computer hangs/reboots

[Falkor]

I built a computer somewhat like 6-8 months ago. a friend installed it windows xp x64, and it was running fine until recently when the computer will just reboot by itself (no blue screens, nothing (and I did activate the blue screens)) or it will just freeze/hang showing a screen full of orange lines (in this situation I was forced to hard reset). there was nothing on the event viewer, either. I upgraded the system to windows 7 just a few days ago, hoping that might solve the issue. the computer already hanged 2-3 times. it was either a black screen or a screen full of blue lines. I checked the event viewer and there appeared a handful of errors. only one was associated to a blue screen and came with a minidump. here's the minidump:

*******************************************************************************
*                                                                             *
*                        Bugcheck Analysis                                    *
*                                                                             *
*******************************************************************************

WHEA_UNCORRECTABLE_ERROR (124)
A fatal hardware error has occurred. Parameter 1 identifies the type of error
source that reported the error. Parameter 2 holds the address of the
WHEA_ERROR_RECORD structure that describes the error conditon.
Arguments:
Arg1: 0000000000000000, Machine Check Exception
Arg2: fffffa800b5298f8, Address of the WHEA_ERROR_RECORD structure.
Arg3: 0000000000000000, High order 32-bits of the MCi_STATUS value.
Arg4: 0000000000000000, Low order 32-bits of the MCi_STATUS value.

Debugging Details:
------------------


BUGCHECK_STR:  0x124_GenuineIntel

DEFAULT_BUCKET_ID:  VISTA_DRIVER_FAULT

PROCESS_NAME:  System

CURRENT_IRQL:  0

STACK_TEXT:  
fffff880`033626f0 fffff800`02ebaa89 : fffffa80`0b5298d0 fffffa80`09d78040 fffffa80`00000007 00000000`00000001 : nt!WheapCreateLiveTriageDump+0x6c
fffff880`03362c10 fffff800`02d9c547 : fffffa80`0b5298d0 fffff800`02e155f8 fffffa80`09d78040 00000002`00000005 : nt!WheapCreateTriageDumpFromPreviousSession+0x49
fffff880`03362c40 fffff800`02d04b95 : fffff800`02e77360 fffffa80`0ad71888 fffffa80`0ad71880 fffffa80`09d78040 : nt!WheapProcessWorkQueueItem+0x57
fffff880`03362c80 fffff800`02c7f161 : fffff880`01087e00 fffff800`02d04b70 fffffa80`09d78040 00000000`00000000 : nt!WheapWorkQueueWorkerRoutine+0x25
fffff880`03362cb0 fffff800`02f15166 : 00000000`00000000 fffffa80`09d78040 00000000`00000080 fffffa80`09d5c040 : nt!ExpWorkerThread+0x111
fffff880`03362d40 fffff800`02c50486 : fffff880`0316a180 fffffa80`09d78040 fffff880`031750c0 00000000`00000000 : nt!PspSystemThreadStartup+0x5a
fffff880`03362d80 00000000`00000000 : fffff880`03363000 fffff880`0335d000 fffff880`04173540 00000000`00000000 : nt!KxStartSystemThread+0x16


STACK_COMMAND:  kb

FOLLOWUP_NAME:  MachineOwner

MODULE_NAME: hardware

IMAGE_NAME:  hardware

DEBUG_FLR_IMAGE_TIMESTAMP:  0

FAILURE_BUCKET_ID:  X64_0x124_GenuineIntel_PROCESSOR_MAE_PRV

BUCKET_ID:  X64_0x124_GenuineIntel_PROCESSOR_MAE_PRV

I tried googling my problem, but I've found nothing (or maybe I don't know where to look). any idea, suggestion and whatnot is appreciated.

[Neat Hedgehog]

First, check your RAM using memtest86. Bad / dying RAM modules are responsible for a lot of cases where a system starts randomly hanging up, or spontaneously restarting, and you typically won't get much feedback on the hangups from the vent viewer in these cases.

