PC Random Freezes

aaronweeks

Reputable
Aug 30, 2014
9
0
4,510
Hello.
A few days ago I've tried to flash my BIOS. I've done everything as it was in the guide, no error messages or anything, no power outage, nothing, but it still failed to boot after flashing.

Sent the PC to service, it was finally booting up (w/ new bios ver), but I've noticed that the CPU constantly got too hot.. Then I've realized that the mofos removed the CPU cooler's screws..
Put back the screws, everything was OK. For a while. Then the PC started to freeze randomly after like 10 minutes of use. (constantly, never hit 20 minutes of use with it)

Did a memtest. I have 2 RAM's in dual channel. The first one was OK, but the second one shown 30-50 errors at test 7 every pass. So I've removed the second one. Thankfully I still have the warranty for it.

My PC was running without problems after that, heavy use (e.g. gaming) for long periods, and everything was fine. For 3 days. Today, it froze again after like an hour of light usage (such as browsing).

My temperatures are all OK, I've reinstalled my windows too, no OC. Also ran an HDD error testing, reported no problems.
I have a pretty old rig, like 3 years old (HDD is 2-3 y more), and I plan to upgrade soon, but I'd be happy if my PC would "hold on" until then.

Oh, and when the PC freezes the screen stays, the audio stutters, and it won't react to anything.
No error message or anything like that.

I have yet to do another memtest for my other memory, but at the last one it did not show any problems after a good while.
Any ideas?
 
Solution
Ok so that blue-screen was because of a page-fault in the part of the memory reserved for the OS. So basically windows couldn't read/write to a section of the memory it had reserved for itself.
This is normally due to a physical ram problem.

And this confirms my suspicion, I'm guessing you have 2 ram slots and the slot the ram occupied during this BSoD was the one where originally the ram module you sent for RMA was located in?

If this is true then that other ram module is very likely still 100% working (Chances of it getting damaged by mobo or cpu are slim). And I assume the other ram module is also 100% working as it passed the mem test in its original slot.

This leaves me mobo or cpu. Now given that your system also freezes when...

Conor17777

Reputable
Aug 23, 2014
111
0
4,710


Possibly hanging on the Hard Drive end?
No idea really, it will be solved through randomly disconnecting things and seeing if it makes a difference, i would imagine it is either your Hard Drive or your CPU has fried itself. God hope its not your CPU but given it was running without a secure heatsink at unfavourable temps it is possible.
 

Selenog

Reputable
Aug 30, 2014
40
0
4,560
It sounds to me like your processor is the problem. Can you test that "problematic" ram in another computer?

I'm guessing your processors memory controller and maybe some other parts are broken (overheated due to incorrectly mounted cooler).
If so you'll have to complain to the PC service you used. The processor manufacturer is not gonna issue warranty for incorrectly mounted cooler.
 

Conor17777

Reputable
Aug 23, 2014
111
0
4,710




I think your CPU is fucked then mate sorry :/ Best bet is to complain to the PC service you used, rather than go for warranty, because Intel will not offer warranty for an incorrectly mounted cooler unfortunately :/
 

aaronweeks

Reputable
Aug 30, 2014
9
0
4,510
Wel, thanks for the replies

I can't test the memory in another PC, since I've already RMA'd it.
I also suspected the HDD at first, because it's really old, but shouldn't that have any other signs? It currently performs just as it performed 1-2 years ago. No unusual noises, no bad sectors, no disappearing files.. I admit, it's slow as crap, but it always was like this. (what can you expect from a 7 year old HDD?)

If it would be indeed the CPU then it doesn't really matter because the warranty has ended..
What I don't understand then is that when I had the other (possibly faulty) memory, I couldn't get 20 minutes w/o freeze, but now, after I've removed it, my PC functioned without any problems for a good while, even under heavy use.

Can a CPU get damaged so soon? It is set to warn at 70 c (TCase of my cpu is >74 c), it warned only once (pc also crashed then), and then I've immediately solved the problem.

I wouldn't wonder at all if the mofos did any other damage to my rig there.. I mean, unscrew the CPU cooler and then close the case as well (my case was open at the time I've sent it).
But I have no idea how to complain for it, how could I prove that they have removed the screws..

I can't really remove modules from my PC, it would be hard to get another component for me now, especially that I don't have a built-in GPU.
 

