Expanding on the concept of the unbalanced-bridge technique, a new system for measuring standard resistors has been developed. This system is capable of comparing any even number (six or more) of nominally equal resistors from 10 ohms to 1 megohm. The precision ranges from about 0.07 ppm at 10 kilohm to about 1 ppm at 1 megohm. This automated measuring system was developed at NIST, and was presented by Ron Dziuba at the CPEM’88.

This system is currently being using at NIST to measure differences between nominally equal four-terminal Rosa and other types of resistors from 1 kilohm to 1 megohm. Six nominally equal resisters, R1 through R6 (as shown in the figure below), are mounted on a mercury-wetted ring stand located in a temperature controlled oil bath. A precision source of known output voltage is connected to opposite corners of the ring at nodes `AA’, dividing the ring into two parallel strings of three resistors. Six measurements are then taken of the small differences in potential between the opposing terminal taps, i.e., Val, Vbk, Vcj, Vdi, Veh and Vfg. The value used for each measurement is the average of the readings with the voltage source in the forward and reverse directions. The voltage source is then moved to `BB’ and six more measurements are taken between opposing terminal taps. Finally six more measurements can be taken after the voltage source is moved to `CC’. A matrix is formed from the three sets of measurements. From this matrix

the values of the unknown resistors are determined based on their comparisons to the known resistor values.

A Basic program developed at NIST runs the entire measurement process. The program controls two low thermal scanners, one that rotates the voltage source and another one that makes the connections between the resistor potential terminals and the DVM. The measurements are recorded and the results are calculated by the computer.

A guarding system is used to reduce errors due to leakage resistance in the DVM and voltage source circuits. The system can be run at night when the environmental disturbances are at a minimum.

At NIST the system is usually set up to operate four six-arm mercury ring stands at the decade values between 1 kilohm and 1 megohm. Also the system has the flexibility to accommodate lower resistance values down to 10 ohms, as well as other types of standard resistors.

The NIST system compares six resistors at a time including three unknowns, two reference units and one check standard. Uncertainties vary depending on the resistance level. Best results occur when comparing 10 kilohm resistors, with a random uncertainty of about 0.07 ppm. At 100 ohms the random uncertainty is about 0.26 ppm and 1 ppm at the 1 megohm level.

Image11.gif (24832 bytes)

Page 10

Home Page Contents (Page1) Previous Section Next Section