PC random freezing problem

[arnes91]

What happens?: PC randomly freezes, the screen freezes for 2-3 seconds, then the monitors turn off and also the lights on the keyboard and mouse turn off and become unresponsive, but the computer still has all fans and lights on. Only way out of it is a hard turn off via the power button, after which it works normally again.

Problem description with details: I bought this computer NEW in July this Year (2020) and as far as I remember the issue is there from the beggining. The freezing happens randomly, mostly once every 3-5 days but sometimes it happens 2 times in 2 days. The problem never happens when I am playing games or doing some high demanding tasks. It happens only on low-demanding tasks, such as web browsing, using my e-mail app, on desktop, on idle or the last couple of times right after a cold boot, sometimes it freezes as soon as it starts transitioning to the desktop, and sometimes after 2-3 minutes. When it happens, after I turn the PC off and on again, it works again normally, usually for a few days, until it happens again. It also never happened when stress testing components such as the CPU or GPU and is impossible to trigger when I want it to happen.

At first it did not bother me much because it was happening so rarelly and I was thinking that it is just a soft software issue which I will find the cause and solution fast, but after trying many things to fix it and it stayed persistent for months, I got worried.

My PC specifications:

CPU: Ryzen 7 2700 8 CPU Cores (16 Threads) @ 3.2GHz up to 4.1GHz (stock without any OC)

GPU: Radeon RX570 xfx 8GB

Motherboard: Asus Prime A320M-K (with latest BIOS update)

RAM: 16GB DDR4 Kingston HyperX Fury 3200MHz (dual channel 2x8GB) (CPU-Z Shows:KHX3200C16D4/8GX – This is written on the RAM HX432C16FB3/8) – using XMP/D.O.C.P. profile 1 for 3200MHz (without it, it runs on 2400MHz)

Storage: 500GB Samsung 860 Evo 2.5"(bought brand new) + 120GB Samsung 840 Evo 2.5" (used in old PC without issue) + 120GB TeamGroup 2,5" (came new with the PC)

PSU: Inter-Tech Argus 620W

Peripherals:

Mouse: HyperX Pulsfire FPS Pro RGB (which I used on old PC without issues)

Keyboard: HyperX Alloy FPS Pro Mechanical Gaming Keyboard (which I used on old PC without issues)

Headphones: Standard, cheap headphones over USB (was using HyperX Cloud II but sold – details bellow)

Printer: HP LaserJet MFP M28-M31 (which I used on old PC without issues)

Monitors: Fujitsu B23T-6 LED – using 3 such monitors all connected via DisplayPort, which I bought used when I got the new PC. Sound output is tranfered via DP to the monitors.

Case: NaviaTec Master which has it's front case USB and Audio connected to the MB. It also has a front case LED RGB „line“ which is connected to the reset button of the case (to change the color and style)

OS:

Edition:                 Windows 10 Pro

Version:               20H2

OS build:              19042.610

Experience:        Windows Feature Experience Pack 120.2212.31.0

With all updates installed also all drivers installed with latest official versions. I have tried using it also with just the standard drivers which windows installs, but the same issue happens no matter which drivers I use.

 

Hardware fixes which I tried/Replacements of parts:

 

GPU: RMA the GPU and they replaced it for another GPU – same model

Motherboard: At first it was MSI A320M which I RMA and they replaced it for a new ASUS Prime A320M-K

RAM: At first it was T-Force 2x8GB of 3200MHz (with XMP enabled), after RMA they replaced it with HyperX Fury 2x8GB 3200MHz (with D.O.C.P. enabled)

Storage: At first it was only 120GB TeamGroup 2,5" (came new with the PC) on which the system was installed and had the issue, after it I bought a new Samsung 860 Evo and reinstalled the system on it, again getting the same issue

PSU: At first it was Inter-Tech Energon 650W after RMA they replaced it with an Inter-Tech Argus 620W

 

Tried plugging in the PC in another power socket to rule it out and the issue still persisted.

 

Until now they have replaced all the parts except the case and the CPU. They, and me also, have stress tested the CPU with Prime95 Stress test, furmark stress test, and also memtest which run about 3 and a half hours without any issues or errors.

 

Software fixes and other things which I tried until yet without any success:

 

-          Reinstalled the system multiple times

-          Reinstalled and tried different drivers

-          Updated the BIOS

-          Tried using high performance in windows power settings

-          Run DISM and sfc/ scannow comands in CMD

-          Turned off Windows automatic driver installs

-          Turned off all power saving options in Windows such as sleep, hibernation ...

