• Please review our updated Terms and Rules here

Computer reboot loop

Tiberian Fiend

Veteran Member
Joined
Mar 1, 2009
Messages
563
Location
Central Florida
My main computer (a quad-core AMD Phenom II) has been BSODing and rebooting during games, lately. It started doing it sooner and sooner, and now the computer gets stuck in a reboot loop while loading both WinXP and Win7 (I have a dual boot setup). I blew all the dust out of it a short while ago, so that isn't the problem. I've also swapped out the video card with one I know to work with no changes, and scanned the memory twice with no errors. I'm looking at the voltages out in my BIOS PC health status, and they've fallen to:

1.42v Vcore
2.88v +3.3v
11.35v +12v

Have these fallen too far below nominal values? Should I look into purchasing a new power supply?
 
Is the BSoD you're getting have the error: PFN_LIST_CORRUPT
If so, then you should upgrade the power supply. I've had to do this, since on one of my WinXP computers, connecting a slave hard drive would cause it to show a PFN_LIST_CORRUPT BSoD. After upgrading it to a 600 watt PS, the problem went away.
 
It's a good rule of thumb that PSU voltages should be within +/-3 percent of nominal. Of course, that can vary by application. Yours are considerably outside of that.
 
General troubleshooting but has anything changed or been installed prior to the problems at hand? Does it only do that under certain load? You can hit F8 before the OS loads and tell it disable reboot on crash or some such option (I think this may have been a service pack specific feature though) if you want some time to look at the actual error. Otherwise it could just be drive corruption/registry corruption.
 
But if it is a continuous reboot loop, you're not going to be able to get into Windows to change those settings. Unplug the CD drive(s) and see if that helps free some power consumption and maybe it'll let you fully into Windows. There, you can disable automatic rebooting. But then I'd recommend using a program called Bluesreenview to view past blue screens on your computer.

Or, try booting from a CD such as UBCD4WIN and access the minidump files that way.
 
I don't remember what the specific problem the BSoD stated was, but it would dump the memory to disk before restarting. I think it might've been a memory error. WinXP and Win7 are both on different partitions, and I can't use safe mode to boot into either (not even command prompt). Win7 recovery does work, but can't find any problems.

Looks like I'm off to buy a new power supply.
 
I've also seen ugly loops like this with a bad RAM DIMM or whatever they're calling the form factor this week. Try a self booting RAM test program or swap your RAM.
 
There are three major things to check when you encounter these types of rebooting issues -- two have been mentioned so far.

Power Supply - due to capacitor aging a PSU that was/is fine under light loads slowly degrades under heavy loads. A good rule of thumb is to use a power supply that's rated 30% more than your total system thanks to this, as after two or three years of use that's what it will drop to in capacity. This is even more true with todays high-end gaming cards where the GPU is sucking as much if not more power than the CPU. You can use a tool like:
http://extreme.outervision.com/psucalculatorlite.jsp

To get an idea what you should have for a supply. I HIGHLY suggest plugging in 30% under 'capacitor aging'.

Memory - RAM failures are the leading cause of BSOD's in brand new systems, but it's rarely an issue if the machine is stable under normal loads and only buckles when pushing it -- like gaming. If it's only having issues when gaming, I'd put that at the bottom of the list of things to check.

Airflow - Is your case REALLY providing enough airflow? Depending on the system a lot of machines (Apple's come to mind, don't even TRY to game on a mac pro) wouldn't know proper cooling if it stripped naked, painted itself purple and hopped up on a table to sing "Oh look at what a big cooling fan I am!". Are your top and rear fans pointed to blow out, front/bottom/side to blow in? Remember, in AND out (I'm shocked how often I see machines where EVERY fan is blowing inward!)... when did you last de-fod your fans? You'd be shocked what a can of compressed air can do.

Do you even HAVE proper fans and vents for the level system? I've been seeing a LOT of people stuffing i7's with 200+ watt ATI or nVidia video cards into cases that have no top vents, anemic rear vents, no side intakes, and barely passable front fans. These are components that can raise the ambient temperature of the room they are operating in 8C or more -- Imagine what's going on inside that little case for temps! Putting them in a case with improper airflow means you're killing them prematurely. If you are worried about the noise of so many fans that's what the larger (180mm+) fans at lower RPM's are for -- either that or you have no business gaming.

It's why I've got my i7 870, GTX 560TI and GTX 260 system (GTX is for dedicated cuda/physx and to drive 2 extra displays) in a TT Element G, with the top and front 200mm fans, rear 180mm fan, and side 230mm fans -- with a TT Frio on the CPU (with arctic silver)... Just as with the power supply I've got a 850 watt TT TR-2 for a system that only adds up to 600 watts.

Could be as simple as your CPU cooling. Are you on a stock cooler with the stupid 'sticky pad', or are you on a real heat sink and fan applied with a decent compound like Artic Silver 5 or Tuniq TX-4? Compound ALONE can be 10C at the CPU. When the compound was applied was it slathered on willy-nilly, or was it applied properly as a micro-thin even coat since the only purpose of compound is to fill the microscopic gaps, NOT to replace actual surface to surface contact! -- the little tiny 3.5 gram containers most compounds ship with should be enough for ten to fifteen CPU's!!! You are NOT FROSTING A CAKE!!! :D

... and that's a big point in terms of stability; excessive heat can compromise the CPU, RAM and video card!

What are your system temps? Coretemp, speedfan -- handy utils.
 
...if it's temperature, just go into BIOS setup immediately after reboot, before the OS gets loaded. Most BIOSes have a temp display--no need for utilities to check.
 
Was getting an idle temp of 42°C in BIOS last night. I've got my computer in an XClio Windtunnel case. It's got two side 220mm fans and plenty of venting at the top. I'm using the heatsink and fan that came with the PIB. My computer is about three years old, and haven't had any problems with it until recently. If a new power supply doesn't help, I'll look at the thermal compound on the CPU heatsink next.
 
Fiend:
What comes to my mind #1 is that you may have a corrupted W7 partition. Barring that, try booting in the safe mode and run the W7 memory check completely. Also, I'm wondering if you have been dabblng with overclocking. You may want to reset your BIOS to the minimums. You say you started to experience BSOD's while gaming, that could be overheating or a number of other closely related things. My checklist would be to ensure that the CPU has a good heat sink seal. Swap out the PSU if posssible and pull half of your memory. However, you don't want to do everything at once. I've been there before, you're not alone so don't panic (yet). If you have a video port on your mobo you might want to pull your discrete card and see what happens. Once you get it to boot properly, try running your rig with the case open. Also, once you get up and running, make sure you have the latest BIOS and chipset drivers. Good luck!
 
It happened to both my Win7 partition and my WinXP partition at the same time. I never overclock. I don't think I was getting a protection error.
If you have a spare HD you might want to pull your system HD and install W7 or XP on that HD. If you're still getting BSOD's after the clean install then for sure it would be a hardware problem that needs to be rung out.
 
Not overly technical but if it was really rebooting from temp, taking the side case apart and letting a fan blow on the system while it's on would probably get you further along. If that doesn't and it truly reboots at the same point every time I would still be suspecting software corruption.

You could also let the system sit in the BIOS for the time it normally takes for it to reboot, although keep in mind that won't really put the system through the same load as running all the drives. Also if load related you could disconnect all the other drives except the hard drive (video card if it's a gaming system would be your biggest power taker though).

Can you tell if it's OS loading based crashing vs timing? Have you been able to boot into safe mode?
 
Back
Top