Almost constant unwanted reboots with oldish PC

[SkyScape]

Hi all,

My PC keeps randomly rebooting. For the past week or so, although it occasionally goes a few hours, and even survived a whole day without a reboot. It also reboots before it can get into windows, literally a second or two (sometimes) after the previous reboot, and also while in safe mode w/command prompt + while running the windows system repair disk, so it does not seem to be a problem with windows itself. To confirm this I've also gone back to previous images of my C:/ drive made when all was fine. I also tried fixing the MBR using the windows system repair disk.

I have not changed my hardware config recently, and cleaning all the gunk that had built up in the CPU fan/heatsink + PSU has not helped, so I'm now thinking that one of my components must have developed a fault.

The main reason I'm thinking this is that it was quite a cheap pre-built PC which I bought after my old self-built PC died a couple of years back (lasted more than 5 years) and I wanted to get up and running fast again, so I don't think it has the best quality components!

I suspect the power supply may be the problem, but I'm wondering if anyone here can suggest other things to try before I start removing or swapping in new components? In the mean time I've ordered parts to build a new PC myself, but I want to try and get this one running smoothly again if possible.

BTW, I don't know if it'll help, but I got this problem report when running windows startup repair (one of the boot options):

problem event name: startuprepairoffline
problem signature 1: 6.1.7600.16385
problem signature 2: 6.1.7600.16385
problem signature 3: unknown
problem signature 4: 1003
problem signature 5: AutoFailover
problem signature 6: 1
problem signature 7: norootcause
os version: 6.1.7601.2.1.0.256.1
locale id: 1033

[DBarber]

could be so many things, post a list of the spec of the machine will help to identify known faults.
Has the machine been stood a while?
Start off with disconnecting any unnecessary hardware like usb devices dvd etc, do you have any bootable diagnostic software like memtest? would be a start and best to start diags outside of windows anyway.

[SkyScape]

Thanks for the reply.

Spec:

• ARIAnet Gladiator Black Micro-ATX Case w/ 500W Power Supply
• AMD (Trinity) A6-5400K Black Edition Dual-Core 3.60GHz Accelerated Processor
• 8GB 1600MHz DDR3 (2x4096) Memory (64GB Max)
• DVD +/- RW Drive w/ m-Disk Support
• 500GB (7200rpm) SATA 6GB/s (SATA III) Hard Drive
• Graphics: AMD Radeon HD 7540D 1024MB GDDR3 Graphics - IGP Supports DirectX 11
• Gigabyte F2A85XM-D3H (A85 chipset)A75 Socket

The problem actually started while I was doing quite a lot of work on the PC (processing hundreds of camera RAW files using Capture 1 Pro and encoding timelapse footage) over a few days, and leaving the PC working overnight after which it would hybernate. The PC was working quite a bit more than it usually would, and being left on/hibernating with out reboot for a lot longer (days) than normal. Normally I reboot every day, shut down for overnight, and only use the PC for browsing with only a bit of image processing, but every once in a while I have lots of heavy processing to do like I did recently when the problem started. That made me think that it could be the PC hardware stressing, possibly from overheating, but now I think it was just a coincidence.

I did run memtest86, which it passed without error, but I think I may have found the problem now.... at least, since I ran a virus scan last night, there have been no unwanted reboots. Kaspersky told me it had found trojans hidden in compressed files, but did not mention that I was infected or that anything had been removed apart from the trojans within the files. So I'm assuming that was the cause, for the moment, and that the issue has been resolved.

Now parts for the replacement PC have started to arrive (although probably won't get the CPU for a while since I took a risk that it would arrive in stock sooner rather than later, which I may end up regretting!), and I didn't hold back on the spec (much), so should have a superfast backup PC that should last me years. A bit annoyed that I ended up buying a new PC for no reason, but at least I'll have a backup, and I'm looking forward to putting the new box together

[shaithis]

I would suspect the el-cheapo PSU that is almost certainly in that case, especially if CPU temps aren't rising very high.

[SkyScape]

Not sure I trust the PSU either. I might just replace it anyway, even though the spare PSU I have is Corsair CX series... at least it's branded! Anyway, still no re-occurrence of the problem, so I guess the hardware is not at fault...