 

BIOS fixes and settings which I tried without any success:

-          Turned off all boot devices except the primary SSD

-          Turned off Global C State

-          Changed Power Supply Idle Control from Auto/Low Current Idle to Typical Current Idle

-          Turned off AMD Precision Boost Overdrive

-          Turned off AMD Core Boost

 

Additional information:

At the beginning I was using HyperX Cloud 2 headset which has its own USB card. After searching for the cause of the issue I quickly saw in event viewer that there is a „HID-Compliant Headset“ driver CRITICAL error and everytime the problem happened the first boot after was slower and those drivers were reinstalling. I googled and found that many users of this headsets have similar issues and it seems to be some kind of driver conflict between them and the High Definition Audio over the DP/HDMI. I solt the headset and reinstalled the system, but the issue was not solved. Tried reinstalling the system to different SSDs but got the same issue. Tried various drivers and still the same. Windows is updated and tried with older and newer driver versions.

 

Now, that I do not use anymore the HyperX headphones there is no specific error in the event viewer. There are the standard warnings, and sometimes, very rarely some error and on the time it happens I just see errors because I did not shut off the computer properly.

I also saw that there was a fast startup error after the issue happens, so I turned of fast startup in windows yesterday, waiting to see if it may be solve the problem? Additionaly I found a post on a forum with a person with the same issue like I have and he says that after using a program called driver booster and updating the drivers with it, the issue never came back. I downloaded and run the program and it found some updates for drivers which were shown installed as 2005-2008, and it updated them to some recent versions. The drivers which it updated were network, printer, USB Mass storage device, HID Keyboard, i 2 times Standard SATA AHCI Controller.

Please find attached the links to the Sysnative BSOD file collection dump/log with also the PERFMON dump/log:
https://drive.google.com/drive/folders/1AtLT39BaYkZAsQnrDCAFf-icOI_axPtk?usp=sharing

[Bombastinator]

Howdy and welcome

I got no really good answers  best I can give you is some more ideas for things to check.  Hopefully someone more useful than me will be along.  They exist.  Have patience

 

the generic most common causes of crash after boot are memory errors or PSU capacity issues.  Considering your hardware 620w should commonly be enough to avoid PSU power stuff but I’ve never heard of that PSU brand.  GPU RMA would have taken care of vram, so that leaves ram.  Ram can be tested for no cost though it is time consuming.  
 

I notice you’ve got an A320 series motherboard though, and those are known to not only sometimes have power issues with larger chips (you’ve got an 8core) but sometimes have issues with fast memory.   
 

There are just a lot of possibles here.  Do you still have the problems when running memory at default speed?  I don’t know whether or not your cpu draws too much for your motherboard or not (perhaps worth checking) 

 

If you read the pinned threads in the troubleshooting section (always a good idea) one of them talks about “dumps” which can apparently give info about exactly happens.  I can’t read them and don’t know how they work, but there are those here who can.

[arnes91]

Hello, thank you for your post

Yeah, the PSU is not a really good one, but it was changed, also the RAM, so there shouldnt be the issue also. I will check out the guides on how to create those dumps which may help solving my issue.

I have also not tested how it works without the XMP/DOCP, because I payed extra for 3200MHz and also want to take full advantage of the CPU, so clocking the RAM down would not be ideally, but I will test it anyway. I have done a full memtest with 3200MHz without any issue.

For now, I am waiting to see if it will happen again after turning off windows fast startup and updating the drivers with driver booster.

I have a checklist of things to try additionally, but I have to go 1 by 1 in order to be sure what caused the issue if I ever find the culprit.

My next steps are:

Load BIOS defaults without XMP/DOCP so RAM runs at lowest clocks
Try without the TeamGroup SSD and use only 1 samsung SSD
Try with a completely different monitor and just 1

The issue is that it tages months to troubleshoot as the issue happens rarely, only once or twice in a week so I have to wait for a few days after every step I take.

[Bombastinator]

Looking at the motherboard VRM tierlist 

your motherboard appears (though I may be wrong) to appear in tier F

[Jurrunio]

2 hours ago, arnes91 said:

  Reinstalled and tried different drivers

including chipset driver?

[arnes91]

13 minutes ago, Bombastinator said:

Looking at the motherboard VRM tierlist 

your motherboard appears (though I may be wrong) to appear in tier F

I can not find it, but I know that it is an entry level. First it was MSI A320M then they changed it for ASUS prime A320M-K. They are also selling the same configurations as mine and did not have issues reported.

2 minutes ago, Jurrunio said:

including chipset driver?

Yes, downloaded it from the motherboard website.

[Jurrunio]

6 minutes ago, arnes91 said:

Yes, downloaded it from the motherboard website.