You can get it here: http://www.memtest86.com/

[Falkor]

I tried using memtest86, but the program will not run. I read that it doesn't work if the system has more than 4 GB, which mine does. on the other hand, I used memtest86+ and it passed 5-6 hours of test. is there any major difference between the two? I will give memtest86 another try, but I don't have much room/time to work with the pc right now (taking out the ram, test it one by one, etc.). will report as soon as I can; in the mean time, I've updated the motherboard bios (wonder if that's going to help).

[chikorita157]

One suggestion is to take out all the RAM modules and leave it so it's under 4GB of RAM and then test each one by swapping it... or if you want, take all of them and put one in and run the memory test and swap it until you find the defective one. This is time consuming, but you can easily find out the ram module that is defective.

[Neat Hedgehog]

oops, memtest86+ is the one I meant to link to, since most PC's have 4gb+ these days (I've got 6gb, myself). No, there's not really much difference that you'd need to worry about, except that memtest86+ runs with over 4gb of RAM.

5-6 hours with no errors is a good sign. Personally, when I'm trying to find bad modules, I like to let it run overnight. Sometimes it will catch things on the third or so pass that weren't problems the first time. Typically 6 hours is plenty of time, though.

If your system RAM really is fine, I'm tempted to suspect your video card RAM, because you said it will sometimes hang with the screen full of colored lines. It could be that your video card RAM is messed up, and so it loses communication with the rest of the system. Now, if it were just a fluke error, the system should be able to restart the drivers and bring it back online, but if it fails, it will typically reboot with no warning or blue screen.

That would also make sense because Memtest86+ would pass with no problems, since it only tests system memory, and not video memory.

What sort of graphics card are you running? Integrated / dedicated? Might not be a bad idea to swap graphics cards for a little while and see how it behaves.

[Claies]

The weird multicolored lines sound like your video card died really hard. Depending on whether your motherboard has integrated video capabilities, take your card out or find a spare video card and see if your PC runs fine with it.

[Falkor]

Originally Posted by Neat Hedgehog

oops, memtest86+ is the one I meant to link to, since most PC's have 4gb+ these days (I've got 6gb, myself). No, there's not really much difference that you'd need to worry about, except that memtest86+ runs with over 4gb of RAM.

5-6 hours with no errors is a good sign. Personally, when I'm trying to find bad modules, I like to let it run overnight. Sometimes it will catch things on the third or so pass that weren't problems the first time. Typically 6 hours is plenty of time, though.

If your system RAM really is fine, I'm tempted to suspect your video card RAM, because you said it will sometimes hang with the screen full of colored lines. It could be that your video card RAM is messed up, and so it loses communication with the rest of the system. Now, if it were just a fluke error, the system should be able to restart the drivers and bring it back online, but if it fails, it will typically reboot with no warning or blue screen.

That would also make sense because Memtest86+ would pass with no problems, since it only tests system memory, and not video memory.

What sort of graphics card are you running? Integrated / dedicated? Might not be a bad idea to swap graphics cards for a little while and see how it behaves.

if I don't have to take my computer apart and test the RAM one by one, it will save me a lot of time.

I have an ATI Radeon 4830 graphic card. Bought it 6-8 months ago together with the other parts of the computer. it has the latest drivers (clean install of the computer, too).

now the problem is, I don't have a spare graphic card (not integrated video, either). I think it's still under warranty; I will try to return/change it as soon as possible, though. kinda wondering, is there a way to test the video RAM so I can be sure that's the issue with my computer?

[Neat Hedgehog]

You could use Video Memory Stress Test 1.7 http://majorgeeks.com/Video_Memory_S...st__d5896.html

I've never personally found anything wrong with my vram using this tool (all the cards I've tested with it were known-good), so I guess I can't say how effective it is. It hasn't given me any false positives, though, so if the tests come back clean on it, the vram is probably ok.

