Memory/Motherboard/CPU/could be anything issues

[Jonah]

Don't be put off by the size of this post, please read it.

So I'll start from the beginning. In January this year I completed my first build with this spec:

Asus M5A97 LE R2.0 MoBo
AMD FX6100 CPU
Corsair CMZ4GX3M1A1600C9 Vengeance 4GB single stick RAM
WD Scorpio Black 160GB 2.5" HDD
Cooler Master Elite 500 PSU
VTX3D ATI Radeon 5450 Graphics Card
A cheap TP-Link PCI Wifi card
Windows 7 Home Premium OEM

Around May the HDD gave up and I replaced it with a nice big 500GB WD Caviar Blue 3.5" and installed windows onto that.

So thats the probably-irrelevant background info, now for the actual problem.

About 3 weeks-a month ago games started crashing at random, with the standard windows APPCRASH message. More apps (Chrome, Explorer.exe) crashed too, and when i logged on not all of my taskbar icons/wallpaper were there. Assuming that this was caused by some sort of malware, i attempted to download some antivirus programs such as Avast and AVG, but the installation programs just crashed. At this point my PC was basically unusable, so i kind of gave up for the night.

The following day windows would BSoD just before or after the login screen. I booted safe mode, but again explorer and tasgmgr continued to crash. Eventually i got the BSoD again, so i decided that my windows was royally screwed. I attempted to boot a Linux distro off my USB drive, but after it crashed it wouldnt boot at all, complaining about the process swapper. I tried running windows again but it BlueScreened again, giving a MEMORY_MANAGEMENT error. Both the swapper and the windows error message point directly to RAM, so I ran both memtest86 and memtest86+ and both reported millions of broken cells. Flashed the BIOS to the newest version, same result. Tried it in a different RAM slot, same result.

After getting the RAM replaced under warranty, I booted Windows Startup Repair and everything worked fine. No memtest errors or any problems inside windows. I continued to feed my PC gaming habit, satisfied because everything was working fine...

Until the day before yesterday.
My TF2 crashed and soon after windows went straight to a MEMORY_MANAGEMENT BlueScreen. Within a minute I was running memtest86+ and sure enough, it again reported errors, though not as many and only in a certain section of the RAM.

I haven't even tried booting Windows again since then, so I will after I post this. What I want to know is this:

- Is this likely to be an issue with the type of RAM, the MoBo, the CPU/memory controller or even the PSU?

- Can software screw up hardware? i.e. Could this have been caused by malware?

- What do i do?

I really don't want to have to replace the motherboard, as that means replacing Windows as well as OEM is locked to the motherboard.

Thanks in advance,
Jonah

-----------------------
Possibly Useful Info:

Steam games: Portal 2, Team Fortress 2, Burnout Paradise, Crysis 2, Forge, Garry's mod, Half-Life 2 Demo, Mirrors Edge, and Portal.

Other games/apps: Toribash, Minecraft (inc. techniclauncher and modpacks), Hamachi, Chrome, Paint.NET, Scratch, Office 2010, Audacity sound editor, Spotify, 7-Zip, DropBox, Greenfoot, iTunes, Adobe Reader, i-Funbox.

I've basically listed all these in case anyone knows of anything in any of them that can mess things up.

Heres some links to files windows says help:
What windows said about the BlueScreen

The minidump mentioned in the previous file
The system info that it also mentions

 

[tylerjrb]

Don't be put off by the size of this post, please read it.

So I'll start from the beginning. In January this year I completed my first build with this spec:

Asus M5A97 LE R2.0 MoBo
AMD FX6100 CPU
Corsair CMZ4GX3M1A1600C9 Vengeance 4GB single stick RAM
WD Scorpio Black 160GB 2.5" HDD
Cooler Master Elite 500 PSU
VTX3D ATI Radeon 5450 Graphics Card
A cheap TP-Link PCI Wifi card
Windows 7 Home Premium OEM

Around May the HDD gave up and I replaced it with a nice big 500GB WD Caviar Blue 3.5" and installed windows onto that.

So thats the probably-irrelevant background info, now for the actual problem.