[DBarber]

if still using the machine for anything secure it is maybe worth following up, could be multipart malware which has stalled because the executable trojans were not uncompressed, who know what other damage it may have done to the OS.
Run a scan with malwarebytes but definitely retire this machine for secure purposes ASAP or at least format and re-install the OS

[SkyScape]

Thanks for the advice. I'd thank you officially, but can't see how to do that!?

Not sure where I stand or what to do now.... it was fine up until last night, when I started getting reboots again. Thankfully they are a lot less frequent than they were before, and I can get things done usually.

I ran a full Kaspersky scan while in safe mode which found nothing new, and I then installed malwarebytes which found the following two files (not in safe mode though), although I'm not sure if they are really threats? :

PUP.Optional.YourFileDownloader, C:\Users\*edited*\AppData\Local\Temp\S7P7sga7.exe.part, , [7bfe5ae9fc8e74c22e6b3ae49e62e11f],
PUP.Optional.Conduit.A, C:\Users\*edited*\AppData\Local\Temp\ct3288691\ism.exe

Going back to Kaspersky, the first scan I ran found these two pieces of malware which are currently quarantined:

heur:trojan.win32.generic - which appears to be an unidentified trojan
trojan.win32.genome.acml

When I search for info on them I find conflicting info. As I mentioned before, I'm still not convinced that either of these have actually been run and that anything is infected. They could just sit there dormant, as long as they are not "clicked" couldn't they? The files that contained them were probably downloaded many years ago - both "infected" files were in folders on my backup drive where I keep old software versions that I think might be useful in future. Ironically, one (KMB_V0.53G-32B-MT.zip) appears to be a CPU diagnostic tool that I downloaded back when I was doing an old build/overclocking.

I'm not sure that reinstalling windows would do anything that going back to a very early image of my C:/ (soon after windows install) wouldn't do? Worst case, the mobo bios might be infected (don't know what this means in terms of the dual bios on my mobo, or even if that makes a difference?) and need to be binned, but I don't know how I can determine weather or not I've been infected?

My Mrs suggested opening up task manager and going through/googling the running processes one by one, which I'm going to try now....

[me-yeah]

if it was your PSU, your PC would just shutdown, ram issues can cause restarts along with PCI cards not in the slot fully, if they actually have a problem you would just get a blue screen, or if they are fully in the slot, you can give the copper a clean with a cotton bud and some witch hazel

[DBarber]

it's possible the original malware infection has been removed already but sometimes the damage it has done has not been repaired and this can cause issues.
make sure your windows setting to "restart on error" is disabled in "start up and recovery" section of windows system properties.
If the end result is BSODs you may get more info as to cause, repeated mention of a particular driver etc may help pinpoint the fault or if entirely random more likely to be a hardware issue.
You haven't mentioned cpu temps, given your description of the fault appearing after a particularly heavy workload an issue i use to come across with older amd cpus is separation of the metal casing from the ceramic round the actual cpu causing problems with increased temps and sudden surges under stress, use something like cpu-z to monitor.
your pc is kept live by two memory addresses, one of which can cause a reboot so it doesn't take much voltage fluctuation to kick that in and doesn't dismiss the cheap psu as a suspect, only using a known good psu and seeing the fault remain will do this.
ps the thanks button is bottom left of post

[SkyScape]

if it was your PSU, your PC would just shutdown, ram issues can cause restarts along with PCI cards not in the slot fully, if they actually have a problem you would just get a blue screen, or if they are fully in the slot, you can give the copper a clean with a cotton bud and some witch hazel

I don't think we can dismiss the PSU as a possible cause just yet as DBarber says. The RAM has already been checked, and no errors found, so I think we can take that off the list of possibilities. I haven't unseated any PCI cards since I got the PC, so I'm not sure how there could possibly be an issue there. Thanks anyway for taking the time to post.... it may be that I have to come back and re-examine the issues you raised if all other possibilities are exhausted!

[SkyScape]

it's possible the original malware infection has been removed already but sometimes the damage it has done has not been repaired and this can cause issues.

