ChipCrunch.com

As of late we have seen a slew of startups that focus on finding digital implementations of what are typically considered analog problems. Examples include Blacksand Technologies and digital amplifiers, Bitwave Semiconductor and programmable transceivers, and WiSpry’s tunable radio frequency (RF) technology. The reasons for this are many: analog circuits are tricky to design, significantly affected by variation and noise, often time difficult to integrate together with other digital circuits especially on the same die, and most likely require special process rules. Last but not least, there a lot fewer analog designers these days than digital designers and finding a really good one can be a challenge in itself.  As such, with price-point and low power operation being the dominant factors in the market currently, digital implementations become more preferable.

With this in mind, Powervation, a fabless semiconductor startup based out of Limerick, Ireland, has chosen to specialize in digital power control. The company was founded in 2006 and is being backed by Scottish Equity Partners, Intel Capital, Venture Tech Alliance, 4th Level Partners and Enterprise Ireland. The company is currently touting a technology that it calls Auto-Control, which is essentially a digital algorithm for DC/DC control circuits. It enables these circuits to sense different conditions such as load, capacitance, and inductance and optimizes the power-control function to compensate accordingly. The digital algorithm replaces the analog compensation loops that were traditionally required to stabilize voltage conversion and transient response. Powervation’s first product, dubbed the Plug-and-Power controller is aimed at power converter solutions in several different spaces. It is expected to become available later this quarter. While the company reveals little about the digital algorithm or its actual implementation, Peter Cark’s column over at EETimes Europe suggests that the initial results looks quite promising. Overall system efficiency gains of up to 30 percent can be expected with a 65 percent reduction in component count when compared with analog implementations.