Is it the same version as the latest one from AMD's site? sometimes drivers from the motherboard site isnt up to date

[Bombastinator]

6 minutes ago, arnes91 said:

I can not find it, but I know that it is an entry level. First it was MSI A320M then they changed it for ASUS prime A320M-K. They are also selling the same configurations as mine and did not have issues reported.

Yes, downloaded it from the motherboard website.

I saw a320m-k in tier f but with no manufacturer associated with it.  Hard to know.  As a general rule a320 is pretty iffy for 8 core chips in general. 6 core is not infrequently the highest that can safely push. My knowledge is limited though

[Guest]

Quote

I have also not tested how it works without the XMP/DOCP,

Would start diagnostics by doing the opposite of this. 

Set all defaults and or clear the cmos. Test it at stock.

 

Then come to a conclusion if the issue still persists, other diagnostic ideas will come to mind. 

 

 

[Guest]

9 minutes ago, Bombastinator said:

I saw a320m-k in tier f but with no manufacturer associated with it.  Hard to know.  As a general rule a320 is pretty iffy for 8 core chips in general. 6 core is not infrequently the highest that can safely push. My knowledge is limited though

You know what was funny Bomb? I see a lot of talk about VRMs and the importance, yada yada..... NewEgg Shipped my B450M-A bundled with 2700X 105w 8 core. 4+1 and big power draw. (also with a set of corsair LPX in the bundle, about 500 bucks couple years ago). 

AnyHoot, I ran the board under a TEC with the 2700X lid-less. Had the chip over 1.5v also.

Testing against B450-I which has 6 IR voltage regulators, I could not see a noticeable difference in cpu performance between the two boards. The only thing that played a role at all defaults where the sub ambient and sub zero temps. 

Later.... The B450M-A is now a daily office PC for my wife and I. It now runs a passively cooled Ryzen Athlon at 3.7ghz. 

Came across an issue, the storage drive kept falling asleep freezing windows for a short period of time. Weird as hell. 

[arnes91]

12 minutes ago, Jurrunio said:

Is it the same version as the latest one from AMD's site? sometimes drivers from the motherboard site isnt up to date

I did not even know that I can find them on AMDs site, lol, thanks for pointing it out. I checked now and see that on the MB website it is Version 2.05.04.352 2020/08/25 but on AMD website it is 2.10.13.408 2020/10/19 so I will download those and install.

11 minutes ago, Bombastinator said:

I saw a320m-k in tier f but with no manufacturer associated with it.  Hard to know.  As a general rule a320 is pretty iffy for 8 core chips in general. 6 core is not infrequently the highest that can safely push. My knowledge is limited though

But it seems strange that no issues appear on high load, when gaming or stress testing, but only on idle or very low demanding tasks.

4 minutes ago, ShrimpBrime said:

Would start diagnostics by doing the opposite of this. 

Set all defaults and or clear the cmos. Test it at stock.

 

Then come to a conclusion if the issue still persists, other diagnostic ideas will come to mind. 

 

 

Yes, you are right. I am just waiting for it to happen again, afterwards I will load defaults (without xmp/docp), clear cmos and see what happens.

[arnes91]

I was just now following the guide to make the dump, the sysnative went without issues but after that I started 

perfmon /report to run PERFMON to collect system data for 60 seconds the freeze happened again!

I will now make both dumps again and post the links to them, I will also reset the bios, which turns off xmp/docp.

EDIT: It froze again right after when I run the sysnative BSOD collection app.

EDIT2: The PERFMON never finishes collecting data, it says 60 seconds but runs now for 3 minutes without the possibility to save a log...

[Bombastinator]

25 minutes ago, arnes91 said:

I did not even know that I can find them on AMDs site, lol, thanks for pointing it out. I checked now and see that on the MB website it is Version 2.05.04.352 2020/08/25 but on AMD website it is 2.10.13.408 2020/10/19 so I will download those and install.

But it seems strange that no issues appear on high load, when gaming or stress testing, but only on idle or very low demanding tasks.

Yes, you are right. I am just waiting for it to happen again, afterwards I will load defaults (without xmp/docp), clear cmos and see what happens.

Point.  I was under the impression they appeared randomly regardless of load.  That’s interesting data though.  Not something I can do much with but someone might.

[Guest]

20 minutes ago, arnes91 said:

I will now make both dumps again and post the links to them, I will also reset the bios, which turns off xmp/docp.

 

This you should just start with first..... just saying.

 

XMP/DOCP is an overclocking profile. You want to rule this out. turn it off then ya?

[arnes91]

19 minutes ago, ShrimpBrime said:

This you should just start with first..... just saying.

 

XMP/DOCP is an overclocking profile. You want to rule this out. turn it off then ya?