About 3 weeks-a month ago games started crashing at random, with the standard windows APPCRASH message. More apps (Chrome, Explorer.exe) crashed too, and when i logged on not all of my taskbar icons/wallpaper were there. Assuming that this was caused by some sort of malware, i attempted to download some antivirus programs such as Avast and AVG, but the installation programs just crashed. At this point my PC was basically unusable, so i kind of gave up for the night.

The following day windows would BSoD just before or after the login screen. I booted safe mode, but again explorer and tasgmgr continued to crash. Eventually i got the BSoD again, so i decided that my windows was royally screwed. I attempted to boot a Linux distro off my USB drive, but after it crashed it wouldnt boot at all, complaining about the process swapper. I tried running windows again but it BlueScreened again, giving a MEMORY_MANAGEMENT error. Both the swapper and the windows error message point directly to RAM, so I ran both memtest86 and memtest86+ and both reported millions of broken cells. Flashed the BIOS to the newest version, same result. Tried it in a different RAM slot, same result.

After getting the RAM replaced under warranty, I booted Windows Startup Repair and everything worked fine. No memtest errors or any problems inside windows. I continued to feed my PC gaming habit, satisfied because everything was working fine...

Until the day before yesterday.
My TF2 crashed and soon after windows went straight to a MEMORY_MANAGEMENT BlueScreen. Within a minute I was running memtest86+ and sure enough, it again reported errors, though not as many and only in a certain section of the RAM.

I haven't even tried booting Windows again since then, so I will after I post this. What I want to know is this:

- Is this likely to be an issue with the type of RAM, the MoBo, the CPU/memory controller or even the PSU?

- Can software screw up hardware? i.e. Could this have been caused by malware?

- What do i do?

I really don't want to have to replace the motherboard, as that means replacing Windows as well as OEM is locked to the motherboard.

Thanks in advance,
Jonah

-----------------------
Possibly Useful Info:

Steam games: Portal 2, Team Fortress 2, Burnout Paradise, Crysis 2, Forge, Garry's mod, Half-Life 2 Demo, Mirrors Edge, and Portal.

Other games/apps: Toribash, Minecraft (inc. techniclauncher and modpacks), Hamachi, Chrome, Paint.NET, Scratch, Office 2010, Audacity sound editor, Spotify, 7-Zip, DropBox, Greenfoot, iTunes, Adobe Reader, i-Funbox.

I've basically listed all these in case anyone knows of anything in any of them that can mess things up.

Heres some links to files windows says help:
What windows said about the BlueScreen

The minidump mentioned in the previous file
The system info that it also mentions

Personally i would say its most likely to be the RAM, i had the exact same issue as you with random crashing games, browsers, blue screens, memtest giving errors and it turned out it was the RAM for me.

as its giving you errors once again id say you could be extremely unlucky as to get an RMA'd stick fail so quick. the memory_management is common with RAM failures but most BSOD's can be hard to follow and seem to hint whats wrong. with my faulty RAM i had many different screens including the memory_management screen.

try the RAM in every slot and run the memory test once again in each slot, could be that the slot itself is faulty. do you have another ddr3 stick to test with, maybe a friends computer or computer around the house? then id suggest this too as you mention you only have a 4gb stick installed its hard to check and a known working stick of RAM is the best to test with.

Id also check the bios settings are all correct, RAM is on the correct latency timings, speeds and voltages mentioned on the label or box. yours should be 9-9-9-24, 1.5v 1600mhz. Also its most important the voltage is correct, anymore that 1.7v it will start causing damage to the Ram so make sure its on 1.5v in the bios.

with regards to it damaging hardware, its very unlikely. The blue screen system itself is a safeguard, as soon as it detects a major problem like dead ram it cuts itself immediately off and gives the error screen so you can interpret whats wrong. this is so the system is not damaged. its kind of like the heat sensor on a cpu it cuts itself off at a certain temperature so it is not damaged.

with regards to everything else, id doubt its the motherboard. if its been running fine and then after a few days/weeks it fails that points to the RAM to me. if it was the cpu or something like a bad memory controller in the motherboard it would fail constantly and would be incredibly lucky for it to run for that period of time if it was faulty before going bad again.

but speaking again its very unlucky to buy a stick, it fails, RMA it then it fails again. but these things have been known to happen.

if no one has any other test and suggestions it could be possible a piece of dirt or debris is blocking a pin on the cpu/motherboard.

