance method, for the same purpose but different from Watt Balance method, has been going on at the National Institute of Metrology for three years. All the measurements were carried out in the atmosphere of an air-conditioned room. The uncertainty is several parts in 104 (k=1) and expected to be improved when superconducting coils and other approaches are used further. The principle and progress are described here.

Traditionally, the most precise way to get the mutual inductance value is by a Campbell calculable mutual inductor. Here, an approach of measuring the mutual inductance at low frequency with compensation method is described. A multi-channel direct digital frequency synthesis source is developed to provide stable and precise phase. The mutual inductance is traced to the value of the frequency and resistance directly. The measurement uncertainty is 0.43 ppm at present and could be improved further.