Now I set the bios to default with XMP/DOCP off, the RAM is now operating at 2400MHz, let's see in the next few days if it will freeze.

Please find attached the links to the Sysnative BSOD file collection dump/log with also the PERFMON dump/log which I made right before and right after a freeze:

 

https://drive.google.com/drive/folders/1AtLT39BaYkZAsQnrDCAFf-icOI_axPtk


Just another question, what are the chances that my monitors are causing the issue?

[Guest]

1 minute ago, arnes91 said:

Just another question, what are the chances that my monitors are causing the issue?

Slim to none.

Let's wait and see what defaults brings you. 

Not sure about the logs, looks like that program gives all details of your PC.

Can you just go into windows event viewer and take a screen shot?

BSODs will also be logged by windows, which I guess you don't get a BSOD, just a freeze right?

[arnes91]

1 minute ago, ShrimpBrime said:

Slim to none.

Let's wait and see what defaults brings you. 

Not sure about the logs, looks like that program gives all details of your PC.

Can you just go into windows event viewer and take a screen shot?

BSODs will also be logged by windows, which I guess you don't get a BSOD, just a freeze right?

I have my hopes up that without the XMP/DOCP it won't freeze, even if that means that I paid something that I can not fully use, but atleast I would not have the freezing.

Please find attached screenshots of the event viewer, i just get those 2 errors + 1 critical after the freeze, and seems they are just errors because of not shutting down correctly.

Some people told me that it can also be because of my power from the socket is not good or uneven, so I just now unplugged everything except my PC and plugged them into another socket so that on 1 socket there is only my PC. But I never had issues with my past PC or any other electronics.

[Guest]

8 minutes ago, arnes91 said:

I have my hopes up that without the XMP/DOCP it won't freeze, even if that means that I paid something that I can not fully use, but atleast I would not have the freezing.
 

A gentleman started a thread just recently with similar issue, but hard crashes. 

 

Perhaps this thread could help? (A 2700X IMC is good for 2933mhz)

edit: And the A320 motherboard is good for 2667mhz, past is an overclock.

It will bring performance, but hopefully better stability. Overclocking is not a for sure thing ever.

 

 

[arnes91]

4 minutes ago, ShrimpBrime said:

A gentleman started a thread just recently with similar issue, but hard crashes. 

 

Perhaps this thread could help? (A 2700X IMC is good for 2933mhz)

It will bring performance, but hopefully better stability. Overclocking is not a for sure thing ever.

 

 

Thanks, I am reading it right now. The best I can do for myself now is wait a few days to see if it will freeze.

I want to ask you, in real life performance, how much is does the difference between 2400MHz and 3200MHz RAM matter, and how much does it affect my R7 2700, because I have read that it needs atleast 3200MHz for full performance?

I need to clarify that I use my PC mostly for my work, which is office work, and web browsing, very very rarely I use it for gaming and if not for any extreme.

[Guest]

33 minutes ago, arnes91 said:

Thanks, I am reading it right now. The best I can do for myself now is wait a few days to see if it will freeze.

I want to ask you, in real life performance, how much is does the difference between 2400MHz and 3200MHz RAM matter, and how much does it affect my R7 2700, because I have read that it needs atleast 3200MHz for full performance?

I need to clarify that I use my PC mostly for my work, which is office work, and web browsing, very very rarely I use it for gaming and if not for any extreme.

Gaming a couple of FPS. For work, probably none noticeable.

Even if you run 2667mhz, that's about a 500mhz over base clock already.(2133) So you'll have performance gains one way or another.

[arnes91]

Thanks everyone for the help. Just want to point out that after turning xmp/docp off the PC does not freeze anymore. I tried testing it with the 3000MHz profile 2, but also turned on PBO and Global C-States and it froze right at the boot after changing settings.

Now I turned off global c-states again and pbo, using profile 2 for xmp/docp but set it manually to 2933Mhz and my PC seems to be working perfectly.

So the issue seems to be fixed and was caused by either or both, the RAM being set over 2933MHz + global c-states and PBO.

[Guest]

6 hours ago, arnes91 said:

Thanks everyone for the help. Just want to point out that after turning xmp/docp off the PC does not freeze anymore. I tried testing it with the 3000MHz profile 2, but also turned on PBO and Global C-States and it froze right at the boot after changing settings.

Now I turned off global c-states again and pbo, using profile 2 for xmp/docp but set it manually to 2933Mhz and my PC seems to be working perfectly.

So the issue seems to be fixed and was caused by either or both, the RAM being set over 2933MHz + global c-states and PBO.

Thanks for coming in with an update. Glad you've gotten it sorted out with a stable system.