RMA the RAM stick and see if you can get another known working RAM stick from somewhere else and test inside the pc. Then if it still fails you can move onto checking the cpu for damaged pins, dirt or debris blocking contacts etc. bad motherboard controllers etc.

but getting that known working ram stick is probably the most importance, if that still fails then its something different. you could RMA the stick and receive a faulty new stick and blame it on something else for example. id say get the RMA all in order and get a working RAM stick to test with.

but due to the errors on memtest, its highly likely the RAM is damaged.

 

[spirit]

I'd start by removing a stick of RAM (remove the suspect stick if you know which stick could be causing problems) and see if you continue to get problems.

If you don't, you know where the problem lies.

RAM is a common cause of blue screens. RAM also fails quite quickly (compared to other pieces of hardware) too.

 

[Jonah]

Personally i would say its most likely to be the RAM, i had the exact same issue as you with random crashing games, browsers, blue screens, memtest giving errors and it turned out it was the RAM for me.

as its giving you errors once again id say you could be extremely unlucky as to get an RMA'd stick fail so quick. the memory_management is common with RAM failures but most BSOD's can be hard to follow and seem to hint whats wrong. with my faulty RAM i had many different screens including the memory_management screen.

try the RAM in every slot and run the memory test once again in each slot, could be that the slot itself is faulty. do you have another ddr3 stick to test with, maybe a friends computer or computer around the house? then id suggest this too as you mention you only have a 4gb stick installed its hard to check and a known working stick of RAM is the best to test with.

Id also check the bios settings are all correct, RAM is on the correct latency timings, speeds and voltages mentioned on the label or box. yours should be 9-9-9-24, 1.5v 1600mhz. Also its most important the voltage is correct, anymore that 1.7v it will start causing damage to the Ram so make sure its on 1.5v in the bios.

with regards to it damaging hardware, its very unlikely. The blue screen system itself is a safeguard, as soon as it detects a major problem like dead ram it cuts itself immediately off and gives the error screen so you can interpret whats wrong. this is so the system is not damaged. its kind of like the heat sensor on a cpu it cuts itself off at a certain temperature so it is not damaged.

with regards to everything else, id doubt its the motherboard. if its been running fine and then after a few days/weeks it fails that points to the RAM to me. if it was the cpu or something like a bad memory controller in the motherboard it would fail constantly and would be incredibly lucky for it to run for that period of time if it was faulty before going bad again.

but speaking again its very unlucky to buy a stick, it fails, RMA it then it fails again. but these things have been known to happen.

if no one has any other test and suggestions it could be possible a piece of dirt or debris is blocking a pin on the cpu/motherboard.

RMA the RAM stick and see if you can get another known working RAM stick from somewhere else and test inside the pc. Then if it still fails you can move onto checking the cpu for damaged pins, dirt or debris blocking contacts etc. bad motherboard controllers etc.

but getting that known working ram stick is probably the most importance, if that still fails then its something different. you could RMA the stick and receive a faulty new stick and blame it on something else for example. id say get the RMA all in order and get a working RAM stick to test with.

The problem with getting a working stick of RAM is that I don't have much money and I don't trust my PC not to brick any friend's RAM I put in it. When i RMA'd the first stick, Corsair sent me a new one and it tested in memtest86+ with no errors. But after a couple of weeks a segment of it seems completely bricked. I cant really see what damages it, it seems highly suspicious for me to get the same error twice in a month, unless the specific model of RAM has issues with my system. If it's that, then I can't see how it would run perfectly for 10 months unless it's some software I recently installed, though I cant think of any that would brick a RAM stick.

And Spirit, its a single stick setup so i cant just remove one

In any case, I'm now gonna look inside for any physical damage on the motherboard and I'll check all the BIOS settings too. Thanks for your help so far, guys

-EDIT-
Checked the BIOS, all seems well, going in-deep now for a look at physical stuff and to swap slots

 

[PCunicorn]

I think its odd you got a stick, it turned out busted, got another stick under warranty and it broke. It could be one of two things
You are very unlucky and managed to get 2 bad sticks
Or the motherboard is messing them up

 

[Jonah]

I think its odd you got a stick, it turned out busted, got another stick under warranty and it broke. It could be one of two things
You are very unlucky and managed to get 2 bad sticks
Or the motherboard is messing them up

That's what I'd figured, and seeing as it took so long for the first stick to break and not long at all for the second, I'd say it was more likely that something in the motherboard has broken and is screwing stuff up.

