Computer Cuts Power (Shuts Down)

[jed]

Graphics card issue will cause that glitch shown on the monitor.
Connect to onboard graphics to see if that functions correctly

I'll try this. Why would it run a game full speed, but not be able to generate the startup splash screen? It usually does load - it always has 4 or 5 of those lines on the top left. Randomly though it won't startup and give a full set of then across the top of the screen.

 

[Toothless]

Not that I recall. Tell me more. What will this scan for?

Everything in the registry i think. Fixed my rigs a few times.

 

[Bill_Bright]

What will this scan for?

"sfc" stands for System File Checker. As noted here, it scans for corrupted Windows system files and, when possible restores them.

 

[jed]

Thanks guys. I am confused though - how could it be a video card issue? My computer didn't boot up 5 minutes ago and got those colors at the top of the screen. Booting up from "Last Known Configuration" worked. What would that have to do with the video card? I feel like all these problems are linked to one issue.

The result was "Windows Resource Protection did not find any integrity violations."

 

[jed]

There it goes again... Random shutdown. Doing a virus scan with Avast and playing League. No overclocking on GPU or CPU.

 

[Bill_Bright]

I think you need to regroup and summarize everything that you have done in an attempt to fix this. Not many are going to want to go back through 6 pages of posts to see where we really are.

 

[jed]

Here's a refresher on the issue:

[quoute]Recently it has been randomly shutting down completely out of nowhere. It's like the power cord gets unplugged from the wall. No warnings or anything, just a 'click' and a dead computer. Reboots up fine with the typical "start normally or in safe mode?" prompt. This can happen both during idle and during gameplay - no rhyme or reason.[/quote]

Other issues:
Won't bootup sometimes (view above picture), have to go back to "last known configuration"
Can't backup PC - shuts down during backup
Can't run CHKDSK of my data/storage drive even after hitting "force dismount"
Can't run antivirus scan - shuts down during scan.

Suggestions from others:
- Dying PSU - still possible
- Unstable CPU and GPU Overclock - ruled out, reverted to stock settings
- Heat / Inadequate cooling - ruled out, reverted to no OC and still dies after hours of idle
- Weak 12v rail
- Open side panel and blast air for cooling - ruled out, did this and still shut down

So here's what I've done so far with no avail.
- Deep cleaned the PC
- Replaced 140mm top fan (it was dying)
- Moved my CPU cooler/radiator from side vent to rear vent, exhaust.
- Disabled ALL overclocks
- Ran "SFC Scan Now" - clean result
- Blasted a fan into the case
- Uninstalled AVG and replaced with Avast (AVG files were preventing safe mode bootup)
- Upgraded to latest GPU drivers

About to try the last resort of the PSU, but that probably won't solve the "other issues". I feel like there's a deeper problem here.

Update:
I just took the PC apart to reconnect all the wires and pins. I also removed the PSU and cleaned if - if was filthy. Upon reconnecting, the computer just clicks once, the fans spin for a split second, and it shuts off. I have verified everything is connected properly... did I somehow just trigger the last breath of my PSU?

 

[Agility]

Hmm not sure if this is adequate but have you considered formatting? (Didn't read through all the posts as it was way too long). Assuming your "Deep Clean" meant clearing dusts on your parts.

Get a mate that lives near you and try swapping parts to check them individually. Other things you can consider looking at are the capacitors on the motherboard (Or your PSU if you don't mind taking it apart to confirm a potential failure of the PSU).

 

[Bill_Bright]

(Didn't read through all the posts as it was way too long).

That's why I asked for a summary which Jed nicely provided.

I do not recommend formatting and reinstalling. That should ALWAYS be a last resort option as it often doesn't fix anything, results in being behind months or years in security updates, you lose all your data and installed programs. Even if it does fix it, you don't learn anything to help prevent recurrence.

More importantly, as Jed noted, his system shutdowns like yanking the power cord from the wall. And the problems appear long before the boot drive is even touched. That clearly is not an OS problem but obviously a hardware issue and most likely power issue.

