Online information source for semiconductor professionals

During our Tuesday (July 29) chat at the company's modest global headquarters in Menlo Park, Signet Solar's founder-CEO Rajeeva Lahri and VP of manufacturing Bijan Moslehi updated me on the company's latest news and future plans, including the status of the ramp of their first production facility in Mochau, outside of Dresden. The chief exec told me that the final acceptance test, or FAT, should be completed on the initial 20-MW single-junction thin-film amorphous-silicon-on-glass line "within the next 90 days," with a subsequent ramp to 45 MW taking place soon thereafter.

"We should be a peak capacity" within that time, he said, and should end the year running volume production "with full capacity." Now, the factory team is going through what the two former semiconductor guys jokingly called "pipe cleaners," which include debugging, tuning up the automation, and establishing baseline manufacturing windows. Although the plant utilizes a single work shift now, a second one will soon be added, with plans for three-a-day, 24/7 operations once they get past the FAT.

Signet's fab is one of the first to feature deployment of Applied Materials' SunFab turnkey line and its 5.72 sq meter, 120-kg slabs of glass (2.2 x 2.6 m), an example of which was displayed at both the recent Intersolar Munich and North America trade shows.Despite rumors to the contrary, Moslehi assured me that the panel is indeed a working unit. "The first module off the line worked within spec at about 6% efficiency, plus-or-minus 10%." The inaugural panel was one of a batch of 10 that were fabricated in May, a mere 10 months after the start of construction on the factory.

Signet plans to move quickly to production of higher-efficiency tandem-junction a-Si panels. Because the floorplan of the fab was designed to eventually accommodate the additional process chambers required for production of the tandem modules, Moslehi expects the ramp to be quicker. Once operational, "line yields will be 90% or better, with at least 80-85% uptimes."

Lahri said that the company hopes to be at 65-70 MW capacity with the more efficient units by the end of 2009. That level of capacity translates into about 140,000 panels of annual output, if one calculates 2000 500-W panels needed for each megawatt of manufacturing capability.

Both the Signet execs put their stamp of approval on the Applied turnkey toolset, although Lahri admitted there were still "lessons to be learned" and enhancements that could be made. "We have shared lead-line data with them, which they are digesting. But we are very pleased with the AMAT tools--they all met their targets.... Applied will continue to be our tool provider." He also mentioned that one of the non-AMAT tools in the turnkey line that has come under suspicion in some quarters--the laser scriber--has been "a pleasant surprise."

One part of the line that "could use some improvements" is in the area of metrology and inspection, according to Moslehi. Lahri cited the need for a "semiconductor mindset" and the implementation of additional in- and offline monitoring systems and statistical process control measures. "Line 1 is running fine, but we still see lots of room for improvements across the board."

Lahri answered the question of how Signet hopes to differentiate itself from other companies employing the turnkey solar-fab strategy in a couple of different ways. First, he said that even with FAT, "there is still alot of engineering left to do. In the strictest sense, nothing's turnkey." For example, the company has its own supply chain, dealing directly with the glass companies and materials houses like Air Products, which provides the silane for the factory.

He also said that Signet will separate itself by having a "faster time to market and lower costs, being among the 'first movers,'" and through research and development efforts. "We are already filing patents in areas like nanocoatings and triple-junction cells, as well as [things like] productivity enhancements and higher efficiencies.

"We want to achieve 12-15% efficiencies using new materials and device architectures....we feel there is lots of room for improvement in micro/nanocrystalline silicon thin films. "With more critical mass, we may look at more disruptive technologies, either internally or through acquisition."

As for the longer-term value of the turnkey approach, Lahri thought it will be "interesting to see how the integrated line model works.... Once the licensees add a few lines, why would they keep with that model? How would it add value in the future?"

The two told me they got a kick out of my term "BAPV" (for "big-ass PV") to describe the huge glass panels they manufacture. Yet Lahri explained that initially, the majority of customers want the half- or quarter-size modules, while a "significant minority" ask for the full-size panels. He also did not discount the possibility of going to a smaller size substrate, say Gen 5 glass, in future factories that might be located in the developing world where infrastructure might limit the ability to transport the larger panels.

Although future plans call for the construction of a second 65-70 MW factory alongside the current plant in Mochau, another several hundred megawatt site in India, and perhaps a US-based fab too as part of what Lehri calls a "long-term regional strategy," the focus right now is on executing the first few factory ramps and moving product in volume to the customers. He believes Signet can get close to or reach about $3.50 per installed panel by 2010. He admits grid parity (more like $2.50 per) "is not one single number" and depends on location, policy, and other factors, but should be attainable by 2012.

In the next few months, the company will put its panels to the test--on its own factory roof. Installation of a 250-KW array should be done within 90 days, according to Lahri. Although both gents pointed out that the system will only provide a few percent of the factory's energy needs, the chief exec joked that "it's time to drink our own Kool-Aid."