[Hardware problem] Random reboots and freezes

[younggotti]

Hi everybody,

some months ago, I built a new entry level PC for my parents. This was the set-up:
PSU XFX 450W Pro450S
Cpu AMD A8 5600K HD7560D 3.6GHz FM2
Mobo Gigabyte GA-F2A88XM-HD3
Ram DDR3 Crucial 8GB / 1600Mhz Ballistix Sport

The PC, even though never used under stress conditions (internet or office applications), has since the beginning had random reboots, both in Windows (after a BSOD) and outside (during the boot or even when in the BIOS menu).
I initially thought that the problem was the motherboard, since I also got random "bios corruption" errors at boot.
So, even if the shop said they tested the mobo and it worked fine for them, I switched to a new one (MSI FM2-A75MA-E35).

After the switch, the PC seemed to work fine for the most part. At the beginning, there were few reboots (without BSOD) in Windows, but I wasn't there so I thought that might be a software problem. After a while, also those few reboots disappeared (at least my mom says so) and for a few months the PC seemed to work fine.

Unfortunately, in the last few days a new (?) problem came up. Randomly, in windows or in the bios menu or even during the boot, the video signal just goes off and the monitor goes to sleep. The fans keep working but, apart from that, the PC looks dead (ctrl+alt+del doesn't work, the caps/num lock indicator lights switch off and don't work any more, etc). I just have to manual reboot the PC and, sometimes, even that doesn't work and I have to manually switch it off and on again.
The problem seems to occurr more often when using cpu-intensive applications but, as I said, it occurrs even when in the BIOS menu. The other strange fact is that sometimes the pc works fine even for a couple of hours; at other times, when the problem occurrs, it just keeps coming again and again and the pc freezes few seconds after each reboot.
I cannot even understand if the problem is due to the same hardware failure as the first problem or if it's something new.

This is the Windows log file after one freeze occurred while in Windows (but I don't think it will help):

Log Name: System
Source: Microsoft-Windows-Kernel-Power
Date: 27/07/2014 09:43:46
Event ID: 41
Task Category: (63)
Level: Critical
Keywords: (2)
User: SYSTEM
Computer: PC-Stanzino
Description:
The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly.
XML event:
<Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event">
<System>
<Provider Name="Microsoft-Windows-Kernel-Power" Guid="{331C3B3A-2005-44C2-AC5E-77220C37D6B4}" />
<EventID>41</EventID>
<Version>3</Version>
<Level>1</Level>
<Task>63</Task>
<Opcode>0</Opcode>
<Keywords>0x8000000000000002</Keywords>
<TimeCreated SystemTime="2014-07-27T07:43:46.863302500Z" />
<EventRecordID>10253</EventRecordID>
<Correlation />
<Execution ProcessID="4" ThreadID="8" />
<Channel>System</Channel>
<Computer>PC-Stanzino</Computer>
<Security UserID="S-1-5-18" />
</System>
<EventData>
<Data Name="BugcheckCode">0</Data>
<Data Name="BugcheckParameter1">0x0</Data>
<Data Name="BugcheckParameter2">0x0</Data>
<Data Name="BugcheckParameter3">0x0</Data>
<Data Name="BugcheckParameter4">0x0</Data>
<Data Name="SleepInProgress">0</Data>
<Data Name="PowerButtonTimestamp">0</Data>
<Data Name="BootAppStatus">0</Data>
</EventData>
</Event>

This is what I tried:
- I updated the mobo bios and the cheapset drivers
- Ram test (memtest86+). I first did it when the first problem occurred and the test was ok. I also tried to use a new ram stick but it didn't help. In the last few days, I tested the ram again. The first time the PC freezed as usual after few minutes. The following day the test went on for a couple of hours before the PC freezed. In all these cases, before the PC freezed, memtest didn't report any error.

I think that, excluding a software and a memory problem, the remaing available options are:
- PSU: these are the voltages reported in windows by the MSI utility (on the right) and by Hwmonitor on the left, whilst running a CPU-stress software. As you see, Hwmonitor (and Speccy as well btw) reports strange values. The ones reported by the mobo utility (which are the same as the ones I get in the BIOS menu) seem ok to me. Which ones are the right ones?