There's also the possibility that it's the main gpu on the graphics card, too, though, so even if the test comes back ok on that, the graphics card could still be at fault. The only reason I suggested the vram in particular was that gpu malfunctions tended to be slightly more dramatic than a screen full of lines and don't always cause the card to stop rendering, whereas that sounds more like the frames are being corrupted in memory which would result in a dropped connection.

[Falkor]

I tried testing the video card with the program you suggested and it didn't show any errors during the first pass. I think it might require a few more passes, though (since I forgot to make it a continuous test, ).

for the time being, I think I'm going to get another video card and see if the same problem persists.

[npcomplete]

It looks very much like bad hardware.

Did you check your cpu's and motherboard's temperature?
is the heatsink tight?
is it relatively clean inside?

Did you fiddle with the machine before? like add or changed cards?
you should also make sure nothing is shorting inside. i.e. no metal (even metal flakes) lying on the board.


For memtest86+, you should let it run for at least 1 full pass, preferably two.

[TCman]

Hi,

It just may be the power supply unit causing the freezes/crashes. An underpowered psu/bad quality psu can cause the symptoms you are having now. What psu do you have? And what are the other specs of the PC, e.g. motherboard, cpu, etc.?

[Falkor]

oops, forgot to put the specs of my computer: (here's the thread, in case someone's was wondering).

• Video Card: SAPPHIRE 100265HDMI Radeon HD 4830 512MB 256-bit GDDR3 PCI Express 2.0 x16 Video Card - $99.99
• Power Supply: CORSAIR CMPSU-550VX 550W ATX12V V2.2 - $89.99
• Case: Antec Three Hundred Black Steel ATX Mid Tower Computer Case - $59.95
• Internal Hard Drive: Western Digital Caviar Black WD6401AALS 640GB 7200 RPM 32MB Cache - $79.99
• Thermal Compound: Arctic Silver 5 Thermal Compound - OEM - $6.99
• DVD Burner: LG Black 22X (CAV) DVD+R 8X DVD+RW 16X DVD+R DL 22X - $24.99
• CPU: Intel Core i7 920 Nehalem 2.66GHz 4 x 256KB L2 Cache 8MB L3 Cache LGA 1366 130W - $288.99
• Motherboard: ASUS P6T LGA 1366 Intel X58 ATX Intel Motherboard - $239.99
• RAM: G.SKILL 6GB (3 x 2GB) 240-Pin DDR3 SDRAM DDR3 1333 (PC3 10666) Triple Channel Kit - $96.99
• CPU Fan:COOLER MASTER Hyper N 520 RR-920-N520-GP 92mm Sleeve CPU Cooler - Retail - $49.99

it's almost the same except that I had to purchase the CORSAIR CMPSU-650TX 650W, instead of the one listed (because it was out-of-stock at the time of purchase).

Originally Posted by npcomplete

It looks very much like bad hardware.

Did you check your cpu's and motherboard's temperature?
is the heatsink tight?
is it relatively clean inside?

Did you fiddle with the machine before? like add or changed cards?
you should also make sure nothing is shorting inside. i.e. no metal (even metal flakes) lying on the board.


For memtest86+, you should let it run for at least 1 full pass, preferably two.

I use speedfan to check the temperatures on the computer and they are usually around 35º-40º C. I don't remember seeing the cpu temperature ever going over 55-60º C (either when I'm doing something intensive or when it's a really hot day).

it's a bit dusty, I guess. but I don't have the tools/time to clean it right now. haven't changed/moved anything inside, either.

I ran memtest86+ for 5-6 hours, more than two full passes, actually.

as I said, the best I can do right now is get another graphic card and see if the same problem persist. I get the feeling it's the graphic card the one that is causing me all these issues.

[TCman]

Hi,

That Corsair TX650W (Seasonic OEM) shouldn't give you problems since it's a high quality unit.

I guess nothing is overclocked and everything is at stock. Overclocking can cause systems to go unstable due to insufficient "juice". Also too tight timings of the system ram can cause freezes/crashes.

