[SOLVED] System randomly shuts down, troubleshooting advice requested

[zer0bit]

I have a PC I custom built several years ago that randomly powers off anywhere from 10 seconds-30 minutes after its powered on and I have been unsuccessful at diagnosing the cause.
I am currently running Windows 10 Pro 64 bit with latest updates and most current bios for the motherboard. The CPU is set to stock settings in bios.
My components are:
Asus Z87I-Deluxe Motherboard Bios 1204
Intel Core i7 4790K CPU running at 100Mhz Bus x44 Multiplier = 4.4 Ghz
2 sticks of 8GB Kingston KXH2400C11D3/8GX DDR3 @ 9-9-9-24-1T timing (Non XMP setting)
2 samsung 250GB SSD running in Raid (Mirrored) off MB
Using the built in Intel HD Graphics 4600 @ stock settings
1000 Watt Gold modular power supply

I'm using a Corsair liquid cooler for the CPU which keeps temps at 22-35 degrees C on the CPU (so overheating should not be a factor, CPU should be able to handle up to 70 C without issues), I also have 3 other static pressure care fans to keep the inside of the case cool.
I do have the latest Asus AI Suite which starts when windows does.

Note: when I initially installed the latest AI Suite, I attempted to run the 5 way optimization (to do a general overclock) when I did it w the Multiplier option it worked as expected and used a multiplier of 48, however when I attempted the other option of increasing the bus speed for the overclock it would get stuck in an endless reboot loop at initial startup which I had to manually reset the BIOS to stock to get out of. However even with the multiplier of 48 it was still having random shutdowns (no blue screen, just powers off) I found that dropping the Digi+ auto power phase control sliders to the minimum seemed to make it more stable for longer periods of time. I have since reverted the CPU settings back to stock and kept the digi+ sliders at minimum, but am still experiencing random shutdowns.

There are no error logs in Windows that identify what caused the shutdown, Device drivers are all up to date and no errors in device manager. There is also no viruses or malware on the system.

This is extremely frustrating and I would greatly appreciate any suggestions for determining the root cause and solutions for this issue.

 

[luckymatt42]

Does it shut off after you've been using it awhile, or does it shut down like you describe even when idle?

If it's only shutting down during use, what kinds of programs/processes are you running when it shuts down? Does it shut down using only certain programs, or no matter what you are doing?

 

[zer0bit]

Does it shut off after you've been using it awhile, or does it shut down like you describe even when idle?

If it's only shutting down during use, what kinds of programs/processes are you running when it shuts down? Does it shut down using only certain programs, or no matter what you are doing?

It shuts down both ways, at idle doing nothing and when using applications such as chrome or kodi, doesn't seem to have a pattern for what triggers it. For example, just at idle I started it a couple times now, it powered off before it got to windows login (5 sec after post), then a second time after about 2 minutes of being logged into windows and idle, CPU temp was at 24 deg C according to AI Suite at that time

 

[luckymatt42]

May sound like a strange question, but what is the temperature of the room the computer is in (when you got that 24C reading?)

I would also recommend getting another CPU temp monitoring software, like AIDA64, Speedfan, HWMonitor, etc. Make sure you're looking at the actual CPU temps, it will be labelled something like "Core#1, Core#2,...etc" and NOT something like "motherboard temp". Come back with the core temps at idle, then while running something that will stress your system (I like RealBench) and report peak temps WHILE the load is running.

 

[zer0bit]

The room temp is about 65 degrees F, so after researching a bit more, I messed around in the bios and disabled the digi+ completely and disabled EPU functions, rebooted to windows and tan the built in 8 core stress test in cpu-z while monitoring temps. Ran it for 20 min and it seems stable, max temp it achieved was 34 C, decided to try overclocking it again and see if it was stable now, running it at 100.2 bus speed and 49 multiplier, using the same method, it’s been running for another 20 min and stable at 45 C.

Idk why the digi+ feature would cause system instability as that’s what it’s designed to avoid. Any ideas?

 

[zer0bit]

So I downloaded real temp and it shows much different temp picture than ai suite.
Aisuite shows CPU temp at 46C, real temp shows cores at 83-86 C. That’s a huge difference, possible aisuite interfering w what real temp is showing?

 

[luckymatt42]

Ya that's what I thought...the temps you were seeing are probably the "motherboard" temp, which is really just the ambient temp around the motherboard sensor, not even close to the core temp. The 83 to 86 sounds about right.

And don't bother messing with the "bus" speed (aka BCLK), had nothing but trouble with it on my Z87.

If I were you I'd reset the bios/uefi to default and start over...don't touch the bclk, and for now just stick to cpu multiplier and core voltage. BUT just run the full test suites with everything at default first (enable xmp if you have it though, that's the ONLY thing that should be non-default).

 

[TechyInAZ]

Well first off, doing all your overclocking in AI Suite 3 is a bad idea. AI Suite 3 has its fare share of issues. You are far FAR better off doing all your overclocking within BIOS and just using AI Suite 3 for fan control and monitoring.

 

[zer0bit]

That’s pretty much what I’m using aisuite for now, just monitoring overall temp, system is stable so far after I disabled digi+ and EPU in bios, did the overclocking in bios, but most of the items r set to auto except max multiplier is at 49.
Those temps should b stable for my processor? 88C was highest registered in real temp per core at full load stress testing. I did activate XMP profile for RAM also when I did the initial stablity testing. The vcore seems a little high at max load tho 1.374v, but it’s been stable, so I’m assuming it’s ok.

 

[zer0bit]

As long as temps and voltage is OK, I’d say a stable overclock of 4.9 is a great success w my liquid cooled CPU and hardware.

I also disabled fan Q control and am running all fans at max speed from bios.

Thanks guys for your help.