PC reboots randomly, no blue screen, no memory dump

[floofyfloof]

Hello,

I started having troubles with my newly built PC (around 3 months old). Since last week it started rebooting randomly, without giving a blue screen. Inspecting the Windows Event Viewer bears no useful information, if not for a message stating that creating a memory dump wasn't possible, so no luck there.

The problem occurs after the PC has been on for a while and then the intervals get progressively shorter. Leaving it off "resets" the situation. This screams temperature issues, but I've been monitoring them and the are *way* within accepted values. Plus, this happens even when the PC is idle, whereas in the past (literally days before the problem showed up) I could play demanding videogames on it for hours on end without an issue.

Here's what I tried to do so far, none of this helped:

* Updated graphical card drivers
* Swapped PSU
* Unplugged storage HDD (which I deemed more prone to faultiness)
* Swapped RAM and tried every slot on the motherboard
* Disabled X.M.P. from the BIOS
* Updated the BIOS from F22 to F24
* Left it idle inside the BIOS and have it shutdown even there, which I think can make me rule out any kind of OS/software/hard disk issue
[EDIT]
* Updated the chipset drivers
* Unplugged the case reset button cable from the motherboard (as per this thread I dug up)

At this point I'm pretty clueless, I think there might be some problem with the motherboard itself, but I'd like some advice for my next steps. I want to state once again that this kind of problem never happened before (that I know of) but started on a very precise day and has persisted ever since.

Below are my system specifications (slashes mark the hardware that I swapped for testing purposes)

* OS: Windows 10
* Mobo: Gigabyte GA-AB350M-DS3H
* CPU: AMD Ryzen 5 1600 3.2GHz (stock cooler)
* RAM: HyperX Predator 8GB 3000MHz DDR4 / Ballistix Sport 8GB 240 MHz DDR4
* GPU: Palit GeForce GTX 1060 stormx 6 GB
* PSU: Thermaltake Smart SE 530W / Seasonic FOCUS plus platinum 650W
* SSD: Crucial MX500 250 GB
* HHD: Seagate Barracuda 2TB

Except for turning on X.M.P. (which I disabled) I am not doing any kind of overclocking

Thank you all very much for your help!

 

[R_1]

My canned random reboot Rant
Random reboots are usually caused by the PSU, the RAM or software AND in that order of likely-hood.
PSU - If you can borrow/swap a PSU for testing. sibling/friend you can swap out the PSU and each system will be testing the other.
RAM - run the system with one stick of RAM see if stability returns if not Test all the RAM with memtest 86 for three passes or overnight. if you get no 0 errors after more than three passes the ram is good. with the random reboots I would suggest running this test after the PSU swap or after the PSU has been cleared.
Software - Drivers or other issues can cause reboots. Boot to a linux distro on a USB drive. mint linux will boot to memtest86. you can run the OS from the USB and await reboot.
end canned rant

you tried different RAM and PSU so that leaves software. was there a Windows update installed around the time it started going nuts? have you tried a restore to a working system state?

 

[terry4536]

The first thing I would look at is the installed update history. If the updates coincide, with your restarts then that is probably the culprit. If it is then you can Windows to schedule the restarts at night for example.

If it isn't update related, you can try updating the device drivers including the motherboard chipset drivers. You already swapped power supplies.

You can try updating the motherboard BIOS, and lastly reinstalling Windows.

 

[schwatzz]

CMOS battery good be dead or the PSU is damaged. Are you plugged into a surge protector?

 

[floofyfloof]

My canned random reboot Rant
Random reboots are usually caused by the PSU, the RAM or software AND in that order of likely-hood.
PSU - If you can borrow/swap a PSU for testing. sibling/friend you can swap out the PSU and each system will be testing the other.
RAM - run the system with one stick of RAM see if stability returns if not Test all the RAM with memtest 86 for three passes or overnight. if you get no 0 errors after more than three passes the ram is good. with the random reboots I would suggest running this test after the PSU swap or after the PSU has been cleared.
Software - Drivers or other issues can cause reboots. Boot to a linux distro on a USB drive. mint linux will boot to memtest86. you can run the OS from the USB and await reboot.
end canned rant

you tried different RAM and PSU so that leaves software. was there a Windows update installed around the time it started going nuts? have you tried a restore to a working system state?

I have not, but at this point it's worth a try. However, it even reboots inside the BIOS (turn the PC on, enter the BIOS, let it idle, still reboots). Shouldn't this rule out software/OS issues?

The first thing I would look at is the installed update history. If the updates coincide, with your restarts then that is probably the culprit. If it is then you can Windows to schedule the restarts at night for example.

