GE boffins build micro-lungs to cool PC innards
Puffing piezoelectronics outpace fatter fans
Researchers at General Electric has developed a new air-cooling system based on human lungs that's half the size of conventional computer fans but just as effective at keeping things chilly.
You just purse your lips and... (click to enlarge)
The prototype uses 40mm by 40mm thin metal plates which are bonded to ceramic piezoelectronics, dubbed Dual Piezoelectric Cooling Jets (DCJ), that flex like bellows when AC current is applied. The 1mm-thick micro-lungs vibrate at between 100 and 175 hertz, producing a 1mm-wide jet of air coming in and out of the device.
As turbulence flows off the sides of the stream, the cooling area is greatly increased and the speed of the jet is such that the same air isn’t reused. The team demonstrated the system by using it to cool what looks like a Lenovo X1 Carbon Ultrabook, and they say the new system uses less than half the battery power of a conventional fan that's double its width.
The prototype design can cool up to 12-18W and uses less than 240mW to do so, and lack of a DC motor or bearings means it may well last longer, Peter deBock, GE electronics cooling researcher, told The Register. It's also quieter, and the frequency of breaths can be adjusted to suit cooling and manufacturing requirements.
But the interior of a PC case can get pretty dusty, and the enclosed nature of the design looks like it could clog up like the lungs of a chain-smoker. The team tested this, however, and it appears that the air velocity is such that the casing stays clean.
"We tested the devices in Mil-STD-810G sand and dust testing environment. They were blasted with really fine grain dust from many different angles," he said. "After four days we had all six devices still working with no problems."
Obviously this kind of system would work well in the tablet and Ultrabook markets, and anywhere else where size counts. But GE isn’t going to be getting into that business, and has licensed the system to Japanese thermal systems builder Fujikura. ®
COMMENTS
Hello, thank you for your comment. Synthetic jets is a family name for devices with a single opening and an oscilating actuator. The GE #DCJ is unique as the dual osciliating plates cancels vibration and the wide slot openening is designed to entrain additional airflow creating a net air flow. It is able to do so in a unique 1mm height and expel air in a planar fashion. I hope this helps.
Test Cycle
"We tested the devices in Mil-STD-810G sand and dust testing environment. They were blasted with really fine grain dust from many different angles," he said. "After four days we had all six devices still working with no problems."
Wow, 4 whole days! :)
I think most tech folks will note that it's not sandy dust that they find in computers but that nasty, sticky, skin-based grey stuff. That's where the testing needs to be.
@PeterdeBock
Thank you for that information Peter. I'm wondering about a couple of points as a result:
With an oscillation frequency of about 150Hz, the wavelength in air will be about 2 metres so any vibration cancellation effects would not be due to air-wave interference at a distance from the source. (I'm not an acoustics expert by any means so I hope anyone with knowledge of this subject will join in....). Do the oscillating plates vibrate in the same direction (one goes into the 'fan' cavity as the other goes out) or do they vibrate in opposite directions (both go in/out at the same time) thus expanding the cavity and then compressing it? Given that a flat plate is vibrating at 150Hz and given the complexities of vibration and sound propagation inside a PC or laptop, I'm surprised that a nearby plate (on the other 'side') is able to cancel sound waves launched by the first plate.
Is the frequency of operation related to the mechanical resonance characteristics of the fan plate assembly, or is it related to requirements due to the characteristics of air flow through the fan?
It seems to be a fascinating and useful development and I look forward to large and silent PCs and laptops one day.
Re: Test Cycle
That's four days consuming as much sand/dust as a device could reasonably inhale in several years I'd reckon.
But yes - the sticky stuff does require other testing.
Hmm.
Or maybe hummmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm.
This would depend how good/bad its isolation/coupling to the surroundings are; low frequency hum is notoriously difficult to keep out of places where you don't want it. Perhaps a three-layer device working in anti-phase to itself?