While it could be leaky motherboard capacitors, those problems are less common these days. You really need to swap in a known good PSU since everything else inside the computer depends on good, clean, stable power.

 

[jed]

That's why I asked for a summary which Jed nicely provided.

I do not recommend formatting and reinstalling. That should ALWAYS be a last resort option as it often doesn't fix anything, results in being behind months or years in security updates, you lose all your data and installed programs. Even if it does fix it, you don't learn anything to help prevent recurrence.

More importantly, as Jed noted, his system shutdowns like yanking the power cord from the wall. And the problems appear long before the boot drive is even touched. That clearly is not an OS problem but obviously a hardware issue and most likely power issue.

While it could be leaky motherboard capacitors, those problems are less common these days. You really need to swap in a known good PSU since everything else inside the computer depends on good, clean, stable power.

Do you think this would work for my system, or is it too small? I currently have 650W.

http://m.newegg.com/Product/index?itemnumber=N82E16817139027&ignorebbr=1

 

[DRDNA]

While it could be leaky motherboard capacitors, those problems are less common these days. You really need to swap in a known good PSU since everything else inside the computer depends on good, clean, stable power.

I would also even before trying a different PSU look at the tops of all your capacitors for any upward bulging, even though it happens less often it still happens. If you dont find any then try PSU swap around.

 

[Bill_Bright]

I'm a little confused because I am seeing way back on page one, many people, including me, suggesting you try a different PSU. At the time, you said you had a 700W Antec PSU in there and now you are saying 650W. So have you already swapped out supplies?

The CX is Corsair's next to bottom line and doesn't have the best reputation. I would not buy one but if I had one laying around, I would use it temporarily for a test. If you system is the one listed in your System Specs, it should have enough power.

 

[DeathtoGnomes]

Have the boot issues been fixed?

Edit: after reading the entire thread, my eyes are bleeding.

I've seen that squiggly line on a monitor before, IIRC its from a borked and dirty driver install.

There could be various windows settings as an excuse for the computer to shut down, like hibernate is enabled. Another is BIOS settings, like the temp monitor is enabled and not set correctly.

That case doenst have a lot of room to install an AIO without it getting in the way of something or other. The best placement is in the top exhaust slot with the fan on the outside of the case to elevate the radiator as far away from the mobo, yet close enough to the rear exhaust fan so that the radiator's push fan is not pulling in hot air and pushing it thru the radiator.


IF the PSU is indeed having heat issues, I give my very Ghetto case mod suggestion : Create a partition from something like cardboard long enough to almost reach the front intake fans installed in such a way that will deflect some of the intake towards the PSU intake.

Or, if you are feeling really froggy , create a vent in the bottom of the case and install another intake fan.

 

[jed]

I'm a little confused because I am seeing way back on page one, many people, including me, suggesting you try a different PSU. At the time, you said you had a 700W Antec PSU in there and now you are saying 650W. So have you already swapped out supplies?

The CX is Corsair's next to bottom line and doesn't have the best reputation. I would not buy one but if I had one laying around, I would use it temporarily for a test. If you system is the one listed in your System Specs, it should have enough power.

Thanks Bill. I'm sorry for the confusion. I have not swapped the PSU. I thought it was 700 but when I took it out to clean it I found it was 650. My dad is bringing a spare today to test. Hopefully I can get it to at least boot.

 

[jed]

I would also even before trying a different PSU look at the tops of all your capacitors for any upward bulging, even though it happens less often it still happens. If you dont find any then try PSU swap around.

The capacitors look good. However, upon further inspection, I found the Northbridge Heatsink to be quite loose. Sounds like tons of different solutions to remedy it including screws and thermal paste, any suggestions?

Edit: After reading more, it sounds like this is the regular design. It has push-pins with springs; it re-seats itself when I stop messing with it... it's just "loose" I guess.

 

[Bill_Bright]

I found the Northbridge to be quite loose

What??? That's pretty much impossible as the chipset is typically surface mounted (soldered) directly to the motherboard. Did you mean the Northbridge's heatsink?

