Thursday 25/11/2004 Exciting times in miniature await, as I write up a story from Georgia Tech where they've made a chip/coil/magnet combo that generates a watt from a device roughly the size of a pound coin. The only drawback is that the magnet has to whizz around at 100,000 rpm -- possibly many times that if you want laptop-sized wattage. But that's OK, because lots of people are building engines of similar size and capability. They'll probably run on gas or petrol, which raises an interesting vision of tiny Esso filling stations with itsy-bitsy pumps set up to refill the road warrior's gizmos. Wonder if they'll sell sandwiches the size of postage stamps.

Despite a century of work, hydrocarbons remain the most efficient and easy way to carry large amounts of energy about -- which is why there's no prospect of electric 747s in the near future. Yet portable electric generators have been too bulky and inconvenient to compete for low power use: anything less thirsty than a light bulb and you might as well use batteries. The micro-engines may change this -- but I have my doubts.

For a start, anything that involves spinning something heavy around at very high speeds is going to be noisy. The Georgia Tech prototype was powered by the same sort of compressed air turbine as dental drills use, and those whine louder than Watford fans. Then there's the small issue of safety: you have an explosive mixture millimetres away from a flywheel that contains a huge amount of energy. If it jams, it will explode and send supersonic red-hot shards of metal over a very wide radius -- not normally a hot sales proposition in pocket equipment. Oh, and have you tried moving a gyroscope twisting at 15,000 times a second?

Finally, there's the small matter of the exhaust. This will be hot. You thought cooling a Pentium 4 was hard? Wait until you've got a steady stream of combustion products desperately seeking release -- again, something normally thought incompatible with personal technology nestling close to the skin.

All these things can be fixed, of course. Good shielding, heat exchangers, proper audio isolation and safety mechanisms are all theoretically possible. You'll end up with something as portable as a car battery, mind: fuel cells, which have far fewer intrinsic design problems, are still years away from commercial acceptance in portable gear.

So let's enjoy the exciting world of microtechnology and whizzo devices -– after all, any story that gives me the excuse to publish the word 'Wankel' is to be cherished -- but don't sell those shares in nickel metal hydride battery companies just yet.