Random crashes even on bios (no error message)

[Luispereira]

Hey guys, about a week ago I finished my build and it worked fine for the first two/three days but the last 3 days it just crashes all the time. On widows there is no errors apart from the unexpected shutdown. Ive already builded from scratch one secound time but still the same. The pc crashes in random situations so can even be on idle, in the bios, while using Firefox...i got the bios runing on factory settings since i had turned de xmp on. And even restored windows to a previous date.

 

Some info on the crashes (3 types):1st the pc just turns off and stays off until I unplug it press the power button and then plug again.

2nd the pc shuts down and reboots.

3th the pc shuts down and the vga/boot/ram motherboard red ezdebug leds stay on while cpu led blinks red. This goes on for a bit until it reboots or sometimes I have to shut it down manually.

 

Cpu and gpu run cool and don't overheat. I'm running a memtest and so far no problems. So I'm really scratching my head has to where this is coming from.

 

Pc parts:

Ryzen 7 3700x

Asus Dual GeForce RTX2070 super evo 8gb oc

Msi b450 tomahawk max

Ram g.skill ripjaws v 1x16gb ddr4 3200 cl16

Psu corsair rm550x 80plus gold 

Phanteks eclipse p400s

Ssd m2 2280 WD blue SN550 500gb 3d nand nvme

Barracuda 2t 7200rpm

 

Edit: in the first days i gamed on the for hours, even on 4k with no problems. Bios is updated to latest version and all the drivers are updated 

 

[Viper9]

Looks like a memory issue, create a memtest86 usb drive and test your RAM: https://www.memtest86.com/downloads/memtest86-usb.zip

[Luispereira]

 

5 minutes ago, Viper9 said:

Looks like a memory issue, create a memtest86 usb drive and test your RAM: https://www.memtest86.com/downloads/memtest86-usb.zip

Im already running that test and so far no erros but the problem is the system wont stay on long enough to finish the test

[Tristerin]

BIOS up to date?  If not, and PBO is enabled, you could be pulling far to much voltage through that chip.  

 

Turn off PBO, still have issues?

[Luispereira]

Just now, Tristerin said:

BIOS up to date?  If not, and PBO is enabled, you could be pulling far to much voltage through that chip.  

 

Turn off PBO, still have issues?

Yes bios is updated. Will try the pbo thing

[Viper9]

6 minutes ago, Luispereira said:

 

Im already running that test and so far no erros but the problem is the system wont stay on long enough to finish the test

In that case, could be a power issue. Can you test with another PSU? 

[Luispereira]

Although shouldn't it handle pbo on default? Since its the fabric setting it should not cause all these crashes, will try anyway and see.

[Luispereira]

1 minute ago, Viper9 said:

In that case, could be a power issue. Can you test with another PSU? 

I dont have another psu, had thinked about that but on idle im way under the 550w and the thing is the pc hasn't crashed while doing stress test and benchs so far, that would be the point of failure for the psu right?

[Luispereira]

21 minutes ago, BlueScope819 said:

Sometimes, that isn't the case. It can just be generally unstable. It could also be a thermals issue. Try and make sure your cooler is fully seated on your CPU.

Yes i have double checked and it is, thr cpu and gpu are running cool and under load never go to any extremes

[Luispereira]

58 minutes ago, Tristerin said:

BIOS up to date?  If not, and PBO is enabled, you could be pulling far to much voltage through that chip.  

 

Turn off PBO, still have issues?

Still crashing with pbo off

[Tristerin]

5 minutes ago, Luispereira said:

Still crashing with pbo off

Whats voltage showing in CPU-Z (main tab) during any sort of CPU stress test?  Just trying to eliminate voltage on CPU from the equation.

[Tristerin]

I just noticed you are running a single DIMM (if you don't know about single vs dual channel with Ryzen you are gimping your CPU with single channel config you are in) - is it in the correct slot in the mobo for a single DIMM?  I.E. on my B450 its the second slot away from the CPU for 1 stick, second and fourth for 2 sticks (dual channel)

[Luispereira]

8 minutes ago, Tristerin said:

Whats voltage showing in CPU-Z (main tab) during any sort of CPU stress test?  Just trying to eliminate voltage on CPU from the equation.

Core voltage 1.34v at 4023mhz with the cpu-z stress test.

