Random Restarts (Startup, gaming, browsing)

I have a problem with my machine I've been battling for months now. It will just randomly restart and I could be doing anything. It has happened when browsing the web, playing games, and even when the motherboard logo shows up during startup.
It never seems to crash when running Prime95 or OCCT stress tests.
What I've done:
Hardware
-Re-installed windows twice
-Purchased new motherboard
-Purchased new CPU cooler (from air into liquid)
-Purchased new GPU (HD7950 to GTX780)
-Purchased SSD
-Monitoring software showed PSU 12v rail running at 10.8v when stressed. Checked 4 pin molex connector with voltmeter and it measured 12.2v and it dropped to 12.1v under stress.
-Changed the case, thought it was a short circuit somewhere.
Software
-Run multiple stress tests which did not cause a crash.
-Ran Prime95 and Battlefield 4 at the same time. No crash
-Updated GPU drivers
-Updated motherboard drivers
-Can't update BIOS, scared computer will restart during the flash. My motherboard revision is 1.01 and
-I've run memtest for 5 hours, but I should do it again

Sometimes, instead of a restart, my display will turn off and any sound playing will begin to loop. PC will be unresponsive.
CPU temps reach only about 35 degrees under full load.

Ouch. A newer, beefier power supply sounds like the solution as it smells like a hardware problem in the power. I would blame the heat if it was just occurring during the last couple of months. Since I have not had anything like this tho, I am guessing. You need to have a rock steady 12V no matter what. If this was a USED power supply, chuck it anyways. I always found that if something was going to fail in a new build, it was always something that I swapped from the old computer to the new one. Memory chips, even a optical drive once! About the only thing I keep on a new build nowadays is the monitor, the keyboard, and the mouse. All the rest is brand new.

My PSU is 750W, and it's Corsair. I've heard they are very reliable and would be annoyed to have to replace it. That should be powerful enough to handle my computer. But the thing I don't understand is why my monitoring software shows different voltages. OCCT showed 10.8v when stressed. HWmonitor showed 11.3v stressed. But with a real voltmeter reading on a molex connector, it showed 12.1 volts. Should I measure the voltage from somewhere else using the voltmeter? If so where?

How often exactly are you getting the restarts? Once a week? Day? Every session?

If they are happening every session (or almost), boot into safe mode and play around with it that way for awhile (ie., do your best to do what you can that might trigger a restart, even though I know you are limited in safe mode).

If you don't get any restarts, then you are definitely having some kind of driver issue. If you do have problems in safe mode, then it is more than likely hardware related. If the problem doesn't happen frequently enough (ie once a week, etc) for our initial safe mode test to be viable, then let me know and we'll move on from there, but we have to start somewhere.

One more thing, are you using any kind of overclocking software (either on the processor or your GPU)? And did you check event viewer for any critical errors?

lastof said:

How often exactly are you getting the restarts? Once a week? Day? Every session?

If they are happening every session (or almost), boot into safe mode and play around with it that way for awhile (ie., do your best to do what you can that might trigger a restart, even though I know you are limited in safe mode).

If you don't get any restarts, then you are definitely having some kind of driver issue. If you do have problems in safe mode, then it is more than likely hardware related. If the problem doesn't happen frequently enough (ie once a week, etc) for our initial safe mode test to be viable, then let me know and we'll move on from there, but we have to start somewhere.

One more thing, are you using any kind of overclocking software (either on the processor or your GPU)? And did you check event viewer for any critical errors?

Click to expand...

The restarts are completely random. Sometimes they wont happen for a whole week. Sometimes the restarts will happen 4 or 5 times in a single day. Usually it restarts more often when I turn the computer on for the first time in the day and try to run a bunch of programs at once. If I leave my computer on all day in idle it is rare for it to restart.

