NONSTABLE BEHAVIOR OF WIDELY USED IONIZATION GAUGES

This is the first in a set of related reports on (1) nonstable behavior of widely used ionization gauges, (2) causes of nonstability and nonreproducibility in widely used Bayard-Alpert (BA) gauges, and (3) a stable and reproducible BA gauge design with approximately a tenfold improvement in both stability of gauge calibration and reproducibility gauge-to-gauge compared to older ionization gauge designs. Thirteen widely used BA gauges and inverted magnetron gauges (IMGs) of eight different designs from five manufacturers were operated along with 29 BA gauges of a new design for thousands of hours each and periodically calibrated against a reference standard traceable to the National Institute of Standards and Technology. The older design BA gauges exhibited changes in calibration ranging from -57 to 72% relative to their initial calibration during the first 4000 h of operation and varied within this range over total operating periods of up to 14 000 h. The IMGs showed changes ranging from -90 to 148% relative to their initial calibration during the first 4000 h. A set of five glass BA gauges, nominally identical except for two different port diameters, differed by -20 to 32% from the average of their initial calibration values after a minimum of 48 h operation, and by -18 to 55% from their individual initial calibration values during 3500 h of operation. Because the new design gauges did not exhibit such nonstability using the same apparatus and calibration procedures and the same operating procedures except for the frequency of degas, our results demonstrate that these older design gauges significantly change calibration during long term use.