Sign in with
Sign up | Sign in
Your question
Solved

Why does my new computer crash? I'm here to learn.

Last response: in Systems
Share
October 25, 2011 10:25:52 PM

I recently built a new computer and I am using Windows XP SP3.

Here is the list of components:
Intel i3-2100 w/OEM heatsink and fan
MSI PH67S-C43
PNY Optima 8GB (2 x 4GB) DDR3 1333MHz
Western Digital SATA 6gb 500gb
Asus DVD re-writable drive
Corsair CX500
Radeon HD 6770 1gb DDR5

Now the issue. Was getting blue screens and now it just shuts down and reboots at random times. I can play video games and it has not crashed during those at all, but it will crash when doing defrag, using mozilla, right after boot up, and sometimes just sitting there.

Any ideas? I can follow directions if you have advice for checking components. I ran memtest86 and each DIMM checks fine after 18 passes.

I have also tried using Windows 7 64 bit and it gave me BSODs even faster than with XP.

More about : computer crash learn

a b B Homebuilt system
October 25, 2011 10:32:00 PM

bad psu, bad mobo, or bad psu and mobo.
m
0
l
a c 136 B Homebuilt system
October 25, 2011 10:33:33 PM

XP 32 bit with 8 gig of RAM?

I would check in BIOS that the RAM is running at its specified speed/ timings and voltage .
If that voltage is greater than 1.5 volt then you should cut the voltage to 1.5 volt and lower the frequency till it runs
m
0
l
Related resources
October 25, 2011 10:33:54 PM

Any way to check PSU of MOBO other than swapping the item out for new item?
m
0
l
October 25, 2011 10:41:17 PM

Windows XP 32 only recognizes 2.99gb but that doesn't seem to be a problem.

I couldn't figure out how to check memory voltage. Would it be under "over-clocking"? I can write down anything that is under question...just ask.
m
0
l
October 26, 2011 12:57:08 AM

I seem to have less issues with crashes when I change the CPU voltage from 1.136 to 1.140 (I read about doing this somewhere and not sure why I've done it or why it might possibly have helped). Any ideas why?
m
0
l
October 26, 2011 1:25:53 AM

casualbuilder said:
bad psu, bad mobo, or bad psu and mobo.

why do you think that? My vote is for memory incompatability http://au.msi.com/file/test_report/TR10_2324.pdf that ram your using isnt on the compatable list for the motherboard. just because it passes memtest doesnt mean it wont cause a BSOD.
m
0
l
October 26, 2011 1:42:40 AM

It could be numerous things. Run ChkDsk prior to start up. Make sure your bios is updated. Double check vid card drivers for the latest version. Try running with just 2GB's (or 4 if 4GB modules) and see if that helps. Try default keyboard drivers for awhile if you have a gaming keyboard. Update firmware on optical drives.
m
0
l
October 26, 2011 1:44:09 AM

Run it the CPU at default settings, that's an obvious one. If it doesn't run at its rated, it won't run under an OC.

I would also check your PSU, it almost sounds like a PSU error. You need a tester. They're cheap.
m
0
l
a c 136 B Homebuilt system
October 26, 2011 1:56:54 AM

iam2thecrowe said:
why do you think that? My vote is for memory incompatability http://au.msi.com/file/test_report/TR10_2324.pdf that ram your using isnt on the compatable list for the motherboard. just because it passes memtest doesnt mean it wont cause a BSOD.


pretty much what I was thinking too .
VERY unlikely to be the mb or psu

The OP needs to go back in to BIOS and find the page with memory frequency, timings and voltage
m
0
l
October 26, 2011 2:03:57 AM

iam2thecrowe said:
why do you think that? My vote is for memory incompatability http://au.msi.com/file/test_report/TR10_2324.pdf that ram your using isnt on the compatable list for the motherboard. just because it passes memtest doesnt mean it wont cause a BSOD.


I didn't even know they had a list of compatible memory. I have just ordered Corsair product# CMX8GX3M2A1333C9 as an replacement.

Thank you.
m
0
l
October 26, 2011 2:14:31 AM

JonnyDough said:
It could be numerous things. Run ChkDsk prior to start up. Make sure your bios is updated. Double check vid card drivers for the latest version. Try running with just 2GB's (or 4 if 4GB modules) and see if that helps. Try default keyboard drivers for awhile if you have a gaming keyboard. Update firmware on optical drives.


I don't remember how to initiate check disk. I should be able to look this up online though.

I also don't know how to check to see if the Bios needs to be updated or how to update it. Any links I could read to learn?

I have run with 4gb only and it still does it.

Standard PS2 keyboard and optical mouse. How do I update firmware?

JonnyDough said:
Run it the CPU at default settings, that's an obvious one. If it doesn't run at its rated, it won't run under an OC.