Note the temperature of the Northbridge is typically (though not always) designed in hardware monitoring as the "system" temp. And if that gets too high, it can certainly result in an unstable system.

If the heatsink is loose, then you need unplug the computer from the wall, touch bare metal of the case interior to discharge any static in your body, Then carefully remove the heatsink. Thoroughly clean it and the chipset IC (frequently touching bare metal again to keep static down) and apply a fresh new layer of TIM (thermal interface material).

If the heatsink is secured by a clamping mechanism and screws, use normal TIM as you would with a CPU. If there is no mechanism to hold the heatsink in place, you need to use "adhesive" TIM like this.

 

[jed]

What??? That's pretty much impossible as the chipset is typically surface mounted (soldered) directly to the motherboard. Did you mean the Northbridge's heatsink?

Note the temperature of the Northbridge is typically (though not always) designed in hardware monitoring as the "system" temp. And if that gets too high, it can certainly result in an unstable system.

If the heatsink is loose, then you need unplug the computer from the wall, touch bare metal of the case interior to discharge any static in your body, Then carefully remove the heatsink. Thoroughly clean it and the chipset IC (frequently touching bare metal again to keep static down) and apply a fresh new layer of TIM (thermal interface material).

If the heatsink is secured by a clamping mechanism and screws, use normal TIM as you would with a CPU. If there is no mechanism to hold the heatsink in place, you need to use "adhesive" TIM like this.

It's this circled unit here (From another persons picture.) My HWMonitor temps are currently reading at idle:

TMPIN0 31C
TMPIN1 61C
TMPIN2 70C

After reading, it actually sounds like you're supposed to be able to wiggle it. It has the push pins with the springs. I can wiggle it but it re-seats itself... I just assumed it was supposed to be tight on there. Regardless, the temps on this board are huge...

Also, newbie mistake with the computer not booting up... I checked the RAM for gunk or dust, and didn't get them seated properly. I can now boot up, so the PSU is working as it was before.

 

[jed]

Have the boot issues been fixed?

Edit: after reading the entire thread, my eyes are bleeding.

I've seen that squiggly line on a monitor before, IIRC its from a borked and dirty driver install.

There could be various windows settings as an excuse for the computer to shut down, like hibernate is enabled. Another is BIOS settings, like the temp monitor is enabled and not set correctly.

That case doenst have a lot of room to install an AIO without it getting in the way of something or other. The best placement is in the top exhaust slot with the fan on the outside of the case to elevate the radiator as far away from the mobo, yet close enough to the rear exhaust fan so that the radiator's push fan is not pulling in hot air and pushing it thru the radiator.


IF the PSU is indeed having heat issues, I give my very Ghetto case mod suggestion : Create a partition from something like cardboard long enough to almost reach the front intake fans installed in such a way that will deflect some of the intake towards the PSU intake.

Or, if you are feeling really froggy , create a vent in the bottom of the case and install another intake fan.

Do you think it's the GPU driver, then? I assume I should safe mode boot and remove the driver and reinstall. Hibernate is not enabled, but I will check the BIOS settings. They haven't changed, but I don't think it'll make a difference (temp monitor) since it shuts off at idle as well when temps are "low". Interesting on the cooler placement - I could try it out. That's an interesting suggestion about diverting air to the PSU... Worth a shot. I am afraid of losing airflow to the rest of the components via the partition - everything runs hot enough as it is.

 

[DeathtoGnomes]

It's this circled unit here (From another persons picture.) My HWMonitor temps are currently reading at idle:

TMPIN0 31C
TMPIN1 61C
TMPIN2 70C

After reading, it actually sounds like you're supposed to be able to wiggle it. It has the push pins with the springs. I can wiggle it but it re-seats itself... I just assumed it was supposed to be tight on there. Regardless, the temps on this board are huge...

