Computer shuts down as if power loss

[SpiritFlyer 987]

My machine has taken on a rather strange fault. I will be online flying with my friends and BAM, my computer shuts itself off. It does not CTD or begins shutting itself down, but one second full power, next dark and silent. It happens at anywhere on the GHZ spectrum, from 4.4GHZ where I operate daily to 5Ghz for FSX. Very Strange. Last evening I unplugged everything and then replugged it all back again - No difference. Any Ideas? Kind regards,

[thefsxflyer 1]

How old is your power supply? I had a problem similar to this with my ps3 a while back. It would just shut off randomly. It got to a point were it would shut off within 5 minutes. I wonder if your power supply is dieing.EDIT: Btw my ps3 was fixed by me replacing the power supply

[Chock 17,769]

Could be a few things, but when I've had that happen with computers, the most likely candidates turned out to be:The Power Supply Unit is on its way out and will only work for a while until there is some sudden demand on the rails, whereupon it gives up. Very occasionally a PSU will give off a slight burning smell if it is suffering from that issue, but not always. Sometimes, placing increased electrical demands on your PSU, by overclocking, adding more RAM or peripherals will be the catalyst for having a PSU start to fail. Or, it could be...Something is not seated properly on the motherboard or a device power socket and when your computer heats up, a metal contact expands and shifts position, that shift causing an electrical contact to break and cut the power. Of course when your computer shuts down, the part then cools down, shifts back into contact and it works again for a while. Everything from RAM to power leads should be checked to ensure the contacts are clean, free of dust and fully seated in their slots, as well as being reasonably clear of other ribbon leads and power leads which might lean against them and cause this issue. A common problem of that kind is when (red) raid leads get moved by something warming up and putting pressure on the raid ribbon lead, making it move; as you probably know, those raid lead plugs are not noted for having a very positive click when you plug the lead in, so they can be easily dislodged. It's sometimes worth taping leads up neatly, or using cable ties to stop that kind of thing happening.Dust on the heat sink vanes can cause a GPU or CPU to hit a temperature where the BIOS will switch things off as a protective measure against melting components. So a good clean out can often be all that is needed to rectify things.Al

[dazz 69]

What are your CPU & GPU temps like?

[FS++ 712]

Two devices that have safety measures buit into them capable of causing these sudden power cuts.PSU when demand is too high (as Chock said)CPU going above a certain temperature (as Dazz is suggesting)I know a Corsair 850 W PSU should be OK, but you never know when you operate the rig at 5GHZ for many hours.Gerald. Rhttp://www.multicrewxp.com

[rhodges 6]

Probably not your problem but It wouldn't reaquire much to check your BIOS settings and see if they have changed.I have my system overclocked but sometime ago for unknown reasons something changed all of my settings (i.e. memory, voltages)back to defaults with the exception of my overclocked speed. Temperature went out of sight and I got the same scenario you describe. It was as if someone pulled the plug out of the wall. Reset my settings back to what they should have been for my overclock and everything has been back to normal for weeks.Good luck:RTH

[Guest firehawk44]

My machine has taken on a rather strange fault. I will be online flying with my friends and BAM, my computer shuts itself off. It does not CTD or begins shutting itself down, but one second full power, next dark and silent. It happens at anywhere on the GHZ spectrum, from 4.4GHZ where I operate daily to 5Ghz for FSX. Very Strange. Last evening I unplugged everything and then replugged it all back again - No difference. Any Ideas? Kind regards,

I had the same problem Stephen. I had my system OC'd to 4.8GHz (thanks to you), ran Prime95, 3dMark11, and other stability testers and had no crashes indicating my system was stable. Then suddenly, two or three days later the system crashes for unknown reasons. In my case I had the Real Temp program running and was monitoring my CPU temps around the time of the crash and saw that the core temps were getting pretty hot (a couple around 145F degrees) before the crash. But I had seen the temps in that area many times before and received no crashes but maybe my air cooling system is not enough for speeds around 4.8GHz for an extended period of time. Perhaps the high temps had something to do with it as I got no BSOD or Event msg other than the Event Viewer indicated the system shutdown and it was not expected. I'm back to the default optimized settings until I get time to do further research. FSX runs very well at the optimized settings and the core temps stay around 82F degrees. Best regards,Jim

[Bob Scott 7,756]

I'd suspect a problem with the voltage regulator circuitry on the mobo. I had a board do the same thing to me, and it turned out the vreg was the culprit. Often, when overclocking, people don't pay proper attention to the vreg temps and cooling...it's usually a passively-cooled section of the mobo that can get quite hot, especially when run at elevated PWM frequencies (often that's a BIOS setting).If it's temp related, adding heat sinks and/or an internal fan to direct airflow over that section of the board can help(usually it can be found as a cluster of "can" style electrolytic capacitors around/adjacent to one or more MOSFET chips). Also, if running with an elevated vreg PWM freq, try turning that back down to default...that setting rarely needs changed unless running extreme high clocks.All that said, if it's doing that soon after booting a cool system, it's probably going to mean an RMA on the mobo.

