Sign in with
Sign up | Sign in
Your question

Somewhat baffling reboot behavior...

Last response: in Systems
Share
November 25, 2010 5:28:48 PM

Hi all,

I'll post my rig's specs later if it'll help, but right now I am just looking for theories. For now, the only important points are:

* I built this computer about 3 years ago; it was a pretty good Intel-based gaming PC at the time. It has worked fine up until a month or two ago. (It's running Win 7 64-bit.)
* Since my wife has been using it, there may have been some warning signs before it really started to decline, which she missed.

Anyway, about a month ago the thing started crashing/ auto-rebooting just after at the Windows 7 logo appeared, but just before the desktop would show up.

I figured it was either a Windows 7 problem or a hard drive problem. Since at that point the system was still fully booting *sometimes*, I used System Restore to see if a recent update was at fault, but this didn't help. Using Safe Mode, etc., made no difference either. I therefore chalked up the problem to either a deeply embedded Windows 7 error, or a beginning-to-fail hard drive. I set it aside for a couple weeks, planning to swap the HD and do a clean install.

Now then... yesterday I replaced the hard drive and did a clean Windows 7 install on it. I also took out everything but the basics (mobo, memory, cpu, hd, dvd, keyboard... I'm using onboard video ATM). Windows installed fine, but when I reached the exact point where the old Windows would crash... I got a crash again. Several reinstall retries resulted in the same thing.

I also tried using just one memory stick, then the other, meanwhile moving them to different slots. Memtest looks ok after an hour or so of activity... I can run it some more but frankly, the sheer consistency of the crash suggests it is not a problem with faulty memory.

So, to recap: I'm using a brand-new, working hard drive; I am clean-installing Windows 7 from a good install disc; and I am getting the exact same problem at the exact same time (in the boot sequence) as before. Apparently some bit of hardware is being polled and causing problems, right before Windows 7 pops up the desktop.

I would be less baffled if the thing was crashing at other times too. But it seems to me that all the key hardware components (cpu, mem, hd, dvd, etc.) are *totally* involved in the install process too. It's not as if the install can happen without these things working. And yet, the install never crashes!

Any ideas on what might be causing this? (The easy answer is to try a different mobo, and I do plan on that, but I have no spares ATM...) Since I know I can eventually figure out the problem (mostly) by swapping hardware and using process of elimination, this question is more for my curiosity than anything else.

Thanks in advance.



November 26, 2010 12:08:30 AM

Man, I'm bafffled about this problem as well, do you have anybody that has similiar specs that could lend you some parts to see what the problem is, My guess though is that it most likely either the mobo or cpu fault that has turned faulty


and yes could you please tell us your build :) 


m
0
l
a b B Homebuilt system
November 26, 2010 12:55:52 AM

Based on the timing of the failure, its either (in increasing order of likelihood) graphics card (ie, mobo), graphics driver, CMOS/BIOS setting error, or psu.

That's the theory not knowing anything about your config.
m
0
l
Related resources
Can't find your answer ? Ask !
November 27, 2010 5:08:31 AM

Thanks for the input, guys. I will post specs when I get a bit more time to tinker...

Twoboxer, based on your list, the CMOS/BIOS and psu are the likely culprits (which is what I was beginning to think myself), as:

* I had the problem both with a dedicated video card and when I stripped it out and just used onboard video, so probably not the graphics card
* Probably not the graphics driver, for same reasons as above (and since this was a clean install, I'd be using the Win 7 vanilla driver, which tends suck, but not to crash)
* I've gotta think that if it was the cpu, the error would come up during the repeated installs... but maybe not. I confess to ignorance of what goes on inside that little metal square.

I do actually have a different psu on hand, so when I get some time I will hook that up and see. Not sure which part I want to be the problem -- mobo or psu -- since the mobo is a fairly decent base for a modest rig and the psu is a pretty nice Silverstone.

Will post back when I have more... thanks again for the replies.
m
0
l
a b B Homebuilt system
November 27, 2010 6:20:32 PM

If its a Silverstone Strider + 750W, its a quality unit. If not, we either haven't found a qualified review or its not a unit we would normally recommend (or I missed it lol).

You want it to be a psu problem :)  - that's the usual source of failure around here. Which is why we are very careful in choosing psus - perhaps more careful than choosing, eg, memory.
m
0
l
December 3, 2010 2:35:29 PM

A bit late, but figured I'd post a follow-up in case anyone was interested.

The problem lies with the motherboard's CMOS battery seating/housing. It was hard to troubleshoot because switching out power supplies, unplugging and unplugging devices, changing the battery itself -- all these seemed to temporarily work, but apparently I was just jostling the battery and causing it to temporarily make good contact. So I would think I had found the bad part, only to have things go bad again after another reboot or two.

The specific problem is that the battery sits in the housing crooked, so it sometimes loses contact. (If I touch the battery, I can feel it shift slightly from one side to the other.) Somehow the resultant loss of BIOS information seems to throw Windows 7 for a loop during a specific part of its boot sequence.

I did a little bending of the battery contacts in the housing, and for now it seems to be OK. It's not a mobo I would use for mission-critical activities any more, but I think it's fine in its new role as the base of a cheap, infrequently-used media pc.

Thanks again for the input!
m
0
l
a b B Homebuilt system
December 3, 2010 9:28:48 PM

And thanks for telling us what caused the problem . . . adding " . . . check battery holder . . . " to Clear CMOS instructions lol.
m
0
l
!