Well I found chocojon investigating the power side of things and noticed the CPU heatsink retention clip wasn't locked. A quick poke from chocojon revealed a plastic nubbin on the socket had broken off and the entire heatsink/fan assembly was hanging off :x
So I guess techinically it was thermal death - a quick whiff of the blackened cpu kinda confirms this :) But root cause was mechanical failure.
Drumm's already back up with a new cpu and a heatsink clamp that attaches to all six (well 5 hehe) nubbins. Hopefully this will stop further explosions :D
Well I'll leave this to chocojon to confirm, but I think he had thermal shutdown at 75degcel. Looking at the damage however the die got hotter than that, it's possible repeated boot attempts drove it over this temp. That's just a guess - I don't *know* how it got that hot.
Now it's got an Athlon 2400 in there - one of the toastiest chips around.. so I'm hoping rikrose will pop some thermal monitoring software on there.
I have temp shutdown in my BIOS and additionally I can run hddtemp (linux software - google for it) to look at the reported temps of my drives through their SMART parameters...
The only other thing I've done recently as chocojon already knows is get shutting down to actually power off the machine. This means if it's a warm day I can log in from home, check the drive temperatures and shut the machine down if it's too warm (although as I hardly use it I now leave it off most of the time...). I compiled ACPI support into the kernel to enable this, and swapped network cards around so that WOL works so I can power it up remotely as well..
Reply
Reply
So I guess techinically it was thermal death - a quick whiff of the blackened cpu kinda confirms this :) But root cause was mechanical failure.
Drumm's already back up with a new cpu and a heatsink clamp that attaches to all six (well 5 hehe) nubbins. Hopefully this will stop further explosions :D
(ooo I like new reply box thingies)
Reply
Reply
Now it's got an Athlon 2400 in there - one of the toastiest chips around.. so I'm hoping rikrose will pop some thermal monitoring software on there.
Reply
I have temp shutdown in my BIOS and additionally I can run hddtemp (linux software - google for it) to look at the reported temps of my drives through their SMART parameters...
The only other thing I've done recently as chocojon already knows is get shutting down to actually power off the machine. This means if it's a warm day I can log in from home, check the drive temperatures and shut the machine down if it's too warm (although as I hardly use it I now leave it off most of the time...). I compiled ACPI support into the kernel to enable this, and swapped network cards around so that WOL works so I can power it up remotely as well..
Reply
Leave a comment