If it isn't update related, you can try updating the device drivers including the motherboard chipset drivers. You already swapped power supplies.

You can try updating the motherboard BIOS, and lastly reinstalling Windows.

I have updated the BIOS but the problem persists, will try updating chipset drivers too. I will try a Windows reinstall, but as I asked above couldn't I rule out a Windows problem since it also happens inside the BIOS itself?

CMOS battery good be dead or the PSU is damaged. Are you plugged into a surge protector?

Mmmh I am not. I have tried two PSUs though and the problem is still there. I suppose there's no way to check the CMOS battery other than replacing it, right?

 

[terry4536]

My canned random reboot Rant
Random reboots are usually caused by the PSU, the RAM or software AND in that order of likely-hood.
PSU - If you can borrow/swap a PSU for testing. sibling/friend you can swap out the PSU and each system will be testing the other.
RAM - run the system with one stick of RAM see if stability returns if not Test all the RAM with memtest 86 for three passes or overnight. if you get no 0 errors after more than three passes the ram is good. with the random reboots I would suggest running this test after the PSU swap or after the PSU has been cleared.
Software - Drivers or other issues can cause reboots. Boot to a linux distro on a USB drive. mint linux will boot to memtest86. you can run the OS from the USB and await reboot.
end canned rant

you tried different RAM and PSU so that leaves software. was there a Windows update installed around the time it started going nuts? have you tried a restore to a working system state?

I have not, but at this point it's worth a try. However, it even reboots inside the BIOS (turn the PC on, enter the BIOS, let it idle, still reboots). Shouldn't this rule out software/OS issues?

The first thing I would look at is the installed update history. If the updates coincide, with your restarts then that is probably the culprit. If it is then you can Windows to schedule the restarts at night for example.

If it isn't update related, you can try updating the device drivers including the motherboard chipset drivers. You already swapped power supplies.

You can try updating the motherboard BIOS, and lastly reinstalling Windows.

I have updated the BIOS but the problem persists, will try updating chipset drivers too. I will try a Windows reinstall, but as I asked above couldn't I rule out a Windows problem since it also happens inside the BIOS itself?

CMOS battery good be dead or the PSU is damaged. Are you plugged into a surge protector?

Mmmh I am not. I have tried two PSUs though and the problem is still there. I suppose there's no way to check the CMOS battery other than replacing it, right?

Incompatible drivers can affect the system at the BIOS level too. For example if the motherboard drivers are corrupted, it could affect the BIOS's ability to find the system drives. It could affect the ability for the motherboard to interact with the BIOS itself.

 

[floofyfloof]

Incompatible drivers can affect the system at the BIOS level too. For example if the motherboard drivers are corrupted, it could affect the BIOS's ability to find the system drives. It could affect the ability for the motherboard to interact with the BIOS itself.

Got it! Will try updating that too.

 

[terry4536]

With a system only 3months old, I doubt that it is the CMOS battery. But it is always possible that it is a dud.

You can check it with a multimeter. You can also check it in HWmonitor. The 3VCC voltage should be around 3.3V

 

[floofyfloof]

was there a Windows update installed around the time it started going nuts? have you tried a restore to a working system state?

The first thing I would look at is the installed update history. If the updates coincide, with your restarts then that is probably the culprit. If it is then you can Windows to schedule the restarts at night for example.

For a moment I thought I had it... I DO have an update installed on the day that the problem started showing, but comparing the *times* of installation (had to use a PowerShell script to get that...) vs the errors from the Event Viewer shows that things don't add up: the update happened *after* the first problem by a margin of several hours, so I don't think that could be the culprit?

With a system only 3months old, I doubt that it is the CMOS battery. But it is always possible that it is a dud.

You can check it with a multimeter. You can also check it in HWmonitor. The 3VCC voltage should be around 3.3V

I think my voltages should be in check. Using HWiNFO64 what i get is

Vcore: 1.284V
+3.3V: 3.344V
+12V: 12.168V
+5V: 5.070V

 

[terry4536]

In HWmonitor, at the top are the voltages for the motherboard. In my version, the "3VCC" is the fourth voltage from the top of that list.

The voltages that you listed are the voltage aren't for the CMOS battery.

 

[jhughey]

Hello floofyfloof,

If all else fails, do more diagnostics with the GPU. It could be failing for reasons other than temperature.

I had a problem with random reboots on my HP HDX 18T Laptop with W7 and a NVIDIA GeForce GT 130M GPU. When I disabled the GPU device drivers or went into Safe Mode, the problem went away.


