Random Reboots


Aug 11, 2003
I am experiencing random reboots on my system and am at a loss of what it could be. The system will sometimes run for 15 minutes and reboot or run for 2 hrs then reboot.

Here is a list of what I've tried so far
1. Clean Install Windows 10
2. Disable Asus Voltage Protection
3. Load optimal settings in BIOS
4. Removed PSU extension cables
5. Reseated Ram
6. Reseated GFX
7. Cleaned out all fans/dust
8. Changed cable on rear of power supply
9. Memtest86 10 tests, cleared 100%

System specs are below, I'm not sure what else to try. This started happening randomly on Windows 8.1 and continued over to 10 so I tried the clean install today to no avail.

Current System:

CPU: I5-3570K @ 3.4 GHz w/ H100I for cooling
MOBO: Asus Sabertooth Z77 TUF LGA 1155
RAM: Corsair Vengeance 16GB (4 x 4GB DDR3 PC3 15000 )
PSU: Corsair HX 850 WATT
HD: Samsung 840 PRO 128 GB
Last edited:
If event viewer is not telling you any thing, you could try raising the VCCSA voltage a little or you could implement XMP profile if your 4 memory modules were a single kit.
The 4 memory modules were not a single kit, I bought 2 x 4 and 2 x 4.

It is weird because in some instances the system will run fine for say 2 hours, then reboot out of no where. Then it may run for 5 minutes and reboot, or even 20 minutes then reboot. No rhyme or reason to the rebooting.
My daily driver is the Z77 Sabertooth, it did have a 3570k, but now 3770. In the beginning I experienced random shutdown/reboot that I thought was power related. I changed out the PSU and installed a "pure sine wave" UPS. That did not fully solve the problem.

Like you I populated all four memory slots with 2 kits, raising the VCCSA voltage was the last thing I did and I did not see that problem for more than a year. Later I switched to running a matched set 32 gb and XMP and no problems.

All four slots populated is technically an OC and needs the extra voltage to be fully stable.

You might consider testing with a different PSU and add some kind of line voltage conditioner. The Sabertooth seems to me to be oversensitive to voltage fluctuations.
Last edited:
How do I raise the vccsa voltage?

I've contacted corsair to do a RMA as the system was rock solid for approximately a year and then this all started a few months back, but I'm annoyed to the point I decided to take action.
While you are there check to see if your memory timings are the same for each of the four modules, at least through the secondary timings. If necessary loosen up the tighter modules until they all match.

I have seen 2 sets of 2 module kits have better matched timing than some single 4 module kits even in the tertiary timings. The problems with running two kits can be; mismatch timing and you can't use XMP profile without modifying the settings, which if I remember correctly wipes out some of the XMP settings, because it was written for only two modules.

CPU's can degrade over time so needing to adjust the voltage (VCCSA) to the on CPU memory controller to maintain stability would not be abnormal.
Tried upp-ing the voltage, no go. Called corsair and rma-ing my psu. If this doesn't help then I'll try I guess motherboard next. Kind of stinks
Tried upp-ing the voltage, no go. Called corsair and rma-ing my psu. If this doesn't help then I'll try I guess motherboard next. Kind of stinks
I ran memtest86 all 10 tests, cleared with 0 errors.

What is killing me is sometimes the system will run 5-10 minutes and reboot. Other times it'll run for hours without any issues then randomly reboot.

I'm hoping the psu swap does the trick otherwise I don't know what to do. It had this issue on windows 8.1 and windows 10 so don't think it is a windows issue. Event log only gives kernal 41 error and that most likely points to the psu. BIOS shows CPU temp 25C, Mobo Temp 29C. I run a H100I water cooler on cpu so don't think that overheating is an issue. I also am not overclosed either!
Pull the video card
Are you running stock speeds on the memory? (1600)?
Everything is stock speeds yes. I'll pull the graphics card once I swap out the psu
Im almost sure that is bad RAM or PSU is half dead. Try using a single RAM in different slots and different RAM blocks.
Your not suppose to play with raising the vccsa voltage in order to make your pc working as it should.
Next take a look for CPU tems.
The ram I ran memtest86 through the whole 10 tests and everything passed.

The replacement PSU will be here Thursday, I will swap out the PSUs and see if that fixes the issue. Going to be a PITA to get all the cables sorted.

If the PSU replacement doesn't fix it then I will look at doing the single ram sticks to see if that fixes it.

CPU temps in BIOS shows CPU temp 25C, Mobo Temp 29C,
Replacement PSU came in and just spent the hour getting everything plugged in. All seems well so far, knock on wood!