On Fri, 16 Jul 2004, Major Variola (ret) wrote:
Does anyone *know* (first or second hand, I can speculate myself) which laptops, if any, can safely go to zero air pressure (dropping from 1 atm to 0 in, say, 1 minute.)
Sorry so late ---but your can-shaped capacitors might not handle the rapid depressurization so well.
Perhaps it's time to challenge the introductory assumption. Why a laptop? There are many various embedded computers available on the market, eg. the one from <http://www.gumstix.com/>. (Question for the crowd: anybody knows other comparable or better Linux-ready affordable embedded computer solutions?) You may like to take such module and seal it in resin in order to shield it from the pressure changes (question for the crowd: would it really work?). Use memory card instead of hard drive; you don't want moving parts that depend on air density. The smaller size and lower power consumption than a laptop has makes many issues, from cooling to powering, much easier; vacuum-proofing and testing of the assembly is potentially simplified as well. I'd also be cautious about the fluorescent tubes for the displays, the glass won't necessarily have to withstand the rapid change in air pressure. The LCDs themselves consist from two layers of glass with a electricalyl-sensitive light-polarizing liquid between them, make sure it won't have tendency to boil or vaporize in vacuum. Optionally, for unmanned operation, do without the display completely. For manned operation, use something like the head-worn see-through <http://www.microopticalcorp.com/> display, located in the operator's pressure suit, and connect it to the computer by a suitable wired or wireless connection. If the system has to go beyond the reach of the atmosphere, you would like to use some sort of radiation shielding, or use a redundant assembly with several computers working in parallel, compensating lower reliability (silicon-on-insulator chips are difficult to find in off-the-shelf setting) with redundancy. You may also prefer to keep critical systems working on lower frequencies, with older-design parts, using bipolar transistors instead of CMOS (which tends to trap charged particles in the insulator layers of the gates, which shifts the gate threshold voltage), and chips with larger structures (so the ionization traces of particles won't affect the chips that much). Protect the content of the memories - large arrays of rad-sensitive elements - with ECC codes. GaAs is also more radiation resistant material than silicon. Again, combine rad-hard design with redundancy for best results. Cooling is a royal bitch. You can't use anything but radiation cooling. I think satellites use a neat trick with pipes containing a wick soaked in a suitable liquid, eg. some freon. The liquid is vaporizing on the hot end of the pipe, condensing on the cold end, and soaking back to the hot end by capillary forces; this is used to bring the heat from the power parts and the sun-facing side of the satellite to the dark side of the satellite, from where it radiates to space. (Question for the crowd: Can thermal imaging be used for scanning the sky for low-orbit satellites? Other question for the crowd: How suitable would be this wick-in-a-tube approach for "ground-level" computers, could it increase the efficiency of heat transfer from the CPU chips to the wings of the heatsinks? Eg. for the purpose of having the computer sealed in an RF-shielded enclosure, with the heatsinks being part of the case, which could eliminate the cooling air inlets?)