[Luispereira]

7 minutes ago, Tristerin said:

I just noticed you are running a single DIMM (if you don't know about single vs dual channel with Ryzen you are gimping your CPU with single channel config you are in) - is it in the correct slot in the mobo for a single DIMM?  I.E. on my B450 its the second slot away from the CPU for 1 stick, second and fourth for 2 sticks (dual channel)

It is on the secound slot from the cpu yes, didnt knew about the problem with single but will read about it

[Luispereira]

This is what happens when it crashes...it stays like this until it manages to reboot 

[Tristerin]

Pull CMOS battery for a little over 30 seconds and reset BIOS, do not boot up with XMP, still crash?

[Luispereira]

17 minutes ago, Tristerin said:

Pull CMOS battery for a little over 30 seconds and reset BIOS, do not boot up with XMP, still crash?

Already have done that and the xmp has been turned off for long, still crashing.

[Tristerin]

11 minutes ago, Luispereira said:

Already have done that and the xmp has been turned off for long, still crashing.

Is this a fresh install of Windows 10?

[Luispereira]

Just now, Tristerin said:

Is this a fresh install of Windows 10?

Yes its a need build with just a week

[Tristerin]

Just now, Luispereira said:

Yes its a need build with just a week

10-4 just confirming it wasn't a drive migrated from another build  

 

At this point I would breadboard the PC on the motherboard box or similar to isolate the piece of hardware that is the issue.

 

Read this, do it, report back:

 

https://forums.tomshardware.com/faq/breadboarding-stripping-it-down-to-the-basics-for-troubleshooting.1848255/

[shark103]

1 hour ago, Luispereira said:

This is what happens when it crashes...it stays like this until it manages to reboot 

That video kinda suggest to me 1 thing with two possible reasons.
1. PSU is faulty but in a weird way (like one power line is started to degenerate) that's why this kind of things happen
2. Long shot, but I've had this weird case on one mobo 2-3 years ago. Do you use the same peripherals (mouse/keyboard) all the time on this machine while testing- I've had an  issue in the past where keyboard/mouse would short out mobo usb port ever so slightly that bsods happened or just straight powerdowns

[aDoomGuy]

1 hour ago, Luispereira said:

This is what happens when it crashes...it stays like this until it manages to reboot 

Judging by the sound it seems to me there is some power failure maybe power drop or something, I can't be sure though just what it seems like with that sound... In specifications for 2070 Super it says that 650W is minimum and I'm not sure the 550W is enough despite that they have margins. Though you said it worked for days so obviously it may not be the issue..

 

I assume you have thoroughly checked all PSU cables inside the PC, that they are connected and not damaged. Check the power cord going to the wall socket for damage as well or try another one if you have one. Also try connecting it directly to a wall socket just to rule out some faulty cables.

 

Have you looked at the eventlog in Windows for clues? How long does it take for the issue to occur when the system is idle at desktop?

[Luispereira]

9 minutes ago, aDoomGuy said:

Judging by the sound it seems to me there is some power failure maybe power drop or something, I can't be sure though just what it seems like with that sound... In specifications for 2070 Super it says that 650W is minimum and I'm not sure the 550W is enough despite that they have margins. Though you said it worked for days so obviously it may not be the issue..

 

I assume you have thoroughly checked all PSU cables inside the PC, that they are connected and not damaged. Check the power cord going to the wall socket for damage as well or try another one if you have one. Also try connecting it directly to a wall socket just to rule out some faulty cables.

 

Have you looked at the eventlog in Windows for clues? How long does it take for the issue to occur when the system is idle at desktop?

Its really random, can take hours after log in or can be just seconds...the event just says that there was an unexpected shutdown and nothing more...im breadboarding right know to try and find the problem

[aDoomGuy]

Just now, Luispereira said:

Its really random, can take hours after log in or can be just seconds...the event just says that there was an unexpected shutdown and nothing more...im breadboarding right know to try and find the problem

Alright. When you are done and if issue still persists, I would advice you to hit the system with load from 3DMark or Superposition or similar and see what happens.. But I'm not sure if you will blow something up...  However if you do that and it insta-crashes as soon there is load on CPU and GPU I think it's safe to say it is power related.

[Luispereira]

12 minutes ago, aDoomGuy said:

Alright. When you are done and if issue still persists, I would advice you to hit the system with load from 3DMark or Superposition or similar and see what happens.. But I'm not sure if you will blow something up...  However if you do that and it insta-crashes as soon there is load on CPU and GPU I think it's safe to say it is power related.

I've done a couple stress tests and played in 4k for several hours without any problem, the strange thing is that it crashes more easily on idle, while I'm on YouTube, ...