PC "Shutting Off" Randomly, No BSOD, Not Under Load

Apr 14, 2018
3
0
10
I have gathered the "critical" events from the system event log, and have them here.

There are no dumps. I set the computer to not automatically restart and to produce a full dump on crash, but there are no dumps there, nor detected by programs like WhoCrashed. This doesn't surprise me as there is no BSODs, but I've had troubles pinpointing the issue from there.

Specs, most of it anyways:
ASUS M4A785M with Phenom II X6 1100T CPU
Corsair GS800
7 GB of RAM, a pair of 2 GB sticks, and a mismatched 2GB and 1GB stick
Micron 1100 2TB SSD (/w OS)
4 other drives, one external
16x PCIe MSI gfx card with specs that I do not know of offhand

Story time below, I want to give all the details of what I have tried and what is wrong without writing too much:

This computer is a bit Frankenstein's monster of my old tower, it has been offline for the most part of the last 3 years. When I last had it working (about a year ago) a (very slow!) 200GB 5200RPM HDD had the OS, and it chugged along ok. I stopped using it after a storm had knocked the power out, and it no longer booted at all (no video, no 'beep' from the speaker installed). I finally fixed it buy replacing the CPU with the current Phenom II with bent pins- I used a small edge of a plastic container to straighten them and they slotted right in. The computer then ran straight for three days before my first restart issue.

When the computer came back online I bought the SSD, so it is brand new and I sure hope it isn't failing. However, due to the motherboard not having UEFI and... another issue I discovered, I couldn't install windows on this disc. Found a solution online that Windows 10 will not install on a SSD larger than ~300 GB. I could have tried drive partitions, but eventually I cloned the old drive and expanded it. The OS was only used for about 3 months total with that old drive, so it just seemed simpler anyways.

I also added more drives, one of which is showing SMART errors- bad sectors that are not spreading. This being part of the reason I had installed the OS on a PoS drive, but unless I'm incorrect, a HDD with a problem shouldn't be causing this unless it's the OS drive.

I am aware that mismatched RAM sticks aren't exactly recommended, however when I had computer running with the old CPU I never had this problem. I made sure to blow dust out from them to be sure it isn't over heating. I'm going to pull it apart soon and take out each stick of RAM, then reseat to be sure. Currently I am running memtest (before I post, 1.5 passes and no errors), but I don't know that RAM issues could cause a computer to shutdown without BSOD.

The GS800 is certainly overkill, and I'm hoping it isn't the dead one. I'm willing to test it with a multimeter- or a specialized tester if multimeters don't actually work that well- but I want to first have a sign that there is something wrong with it. The symptoms are right, however these restarts happen sometimes with only discord or a single chrome window open, not under load. I've had it happen during gaming, but with games I would consider lightweight- it has never rebooted during GTAV or DOOM (yet), just Eternal and some small indies. It has so many cords that I try to have placed out of the way, but I'm using most, if not all of the SATA power ports. Perhaps I should move the OS drive somewhere else, but all the drives are pretty much pulling the power wires taut.

Some other threads mentioned updating BIOS, and I looked and found that my current version is 1006, and ASUS's most recent was 1011- however 1011 came from 2012, and I was also told only to update BIOS if I'm sure I need to. Considering this issue never happened before, I do not know if I should bother or not.

The crash has not been consistent. I left my computer on for 16 hours yesterday (most of the time using it) and never had a crash. However today I've had 4 already. I leave it for a bit afterwards, weary if it is heat or anything, but HWMonitor doesn't show anything unusual- in heat, anyways. I'm confused as to what I should see for voltage and clock speed. My 6 cores all regularly run at ~3.7 GHz instead of 3.3GHz the processor says it runs at. I don't know at all what numbers I am to find in the voltages section.

I would appreciate help with this, I am used to computer crashing and trying to find info on the BSOD dump to solve it. I don't know what information can be gleaned from all the event logs, and I don't know if there was a way to see what the temperatures were at when the computer last crashed. After the memtest I will proceed to reseat unless I receive further guidance. And thanks!

Also, this forum absolutely insists that I tag this post with "Blue Screen", even though that's precisely what I don't want- oh, well.
 
Apr 14, 2018
3
0
10
So I gathered more of the logs, still with a wide... berth of other crash times. Sometimes on startup, sometimes on shutdown. A strange change is the the crashes stopped shutting down the computer completely, the screen would go black but the case lights and fans would stay on, then it just started restarting again. All this without a change of settings.

So Memtest came back with multiple passes fine, chkdsk is fine, and I checked the voltages on the PSU with a multimeter (I realize that this doesn't mean too much, but since the crashes weren't under load I figured it meant a bit more). The only thing I didn't do was update the BIOS, and I was worried the power would cut off while flashing.

Once again, I always had the computer running for 16-18 hours a day over the weeks of testing it. But now there is a new issue, and it is, unfortunately, the same issue I was having with the last CPU. I thought the CPU was causing it- and maybe it is, as I may have broken it- but the tower is no longer booting. Powering up, no beep(s), no video. I have taken the RAM out, then took the video card out... and I had it boot (single beep, then many beeps for no RAM) the first time the video card was out, so I shut it off, replaced RAM, and no boots again.

However, I believe I found the culprit: https://imgur.com/gXr3MIn
The PSU is fine- I mean, I tested all of it with a multimeter. But that burnt bit is untestable. I put my finger on the CPU during startup and it didn't heat up at all, also making me assume. That is, unless I fried both CPU's for this machine. Since I can't feasibly replace this port- I would try it if anyone could point me out to a place to test or replace motherboard connectors- I guess I'm looking for a new motherboard.

Some of the symptoms are still original and hard to diagnose, so I hope a new motherboard will solve all of it. I know the PSU isn't causing the booting issues, however the PSU may still have caused the original issues of crashes... but since it wasn't crashes (only) under load, and the PSU passed the multimeter test, I got my hopes way up.
 
Apr 14, 2018
3
0
10
After cleaning the ATX12V (4-pin) contacts with rubbing alcohol and contact cleaner. I saw little improvement when I also saw that one of the contacts on the power supply side seemed to by further into the housing than the others. I took a pair of pliers and wiggled it for a bit, and it popped into place. I honestly did not expect it to work, or I would have to secure it or splice on a new one... but it did. It worked.

Three days later, no crashes. It appears either the corrosion before or after the displaced contact caused the power-off crashes. At least, that's what I'm hoping. Maybe that part of this PC's problems are over.

Edit: I spoke too soon! Augh!

Three days of running and the power crash happened, again. Using Chrome, streaming video, not incredible load, but not empty, either. I'll post again if I see anything too different, but I assume the original issue just isn't fixed. Must have to get another power supply!
 

TRENDING THREADS