I am not overclocking at all. GPU is also stock speeds. I have checked event viewer and I always get the same critical error which is Kernel Power event ID 41 Category 63. In the details all parameters are 0.
- <Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event">
- <System>
<Provider Name="Microsoft-Windows-Kernel-Power" Guid="{331C3B3A-2005-44C2-AC5E-77220C37D6B4}" />

<EventID>41</EventID>

<Version>3</Version>

<Level>1</Level>

<Task>63</Task>

<Opcode>0</Opcode>

<Keywords>0x8000000000000002</Keywords>

<TimeCreated SystemTime="2014-08-08T06:08:39.977495700Z" />

<EventRecordID>1635</EventRecordID>

<Correlation />

<Execution ProcessID="4" ThreadID="8" />

<Channel>System</Channel>

<Computer>James</Computer>

<Security UserID="S-1-5-18" />

</System>


- <EventData>
<Data Name="BugcheckCode">0</Data>

<Data Name="BugcheckParameter1">0x0</Data>

<Data Name="BugcheckParameter2">0x0</Data>

<Data Name="BugcheckParameter3">0x0</Data>

<Data Name="BugcheckParameter4">0x0</Data>

<Data Name="SleepInProgress">0</Data>

<Data Name="PowerButtonTimestamp">0</Data>

<Data Name="BootAppStatus">0</Data>

</EventData>


</Event>

Windows Kernel event ID 41 error in Windows 8.1, Windows 8

Definitely seems like a power supply problem or incorrectly attached/damaged cables?
Do you have a surge protection? You should consider checking the household electrical wiring.

TairikuOkami said:

Windows Kernel event ID 41 error in Windows 8.1, Windows 8

Definitely seems like a power supply problem or incorrectly attached/damaged cables?
Do you have a surge protection? You should consider checking the household electrical wiring.

Click to expand...

Hmm I've re-done all the cable management multiple times, never noticed any damaged cables. Could be something inside the powersupply. I don't think my powerboard is surge protected. If the electrical wiring was faulty, then wouldn't the peripherals connected to the wall outlet have turned off for a split second?

vancamp said:

TairikuOkami said:

Windows Kernel event ID 41 error in Windows 8.1, Windows 8

Definitely seems like a power supply problem or incorrectly attached/damaged cables?
Do you have a surge protection? You should consider checking the household electrical wiring.

Click to expand...

Hmm I've re-done all the cable management multiple times, never noticed any damaged cables. Could be something inside the powersupply. I don't think my powerboard is surge protected. If the electrical wiring was faulty, then wouldn't the peripherals connected to the wall outlet have turned off for a split second?

Click to expand...

This error is just telling you that power was lost to the kernel abruptly (ie., the automatic restarts), and that Windows was not shut down completely.

If I am understanding correctly, it seems that the problem has persisted through 2 different motherboards (old and new), 2 clean installs of the OS, 2 different GPUs (old and new), and 2 HD's...unless you are saying that the problem started AFTER you made all the hardware upgrades, then my best guess would be either your system memory, or the CPU - as these are the only two critical pieces of hardware that are consistent throughout your upgrades.

With that in mind, how much RAM do you have? How many DIMM banks are occupied? I would remove all the sticks except for the bare minimum (ie., if you have 4 sticks of RAM, remove 3). Use the computer for awhile/try to recreate the problem. If everything seems ok, add one more stick, then use computer, add another stick, etc etc.

If you do get a crash right away when you are using the bare minimum (ie., one stick), then pull that one, label it, and try another. If it continues to crash no matter what stick you use, then it's more than likely your CPU is the culprit, with a slight chance it could be the PSU. At that point, I would go buy a new PSU, keep the receipt, install it, and march ahead with your 'testing'. If it still crashes, pull the new PSU, put your old one back in, and return the new one for a refund - if you get this far, and eliminate all those variables, I really think the problem is the CPU.

EDIT: (Do this first): After taking a look at your attachment, I noticed that you have "origin.exe" in your startup folder. I did a little search on Google and got quite a few hits of people getting BSOD's while using and/or launching origin.exe. Go into your startup folder and disable it. If there is a control panel for this app, make sure you turn off "start with Windows" or similar. Restart, and verify in task manager that origin.exe is not loaded and running in the background. Use your computer normally and see if you have any problems...

vancamp said:

My PSU is 750W, and it's Corsair. I've heard they are very reliable and would be annoyed to have to replace it. That should be powerful enough to handle my computer. But the thing I don't understand is why my monitoring software shows different voltages. OCCT showed 10.8v when stressed. HWmonitor showed 11.3v stressed. But with a real voltmeter reading on a molex connector, it showed 12.1 volts. Should I measure the voltage from somewhere else using the voltmeter? If so where?

