[ Date Index ] [ Thread Index ] [ <= Previous by date / thread ] [ Next by date / thread => ]
On Saturday 12 December 2009, John Horne wrote: > On Sat, 2009-12-12 at 12:44 +0000, Gordon Henderson wrote: > > I've seen a 1U server get hot enough to emit copious amounts of smoke and > > melt itself internally. No real flames though. > > I had a Sun E250 at work which caught fire internally - not sure what > exactly but it may well have been the psu. Smelt awful and with > flames :-) Fortunately I was there at the time - the smell was the > giveaway of something not being right. Fortunately also that it burnt > itself out quite quickly as the computer room had automatic halon > extinguishers at the time (i.e. when they go off they remove the > oxygen!). > > > > John. > > -- > John Horne, University of Plymouth, UK > Tel: +44 (0)1752 587287 Fax: +44 (0)1752 587001 It's surprising how often smell, hearing and (very cautious) touch come into hardware faultfinding. Hearing tends to be fans now but it used to be the crackle and buzz of EHT flashover. I love the smell of ozone in the morning. And at one time if you didn't get a belt every week from the HT supply you were losing your touch. (OK, OK, but it is early in the morning.) 240 volt mains is nasty but you give 400 volt 3 phase respect. Modern circuit boards? Easy peasy. And software is just that. George -- The Mailing List for the Devon & Cornwall LUG http://mailman.dclug.org.uk/listinfo/list FAQ: http://www.dcglug.org.uk/linux_adm/list-faq.html