View Full Version : This PC keeps just powering off. No warning. Done it since new 6yrs ago.

04-12-2012, 06:47 PM
This PC was built for me by a TradeMe trader some 6 yrs ago. Ever since I loaded the OS it's had this problem I can't seem to solve.
It just turns off. No warning at all. Does it randomly too. Last night I set it doing an Avast 'boot-time' scan - ran all night & was going fine when I got up.
Then i started to use it & check stuff online - it just turned off. I waited then rebooted it & all went OK.
There seems no rhyme or reason why it powers off. It gives no warning - just like there's a power cut (but there isn't).
1. Could it be software driven?
2. Could it be a CMOS setting?
3. Could it be a faulty Power Supply?
4. Could it be a trojan or virus? (I've scanned everything to 'death')
5. Could it be triggered by inactivity or something else?
Some days, it goes all day without powering off, then other days it does it 3 or 4 times.

I've reloaded the OS (repaired it & reinstalled it many times over the last 6 years) - no change.
I've even loaded the OS on another partition & left the old OS derelict (unused). No change in the behaviour.
It now boots to F drive & C drive is unused - no change - it still powers off randomly.

Has anyone come across this problem before?

04-12-2012, 07:47 PM
Could be overheating - not proper installation of fan/thermal compound. Check the temperature by downloading Speccy - http://www.google.co.nz/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&cad=rja&ved=0CDMQjBAwAQ&url=http%3A%2F%2Fwww.piriform.com%2Fspeccy%2Fdownl oad&ei=X5q9UJWdGKWwiQeJy4CADA&usg=AFQjCNFwIvGjUbiZD0e9df-ytD8lQC3JGw&sig2=iKWvlItyalhPq1dTouk0ig

04-12-2012, 08:32 PM
heat (as above), psu or motherboard imo.

04-12-2012, 08:47 PM
To power off suddenly without shutting down generally is not a software issue which is more likely to cause a sudden reboot, lockup, crash, or shutdown properly. It's not impossible but it's unlikely.
More likely is a hardware issue like overheating as Sahilcc7 suggested or a faulty PSU as you mentioned. I've also seen issues like that when there is a compatibilty problem between two pieces of otherwise faultless hardware - in my case a motherboard and graphics card that would not work properly together but would both work fine with other hardware.

Start with the temperatures, if that goes nowhere try a Memtest and see if you can borrow a PSU from somewhere to try. The biggest problem with diagnosing a single machine is no parts to swap around.
It may sound like a cop out to some but simply doing a bit of trial and error swapping parts around can prove a fault much more quickly and conclusively than days worth of diagnostics and testing and because it takes less time is also the cheapest way to do it.

Other things you can try are:
updating to the latest BIOS and drivers for everything,
removing any unecessary hardware,
checking all the connections are secure inside the case and there are no stray bits of metal anywhere or shorts between the motherboard and case.
If you have more than one stick of RAM try removing all but one at a time.

04-12-2012, 10:33 PM
PSU might be overloaded too, coming from TradeMe who knows what they used.

04-12-2012, 10:34 PM
Sounds like it was never stress tested by the place who built it, if it is hardware, it should have been picked up. Im wondering why you have put up with it for 6 years before trying to fix it. From the symptoms described, it could be a number of things, but yeah, as suggested, start with checking memory and PSU. If not those, I would suspect a faulty motherboard.
Good luck, hope you get it sorted, it must do your head in dealing with it all this time.

04-12-2012, 10:59 PM
I'll second a power supply / issue, or overheating CPU / motherboard.
Quickest way to rule out overheating:
Download Speccy (It's free) and run it. Then, click on File --> Post Snapshot.
Post your results to us here and we'll let you know :)

If it's not that, it's highly likely the power supply.

05-12-2012, 05:49 AM
If I might continue on with Chilling_Silence direction, which sounds worth trying while you have no idea what it could be, what is the power rating of the power supply, and some of your hardware e.g. number hard drive & CDROM drive etc, I doubt if you would have a 200 Watt power module and also a 100 Watt sound card in there, then you might have trouble. (sorry my sick humor again)

My Windows 98se machine randomly used to cut out just like you said, it was a heat problem, built up of dust in the heat fins, so it needed a blow out. however the difference was once it got over the warming up period half a hour it was more likely to keep going, up until that half hour it used to drive me nuts. But it still gave trouble after that warm up.

So one could ask "what do you think of the CPU fan?" is it a cheap looking thing? and have you tried taking the cover off your computer and blowing a external fan into the computer to see how that goes.

05-12-2012, 07:35 AM
Overheating is pretty specific.
Generally it would boot OK, then once heat is high, then shut off. Restarting at that point wouldn't help, it would shut off again straight away - until it's cooled down.