Joe

 

[floofyfloof]

In HWmonitor, at the top are the voltages for the motherboard. In my version, the "3VCC" is the fourth voltage from the top of that list.

The voltages that you listed are the voltage aren't for the CMOS battery.

This... is really bizzarre. I can't seem to find the 3.3V, etc readings in HWmonitor. Could it be due to some sort of incompatibility with my motherboard? (similar situation in this thread).

HWmonitor readings

However I get this with HWiNFO64

HWiNFO64 readings

are those the ones I should be looking at?

 

[floofyfloof]

Hello floofyfloof,

Do more diagnostics with the GPU. It could be failing for reasons other than temperature.

I had a similar random reboot problem with my HP HDX-18T Laptop with W7. When I disabled my NVIDIA GeForce GTX 130M Graphics or went into Safe Mode (which did the same), the problem went away.

Unfortunately, the graphics card was soldered to the motherboard. After a futile effort to locate a replacement motherboard, I was lucky to locate an almost identical HDX-18T and transferred my O/S SSD and Data HD to it.

Joe

Hi Joe,

what do you suggest? Should I be trying some sort of stress test and see if that coincides with crashes? Would I be at risk of (further) damaging the GPU if that were the case?

Since my CPU doesn't have an integrated graphics processor I can't really unplug/disable my GPU to see if that's the problem. I guess I could come by a spare one eventually but I'd prefer to not go that way if it's not completely necessary.

 

[terry4536]

In HWiNFO64, the CMOS battery voltage is listed as VBAT. On your link it is listed as 3.216 V . It is a bit low. I've read to replace it if it is below 3.2V.

As to why HWmonitor isn't reporting it, I would assume that the motherboard isn't fully compatible with HWmonitor. Or perhaps there is some variation between motherboard manufacturers.

My version is 2018 Version 1.35.0, driver version 146 .

 

[jhughey]

Hello floofyfloof,

Do more diagnostics with the GPU. It could be failing for reasons other than temperature.

I had a similar random reboot problem with my HP HDX-18T Laptop with W7. When I disabled my NVIDIA GeForce GTX 130M Graphics or went into Safe Mode (which did the same), the problem went away.

Unfortunately, the graphics card was soldered to the motherboard. After a futile effort to locate a replacement motherboard, I was lucky to locate an almost identical HDX-18T and transferred my O/S SSD and Data HD to it.

Joe

Hi Joe,

what do you suggest? Should I be trying some sort of stress test and see if that coincides with crashes? Would I be at risk of (further) damaging the GPU if that were the case?

Since my CPU doesn't have an integrated graphics processor I can't really unplug/disable my GPU to see if that's the problem. I guess I could come by a spare one eventually but I'd prefer to not go that way if it's not completely necessary.

No, mine was just a simple-minded suggestion. If you do not have alternative graphics to fall back on, it won't work.

 

[floofyfloof]

Tried updating the chipset drivers and disconnecting the case reset cable on the motherboard (it was mentioned in this thread, and as strange as it sounds I guess it couldn't hurt) and still, no tangible improvement.

In HWiNFO64, the CMOS battery voltage is listed as VBAT. On your link it is listed as 3.216 V . It is a bit low. I've read to replace it if it is below 3.2V.

As to why HWmonitor isn't reporting it, I would assume that the motherboard isn't fully compatible with HWmonitor. Or perhaps there is some variation between motherboard manufacturers.

My version is 2018 Version 1.35.0, driver version 146 .

I understand, I'm just a bit wary about identifying the CMOS battery as the culprit because I have never experienced any kind of datetime/BIOS settings loss (I have been inside the BIOS an awful lot lately and I would have noticed it ^^") and I assume those should be the first symptoms of a faulty battery.

 

[terry4536]

Tried updating the chipset drivers and disconnecting the case reset cable on the motherboard (it was mentioned in this thread, and as strange as it sounds I guess it couldn't hurt) and still, no tangible improvement.

In HWiNFO64, the CMOS battery voltage is listed as VBAT. On your link it is listed as 3.216 V . It is a bit low. I've read to replace it if it is below 3.2V.

As to why HWmonitor isn't reporting it, I would assume that the motherboard isn't fully compatible with HWmonitor. Or perhaps there is some variation between motherboard manufacturers.

My version is 2018 Version 1.35.0, driver version 146 .

I understand, I'm just a bit wary about identifying the CMOS battery as the culprit because I have never experienced any kind of datetime/BIOS settings loss (I have been inside the BIOS an awful lot lately and I would have noticed it ^^") and I assume those should be the first symptoms of a faulty battery.