Click to expand...

Heh, I have Corsair 750W power supplies in all four of my desktops and they all are rock solid. Is it the problem? I donno. I usually have an extra power supply anyways as a spare. Wait for a good sale and scarf up another one or if there is another desktop in the house, temp swap it if you think the power supply is the issue. Frankly, no way really to know if it is the problem until you can swap it out to eliminate it. You have replaced damn near everything else tho and I cannot really see how a software bug could be so random..

Voltage you are measuring on PSU itself does not have to correspond with voltages CPU is getting, power circuit om the MB itself handles that and BIOS regulates it automatically if it's set so. What I would do is to manually set processor Vcc in the bios to default for it or maybe even a touch above, turn off any "Turbo" functions and power saving modes in BIOS and turn off AMD's C&Q software and try like that.
There are similar voltage settings for chipset and memory. There should also be some LLc settings in the BIOS that is supposed to limit voltage to the processor and if set too low when processor demands more power even for a microsecond it would shut it down.

My guess is a defective PSU since it's the only part you have not replaced so far. It does not matter that it has enough power (750 W) if it is defective. Even if the PSU only fails for a milisec or so and is only short of a few mV this leads to immediate restart/ shutdown of the pc. Thus my advice is to get a new PSU.

Hi guys,

Thanks for all the responses so far, but I have just encountered a new problem. I was just browsing the internet as usual and suddenly the computer restarted itself. But instead of going through POST the CPU LED on the motherboard instantly turns on and stays lit. I don't understand how the PC can be working for a whole week, suffer one restart and now the computer wont even boot. I turned it off from the switch (case buttons unresponsive) and removed 1 RAM stick. Computer turns on and instantly ( I mean as soon as the power button is pressed) the LED for CPU turns on. Repeated the process with the other RAM stick, same issue. When I boot it, there are two fans connected to the motherboard which do not turn on. My other case fans that are connected directly to the power supply do turn on.

I have encountered this issue before when I had my previous motherboard. I just unplugged all connections and plugged them back in, after doing this several times the PC would work again. I think I'll try that now.

Honestly I think it's my CPU that is playing up.

Did you try what I suggested ? Auto settings are wrong many times.
PSU, it does not have to malfunctioning or too low power all together, Which exact model is it. Check the sticker on it. Some have 2 or more branches with their own limits. One for mainboard and another for video cards etc. Total power of the PSU can look to be enough but one branch may not have enough power to handle everything, that is usually for second branch for GPUs etc. Also, total power of PSU is only peak power and is not supposed to run that high all the time.

CountMike said:

Did you try what I suggested ? Auto settings are wrong many times.
PSU, it does not have to malfunctioning or too low power all together, Which exact model is it. Check the sticker on it. Some have 2 or more branches with their own limits. One for mainboard and another for video cards etc. Total power of the PSU can look to be enough but one branch may not have enough power to handle everything, that is usually for second branch for GPUs etc. Also, total power of PSU is only peak power and is not supposed to run that high all the time.

Click to expand...

No, I did not try that yet, but I will try it as soon as I can get my PC to boot again. At the moment it wont boot into BIOS no matter what. My PSU is a Corsair GS700.

Seems like that PSU got pretty good rating: Corsair GS700 Bronze Power Supply Review | Hardware Secrets but you might have a defective one. Try resetting BIOS to default before you attempt another start.

CountMike said:

Seems like that PSU got pretty good rating: Corsair GS700 Bronze Power Supply Review | Hardware Secrets but you might have a defective one. Try resetting BIOS to default before you attempt another start.

Click to expand...

Okay I managed to get the PC to boot into the BIOS, but I'm not exactly sure what I'm looking for, or how to navigate in this BIOS. I am using an ASUS Saberooth 990fx. I would very much appreciate some help to navigate, thanks.

Grab the manual first: Motherboards - SABERTOOTH 990FX - ASUS

CountMike said:

Grab the manual first: Motherboards - SABERTOOTH 990FX - ASUS

Click to expand...

Yep I've got a hard copy next to me.