[Guest jahman]

Sounds like a temperature problem. As time goes by due to dust/lint accumulation:1. Your fan blades (incl. internal PSU) become less efficient,2. Fan intake/exhaust RF shielding grille holes get smaller,3. Fan axles grime-up and turn slower, and4. Heat sink fins become less efficient.All can contribute to a PSU/CPU/GPU overtemp condition that can trigger a PC shutdown.Use an electronics aerosol duster to clean your PC.Note the problem might not be temperature-related. Even so, ruling our temperature issues is the first step.Cheers,- jahman.

[ZachLW 37]

Strange, Stephen. You seem to be the third (?) person on these forums with a SB system to have this issue.

[nikaslan 9]

I had exactly the same issue when my GPU (5770 back then :Tounge:) was reaching 100C...I found it using the Windows event viewer which reported thermal shutdowns in every single case.

[Superpilotv2 2]

Strange, Stephen. You seem to be the third (?) person on these forums with a SB system to have this issue.

I´m the fourth person. Had the prescribed situation two times since I have this build. I think it´s my bad PSU with only 420W but I don´t have the money for a better one, so I´ll live with that PSU for some time.

Strange, Stephen. You seem to be the third (?) person on these forums with a SB system to have this issue.

[SpiritFlyer 987]

Thanks Taylor, Al, Dazz, Gerald, RTH, Jim, Bob, Jahman Zach, Nikaslan and Steffen,Hope I did not miss anyone! That is a wealth of information and possibilities. Thanks muchly! Here is some more information on the system itself:CoolingThis machine case is well ventilated with 9 fans, ranging from 230mm to 120mm with good directional air flow as follows:Exterior push/pull front toward back:1 x 230mm LED fan1 x 120mm LED fanInterior push/pull front toward back:3 x 120mm LED fans*Exterior side push to motherboard:1 x 200mm LED fanExterior rear exhaust1 x 140mm LED fanExterior upward exhaust2 x 200mm LED fanIn addition: The PSU's 1 x 120mm fan draws from the bottom exterior and exhausts at rear. The 580 GTX GPU draws from interior and exhausts at rear, assisted by a dedicated 120mm feeder fan* and specialty back-plate and flow-through exhaust fittings.All fans continuously run at maximum speed either manually and/or under BIOS turbo settings. The GPU has an aggressive fan profile as modified and controlled by MSI Afterburner.Prevailing Conditions and Operating TemperaturesWith a 25c ambient room temperature the maximum temperatures reached by the CPU at 5Ghz under extreme Prime95 load is about 70c, with the motherboard only reaching about 35c (according to the AI Suite sensor). When under a full load with a custom fan profile, the GPU (900Mhz) reaches + or - 55c. At idle the CPU is at 32c, the MB at 29c and the GPU at 38c.FSX Load TemperaturesWhen under sustained FSX load the CPU is at max 63c with the GPU at 55c and the MB at 31c. At no time during any tests are there any signs, smells or indications of excessive heat. To me this would indicate that overheating of any primary components is likely unless it is hidden as suggested by both Al and Bob. Which reminds me: Bob, which BIOS setting were you referring to?CleaningThe interior is well carefully cleaned and maintained for the most part although some places would have to be disassembled to provide cleaning access. Even those places are blown out every couple of weeks or so. The whole unit is only 3 months old in any case with the motherboard only a month or so due to the B2 replacement program.Conditions of BlackoutThe blackouts usually occur after about 45 minutes of FSX flight or so. When it happens I can push the start button as soon as I can reach it and it will fire up exactly as it should. There is never any indication what happened in either the visible MB LED code, or in the "Event Manager" logs.Power SupplyIn order to be certain that I am getting a consistent and steady power supply to the PSU, I installed a surge protector and backup battery last night. From this point I can more or less assume that any fault will be inside the computer and not out. That is a start! As far as the quality and output of the PSU, I think it is well within prudent guidelines for this kind of setup. I don't know of any way to test it without having to tear the computer down and send it off, not an acceptable option unless a last resort.So that leaves me back to my opening post. Does any of the above information bring any more suggestions to mind? Thank you for your generous help!Kind regards,

[ryanbatc 13,047]

In order to be certain that I am getting a consistent and steady power supply to the PSU, I installed a surge protector and backup battery last night. From this point I can more or less assume that any fault will be inside the computer and not out. That is a start! As far as the quality and output of the PSU, I think it is well within prudent guidelines for this kind of setup. I don't know of any way to test it without having to tear the computer down and send it off, not an acceptable option unless a last resort

Be careful with the battery backup. I just discovered the one I was using isn't very compatible with my PSU. If your PSU has Active PFC make sure the UPS states that it uses true sine wave output. Pretty much all of the cheaper models do not. Anyway mine was causing my PC to hibernate because the PSU was thinking the battery was faulty every so often.

[SpiritFlyer 987]

Be careful with the battery backup. I just discovered the one I was using isn't very compatible with my PSU. If your PSU has Active PFC make sure the UPS states that it uses true sine wave output. Pretty much all of the cheaper models do not. Anyway mine was causing my PC to hibernate because the PSU was thinking the battery was faulty every so often.

Hi Ryan, :Nail Biting: Is this one alright?NEWEGG.CA I know it is on the small side, but I am more interested in the surge protection than running it on the battery if the power completely goes out.Kind regards,