Zum Inhalt springen

Accurate Prediction of Vacuum Capacitor Lifetime Reduces Unplanned Downtime by 80%

Veröffentlicht Juni 22, 2022 von Andrew Merton

Vaccum

The failure of any key element or subsystem in a semiconductor manufacturing facility has the potential to bring the process to a complete standstill and/or to force costly wafer scrap. A key consequence of such failure is an associated increase in the cost of operations. While rarely reflected in the upfront Capital Expenditures (CapEx) pricing, from a Total Cost of Ownership (TCO) perspective, manufacturers typically find that unplanned downtime, raw material wafer costs, unscheduled repairs and the purchase of spare units can be significant contributors to Operating Expenses (OpEx).

With today’s complex semiconductor manufacturing processes, operators are looking for ways to identify not only the potential points of failure but also ways to predict when a failure is most likely to occur. Armed with this information they can then proactively address problems before they lead to costly, productivity-impacting downtime.

Collecting and analyzing equipment data to create meaningful intelligence on which to base predictive maintenance has not always been easy. Now, with PowerInsight by Advanced Energy™, operators can obtain actionable insight from critical power delivery systems that will allow them to maximize performance and cut costs.

Take, for example, the need to predict the lifetime of the vacuum capacitors employed in many of today’s impedance matching networks. Thanks to AE’s proprietary algorithm and years of experience, customers can accurately predict when their vacuum capacitors need refurbishment, allowing them to reduce unplanned downtime by as much as 80 percent.

Determining the Remaining Life of Vacuum Capacitors

As a critical element in the plasma process, the impedance matching networks used for semiconductor manufacture must minimize process variability and deliver high levels of system uptime. One potential cause of failure in a matching network is the end of life of the vacuum capacitors that provide impedance variation. These components have a finite lifespan and will need to be replaced periodically during a network’s useful operating life.

It is difficult to determine how long a vacuum capacitor will last. According to Advanced Energy’s research, capacitors have a wide lifetime variation due to the almost unlimited number of unique tool setups and processes. The actual lifetime of a match network capacitor is dictated by usage pattern and application rather than being solely dependent on time or energy. Vacuum capacitor manufacturers provide an expected lifetime given in number of cycles across its capacitance range, which is traditionally very difficult to measure. Finding a way to predict this capacitor lifetime would deliver significant benefits by preventing unplanned downtime and costly maintenance of the wafer processing chamber and the match network. With AE’s PowerInsight, such prediction has become possible.

PowerInsight is a monitoring and analytics platform that has been specifically designed to bring actionable intelligence and insights to critical power delivery systems and allow operators to maximize performance, cut costs and improve yields. This platform can access more than 50 fields of internal processor and control data from Navigator® II, a rapid, accurate and reliable digitally-tuned matching network that delivers highly accurate RF plasma control.

In developing the solution, Advanced Energy’s in-house reliability lab conducted highly accelerated life testing to validate mechanical wear and failure models from the capacitor manufacturer. This testing provided the input to create the predictive models and analytics that would be used as the basis for determining capacitor life. Using data collected from Navigator II, PowerInsight uses proprietary algorithms to observe details of vacuum capacitor movements and compare them with historical usage patterns. This allows an estimation of the ‘consumed life’ of the capacitor. Using this estimation, coupled with the supposition that the observed usage patterns will be maintained, PowerInsight can project how much time and/or throughput remains until the unit should be scheduled for refurbishment.

For many Advanced Energy products, including the Navigator II, PowerInsight uses a ‘plug-and-play’ hardware device with integrated edge computing capacity, known as the Explorer. In addition, the Ascent® SMS AP-10 is the first AE product with integrated PowerInsight capabilities. By connecting the Explorer to a browser, customers can immediately see data flow and customized analysis chart. An easy-to-use dashboard displays the percentage of life left for each vacuum capacitor, as well as an estimation of days left until a refurbishment will be required on the unit.

Customer Deployment

In the past, customers were faced with making refurbishments upon failure due to the inability to forecast vacuum capacitor lifetime. Now, thanks to AE’s proprietary algorithm and years of experience, they can take a highly proactive approach and accurately predict when their unit is at the recommended manufacturer refurbishment interval for the vacuum capacitors. This allows customers to maximize the useful life of their vacuum capacitors and reduce unplanned downtime by as much as 80 percent.

Teilen Sie

Andrew Merton

Advanced Energy
Andrew is a Data Scientist with PowerInsight at Advanced Energy ®, responsible for defining data acquisition & analysis requirements and to develop algorithms & metrics to create actionable intelligence from field data to guide diagnostic activities and to estimate asset performance & health with respect to preventative maintenance schedules; he leads a team, working with internal and external customers, to support analysis across a large suite of Advanced Energy product offerings including RF and DC generators, RF Matching Networks and Remote Plasma Sources. He holds a Masters of Mechanical Engineering from the University of Colorado at Boulder and Doctorate of Statistics from Colorado State University at Fort Collins. In his free time, Andrew likes to hike, travel and enjoy good food & wine.
Mehr Beiträge von Andrew Merton

Verwandte Beiträge

durchsuchen
Kategorien A-Z
Tragen Sie sich in unsere Mailingliste ein
Abonnieren
Neueste Beiträge
View on X
Code added on 30 Aug 24

Looking for the right solution? Let's chat!

Chat Now Just Browsing