Is M.2 Causing System Crashes? Is RAM Unstable?

Status
Not open for further replies.

ericslaff

Prominent
Jan 9, 2018
12
0
510
I just completed my first build and am having a strange issue with machine crashes. The crashes occur at random times during random tasks, anywhere from 10 minutes after startup to 5 hours later. All temps are low (35-45ish), all drivers are up to date, and I'm not overclocking. When the crashes happen the screen freezes for about 15 seconds then just goes black, but all the fans in the case are still spinning. After the crashes the reset button works but the OS never boots. I have to do a full shut down and restart twice before the OS boots, the first shut down just reboots a blank screen without even a BIOS POST. I also noticed that after the crash the HDD light on the case blinks continually. When the PC is working properly the HDD case light rarely blinks.

This is not a gaming rig, it's for work, so it's really important I have no down time.

My specs are:
Ryzen 5 1600X
ASRock AB350M Pro4
G.SKILL TridentZ (2x16GB) F43200C16D-32GTZSK (set to 3200 with MOBO XMP Profile)
ADATA XPG SX7000 (512GB) M.2 2280 PCIe Gen3x4
EVGA GeForce GTX 1050
Corsair 450 watt (CX 450M)
Arctic Alpine 64 Plus CPU cooler
2 TB Western Digital Purple (for backup and archives)
Windows 10 Pro

(Please don't critique my component choices, it's already been bought and built)

Any help or advice you can provide is greatly appreciated!

Thanks!
 
I suggest you read page 24 of the motherboard user manual.
DDR4 UDIMM Maximum Frequency Support

Until you become more adept at OCing, you may have to settle for less aggressive memory speeds. Updating BIOS can help too. You may need to manually set memory parameters instead of using XMP profiles.
 

ericslaff

Prominent
Jan 9, 2018
12
0
510


I did flash the most recent BIOS so I'm all good there. I don't need to OC the RAM, I just figured XMP was a safe quick way to get a little boost. I don't know what RAM instability acts like. Could it really run fine for 5 hours then just crash because of the RAM setting?
 

ericslaff

Prominent
Jan 9, 2018
12
0
510
I went into the BIOS this morning and reset the RAM speed to see if this was the issue. I turned off XMP and set it to Auto (2134mhz). At about 1.5 hours into the day I had another crash. I'll disconnect the Purple drive now and report back if anything crashes, though I'm not sure how that drive would cause any issues, since it just has backup files... no OS, driver, or other run files.

One strange thing happened during the first crash but I just brushed it off as a quirk. When I went into the BIOS somehow the Purple drive was set as the default boot drive and the M.2 drive wasn't listed. Though that quirk didn't happen again with any subsequent crashes.
 

ericslaff

Prominent
Jan 9, 2018
12
0
510


After disconnecting the Purple drive I had another crash about an hour after rebooting. Any other thoughts? Something with Windows 10?
 

ericslaff

Prominent
Jan 9, 2018
12
0
510
One final thought... On my initial boot-up I accidentally discharged a small static shock to the GPU while it was running (like I said, it's my first build). The screen image got weird and had horizontal lines running across the image. After I restarted the PC, the screen image was fine and everything seemed good. The crashes started a day later. Could damage to the GPU be causing my crashing issues or is this another red herring?
 

ericslaff

Prominent
Jan 9, 2018
12
0
510
I noticed one more weird quirk. Twice now I shut down the PC from the OS. It appears that the OS properly shut down, but the power LED stayed on and all fans were still running. It seems like the OS exited but the hardware remained on. Could I have wired something on the MOBO header wrong that is causing the crashes?
 

ericslaff

Prominent
Jan 9, 2018
12
0
510
Yes it did a complete shutdown from the command prompt. So are my crashes and incomplete shutdown something wrong with the OS settings or drivers?

It's now taking 3 hard reboots to get the BIOS POST and OS to load.
 
There's nothing wrong with the way Windows 10 behaves; it hibernates so that boots are faster. The drawback is that it doesn't reinitialize everything.

Your issue probably is the motherboard, but test with another PSU and GPU to make sure. Any older PSU and GPU is fine for that type of test; you only need to make sure the system always POST and boots when powered up.
 

ericslaff

Prominent
Jan 9, 2018
12
0
510


Unfortunately I don't have access to a spare PSU or GPU. I spent all morning uninstalling and reinstalling all the drivers I could think of. Asrock released a new BIOS yesterday that I flashed as well. No improvements so I jumped the CMOS. It crashed again so I plugged in the MOBO buzzer. I got an error code with 5 short beeps, a pause, then 5 more short beeps. Is this a 5 beep error code or a 10 beep error code?
 

ericslaff

Prominent
Jan 9, 2018
12
0
510


I bought a new PSU (it's an easier fix than MOBO). Since this is a work PC I need immediate uptime, so buying a PSU made more sense than returning, shipping, and waiting for a replacement. I got a Thermaltake Smart 600W (over power for my needs but all that was available last minute). I ran some benchmarks and stress tests and so far no crashes... However, I did restart my PC and I got the same beep code error. After doing a hard reboot I got a new beep code, 3 short beeps. I did another hard reboot and it went to BIOS POST then to OS as normal. I won't know if the new PSU solved the crash issue until I run it for a few more hours, but the beep codes are concerning. I'm also not clear on how the brand new 450W PSU could have been causing the crashes. I'll update if anything changes.

BTW, thanks for the support and quick responses.
 

ericslaff

Prominent
Jan 9, 2018
12
0
510


I'm bringing it to a shop today. I just spent so much time and money that it's tough to "give up". I was hoping I could finish the project.
 
Status
Not open for further replies.