- overheating CPU: as you can see from the same screenshot, also for the cpu temp readings I get a huge discrepancy between hwmonitor (and speccy) on one side and the mobo utility (and the BIOS) on the other. Anyway, I don't think overheating is the problem (sometimes the problem occurrs just after a cold start-up).

- cpu failure

- mobo (again???)

Any ideas?

Thanks a lot.

 

[ICIT2LOL]

Hello yonggotti mate run the HW Info and check the PSU voltages with the sensor button.
That is juts for starters as I am not familiar with that brand.

If they come up crap or not then try this but please use a digital meter as in the tutorial as the analogue ones inject 9v into the system to check stuff and needleless to say the 3.3 and 5v would not appreciate that!
http://www.sevenforums.com/tutorials/301799-psu-test-dc-output-voltage.html

If the volts are there then that is one thing out of the way.

I ma sort of assuming you have run them but if not - these in safe mode if necessary

http://www.sevenforums.com/tutorials/1538-sfc-scannow-command-system-file-checker.html

http://www.sevenforums.com/tutorials/433-disk-check.html < use the /f option in Option 2 ifnecessary

 

[essenbe]

Hi younggotti, I would try what Icit2LOL suggested for actual readings. However, as you can see from the bios, you are getting warnings about your PSU. The most correct readings will come from BIOS. Readings from within windows are software trying to read sensors on the board. The low readings and some of the high readings are out of specs. ATX specifications are +/- 5% of the rated values, meaning for example the +12V must stay within 11.4-12.6 to be within specs. If I am reading your bios readings correct your +12V is reading a Low of 11-11.9 and a high of 12.1-13V both of which are out of specs. If I am reading it wrong, then ignore this. But you may want to check your manual for the readings in the bottom left of the bios page. But, you BIOS readings are more accurate than software readings. It is also reading your CPU utilization as 100% and CPU temp at 61C in one part of the bios. So, you should check your manual and see exactly which readings are which. The top says CPU tem as 61C the bottom says 32C.

 

[essenbe]

How about running this program and upload all files in your next post http://www.sevenforums.com/bsod-help-support/96879-blue-screen-death-bsod-posting-instructions.html

 

[younggotti]

Hi,

thanks both for your reply.

@ICit2lol
I'll run the HW info and post the results.
Unfortunately, I don't have a multi-meter. What I can do is connecting mobo+cpu+ram to another psu and see what happens.

@essenbe
I see that the screenshot I posted before was not completely clear. I modified the previous message with a new one. The red squared values are the Hwmonitor ones, whilst the yellow ones are the MSI utility ones (which match the BIOS reported ones).

As you can see, the yellow ones look ok: +12V is 12.547V

CPU utilization is 100% because I was running a cpu stress tool. The cpu temp was 61C (141F)

I'll run the BSOD dump file.

 

[younggotti]

Hi, this the HW Info screenshot with the pc running prime95 (100% cpu utilization):

 

[younggotti]

And here is the BSOD debug file

 

[ICIT2LOL]

Mate you need to do what essenbe says and check the volts in the BIOS or use that other tutorial. The multimeters you can pick up very cheaply they do not have to be a really high need device. I bought one recently to use at my partners place for $10!

Now having said that looking at the Fintek volts I think the HW Info must be well out of whack or that PSU is indeed so far up that creek it is not going to come back. Those readouts are not acceptable in any way shape or form because as essenbe states those rails must be within that very small percentage of the required rail voltage ie 3.3v 5v and +12v if they are not then there is the possibility of real damage to the board and components or simply no or intermittent functioning of the machine.

When you have time this might be something worth a quick read of to see what we mean.
Voltage Rails - Power Supply 101: A Reference Of Specifications

 

[essenbe]

younggotti, You may consider a clean install of the OS. You have AMD sata drivers, Nvidia Raid drivers, LSI Logic drivers and Intel Raid drivers in the system.

