Data mining to detect ion source failures in varian VIIsta implanters

Ion source replacement is the most frequent maintenance activity on an ion implanter, impacting both productivity and consumable costs. Quality of parts and quality of build are essential to maintain proper beam quality throughout a normal projected source lifetime. Early detection and intervention of a source that is prone to premature failure will enable service engineers to complete corrective maintenance before an unplanned failure surfaces during a process step. This paper shows how data mining was used to detect the failure of indirectly heated cathode (IHC) type ion sources used in Varian implanters.