Do you mean damage to software/OS? Surely if that is the case, then going back to a previous C:/ drive should fix it.... I'll try restoring an old image since I have not done so since I started scanning with Kaspersky.

make sure your windows setting to "restart on error" is disabled in "start up and recovery" section of windows system properties.
If the end result is BSODs you may get more info as to cause, repeated mention of a particular driver etc may help pinpoint the fault or if entirely random more likely to be a hardware issue.

I'm running Win7, and the closest thing to the setting you suggest that I can find is "on system failure... automatically restart", which I've un-ticked now. I should also mention that in perhaps 1/10 crashes the PC screen just goes black and the PC hangs there without re-booting, making me have to turn it off at the PSU switch. For some reason pressing/holding down the power on switch does nothing, even though I've set it to "soft" power down in BIOS. No change since I last replied - getting 1 or two reboots a day vs the 20-50 which I was getting before.

You haven't mentioned cpu temps, given your description of the fault appearing after a particularly heavy workload an issue i use to come across with older amd cpus is separation of the metal casing from the ceramic round the actual cpu causing problems with increased temps and sudden surges under stress, use something like cpu-z to monitor.

Not sure what I can use as cpu-z doesn't work with AMD CPUs. Anyone? Think I'll stick to Intel in future!

your pc is kept live by two memory addresses, one of which can cause a reboot so it doesn't take much voltage fluctuation to kick that in and doesn't dismiss the cheap psu as a suspect, only using a known good psu and seeing the fault remain will do this.

Agreed.... still have a few things to try and if that does not do the trick I'll try replacing the PSU.

ps the thanks button is bottom left of post

Found it now. Didn't realize that it would only appear after I'd made a certain number of posts!

Thanks for your assistance once again.

[peterb]

I have know a faulty SATA plug cause problems. The housing can develop a hairline crack causing intermittent connections, which can cause reboots. As the computer warms up, expansion makes the connection slightly better.

It is unlikely, but it might be worth checking the cables. It doesn't take long and worth it if only for elimination purposes.

As this does seem heat related, it is worth removing and re-seating all plugs and plug in components, if you haven't already done so.

[SkyScape]

Thanks. I'll give it a try.

[DBarber]

an image maybe helpful once the fault is confirmed but for complex reasons may not be definitive, maybe better trying a bootable dvd/usb alternate OS for testing.
yup auto restart is the one.
sorry thinking wrong on the cpu-z(not for the incorrect amd cpu reason), got a cold and the sneezing is driving me loony, was more interested in temp monitoring such as speedfan or maybe your mainboard has its own software, check manuf website.

[SkyScape]

I think the problem may be resolved!

The screen went black, PC hung (still running but unresponsive), so I powered off and on at the PSU switch. Then tried to power on the PC and the lights/fans came on for a couple of seconds, then everything died. Further attempts to boot had no effect whatsoever, so I replaced the PSU, and it's now working. No reboots so far, but it's only been running 30 mins.

The old PSU (win power sl-8500btx - ugh!) weighs less than 1/2 what the replacement CX-500 weighs.... what a piece of crap!

[iranu]

I think the problem may be resolved!

The screen went black, PC hung (still running but unresponsive), so I powered off and on at the PSU switch. Then tried to power on the PC and the lights/fans came on for a couple of seconds, then everything died. Further attempts to boot had no effect whatsoever, so I replaced the PSU, and it's now working. No reboots so far, but it's only been running 30 mins.

The old PSU (win power sl-8500btx - ugh!) weighs less than 1/2 what the replacement CX-500 weighs.... what a piece of crap!

Glad to see you got it sorted. You got away with it this time! It's quite remarkable the difference in weight between something decent and something crap. When I started overclocking I had a generic featherweight PSU, who's fans started to die. Replacing with a 480w Tagan showed me the error of my ways.

The PSU is the most important component of a PC. As soon as I saw your "case with generic 500w PSU", alarm bells rang. To others lurking/reading, don't ever skimp on the PSU. A branded and certified unit that has had good reviews is well worth the money, peace of mind and lack of hassle/troubleshooting. You don't have to spend megabucks or go with ludicrously high wattage PSUs either.

Jonnyguru is a good site that tests PSUs thoroughly.