 

[ICIT2LOL]

younggotti, You may consider a clean install of the OS. You have AMD sata drivers, Nvidia Raid drivers, LSI Logic drivers and Intel Raid drivers in the system.

unfortunately youngotti as essenbe has pointed out it does seem to be awfully clogged with drivers

 

[younggotti]

Hi guys,

yes, the voltages shown by HWInfo look crazy. But, the ones shown by the MSI Windows tool and the ones shown in the BIOS are within the ATX range. Multimeters are really cheap, but I think that the quickest, easiest and most reliable way to verify if the cause of my troubles is the PSU is using another PSU I have and see what happens. I'll do it this weekend and let you know.

and yes, the windows setup is really a mess Months ago, thinking that it could be a software related problem, I tried installing different OSs and each time I didn't pay much attention to get a clean setup. I know that the problems I have now don't depend on that (freezes happen also in the BIOS menu or while performing a memtest), but as soon as the hardware problem is fixed I know I'll have to make a clean install.

 

[ICIT2LOL]

Well I guess the finger is pointed at the PSU for me or maybe the hard drive but in any case that is a start.

I am going to look into that HW Info program as I have been using it and linking folks to it for some time and it seems to me that there are a few errors starting to pop up. Shame really because I do like how you can dig right down into components details for example the RAM sticks and GPU.

 

[younggotti]

Little update:

yesterday the pc seemed to work fine for a couple of hours, but I didn't put much hope in that since, as I said the freezes seem to have a random behaviour.

Today I was able to connect a different PSU to the system. I let the pc run for some time both in idle and under stress and it seemed to work fine. So I was starting to hope I had found defective hw component.

As a double check, I connected the defective (?) PSU to a different PC (similar in terms of watt consumption), "hoping" to get a freeze or a re-boot. But, again, also that PC seemed to work just fine.

I've then reinstalled the "original" defective PC and let memtest 86+ run for 10 hours. The PC didn't freeze and no errors came up. Just now I'm having Prime95 and Furmark run at the same time and the PC is working fine.
The only things I notice are:
- CPU temp (as measured by the mobo producer windows tool) is quite high (75 C / 167 F) but I don't think that's an extreme value, considering I'm stressing the CPU and the integrated GPU at the same time
- PSU voltages (as measured by the mobo producer windows tool): the 12V is quite high (around 12.3V with peaks of 12.55) and the 5V is quite low (around 4.96 with dips at 4.88). Anyway they are within the ±5% range and seem consistent with the results of this review:
XFX Core Edition 450W Review

To sum things up, the PC now seems to work fine but I have no idea why.

 

[essenbe]

Whatever was causing it, I hope it continues to run well. Let me know how it does. But, those temps ar pretty high for your CPU.

 

[ICIT2LOL]

As essenbe has said the temp is way too high and exceeds the makers specs in fact see this
AMD A8-Series A8-5600K - AD560KWOA44HJ / AD560KWOHJBOX max temp according to this site 74C!!

I would be taking a very close look at the cooling system and renewing the themal compounds ASAP.

 

[younggotti]

You're both right, 75 C is out of spec.

I got 75C in a really extreme situation (CPU and GPU stress software running at the same time). In idle the cpu temp stays at around 40C. If I stress test just the CPU the temp doesn't rise above 65C. Anyway, as you said, just to be sure I'll have to renew the thermal compund.

Thanks both, mates

 

[ICIT2LOL]

Just out of interest what cooler have you on the CPU I am not really a fan of the stock ones these days and for a little outlay even one of the liquid coolers is a good option. That is depending on ones budget of course.

 

[younggotti]

I have the stock cooler and I'm not a big fan either (it's very noisy)

 

[essenbe]

Consider a third party cooler. You can buy a new one for little money. I have a box of stock ones. I never use them. I have no idea how I managed to miss this.

Please fill out your System Specs​

   Information

Your System Specs will help us to help you, and doing it in this manner will make them available to all helpers in every post and keep us from hunting for them. We ask that you fill them out in as much detail as possible including Desktop or Laptop, Model number if it is an OEM computer and all components with the Manufacturer and Model number if possible.

If you will go to your last post and click the 'System Specs' in the bottom left of the post, you will find a link to update your system specs. Please fill those out in as much detail as possible, making sure to click save at the bottom of the page. If you would like to know what we would like, you can click 'My System Specs' at the bottom left of this post to see mine. If you do not know what your components are, this will help you accomplish this task. http://www.sevenforums.com/tutorials/180324-system-info-see-your-system-specs.html​

 

[ICIT2LOL]

Yes youngotti as essenbe says mate because the specs at present open out to the machine running 8.1 (down in the left hand side of your post screen in small grey letters)