So unless it has that kind of pattern (and I doubt it would last 6 years either) it isn't heat.
What PSU has it got in it?

05-12-2012, 08:36 AM
sorry pctech I am not trying to pull your nose? that is how it happened for me, when I applied thermal paste inbetween CPU & heatsink no more problem. At one stage I had the power supply apart _ a waste of time that was

05-12-2012, 09:41 AM
At one stage I had the power supply apart

Why on earth would you do that!? :eek:

05-12-2012, 10:21 AM

05-12-2012, 10:53 AM
Why on earth would you do that!? :eek:

Why wouldn't you do that?

05-12-2012, 11:04 AM
dont teach anything at school these daysYep, we live in a throw-away society, if it dont work, dont bother pulling it to bits to see why, you just throw it away and buy another.

05-12-2012, 01:00 PM
Why wouldn't you do that?

The risk of being zapped by 230 volts of residual electricity, even when unplugged...

05-12-2012, 01:08 PM
Why wouldn't you do that?

Because 99.99% of people dont have the skills to warrant opening up PSU's , just for a poke around/stab in the dark.
If you dont have spare PSU to try instead, It would be a good guess that you dont have the skills to try & fix a PSU

05-12-2012, 01:27 PM
My original comment was a bit of a dig aimed mostly at pcuser42's reaction after reading Frank's comment, but reading his other posts especially this one http://pressf1.pcworld.co.nz/showthread.php?127628-Support-the-moderators-Dont-hand-me-(or-others)-abuse-it-does-not-wash I suspect he knows what he's doing.

For anyone else... the charge in the primary filter capacitor(s) will likely have gone by the time you open the PSU anyway due to the bleeder resistors across them. If the resistors have gone open circuit or are not present then you may have a problem, but only if you're silly enough to go poking your fingers on the primary side components without checking the voltage first.

If there is still a charge present, simply drain it with a 10kOhm or so 5-10 watt resistor and verify with your voltmeter. Don't just use a screwdriver etc to short the terminals or you can damage the capacitor, or weld your screwdriver to the leads.

Following such simple precautions and it's not dangerous.

Yes most people wouldn't have a clue how to repair one but even so, if they do feel the need to have a look (and some people will!) they should know how to do it safely, in my opinion.

05-12-2012, 04:29 PM
I'd check the CPU cooling fan to make sure it is clear of dust.

I had this happen to me, and when I cleaned it, no more problems.

05-12-2012, 05:42 PM
One common symptom of overheating is when it shuts down, and it wont start right away. There are built in safety devices that wont allow the PC to restart again until the overheated CPU has cooled enough.

If you have to leave it for a few minutes, then its overheating ( usually)

If it starts right away, could be a faulty PSU, or some other fault.

08-12-2012, 07:15 PM
Many thanks to all for their input on this issue.

Yes, it's been a 'pain' of an problem to deal with - especially when it shuts down when I'm in a hurry to get something done before I rush out the door to work!?!
Of course some days it runs all day without any shutdown issue....

I have Speccy loaded so will run it later & post the results here.

When it shuts down, I can restart it almost immediately & it often will run for 10 - 12 hrs no problem.

Again, many thanks to all.

08-12-2012, 07:30 PM
Operating System
MS Windows XP Professional 32-bit SP3CPU
Intel Celeron 331Prescott 90nm TechnologyRAM
3.50 GB Dual-Channel DDR2 @ 199MHz (5-4-4-12)Motherboard
Gigabyte Technology Co., Ltd. 8I945GMF (Socket 775): 37 °CGraphics
VX2433wm (1920x1080@60Hz)Intel(R) 82945G Express Chipset FamilyHard Drives
488GB Western Digital WDC WD5000AACS-00ZUB0 (SATA): 33 °C391GB Western Digital WDC WD4000AAKS-00YGA0 (SATA): 38 °COptical Drives
TSSTcorp CD/DVDW SH-W162CAudio
Realtek High Definition AudioOperating System
MS Windows XP Professional 32-bit SP3Windows Update
AutoUpdate: Notify prior to downloadFirewall
Firewall: EnabledCompany Name: COMODODisplay Name: COMODO FirewallProduct Version: 3.9Antivirus
Antivirus: EnabledCompany Name: AVAST SoftwareDisplay Name: avast! AntivirusProduct Version: 5.0.117441986

