Hot hot hot
May. 25th, 2005 08:40 pmMy computer has been running really hot lately. It's running reliably enough, but I'm a bit concerned about how long it will keep working properly with the core running at 59C. I took the side off the case this evening and blew some of the dust out of the fans. With the side sitting off it's running at 51C now - which is still about 5-10 degrees warmer than I'd like to see it running at, but that's better than 59.
All the voltages are a little out too - which would cause it to run warm. I'll have a look at the BIOS to see if I accidentally boosted one of them when I was thinking of over-clocking awhile back, but I'm concerned that the power supply might be going. It's the (fairly) cheap one that came with the case, and right now it's almost hot enough to fry an egg on it. What's it going to do when I drop in another hard drive this weekend? =(
[Edit: I bumped the voltage down a couple of notches in the BIOS, and now it's registering exactly on spec. The temperature has dropped a couple of degrees as well to 48C. That's still a bit hotter than it used to run, but it's an improvement. We'll see how this holds up.]
All the voltages are a little out too - which would cause it to run warm. I'll have a look at the BIOS to see if I accidentally boosted one of them when I was thinking of over-clocking awhile back, but I'm concerned that the power supply might be going. It's the (fairly) cheap one that came with the case, and right now it's almost hot enough to fry an egg on it. What's it going to do when I drop in another hard drive this weekend? =(
[Edit: I bumped the voltage down a couple of notches in the BIOS, and now it's registering exactly on spec. The temperature has dropped a couple of degrees as well to 48C. That's still a bit hotter than it used to run, but it's an improvement. We'll see how this holds up.]
no subject
Date: 2005-05-26 01:45 am (UTC)no subject
Date: 2005-05-26 01:57 am (UTC)The fan died in
no subject
Date: 2005-05-26 02:35 am (UTC)no subject
Date: 2005-05-26 03:28 am (UTC)no subject
Date: 2005-05-26 12:46 pm (UTC)Most are rather optimistically rated, particularly the cheap ones, and when they get close to max they tend to get unstable...
Others... are crap all the way through...
I still remember with dread the old Olivetti M380 machines we had at the office... (386DX/20MHz, and that was the good part)
After about two years of use the insulation on a coil melted off and it shorted out...
Happened to at least four, maybe five of the buggers before we did the sensible thing and scrapped them.
no subject
Date: 2005-05-26 04:29 pm (UTC)Power supply regulator drift can cause unexplained heat problems. Check your power supply outputs with a known-good digital meter:
+12v Line - 11.75 to 12.25v
+5v Line - 4.9 to 5.1v
+3.3v Line - 3.22 to 3.38 (most spec 3.25 to 3.35, but that's a bit tight)
Processors have 3.3v IO and core voltage sections; increases in the 3.3v line can affect processor temps, though not as drastically as core voltage changes will. Most onboard core voltage regulators are pretty crappy, they often vary up to +/- 10% of setpoint and may drift over time (capacitor aging).
no subject
Date: 2005-05-30 06:05 am (UTC)