U.S. intercomparison of Josephson array voltage standards

In the Summer and Fall of 1993, ten laboratories participated in an intercomparison of their 10 V Josephson array standards. The transfer standards were circulated in a "daisy" pattern to minimize the time difference between the pivot laboratory and participant laboratories, thus minimizing the effect of long-term transfer standard noise. The results were encouraging. With the assumption of a group mean equal to 10 V exactly, all laboratories were within 25 ppb of the group mean. Errors this small are more than adequate for supporting any existing equipment. Further analysis of data shows very small systematic errors, most likely caused by thermal emfs generated by above ambient temperatures on the transfer standard output connectors. We describe the experiment, present the results, and discuss implications for future round robins and laboratory proficiency testing.>