From k.buzzard@ic.ac.uk Sat Sep 9 17:35:26 2000 Received: from judy.ic.ac.uk (judy.ic.ac.uk [155.198.5.28]) by abel.math.harvard.edu (8.9.3/8.9.1) with ESMTP id RAA14754 for ; Sat, 9 Sep 2000 17:36:17 -0400 (EDT) Received: from juliet.ic.ac.uk ([155.198.5.4]) by judy.ic.ac.uk with esmtp (Exim 2.12 #1) id 13XsHX-00030q-00 for was@math.harvard.edu; Sat, 9 Sep 2000 22:35:27 +0100 Received: from geometry.ma.ic.ac.uk ([155.198.192.7]) by juliet.ic.ac.uk with esmtp (Exim 2.12 #1) id 13XsHf-0002PF-00 for was@math.harvard.edu; Sat, 9 Sep 2000 22:35:35 +0100 Received: from blair ([155.198.192.160] helo=blair.ma.ic.ac.uk ident=kbuzzard) by geometry.ma.ic.ac.uk with smtp (Exim 1.90 #1) id 13XsHW-0003oC-00; Sat, 9 Sep 2000 22:35:26 +0100 Date: Sat, 9 Sep 2000 22:35:26 +0100 (bst) From: Kevin Buzzard X-Sender: kbuzzard@blair.ma.ic.ac.uk To: was@math.harvard.edu Subject: Keyboard problem---fixed (I think) (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Length: 6001 Status: R X-Status: N And here's my solution email. Since then I've been switching the laptop off most nights andthe problem hasn't recurred. By the way, I found this old email using the following procedure: 1) grep screw Mail/* and then spotting that the mail was there, in sent-mail-aug-2000 2) grep -A20 -B20 screw Mail/sent-mail-aug-2000 | less which told me the date I'd sent the email and then using pine to find it and forward it. The only reason I mention all this is that I would have no idea how to do it using, say, a windows mailer. Kevin PS on reading the old email I see I am vague about some things that I don't know what they are, you could of course just edit my email or clarify it if you liked: this email is in the public domain and not copyrighted so do what you like with it :-) ---------- Forwarded message ---------- Date: Wed, 12 Jul 2000 16:57:36 +0100 (bst) From: Kevin Buzzard X-Sender: kbuzzard@blair.ma.ic.ac.uk To: was@math.berkeley.edu Subject: Keyboard problem---fixed (I think) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Keywords: X-UID: 141 OK, so I took those magic 11 screws out of my computer. I did this by buying a screwdriver, getting Tamzin to look after Joel, turning my computer off, closing the lid, removing the battery, turning the computer upside-down so that I could see the 11 screws, and then starting. The orientation of the computer when I turned it upside-down was such that the words "LOCK" and "UNLOCK", referring to the places which lock/unlock the battery, were upside-down. I dismantled from left to right, so the first screw I removed was quite near the headphone/mike outputs, the ninth was by the button which rotates but doesn't do anything, and the tenth was by the power-on switch. This brings us to the 11th, which floored me for a while because I couldn't get it out for a good few minutes! The 11th screw is the one near the ethernet-or-whatever-it-is input and the fan. I didn't want to push down too hard because I didn't want to smash the screen of the monitor; on the other hand, I seemed to be destroying the screw at the beginning---ruining the screw head instead of rotating it, which didn't look so great. After a while I got it to budge but even then it was non-trivial to get it out: even after the initial movement, the screw did not move freely. After some coaxing I got it out. The screw was clearly different from the others: it was blackened at the bottom and was obviously slightly damaged. Having done all this I was interested in the state of the keyboard so I kept the computer turned over and opened up the monitor. I don't recommend this: the monitor opened and the keyboard fell out onto the table, together with a piece of metal. Having cursed my stupidity at not guessing that things perhaps would not be as joined together as usual after having removed 11 screws, I decided not to take the back of the computer off, but to leave it in the state it was. I turned the computer over and worked out what had happened. The piece of metal appeared to be a heat sink and the word "UP" was written on it; I was now sitting with the computer opened and oriented as if I were using it. I put the piece of metal back so that the word UP was face up and upright :-) and it fitted nicely over two chips. Now the computer looked normal, except the pieces of plastic on the sides were kind of falling out, and the keyboard was off. My first piece of relief was that I realised that I had misguessed what was happening: underneath the delete and backspace keys there was *nothing* important at all that I was continually bashing whenever I pressed them; there was just plastic/metal/whatever it is. So I realised that I probably hadn't (yet) done the computer any damage unless I'd damaged the keyboard (I was worried that when the keyboard was "loose", I was smashing into a chip or something, whenever I pressed the delete key). >From the bottom of the keyboard it was clear that the troublesome 11th screw was holding the keyboard _down_ and that in fact the reason the keyboard was "giving" slightly at the top RH corner was that it was about 1mm too _high_. I then simply put everything back together, exactly as it had originally been, except that I put the damaged screw in the unique hole which could reasonably be called "the middle hole" and I put the screw from the middle hole back into the RH centre hole, the one near the fan. Then the keyboard was back to normal and the computer seems to work fine, and the problem looks solved to me. My conclusion: leaving my VAIO on for about 10 days straight computing modular forms and hence processing _hard_ has kept the fan on "high power" and has kept the computer running slightly hot (start up a long and complicated magma calculation, leave it for a while, and then kill it and listen: the speed at which the fan is going decreases noticeably!). Not only did we have this 10 day marathon; I have basically left my VAIO on doing very CPU-intensive things ever since the 26th June or so, which was about 3 weeks ago, only rarely switching it off---usually it's on all day and all night. The resulting heat seems to have mildly damaged this 11th screw with the result that it wasn't holding the keyboard down correctly any more. Perhaps in retrospect, if I'd known this was the problem, I would have just removed this one screw and put in a new one. It frightens me slightly though: if this is the damage caused to a screw by leaving the computer on for 3 weeks, what else have I damaged? There is in fact no indication that I've damaged anything else, and the screw was only mildly "charred", it wasn't completely melted or anything, but it was distorted enough to make it hard to remove. I am now in a slight quandary: I really want to leave my VAIO on all the time, but I don't want to damage it any further. What do you think? Kevin