Intel Celeron 331
Cores: 1Threads: 1Name: Intel Celeron 331Code Name: PrescottPackage: Socket 775 LGATechnology: 90nmSpecification: Intel(R) Celeron(R) CPU 2.66GHzFamily: FExtended Family: FModel: 4Extended Model: 4Stepping: 9Revision: G1Instructions: MMX, SSE, SSE2, SSE3, Intel 64Virtualization: UnsupportedHyperthreading: Not supportedFan Speed: 2410 RPMBus Speed: 133.0 MHzRated Bus Speed: 532.0 MHzStock Core Speed: 2666 MHzStock Bus Speed: 133 MHzCaches
L1 Data Cache Size: 16 KBytesL1 trace cache: 12 KuopsL2 Unified Cache Size: 256 KBytesCore 0
Core Speed: 2660.0 MHzMultiplier: x 20.0Bus Speed: 133.0 MHzRated Bus Speed: 532.0 MHzThread 1
Memory slots
Total memory slots: 4Used memory slots: 3Free memory slots: 1Memory
Type: DDR2Size: 3584 MBytesChannels #: DualDRAM Frequency: 199.5 MHzCAS# Latency (CL): 5 clocksRAS# to CAS# Delay (tRCD): 4 clocksRAS# Precharge (tRP): 4 clocksCycle Time (tRAS): 12 clocksBank Cycle Time (tRC): 1 clocksPhysical Memory
Memory Usage: 21 %Total Physical: 3.24 GBAvailable Physical: 2.55 GBTotal Virtual: 2.00 GBAvailable Virtual: 1.90 GB

Manufacturer: Gigabyte Technology Co., Ltd.Model: 8I945GMFChipset Vendor: IntelChipset Model: i945P/PL/G/GZChipset Revision: A2Southbridge Vendor: IntelSouthbridge Model: 82801GB (ICH7/R)Southbridge Revision: A1System Temperature: 37 °C

Hard Drives
Manufacturer: Western DigitalForm Factor: GB/3.5-inchBusiness Unit/Brand: Desktop/WD Caviar®RPM/Buffer Size or Attribute: IntelliPower® with 16 MB cacheInterface/Connector: SATA 3 Gb/s with 22-pin SATA connector/SATA 1.5 Gb/s with 22-pin SATA connector (Mobile)Heads: 16Cylinders: 16383SATA type: SATA-II 3.0Gb/sDevice type: FixedATA Standard: ATA8-ACS48-bit LBA: SupportedSerial Number: WD-WCASU4421864Interface: SATACapacity: 488GBReal size: 500,106,780,160 bytesRAID Type: None

Manufacturer: Western DigitalForm Factor: GB/3.5-inchBusiness Unit/Brand: Desktop/WD Caviar®RPM/Buffer Size or Attribute: 7200 RPM with 16 MB cacheInterface/Connector: SATA 3 Gb/s with 22-pin SATA connector/SATA 1.5 Gb/s with 22-pin SATA connector (Mobile)Heads: 16Cylinders: 16383SATA type: SATA-II 3.0Gb/sDevice type: FixedATA Standard: ATA8-ACS48-bit LBA: SupportedSerial Number: WD-WCAS84495662Interface: SATACapacity: 391GBReal size: 400,087,375,360 bytesRAID Type: None

Optical Drives
Media Type: CD-ROMName: TSSTcorp CD/DVDW SH-W162CAvailability: Running/Full PowerCapabilities: Random Access, Supports Removable MediaConfig Manager Error Code: Device is working properly

08-12-2012, 09:16 PM
You tried a BIOS update? Might be some weird bug.

Would explain why the problem has always been there.

Billy T
09-12-2012, 03:30 PM
Yep, we live in a throw-away society, if it dont work, dont bother pulling it to bits to see why, you just throw it away and buy another.

I'm with Frank here, I'll check any failure to see if I can fix it, and more often than not, I can.

There is a certain perverse pleasure in beating the machine, and 90% of faults have a reasonably simple cause and are fairly straighforward to diagnose, it is the other 10% that give you shaky hands and a booze problem! I go as far as to cut open sealed units, and the most common cause of electronic failure is bad soldering or poor crimping. Tools, lawnmowers, appliances, clocks.........you name it, I'll have a go.


Billy 8-{)

14-12-2012, 06:47 AM
This is the first time I come back since I made my last posting to the thread

The risk of being zapped by 230 volts of residual electricity, even when unplugged...

Yes there is always a risk of getting a shock from charged electrolytic capacitor the electrical regulations used to say something like "all high voltage capacitors should have a suitable bleeder resistors across them so when the power is removed then there is prompt discharge within a reasonable time"

Normally I have found that if there is a fault in the power supply and the appliance does not work at all, then BEWARE the electrolytic capacitors on the primary part of the supply could be left charged up, otherwise if the fault is on the secondary part of the chopper transformer or after then the mains filter capacitor gets discharged just after the power is removed.

By the way ALL ELECTRICAL SHOCKS MUST BE REPORTED TO THE ''ELECTRICAL WORKERS RESIGNATION BOARD'' (found in the telephone book under that name) but I think this would many apply to unopened appliance.