Also, newbie mistake with the computer not booting up... I checked the RAM for gunk or dust, and didn't get them seated properly. I can now boot up, so the PSU is working as it was before.

No the Northbridge heat sink shouldnt just wiggle, it can move but it should be a bit difficult to do so. It should feel like pushing large heavy furniture across a carpeted room, any less and your thermal paste hardened up and cracked or if its a thermal pad, its damaged. The Northbridge could contribute to your woes, not sure exactly sure.

Do you think it's the GPU driver, then? I assume I should safe mode boot and remove the driver and reinstall. Hibernate is not enabled, but I will check the BIOS settings. They haven't changed, but I don't think it'll make a difference (temp monitor) since it shuts off at idle as well when temps are "low". Interesting on the cooler placement - I could try it out. That's an interesting suggestion about diverting air to the PSU... Worth a shot. I am afraid of losing airflow to the rest of the components via the partition - everything runs hot enough as it is.

You wont lose air flow, unless you completely block it off. The diversion is only to direct 1 intake fans air flow towards the PSU, directing it as close to the PSU without blocking it from the rest of the case. Its pretty simple do do. If could make you a drawing I would.

 

[AceTheBass]

sound like bad capacitors in psu not worth fixing if you have no knowledge on it or if its a cheap supply but if you find caps in it wit domed tops its bad caps are for power management they stabilize power outputs by first charging say 12v and if there is a sudden power drop say 12 drops to 10 the cap will replace that for a short time depending on its capacity or if the voltage rises to 14 it will charge to that so it the capacitor goes bad that can lead to you pc dying especially if your gaming because it requires more power. so if you have the know how fix it if not buy a new one.

 

[DeathtoGnomes]

sound like bad capacitors in psu not worth fixing if you have no knowledge on it or if its a cheap supply but if you find caps in it wit domed tops its bad caps are for power management they stabilize power outputs by first charging say 12v and if there is a sudden power drop say 12 drops to 10 the cap will replace that for a short time depending on its capacity or if the voltage rises to 14 it will charge to that so it the capacitor goes bad that can lead to you pc dying especially if your gaming because it requires more power. so if you have the know how fix it if not buy a new one.

These days its cost efficient to buy a new PSU, a one year's worth of warranty is worth 25%(-ish) of that cost if you are paying over $150 (US) for said PSU.

 

[Bill_Bright]

After reading, it actually sounds like you're supposed to be able to wiggle it.

No, that is not how it is intended.

Those type spring mounts only wiggle because the downward pressure is applied by springs. And springs are used to apply even pressure. Once you apply a proper layer of TIM and mount the heatsink, you leave it alone. If you wiggle it after that, you will break the cured bond of the TIM and that then requires you remove the heatsink, clean the mating surfaces and apply a fresh new layer of TIM. If you don't break the cured bond, the TIM last and remain effective for years upon years. But as soon as the cured bond is broken, insulating air gets in between the mating surfaces and that is bad.

 

[DeathtoGnomes]

I had to correct the spelling on what I said, you shouldnt be able to just wiggle it as Bill said. IF your thermal paste/TIM is the curing type, its the same as a thermal pads, if they wiggle they more then likely "broken loose" and should be replaced as soon as possible.

 

[jed]

I've got a spare PSU in there but it doesn't have the right connections for my 7950. Right now it's just sitting there running with blank screens. MoBo does not have onboard GPU. If it shuts off again then we'll know the PSU is probably not the issue.

Using PSU calculators, one tells me 450 watts while the other told me 815 watts... could you guys give your input on what would be sufficient? I am looking at new units in case the PSU is the issue. Right now I have a 650 and for the first 2 or 3 years that seemed to work well enough. Would that 600 to 700 range be good?

 

[Mr.Scott]

I've got a spare PSU in there but it doesn't have the right connections for my 7950. Right now it's just sitting there running with blank screens. MoBo does not have onboard GPU. If it shuts off again then we'll know the PSU is probably not the issue.

Wait....you have no onboard GPU, and you have no power connected to your 7950, so how did it even boot?