Computer Powers Down and Restarts on Cold Boots

DucktapeEngineer

Distinguished
Aug 20, 2011
3
0
18,510
Hello,

I've just finished my new build and noticed an odd bit of behavior when the computer starts up. It seems when I either shut down windows entirely, or restart from the BIOS, the computer will shut down as it should, but upon powering on (either by the button in case of shutdown, or just the restart from BIOS) it will turn on, the fans spin up, and then you can hear the power supply shut itself off just as the DVD drive spins up and everything goes dark in the case. This all happens in about 3 seconds...and after another 3 seconds, the computer powers back up, fans spin up, and everything works as normal. I haven't tried any overclocking or anything yet, just making sure the system checks out and everything, but this little nuance makes me wonder if I've done something wrong. Anyone got any ideas? Specs are as follows:

CPU: Core i7-2600k
Mobo: ASUS P8P67 WS Revolution
GPU: Gainward GTX 580 Phantom 3GB
RAM: 16GB Corsair Vengeance (4x4GB) 9-9-9-24 1600MHz upped the RAM to the 1600MHz in the BIOS from the mobo default 1366)
PSU: Corsair AX1200
OS: Windows 7 Ultimate
HDD: Hitachi Deskstar 7K3000 2TB

I've got an NZXT Sentry Mesh running 2 Antec BIGBOY 200's on a channel, two Coolermaster 120mm's on the Hyper 212+, a 140mm, and then the default 200mm and 230mm on the HAF X case, just to make their use of power known.

Not sure if it helps, but according to the ASUS AI Suite II my voltages don't look too far off, but I've never dove into them before this build so I'm not 100% sure how far off they should be.

Vcore: 1.048V
+12V: 12.288V
+5V: 5.080
+3.3V 3.360

I know the Vcore is down, but either the mobo or the AI Suite seems to be throttling my CPU until it's needed, temps are at 20 Celsius on idle right now, and the only game I've installed so far has only brought that up to around 35 C.

OK...I know it's a wall of text, but I tried to include all the info I can. If there's any other info you need, just ask. I just want to make sure nothing is faulty and the strange restarting behavior isn't some sort of precursor of doom.
 
Solution
That sounds like behavior common to Asus P67 boards. Make sure *everything* is set to stock . . . reset cmos, reload optimized defaults, leave the memory alone for now, disable or better yet remove the AI suite.

That *should* stop the recycling. If it does, start setting things up the way you want them one step at a time so if the problem recurs you know what caused it.

If it does not, run Windows Memory diagnostic, and or Prime95 monitoring temps with CPUID's Hardware Monitor. If Windows detects a memory error, or Prime95 gets a rounding error, you have some form of memory issue.
That sounds like behavior common to Asus P67 boards. Make sure *everything* is set to stock . . . reset cmos, reload optimized defaults, leave the memory alone for now, disable or better yet remove the AI suite.

That *should* stop the recycling. If it does, start setting things up the way you want them one step at a time so if the problem recurs you know what caused it.

If it does not, run Windows Memory diagnostic, and or Prime95 monitoring temps with CPUID's Hardware Monitor. If Windows detects a memory error, or Prime95 gets a rounding error, you have some form of memory issue.
 
Solution

DucktapeEngineer

Distinguished
Aug 20, 2011
3
0
18,510
Thanks for the help, I *THINK* that info has lead me to a solution. Here's what I did:

I ran the Windows Memory Diagnostic, get no errors.

Then I ran Prime95 for 8 tests (not very long, I know) and got no errors.

Then, I removed the ASUS AI Suite II, then the demon showed itself. Upon restarting to finalize the changes, I get TWO restarts before the PC will start to POST, then I get an error when it shows American Megatrends... "Overclock failed!" Sounded a bit odd since I hadn't been trying to overclock, so I enter setup to load the optimized defaults as also suggested. Apparently the AI Suite had pumped the CPU multi up to 44 so I was running 1 GHz over the rated 3.4GHz when in turbo mode, and the RAM had been changed to a few MHz above 1600, something like 1630 or 1640MHz. Anyway, I reset all of the options back to defaults and restart, no strange behavior. I let it boot into the Windows 7 login screen 3 times with no problems. I then changed the memory to 1600MHz and rebooted another 3 times with no issues.

I only have a couple of questions now. Firstly, my multiplier stays at 38 now (the turbo rating for the 2600k) leaving me at a constant 3.8GHz. Will this throttle down if the temps get too high or the Vcore climbs too high? And also, I'm paranoid over temps as it gets rather hot here in the summer, and that's just around the corner down under. Will HWMonitor work well enough to run on the 2nd monitor when doing anything stressful just to make sure I'm not heading toward a melt-down scenario, or should I look into something else?
 
Yes, you can rely on Hardware Monitor. You can also have Coretemp or Realtemp load at boot so current temps are displayed on your task bar. Frankly, just run Prime95 for a half-hour, and then toss in Furmark for 15 min, and you'll most likely see your temps are fine (Vidcard prolly hovering at 85C, and core temps under 70) under max exposure. Then you can relax lol.

If your multipler is staying at 38, then you are either running cool enough for it to sit there, or your BIOS is saying no turbo, run at 38. Watching it (CPU-Z) while running Prime95 as suggested above *should* show it dropping down under load. Although I seem to recall my Asus Deluxe/2600k (see my config) running at 3.8 all the while lol.

Yes, it should throttle down if things get out of control, but that's unlikely to happen.

I personally have had *many* "overclock failed" messages. Went researching and found many others who did too. The "fixes" are all different, and I'm not certain you will ever eliminate them until the BIOS gets fixed. Fortunately, the board seeems robust enough to recover.

Here's my *THEORY*: All the auto settings, mem checking and auto-setting, auto-OCing in the BIOS combined with the AI suite . . . well, there's a couple of bugs somehwere in that morass. BIOS decides it needs to make a change to optimize/correct for something, and gets screwed up while on its mission. It then gives the message, and resets stuff, and you boot OK.

When you've got $1k to $2k in your self-build and it fails to boot smoothly . . . its a bit disconcerting.
 

WillOnline

Honorable
Aug 7, 2012
2
0
10,510
Same happening with me same processor same chipset, everything works but it just does that for some wired reason. PSU just like cuts out and the restarts and boots great with no errors