Internal Comparison of a Large Group of Standard Cells

Equipment is described for comparing any number N of standard cells which constitute a standard of electromotive force. The scheme is economical, needing only 2N measurements. The cells are disposed in a ring, and there is simple switching to compare each cell with its four neighbors. The calculations are done by a program on a desk calculator. They give the value of each cell with respect to the mean of the group. They also give the rms error, which is typically less than 0.02 μV for one cell in a group of 22. There are safeguards for identifying a spurious measurement and a false keyboard entry.