If nothing is overclocked and system memory is not over-optimized which can cause the issues, then there's no overheating, it also seems no faulty memory modules are installed, while the psu is fully sufficient for the whole pc system, then I think the videocard, the HD4830, somehow has become defective/faulty. Just like you said, rma it as soon as possible.

Hope this can help you,

Good luck!

[Mukki]

Just because you've already tested everything else without any failures showing up - although it's very unlikely to be the problem, there's no harm in playing it safe and taking a few hours to run HDD diagnostics.

Download and burn the ISO to CD, boot to it, run the quick test - if it passes, run the extended.

Again, if you're getting video issues, it's most probably not going to be a HDD-related problem, but it's always good to know that the drive's in good condition - especially if you rely on that drive a lot.

[Wervy]

Hirens Boot CD will provide any tools you would need for testing for a problem outside of windows and some in.

If you cant nail it down you might want to get yourself a stress test of some sort (there are free ones if you google around) and just run the tests one by one on each piece of hardware to see if you can make it reboot.

[Falkor]

got myself an ATI HD 4350 and I swapped it with my ATI HD 4830. The computer seems fine for now. haven't encountered any problems yet.

I'm trying to RMA the card, now. sapphire's customer service is not very intuitive, I would say. let's see how long it takes...

thank you guys for all the help so far!

[Falkor]

ok, Sapphire sent me another HD 4830, which is apparently an used card (I asked for a replacement). I put the video card in, along with the HD 4350 that was already inside my computer. everything seemed to work fine---but that was yesterday. today, I woke up, booted up the computer and went to the bathroom. when I came back there was no signal on the monitor.

took out the HD 4350 and left the HD 4830 as the only video card inside my computer; the problem persisted. it would POST and get to the windows logo, and then shut down. When I try to start up the computer again, there would be no signal on the monitor. I've been switching between the two video cards. the system works fine with the HD 4350, but it will not work with with the 4830. I took a look at the Event Viewer and it mentions an overheating problem. After careful inspection on the video card, a red light is on, the one associated with overheating. the strange thing is, the fan is still spinning. When I checked the temperatures yesterday (when the card was working fine), they were around ~30 ºC.

I tried googling the issue and there seems to be people with a similar problem, which oddly enough, appears to be associated with Windows 7 (the OS I'm using) and a possible conflict of drivers. I remember getting to my desktop in safe mode or in normal mode, though it would be for a short while since the computer shuts down. unfortunately, it didn't cross my mind to check the temperatures...

it seems to be an overheating problem, but the card was working fine yesterday and the fan is still spinning. I will try to get into safe mode and check the temperatures, that's one option, I think.

do you guys have any thoughts on this?

I will not buy anything from Sapphire, ever again...

[Neat Hedgehog]

Did you uninstall the ATI drivers prior to using the new card? I don't believe this would (normally) cause a false-positive about overheating, but it would be a good idea.

For now, you could assume that yes, your card really is overheating. To make sure this isn't the problem, you can manually adjust the fan speed up to around 80% (more than enough to keep a card from overheating) and try running it for a while.

If it isn't really overheating, then it's likely either a driver issue, or the card's got a screw loose.

[Kurz]

Uninstall the old drivers.

Are you planning to Crossfire them?
If not just keep the 4350 out of there.
Install only the 4830 with the appropriate drivers.

[Falkor]

sorry for the late reply...

I must admit that I didn't uninstall the drivers at first, because the two cards use the same display drivers for windows 7 (and they were working fine together). I tried to do what you guys suggested: uninstalling the drivers and installing them again; but now, when I turn on the computer, it will not POST (no signal on the monitor either) and the red light I mentioned before will just stay ON.

I get the feeling this graphic card I've received is defective (with the one I originally had the system will at least POST and get to Windows, and I haven't made a whole lot of changes to the computer, so to speak). anyway, I talked to the guys on tech-support/RMA, and they seem willing to replace my card again. hopefully I have a better luck this time.

edit: fixed the wording