A reliability engineer’s dream comes along

by Debra Vogler, senior technical editor, Solid State Technology

Jan. 26, 2009 – In a former life within the semiconductor standards and reliability engineering community, I advocated for the adoption of well-known reliability and maintainability engineering practices that have been used for decades in the aerospace and defense sectors. Only in the last few years, though, have some of these principles developed legs in the semiconductor equipment community — e.g., greater use of standard parts and commonality of designs, conducting FMEA/FMECA (failure modes and effects analysis/failure modes, effects, and criticality analysis) and using HALT (highly accelerated life testing) and HASS (highly accelerated stress screening). While not the only company adopting these principles, Crossing Automation’s entry into the ring comes at one of the most challenging of economic times for the industry and it was willing to provide supporting data to back up its claims of a comprehensive reliability program. Some of that data is cited below.

The company recently launched its ExpressConnect family of what it calls modular building blocks of automation components for vacuum wafer handling systems. Touting its approach as “configuration neutral,” the company says the system provides a 20%-70% smaller footprint with a throughput of up to 250wph.

The product family consists of five building blocks:

– Trans-Center (wafer load-lock),
– Multi-Wafer Buffer (a 4-wafer transfer module),
– Dual Wafer Load-Lock,
– Shuttle-Lock (to manage all wafer transfers in a system), and
– SEC-1000 (a common controller to drive the other sub-systems).

These modules can be connected together in different configurations to achieve specific system-level requirements; end users retain control over proprietary configurations that enable market differentiation on the basis of process.

Larry Dulmage, VP of marketing at Crossing Automation, told SST that HALT is conducted on critical mechanical sub-systems via highly accelerated duty cycles and motion profiles; these tests are conducted in vacuum and replicate the loads. “Electrical components (e.g., the controller) are subjected first to a HAST (highly accelerated stress test), including accelerated thermal cycles to demonstrate the forecasted MTBF [per the Coffin Mason criterion],” he explained (Figure 1). In addition, the controller was subjected to combined progressive thermal and vibration shock to determine the destruct limits (Figure 2). Long term reliability testing is still being conducted with mechanical cycling beyond HALT to mechanism wear-out.

Objective 1: Demonstrate 100k hours MBTF via thermal cycling. According to Coffin-Manson method, 26 cycles are needed to demonstrate and MTBF of 100k hours. The low temperature for the thermal cycle is 5C and the high temperature is 45C. Each temperature dwell cycle is 3 hours. This means 26 cycles in approximately 3 days. (Temperature measurements at 11 locations in the controller) (Source: Crossing Automation)
CLICK HERE for larger version

Reliability and maintainability must be designed into a product as well as maintained throughout the manufacturing process and field deployment, so the company conducts FMEA/FMECA as a part of its design for reliability methodology. “The focus is on all critical modules that have significant reliability risks associated,” Dulmage said. “The analysis is conducted early in the design process, critical design reviews are used to lower the risk priority of the identified failure modes.” Additionally, reliability risk for non-critical subsystems is done through vendor data and selected engineering analyses to verify overall structural integrity.

Objectives 2-3-4: Lower and upper temperature limits and rapid thermal cycling (temperature measurements at 11 locations in the controller). (Source: Crossing Automation)
CLICK HERE for larger version

According to Dulmage, the common fab practice of using the latest tools for critical applications and the older tools for non-critical applications has worked in the past, but it’s also expensive. Another model the company wants the industry to consider is shifting non-critical tools to less expensive applications. Some of the critical tools could be put toward other solutions that are more productive and cheaper to build than circular cluster tools. “Out of a very few parts, you can build lots of different solutions…building a process-specific tool, including the automation, that hits not only a throughput number, but also a cost number and a results-on-wafer capability,” said Dulmage (Figure 3).

Examples of system configurations using the standard building blocks. (Source: Crossing Automation)
CLICK HERE for larger version

The company also offers a cost-effective way to repair a system — and enhance cost-of-ownership — by exchanging parts. “By decoupling all the mechanisms, we make it much simpler from a repair perspective,” Dulmage told SST. The company has an exchange program for three primary components (the equipment controller, the z-theta mechanism, and the extension mechanism). Dulmage says that the company will provide replacements within 24hrs.

All manufactured parts, including the refurbished ones, undergo a detailed manufacturing characterization testing, which includes an accelerated burn-in testing (equivalent to HASS). “Data from manufacturing testing and refurbished parts is stored to drive reliability improvement programs and derive reliability estimates of field replaceable units [FRU],” said Dulmage. Replacement of the FRU is based on the number of cycles accumulated and on the measurement signatures acquired via the automated test capability of Crossing Automation’s remote client GUI. — D.V.


Easily post a comment below using your Linkedin, Twitter, Google or Facebook account. Comments won't automatically be posted to your social media accounts unless you select to share.