The new gremlins of refurbishing complex step manufacturing tools for user-defined computerized automation

More than a decade ago, mainstream semiconductor manufacturing tools reached an impasse in technological standards and economical maintenance, and studies to find long term solutions to improve tool operation and outputs prompted the semiconductor industry into risk assessments of whether to upgrade existing equipment or pay the enormous cost of new, technologically advanced and computer ready tools. This assessment was approached with great care due to the investment in banks of duplicate older machines doing large volumes of repetitive processes. The innovations of microcontrollers, PLCs and the PC platform led to cheap, accurate solutions for machine conversion to near state-of-the-art automation. Expertise to train and learn to implement the conversions appeared to be within the scope of existing engineering and technical abilities and, in some cases, reasonable outside specialists in rebuild technology. The only favorable asset for the older fixed tools was that user reconfiguration possibilities were very limited and most tools were operated with OEM-created recipes and process steps or hard-wired hardware sequencing. However, evolutionary and later generation developments created complexities which are often beyond maintenance and user ability to control. Repetitive and automated programs created over several generations of hardware upgrades and changes in process tolerances create subtle, disruptive quirks that defy even the experienced user. This paper describes the scope of the risk assessments required for such an automated system to contain and minimize the potential for failure of user programmable sequencing machines