Selenog

Reputable
Aug 30, 2014
40
0
4,560


Know that they might very well send it back saying it works.



You can still make a claim the PC service made a mistake, that's not warranty but service.



It is possible for the memory controller damage to be only noticeable when it is being used (aka with the extra memory module inserted). It is also possible for a cpu to have for example 1 instruction that is broken and if it isn't a commonly used one then it might take very long before some program uses it. Just speculation of course but it's a possible explanation.



The temperature of your cpu is only measured as a specific point, without heatsink to spread the heat it is possible for a local heat buildup to damage part of a chip before the cpu throttling kicks in.



Are you sure you don't have a built-in GPU, most motherboards have a (very simple) build-in vga solution. Might not even do 1080p but it will give you a visual output.

You can also remove possible extra hard drives or media drives (cd/dvd/bluray).



So no BSoD or pc turns off? Sounds a bit like a graphics issue though that would be a strange coincidence unless the PCI driver in the CPU is somewhat broken as well. Not sure, just speculation.

And also make sure your bios has default settings, didn't occur to me yet because of the bios-update but since those service guys don't seem capable make sure its on defaults.
 

aaronweeks

Reputable
Aug 30, 2014
9
0
4,510



No, it's just like a lag (that won't end).

I've found nothing unusual in bios, but still set it to defaults and reconfigured it.
No freezes so far, although I only had one in the past 3 days (after I've removed the other RAM.)
 

aaronweeks

Reputable
Aug 30, 2014
9
0
4,510
Just got another freeze when watching youtube.

Shortly after another freeze, but an unusual one. Everything was crashing at first, then after like 5 secs a freeze, but I could move my mouse. Nothing happened for 5 minutes, had to hard reset, but there was occasional (not constant) buzzing from the HDD.
Is this indeed a dying HDD?
 

Selenog

Reputable
Aug 30, 2014
40
0
4,560
And was your memory compatible with your motherboard? (check manufacturer of motherboard's website)

If it was then it should have been without bios update (on specified speeds ofc). This leads me to think something might have been already wrong.
 

Selenog

Reputable
Aug 30, 2014
40
0
4,560
That leads me to think the original problem is still there and only got worse. It is possible there is now a second problem but that seems improbable.

So the problem should be ether cpu, mobo or ram.

You already tested the ram and are now running on one ram module if I understood correctly. If you could please try to move that one ram module through each different slot and see if you manage to get a freeze.
 

aaronweeks

Reputable
Aug 30, 2014
9
0
4,510
Got 3 more BSoD's after moving memory to another socket, the first one was also associated with ntoskrnl, another one when I was trying to upload it (before that the browser repeatedly crashed), the third one was at login screen.
 

Selenog

Reputable
Aug 30, 2014
40
0
4,560
Ok so that blue-screen was because of a page-fault in the part of the memory reserved for the OS. So basically windows couldn't read/write to a section of the memory it had reserved for itself.
This is normally due to a physical ram problem.

And this confirms my suspicion, I'm guessing you have 2 ram slots and the slot the ram occupied during this BSoD was the one where originally the ram module you sent for RMA was located in?

If this is true then that other ram module is very likely still 100% working (Chances of it getting damaged by mobo or cpu are slim). And I assume the other ram module is also 100% working as it passed the mem test in its original slot.

This leaves me mobo or cpu. Now given that your system also freezes when you have the ram module in the working slot I'm inclined to think your cpu is the problem.

Here is what I think happened:

    ■ cpu's memory controller (or more unlikely mobo memory lanes) got somehow damaged
    ■ memory didn't work anymore on the specified speeds
    ■ you did a bios update but the pc crashed
    ■ PC service fixed the bios but did a bad job mounting the cpu cooler
    ■ cpu overheated and was further damaged


Though I'm not 100% sure if I had to guess with the information provided the problem is your cpu.

Though like I mentioned before often the mobo has a simple vga solution so you might want to try to remove your graphics card and see if the problem continues, though it's very likely it will.
 
Solution

aaronweeks

Reputable
Aug 30, 2014
9
0
4,510
Well it can be possible, but the memory didn't work at specified speeds since the beginning, only that I was not aware of the bios update then.

I'll upgrade my rig soon, and I'm curious what will they say about the other ram module which I RMA'd.

I'm selecting your post as the solution for your time, thanks
 

TRENDING THREADS