You may not notice the time loss. Most modern motherboards are set to update the time from the internet now. Before the internet you ha to set the time on the computer yourself just like a wall clock. Then you would notice the time being off.

When I've run into this, you loose the BIOS settings. For instance your memory settings or your CPU overclock will be gone every time you unplug the computer. But it can cause other problems before it gets to that stage too.

 

[terry4536]

On the flip side, a replacement lithium battery (most likely a CR2032, but check your manual) is only about $3.

But my best guess is a driver update, BIOS update, and fresh install of the operating system. After that replacing the motherboard is the best route (and possibly the power supply).

 

[R_1]

if you live near a dollar store, you can get a CR2032 3.3V lithium there if available

 

[floofyfloof]

On the flip side, a replacement lithium battery (most likely a CR2032, but check your manual) is only about $3.

But my best guess is a driver update, BIOS update, and fresh install of the operating system. After that replacing the motherboard is the best route (and possibly the power supply).

if you live near a dollar store, you can get a CR2032 3.3V lithium there if available

I will try swapping the CMOS battery and clean installing the OS, I already updated the BIOS and drivers but to no avail.

However, something weird happened last night: basically I've had to crashes/reboots that happened *precisely* as I unplugged and plugged my monitor in. These two crashes were ~2 hours apart and I noticed only after (was connecting the monitor to another laptop so I couldn't notice the first crash). After noticing that I quickly unplugged and plugged my monitor again... lo and behold! Another reboot. This happened approximately 5 minutes after the last one, something that hasn't happened before. However, after these 3 very timely crashes I haven't been able to replicate.

Still, this seems a very odd coincidence and since I managed to come by a spare GPU I'll try testing that first.

It still rings very weird to me: am I right in assuming that it could point to a GPU issue or another clue that it is power/mobo related? I am no technician but I assume that plugging and unplugging the monitor while the PC is running could create a power surge and mess things up a bit.

 

[terry4536]

On the flip side, a replacement lithium battery (most likely a CR2032, but check your manual) is only about $3.

But my best guess is a driver update, BIOS update, and fresh install of the operating system. After that replacing the motherboard is the best route (and possibly the power supply).

if you live near a dollar store, you can get a CR2032 3.3V lithium there if available

I will try swapping the CMOS battery and clean installing the OS, I already updated the BIOS and drivers but to no avail.

However, something weird happened last night: basically I've had to crashes/reboots that happened *precisely* as I unplugged and plugged my monitor in. These two crashes were ~2 hours apart and I noticed only after (was connecting the monitor to another laptop so I couldn't notice the first crash). After noticing that I quickly unplugged and plugged my monitor again... lo and behold! Another reboot. This happened approximately 5 minutes after the last one, something that hasn't happened before. However, after these 3 very timely crashes I haven't been able to replicate.

Still, this seems a very odd coincidence and since I managed to come by a spare GPU I'll try testing that first.

It still rings very weird to me: am I right in assuming that it could point to a GPU issue or another clue that it is power/mobo related? I am no technician but I assume that plugging and unplugging the monitor while the PC is running could create a power surge and mess things up a bit.

That would indicate to me that the power supply is detecting a fault at those occasions. It probably isn't the best practice to unplug the monitor without shutting the system down, but it doesn't usually cause the system to crash. Generally the only thing that happens is that you get a notification that a device was added or removed.

But your system isn't normal right now. It is unstable, so perhaps it doesn't take much to trigger a crash.

But you could switch out the power supply or check the cables again.

 

[terry4536]

Look for anything that may be grounding the motherboard to the case. You could try breadboarding the motherboard to isolate any grounding issues and see if the system stabilizes.

If you want to try that here are some breadboarding instructions..

http://www.tomshardware.com/faq/id-2176482/breadboarding-stripping-basics-troubleshooting.html

http://www.tomshardware.com/forum/262730-31-breadboarding

http://www.tomshardware.com/faq/id-1753671/bench-troubleshooting.html

Read through them and pick one to follow. This is generally used to isolate the problem (generally the motherboard).

 

[floofyfloof]

Look for anything that may be grounding the motherboard to the case. You could try breadboarding the motherboard to isolate any grounding issues and see if the system stabilizes.

If you want to try that here are some breadboarding instructions..

http://www.tomshardware.com/faq/id-2176482/breadboarding-stripping-basics-troubleshooting.html

http://www.tomshardware.com/forum/262730-31-breadboarding

http://www.tomshardware.com/faq/id-1753671/bench-troubleshooting.html

Read through them and pick one to follow. This is generally used to isolate the problem (generally the motherboard).

Thank you very much! It definitely is very interesting and useful information! Will read through these threads