I would also check your PSU, it almost sounds like a PSU error. You need a tester. They're cheap.


I have been running it at default settings, I don't care to OC anything. I changed the voltage because I read somewhere that it might prevent the crashes.

I ran this PS on my last setup and had no issues. It was an AMD 64 3200+, 1gb of PC2700, Asus K8V-SE Deluxe, and Radeon X1300. So do you think it's possible it's still bad?
m
0
l
October 26, 2011 2:31:18 AM

Your default CPU voltages should be fine. In fact, go into bios settings and go ahead and just set back to default. The issue is 99% likely to be elsewhere.

ChkDsk is run via the Windows Installation disk in the recovery option. You have to use the command prompt and type chkdsk /r

It can take hours, mostly depending on size of disk.
m
0
l
a b B Homebuilt system
October 26, 2011 10:29:00 AM

iam2thecrowe said:
why do you think that? My vote is for memory incompatability http://au.msi.com/file/test_report/TR10_2324.pdf that ram your using isnt on the compatable list for the motherboard. just because it passes memtest doesnt mean it wont cause a BSOD.



I had my old rig's MoBo go bad, same issues. Was going off of personal experience. Still using the same ram i originally had in it. Also, just because the RAM isn't listed on the "compatibility" list, doesn't mean it wont function fully with the MoBo. Again, personal experience was the source for my response. I apologize for giving wrong information, im still learning too.
m
0
l
October 26, 2011 2:14:03 PM

I ran chkdsk and it came out fine. (Took about 50-60min.)
m
0
l
October 26, 2011 2:17:30 PM

casualbuilder said:
I had my old rig's MoBo go bad, same issues. Was going off of personal experience. Still using the same ram i originally had in it. Also, just because the RAM isn't listed on the "compatibility" list, doesn't mean it wont function fully with the MoBo. Again, personal experience was the source for my response. I apologize for giving wrong information, im still learning too.


It's likely that not all the compatible memory is on that list, but since I had just bought all this stuff I figured now is the best time to remove memory from the "possible issues" list.
m
0
l
October 27, 2011 5:25:44 PM

schmiddr2 said:
I didn't even know they had a list of compatible memory. I have just ordered Corsair product# CMX8GX3M2A1333C9 as an replacement.

Thank you.



Compatible memory lists are only memory that is guaranteed to work which the motherboard maker has tested. It is certainly not even close to being all inclusive. About any memory modules that are within JEDEC spec should do. The biggest thing to watch for is ECC vs Non-ECC. Don't buy ECC for most boards. There is also software and hardware (built into the memory modules) ECC. It can get confusing.



http://www.jedec.org/about-jedec/member-list

Something interesting perhaps?
m
0
l
October 27, 2011 5:26:53 PM

schmiddr2 said:
I ran chkdsk and it came out fine. (Took about 50-60min.)


Then you've pretty much eliminated most boot sector errors. Good work. :)  Its likely a hardware issue then. I would guess that it could be your northbridge overheating. Your mobos Intel H67 cooler looks pretty weak. I've had similar issues on a board before, and once I replaced the OEM heatsink the system worked fine. Try putting a fan blowing directly on it, and/or using some AS5 and re-seating it. Make sure to use a credit card and scrape it smooth over the top of the chip before re-seating. Its not glued on with adhesive is it?
m
0
l
a b B Homebuilt system
October 27, 2011 5:27:30 PM

^ +1

Great site JonnyDough
m
0
l
a c 91 B Homebuilt system
October 27, 2011 5:50:57 PM

schmiddr2 said:
I seem to have less issues with crashes when I change the CPU voltage from 1.136 to 1.140 (I read about doing this somewhere and not sure why I've done it or why it might possibly have helped). Any ideas why?


That's the problem - you're running an i3-2100. Unless it's an Intel CPU with a K next to the model number, it cant be overclocked or even remotely messed with. Reset your motherboard to default settings in the BIOS and run a system restore, that should fix it.

I'm actually surprised your motherboard didn't report overclocking errors when you start it up.
m
0
l
October 27, 2011 6:25:11 PM

JonnyDough said:
Then you've pretty much eliminated most boot sector errors. Good work. :)  Its likely a hardware issue then. I would guess that it could be your northbridge overheating. Your mobos Intel H67 cooler looks pretty weak. I've had similar issues on a board before, and once I replaced the OEM heatsink the system worked fine. Try putting a fan blowing directly on it, and/or using some AS5 and re-seating it. Make sure to use a credit card and scrape it smooth over the top of the chip before re-seating. Its not glued on with adhesive is it?


I have checked the CPU temp on reboot (in bios) and it stays under 50c; is that ok? I am using AS 5. The chip was clean and I don't know what you mean by adhesive but as far as I know it doesn't have any on it.

