BIOS Troubles

[Supertrek32]

I've been having the strangest issue with my new computer and can't for the life of me find an answer. This problem occurs whenever I boot my computer after having it turned off for 15 minutes or so. Shutting down and starting the system back up right away gives me no issues.

The system will start up, but monitor stays dark and no keyboard lights show up. After 5 seconds or so, it will turn off, back on, and I'll get an overclock failed prompt. I then press Delete, save and exit (it remembers my settings), and the computer works fine.

I have nothing overclocked. The only non-default setting is raising the memory voltage to 1.65, which is spec for the memory (OCZ 1333). Changing Uncore/Memory multipliers (for 1066 vs 1333) has no effect on the issue.

BIOS reports CMOS battery as giving 3.12 volts, so that doesn't appear to be the issue. I've also tried updating BIOS, but that didn't fix it either.

Some posts I found pointed to an inadequate power supply, but at the same time 630W (50A on 12v rails) should be plenty for i7 and a single 4870.

I also can't use sleep mode. Resuming from sleep quickly works fine, but if I wait the 15 minutes or so the system fails to start.

Once I do get the system running, it's 100% stable. memtest86+ found no errors, and Prime95 ran for 12 hours without error.

PSU:
http://www.newegg.com/Product/Product.aspx?Item=N82E16817182185
Mobo:
http://www.newegg.com/Product/Product.aspx?Item=N82E16813186170
Memory:
http://www.newegg.com/Product/Product.aspx?Item=N82E16820227370

i7 920
Radeon HD 4870

Any ideas on what could be causing this or how to fix it?

 

[dndhatcher]

Have you checked for reviews of the PSU? Rosewill has made some really bad ones.

It could also be the CMOS battery (regardless of what it reports when power is applied). The "15 minutes after shutdown" really makes me suspect the CMOS battery.

Could be from superior foxconn engineering.

Also in bios there may be a "wait for HDD" setting you could try to make it wait for the HDD to spin up (what kind of HDD you got?) though I doubt that is it.

Might try resetting bios to factory default then setting the one memory voltage setting you need and see if that resolves it.

 

[pjmelect]

My vote would be for a faulty power supply, substitution is the only test.

 

[Supertrek32]

Have you checked for reviews of the PSU? Rosewill has made some really bad ones.

It could also be the CMOS battery (regardless of what it reports when power is applied). The "15 minutes after shutdown" really makes me suspect the CMOS battery.

Could be from superior foxconn engineering.

Also in bios there may be a "wait for HDD" setting you could try to make it wait for the HDD to spin up (what kind of HDD you got?) though I doubt that is it.

Might try resetting bios to factory default then setting the one memory voltage setting you need and see if that resolves it.

Scanning the PSU reviews, the bad reviews were mostly DoAs (which mine obviously isn't).

CMOS battery wasn't it, I tried replacing it with the one from my old computer and it still happened. (Although at this point I had been thinking the old one could have gone bad in the 3 days it hasn't been in use and I was the victim of a terrible coincidence... Stranger things have happened...)

Foxconn... same thing as the PSU, bad reviews were DoAs. I'm a poor student and need to save as much money as possible.

I didn't think it would be the HDD spin up delay since the first (non-working boot) didn't even post and it ended up not working right anyway).

Like I said, I put everything to default except for the memory voltage when I found the problem.

-----------

Now that the trouble-shooting failures are out of the way... I suppose it would be a good time to say the trouble-shooting success?

It would appear the board wasn't giving itself enough juice. After a day of testing (and playing AC2 on 360 for 30 minutes between tests) the problem seems to be fixed. Bumping the "X58 IOH Core Voltage" from 1.1v to 1.12v seems to have fixed it. I won't know for sure until tomorrow (I want to leave it off all night to be sure), but I've tested it twice now without an error.

Just incase it needs more voltage (I'm at least on the right track now), what's the highest-but-still-safe voltage I should give the IOH?

 

[dndhatcher]

I would go looking at x58 board overclocking reviews and i7 920 overclocking reviews to find some with CPUZ screenshots that show the voltages they used.