I think my 920 died

[ioni]

Came home today and my 920 is dead. My BIOS screen greeting me with a failed to overclock message. Removed the overclock and I just ended up in an infinite reboot cycle. What would be a good CPU to upgrade to for ~$200? I was thinking a 4690k since Microcenter has it on sale right now.

 

[MongGrel]

http://forums.anandtech.com/showthread.php?t=2335636&highlight=l5639

http://forums.anandtech.com/showthread.php?t=2335636&highlight=l5639&page=70

 

[ioni]

Not really sure what you're getting at with those links.

 

[ehume]

Do you know that what failed was your CPU? Could the mb have failed?

 

[CropDuster]

http://forums.anandtech.com/showthread.php?t=2335636&highlight=l5639

http://forums.anandtech.com/showthread.php?t=2335636&highlight=l5639&page=70

Not really sure what you're getting at with those links.

Ebay. X5650 if your mobo is still fine.

 

[ioni]

I'm not sure there's a good way tell if it was the mobo or the cpu that failed. Either way, 1366 is long dead. So replacing either individually would be as expensive as getting something new and it'd probably also be used.

edit: Wow, that's awesome about the x5650! Not just to figure out how to tell if it's my mobo or the cpu. Hopefully it's just the cpu.

edit2: I may have jumped the gun. I didn't think RAM would cause the error I saw "Failed to overclock," but removing 2 of my 3 DIMMs seems to have stopped the reboot cycle. So maybe I just need new RAM? I ran the Windows RAM test utility just a few days ago and it didn't detect any errors. I guess it doesn't pick up everything?

 

[MongGrel]

Yeah, if it's the Mobo start scratch I guess, if not, X5650's pretty popular for extending a bit atm.

 

[CropDuster]

I'm not sure there's a good way tell if it was the mobo or the cpu that failed. Either way, 1366 is long dead. So replacing either individually would be as expensive as getting something new and it'd probably also be used.

edit: Wow, that's awesome about the x5650! Not just to figure out how to tell if it's my mobo or the cpu. Hopefully it's just the cpu.

edit2: I may have jumped the gun. I didn't think RAM would cause the error I saw "Failed to overclock," but removing 2 of my 3 DIMMs seems to have stopped the reboot cycle. So maybe I just need new RAM? I ran the Windows RAM test utility just a few days ago and it didn't detect any errors. I guess it doesn't pick up everything?

I'd try resetting the CMOS and get back to stock BIOS settings. Maybe download memtest86 and see if you can figure out which RAM stick is possibly bad.

 

[ioni]

Not really sure what to think the issue is anymore. I tested each RAM stick and one caused my PC to consistently reboot after about 1-2 minutes from getting to the desktop. So I stuck the other 2 back in and was stable at idle for an hour. I leave the room for a second, come back and it's rebooting again! Back to one stick, another random reboot, then stable for about 30 min now.

I'm hoping it is the RAM. Planning to buy a new set tomorrow. I don't think it's the PSU, a 2 year old Antec HGC 620M, which has pretty good reviews.

 

[ehume]

What if it's not the RAM, but something in the first RAM slot? Can the mb boot with the RAM in the second slot?

 

[Brekyrself]

My observations with a 1366 board:

1. Verify all cpu / memory settings in bios along with the voltages. I've seen random settings change after crashes.

2. Check all power connectors for any burnt leads. Twice I have had the same issue you describe. First time a pin on the PCIe connect burnt up, 2nd time it was on the main board connector. I started to see low voltage on the 12v and 3.3v rails plus they would fluctuate with or without load.

 

[VirtualLarry]

Came home today and my 920 is dead. My BIOS screen greeting me with a failed to overclock message. Removed the overclock and I just ended up in an infinite reboot cycle. What would be a good CPU to upgrade to for ~$200? I was thinking a 4690k since Microcenter has it on sale right now.

Clear CMOS, possibly replace the CMOS battery?

 

[ehume]

My, we think all of our stuff as just plug and play. But clearly when it gets a little old, electromechanical things go wrong. Time for a teardown? Pull every item off and take the mb out. Scrub every socket (new toothbrush), every card and plug and removable chip. Reassemble with a new mb battery. All that is free. At the end maybe it won't work and it will be junk. But it will be clean junk.

 

[VirtualLarry]

Scrub every socket (new toothbrush)

Except for LGA sockets...

 

[ehume]

Except for LGA sockets...

You have a point. I never think of them as sockets because they contain male terminals, but those terminals are inside a wall, and the collection is called a socket.

I would definitely inspect it, though, and maybe even go over it with a soft brush. I bought some paint brushes for general cleaning. And I bought some cosmetic brushes for my camera lens (it is v-e-r-y dusty where I live). I think the latter would be appropriate for a CPU socket.

 

[soccerballtux]

I'd try resetting the CMOS and get back to stock BIOS settings. Maybe download memtest86 and see if you can figure out which RAM stick is possibly bad.

yeah I've had this happen many times, just clear the CMOS. and like you said, maybe memtest

 

[ioni]

This was a really weird ride. Recap, BIOS screen had an error on startup, "Failed to OC CPU." I removed the OC. System would reboot after about an hour (happened multiple times). I started testing individual sticks of RAM. With stick A, system would reboot after about 2 minutes, consistently. With stick B, system would reboot after about 15 minutes, consistently. With stick C, system would reboot after 30-40 minutes. Happened twice in a row.

I started to wonder if dust was the issue since my PC was pretty dusty. Bought some air and dusted it out. With all 3 sticks on RAM in, the system stayed up for a few hours before I shut it off before going to bed.

Next day I ran memtest86 on stick A for 10 hours. 2 errors. Ran memtest86 on stick B for 2.5 hours (6 passes) No errors. Ran memtest on stick C for 7 hours. No errors. Ran memtest on stick A again for 9 passes (4 hours). No errors.

So I can only conclude that the problem was dust. Though I never would have thought dust could cause such an issue. It's not like my PC was absurdly dusty either. It had been about a year since I last dusted it out.

But since I bought some RAM assuming my RAM was the issue, I'm wondering if I can use my new RAM with my old RAM. My old RAM is 1.65v and my new RAM is 1.5v. Is it ok to use DIMMs with mixed voltages?

Edit: Just to clean this issue up, I finally found out that the problem was my UPS! Plugged my PC directly into the wall and all is fine now. I would have thought a UPS would have lasted longer than 1.5 years