Oops. I obviously spoke too soon congratulating the engineering teams at CERN's Large Hadron Collider near Geneva for powering up the project successfully.
Now it appears that the LHC will be out of action for at least two months after a failure caused up to 100 of the LHC's super-cooled magnets to heat up by as much as 100 degrees C.
The fire brigade had to be called out after a tonne of liquid helium leaked into the tunnel.
The latest setback came a day after the LHC's beam was restored after engineers replaced a faulty transformer that had hindered progress for much of the previous week.
Well nobody said science and engineering was easy. Building the LHC has already taken about 13 years so a few months delay is hardly a blip in the development process.
Let's face it some breakthroughs can take decades to perfect. The point was highlighted at last week's Embedded Power Conference in California by Michael A. Briere, formerly CTO at International Rectifier (IR) and who is now an independent consultant for the company. Briere pointed out that although silicon based power devices had moved on in leaps and bounds during the 30 years since IR unveiled its Hexfet devices, improvements have largely leveled out in recent years while the cost of advancements has increased markedly.
As a result Briere proposed that it was certainly time for a new breakthrough technology and gallium nitride power devices could be the answer. IR plans to begin shipping GaN as early as 2009 while a host of other companies are developing solutions including Sharp, Oki, Matshushita, Fuji Electric, Toshiba, Toyota and STMicroelectronics.
The use of GaN-on-silicon epitaxial technology will be key to a successful roll-out. The new devices promise performance improvements up to ten times that of existing devices with the main applications including AC/DC and DC/DC conversion, as well as Class D audio.