-update-
Been in the PC, found a shitload of dust under the cpu cooler edges (i probably need an exhaust fan) but no visible damage anywhere, and moved the RAM to another slot. Lets see what happens in Memtest
-update2-
Still errors - the failing address is at 1657.6mb at each error, with a slightly variable code starting such as 000679ac040 (the last 4 characters vary)

 

[tylerjrb]

That's what I'd figured, and seeing as it took so long for the first stick to break and not long at all for the second, I'd say it was more likely that something in the motherboard has broken and is screwing stuff up.

-update-
Been in the PC, found a shitload of dust under the cpu cooler edges (i probably need an exhaust fan) but no visible damage anywhere, and moved the RAM to another slot. Lets see what happens in Memtest
-update2-
Still errors - the failing address is at 1657.6mb at each error, with a slightly variable code starting such as 000679ac040 (the last 4 characters vary

it is most likely the RAM and not the motherboard believe me, they don't run like that.

it wouldn't fail, put a new RAM stick in runs fine for a few weeks and fail again. if the motherboard went it wouldn't even turn on, post etc even if a small part went. and if it messed up a very very minimal amount it wouldn't work fine for a few weeks after, a new ram stick was installed. on id say you've been very unlucky with the RAM. if it was a bad memory controller it would be blue screening all the time and not run fine for a few weeks, it wouldn't be fixed by a new RAM stick.

its still best to try a working stick of RAM, wear an anti-static band if needed.

Also check your bios, the only way your motherboard would damage the RAM is if it failed which it hasn't. or giving it too much voltage.

check that in the bios the voltage, timings and speed match up. otherwise if you put another new stick in on 1.8v it will kill that pretty quickly too.

like mentioned above, Ram is the most common hardware to breakdown. Ive had it go in my old pc and in my new pc.

as you have ran the tests and they fail on the exact same mb each time, codes are pretty much the same. that memory chip in the RAM stick has probably died.

try this go into control panel and onto system information. does it read 4.0gb installed or 4.0gb installed 3.7 available?

 

[Jonah]

it is most likely the RAM and not the motherboard believe me, they don't run like that.

it wouldn't fail, put a new RAM stick in runs fine for a few weeks and fail again. if the motherboard went it wouldn't even turn on, post etc even if a small part went. and if it failed it wouldn't work fine for a few weeks after. on id say you've been very unlucky with the RAM. if it was a bad memory controller it would be blue screening all the time and not run fine for a few weeks, it wouldn't be fixed by a new RAM stick.

its still best to try a working stick of RAM, wear an anti-static band if needed.

Also check your bios, the only way your motherboard would damage the RAM is if it failed which it hasn't. or giving it too much voltage.

check that in the bios the voltage, timings and speed match up. otherwise if you put another new stick in on 1.8v it will kill that pretty quickly too.

like mentioned above, Ram is the most common hardware to breakdown. Ive had it go in my old pc and in my new pc.

as you have ran the tests and they fail on the exact same mb each time, codes are pretty much the same. that memory chip in the RAM stick has probably died.

try this go into control panel and onto system information. does it read 4.0gb installed or 4.0gb installed 3.7 available?

All the BIOS voltages and timings are correct.
System info says...

Installed Physical Memory (RAM) = 4.00gb
Total Physical Memory = 3.90gb
Available Physical Memory = 2.29gb

I'd say you're right.

-hmm-
Windows still hasn't crashed, I reckon it's because it recognised that that part of the RAM was dead. Though theres still a problem, do you think I'd be able to get by on this for a while without it becoming more damaged?

 

[tylerjrb]

All the BIOS voltages and timings are correct.
System info says...

Installed Physical Memory (RAM) = 4.00gb
Total Physical Memory = 3.90gb
Available Physical Memory = 2.29gb

I'd say you're right.

