The two main challenges to operating a programmable Josephson voltage standard (PJVS) on a cryocooler are the available cooling power and the temperature oscillations of the cold head. We overcame these challenges and successfully operated a PJVS circuit on a cryocooler by employing one supercritical helium buffer that damps the temperature oscillations, developing a new cryogenic package that increases the thermal conductivity between the chip and the cold head, and increasing overall device performance with fabrication improvements. A 1.32 mA step width of the quantized voltage produced with all of the subarrays of the PJVS circuit biased was achieved at an operating temperature of 4.3 K. The quantum accuracy of the PJVS is maintained at temperatures up to 4.8 K. This result was obtained with a cryocooler that employs a 3 kW water-cooled compressor to produce at the chip about 270 mW of net cooling power at 4.3 K.
An ac-Josephson-effect determination of
he with significantly improved accuracy is reported. The precision of the measurement is determined by uncertainties associated with the comparison of a Josephson-device voltage with the emf of an electrochemical-standard-cell voltage reference and is about 3 parts in
108. This precision was made possible by use of Josephson devices at voltages above 10 mV and design and construction of two special voltage-comparator instruments. The fabrication and operation of the Josephson devices and the design and performance of the voltage comparators are discussed. The 3/
108 precision represents the precision with which a drift-free and readily reproducible Josephson voltage standard can be realized in practice using the techniques developed for these experiments. The accuracy of the final result is about 12 parts in
108 and is determined primarily by uncertainties associated with the stability of the local electrochemical voltage standard and with establishment of the relationship between the local volt and the volt maintained by the U. S. National Bureau of Standards. Significant improvements in the maintenance of the local-voltage standard which contributed to reduction of the final uncertainty to this value are discussed. During the course of the experiments, the Josephson frequency-voltage relation was shown experimentally to be independent of magnetic field, temperature, and Josephson-device bias voltage or induced step number to within the accuracy of the final result. The final experimental result and its one-standard-deviation uncertainty are $\frac{2e}{h}=(483.593718\ifmmode\pm\else\textpm\fi{}0.000060)\frac{\mathrm{MHz}}{\ensuremath{\mu}{V}_{\mathrm{NBS}69}}$ (0.12 ppm) referred to the volt as maintained by the U. S. National Bureau of Standards after January 1, 1969. This result is in excellent agreement with the earlier, less accurate result of Parker, Langenberg, Denenstein, and Taylor, which played an important role in the 1969 adjustment of the fundamental physical constants by Taylor, Parker, and Langenberg. It is in reasonable agreement with values recently reported by several other groups. The significantly improved accuracy of the present result makes possible a small improvement in the accuracy of the derived value of the fine structure constant and clears the way for a larger improvement through more accurate determination of the proton gyromagnetic ratio.