Auto shutdown at boot--overheating?

[Squidboy]

Today, after running my Hackintosh for about 5 hours I had a kernel panic and when I rebooted the machine would shut itself down during the boot. Sounded like a heat issue so I removed the CPU fan (came with the CPU--Intel Core I3-2105 Processor) and the fan wasn’t properly seated. So, I pull it out, wipe off the heat paste, wipe off the CPU, apply a rice-sized bit of goop, and put the fan back. Now the fan looks right.

I reboot, and it shuts itself down again. I re-do the heat paste. Reboot and it shuts down. I go away for a couple hours, I come back and go into BIOS to check the temp. Initial reading is 36c (room temp, maybe 23C), heat goes up over 10 minutes to 51C. Seems a bit hot, but not auto-shutdown hot. I boot to my installation USB stick and I get to the installer. I try to boot from my HD and it shuts down at the spinning wheel.

Does it make sense that I can get into BIOS and boot from my USB drive but that heat is causing my HD to fail at boot? Is there something I need to “reset” after having a heat-caused shutdown?

I’m going to try applying the thermal paste again...but if that was the problem, wouldn’t it shutdown if running off the USB stick?

 

[powerpcg5]

after running my Hackintosh for about 5 hours ..... and the fan wasn’t properly seated.

Maybe you damaged your CPU already.

 

[Squidboy]

UPDATE: I cleaned off the old, and reapplied new thermal paste and after watching BIOS for about 20 minutes it hit 43C, which is better. Still, I get an automatic shutdown at the spinning wheel if I try to boot from the installed disk. I booted from my install USB and ran disk utility and that didn’t find any problems.

The Hack has only been running for a couple days and I think the only software I added was Chrome (an SSD holds the OS/applications, a “regular” HD houses documents), so reinstalling the OS wouldn’t kill me, but if there is some fix that would save me that bother (especially if something similar happens again) I’d sure like to hear about it.

TIA

 

[Squidboy]

powerpcg5, I was hoping no one would say that, but I guess it is a possibility. It’s only been running a couple days though, I’d be kinda surprised if the first indication of a heat issue was the CPU dying.

 

[Squidboy]

I haven’t run the installer past the point of using Disk Utility, but if the CPU was dead, would it have even gotten that far?

 

[Squidboy]

UPDATE

After letting it sit overnight I decided to reinstall the OS and see if that did anything. I removed one RAM chip so I just have 4 gig in. Reformatted the drive and started to install and about 8 minutes in, I get a screen I never saw before “Install Failed....the installer encountered an error that caused the installation to fail...”

So, running a little hot, but not hot enough to trigger a shutdown, for a couple days might have really damaged the CPU? Seems a bit off, it’s not like I was overclocking. I’ve seen other machines run hotter for longer without such drastically reduced life, but maybe they were just lucky (or I’m very unlucky).

No one around me carries a decent selection of parts so I’ll have to order it. I’ll update when I get the new CPU in.

 

[Squidboy]

The new CPU arrived yesterday. I was better about getting the fan/thermal paste right this time and after about a half hour in BIOS I was stable at 34 degrees.

Then I tried to run Unibeast.

I kept getting installation errors during....uhm....installation, so I did some poking and saw that it might be RAM, so I replaced the one chip that was in with the other chip I removed for installation. That worked and I was able to get it loaded. After a little tinkering with MultiBeast (with a couple kernal panics on restarts along the way) I was able to get graphics to work right (see viewtopic.php?f=15&t=65318 for more on that), as well as Ethernet and audio. For one glorious moment, all was right.

Then I shut down so I could hook my data HD up (I have OS and applications on an SSD and my Home folder on a regular HD; I removed the regular HD incase that was part of the installation problem I was having). Rebooted to a kernel panic, and another, and another..... Ran Unibeast again, thinking that if I got another installation error that maybe the problem was still with RAM...and I got another installation error. Exchanged RAM again, cleaned the chip and socket with some canned air, removed data HD, still get the installation error.

It was so nice when it lasted.

So, now I’m thinking that when I as running hot maybe I also damaged the RAM. So I have a replacement on order.

 

[powerpcg5]

Boot and run memtest before any installation. Will save you a ton of trouble.

 

[Squidboy]

Thanks, getting that ready in anticipation of another attempt.

 

[Squidboy]

Finally got the replacement RAM and I'm up and running again. Still hard to believe that I so quickly killed the CPU and two RAM sticks, but now I'll know better forward.