-hmm-
Windows still hasn't crashed, I reckon it's because it recognised that that pert of the RAM was dead. Though theres still a problem, do you think I'd be able to get by on this for a while without it becoming more damaged?

it would still try to access it if needed. it would start crashing once it runs into that dead chip/s. hense why if you use a browser, game, even run memtest it uses more RAM and will run into that chip/s causing it to crash. you could get away with it for a while on small programmes but it may crash if it starts using that chip. id say its up to you, you can still use it until a new stick comes but if its blue screening all the time that's when you increase the chance of damage. running low usage programmes may reduce that. just running maybe 1 program at a time will reduce it but it may eventually catch up.

it has to be exact in its take and give commands so to speak. say if the cpu gave it 01001001 and it gave it back as 0111000. dont know how binary codes work lol but just as an example.

it can give the cpu back a different command than what the cpu gave it. this could mean something completely different and causes it to crash to stop damage.

 

[Jonah]

Well, my problem now is that I don't want to buy more RAM if something is just going to brick it again. I can't understand how it would work absolutely fine for 10 months then break, then a second stick works fine for a couple of weeks and breaks. But if the motherboard was damaging the RAM then surely it wouldn't wait almost a month before the new stick cuts out?

Maybe it's a more fine problem on the board, like a connection that doesn't always work properly, that occasionally bricks sends a wrong signal or the wrong amount of power, that can ruin the RAM.

This is just me thinking aloud here, any plausible explanation will be appreciated.

 

[tylerjrb]

Well, my problem now is that I don't want to buy more RAM if something is just going to brick it again. I can't understand how it would work absolutely fine for 10 months then break, then a second stick works fine for a couple of weeks and breaks. But if the motherboard was damaging the RAM then surely it wouldn't wait almost a month before the new stick cuts out?

Maybe it's a more fine problem on the board, like a connection that doesn't always work properly, that occasionally bricks sends a wrong signal or the wrong amount of power, that can ruin the RAM.

This is just me thinking aloud here, any plausible explanation will be appreciated.

don't forget its on RMA, as its broken there's nothing you have done to damage it and its well within your rights to send it back for a replacement.

id say you have been very unlucky, as for the board breaking it. i very much doubt so, if it was sending random voltage spikes it would most likely damage other components too. don't forget this would also most likely damage the board and make it fail.

i'm unsure as to say if they can spike at all, other than an electricity spike the reason it blows is because it frys everything, capacitors etc. if it were to spike im pretty sure it would do the same thing.

send the Ram back for a replacement and if it goes again, you know there's something else wrong.

but again id doubt it would run fine for 10 months, then decide to kill the Ram. replace and take a whole 3 weeks for it to damage it again. pc's send millons, billions of electrical signals in a very short space of time. and its very likely if the motherboard was damaged it would be sending random spikes pretty much all the time due to the amount of electrical signals sent like i say millions etc in a short time frame even if you got the spike once every 1 million.

 

[johnb35]

If I were you I would change out the psu to something better.

 

[Jonah]

Well I'll RMA it again and see what happens, but it has to be sent to Holland for replacement so it'll be at least a week before i have a new stick :/
Thanks again for your help and I'll post back if any more details come up

 

[Jonah]

lets revive the thread then...
I have just received a new stick from Corsair, looking at the box it seems to have taken so long because its come from Hong Kong this time, not the Netherlands. And great news! I now have another dilemma for you guys!
What I think is best to do is reinstall windows. I don't know if this will help but surely it can't hurt. I want to be able to save my steam games and most of my personal data, as well as my stash of install files. I know I'm being paranoid here, but if this was in some way caused by some sort of hardware-screwing-up-level malware, if such a thing exists, is it safe to keep these?
And another (kind of unrelated) question - is it safer to keep my windows and personal files on separate disks, say if I bought an SSD for windows and kept my data/games on the 500GB HDD, if windows breaks, surely this means my data is safer?

 

[StrangleHold]

What happen when you installed the new stick? You sure the memory voltage/timing and speed is set right?

 

[Jonah]

What happen when you installed the new stick? You sure the memory voltage/timing and speed is set right?

Haven't installed the new one yet, i want to be sure of what to do first. But I'm sure it was all fine before- 1.5v, 9-9-9-24, 1600mhz