What is the "mobos Intel H67 cooler"?
m
0
l
October 27, 2011 6:32:17 PM

g-unit1111 said:
That's the problem - you're running an i3-2100. Unless it's an Intel CPU with a K next to the model number, it cant be overclocked or even remotely messed with. Reset your motherboard to default settings in the BIOS and run a system restore, that should fix it.

I'm actually surprised your motherboard didn't report overclocking errors when you start it up.


It actually seems to have less issues at this voltage. Hasn't crashed since I changed it to 1.140V.

In the bios it says do not set the voltage above 1.140V.

The reason I did this is that I read somewhere that it helped someone else with a similar setup. Not sure why they did it or why it helped, but it seems to work.

Any ideas why?
m
0
l
a c 78 B Homebuilt system
October 27, 2011 7:09:41 PM

I read through the thread and didn't see that the OP posted his PSU specs.

OP, if you didn't, could you please do so?

You could have the Dodge Viper of Video Cards and Processors and if you have a Dodge Neon engine in it you won't be going 0 to 60 in 4 seconds.
m
0
l
October 27, 2011 7:25:52 PM

It's in the first post. Corsair CX500. I believe it does close to rated, 500w.
m
0
l
a c 78 B Homebuilt system
October 27, 2011 7:46:31 PM

My fault, I guess I missed it. It is 500 when the manufacturer recommends at least 450 so it may be good enough.

Did you try putting your specs in the power requirements calculator?

http://extreme.outervision.com/psucalculatorlite.jsp

If it crashes when it is just sitting there, I would guess it has to do with driver problems.
m
0
l
October 27, 2011 8:15:52 PM

I had this same issue with an Radeon HD 5670, which maxes out around 60W.

The calculator says I can get by on 244W with the HD 6770.
m
0
l
a b B Homebuilt system
October 28, 2011 10:47:32 AM

schmiddr2 said:
I have checked the CPU temp on reboot (in bios) and it stays under 50c; is that ok? I am using AS 5. The chip was clean and I don't know what you mean by adhesive but as far as I know it doesn't have any on it.

What is the "mobos Intel H67 cooler"?





BIOS temps are always hotter than idle temps in Windows. However, LOAD temps are usually a little higher than BIOS temps. If your system runs better on 1.14v, then that is a good place to start for stability. If you OC'd your chip, which any chip can be OC'd, you may have to increase your cpu's voltage to stabilize it. Just because it isn't a "k" series doesnt mean it cant be OC'd. Keep your temps low, and watch how much voltage you are giving it. Voltage will instantly fry a chip, and possibly other components.
m
0
l
a b B Homebuilt system
October 28, 2011 10:49:11 AM

Your PSU is more than adequate btw...isnt that i3 2100 a 65W chip?
m
0
l
October 28, 2011 7:04:50 PM

The only change I have made in the bios is the CPU voltage. 1.136V to 1.140V .

casualbuilder said:
Your PSU is more than adequate btw...isnt that i3 2100 a 65W chip?


I figured it was enough. And part of the reason I chose the chip was because it's 65W.
m
0
l
October 29, 2011 5:17:48 AM

Its actually more important to have enough amperage on the rails than to have "X watts". I could go into more detail, but there's enough info online. Google it.

schmiddr2, as far as what I was talking about with the H67 cooler:

On your motherboard there are processing chips that control things. One of these is underneath that silvery/gray block which resides at the corner of your PCI-E x16 slot and your RAM slots. It has a blue thing across it and says "MSI". Kind of hard to miss.

Anyway, stick a little fan blowing across that, and reset your bios to default already. :p 
m
0
l
a c 78 B Homebuilt system
October 29, 2011 12:35:38 PM

The manufacturers suggest a certain wattage of PSU for a reason, because they think your PC will work best if that is what you use with it.

They aren't in league with PSU manufacturers to upsell products.

The calculator is really good, but it isn't perfect. For instance, it doesn't have my processor listed at all (AMD Phenom 2 x4 840).

A lot of the killer is the capacitor aging drop down box, if you ignore that you will get wattage requirements way below what you get if you use it. Your 244 quote makes it sound like you set aging to 0 which is pretty unsafe unless you only want to keep your PSU around for a year or less before buying a new one.

Put even a reasonable amount of aging on there like 25% and it will show much higher numbers for required wattage, a lot closer to the manufacturer's recommended numbers.
m
0
l
October 30, 2011 3:47:02 PM

I think I made a mistake explaining the CPU voltage change. Dang.

What I changed it from is Auto. And when set to auto the voltage varies. It can be at 1.1V or 1.136V or whatever, but it always crashes when set at auto. Why does this happen and what should the Voltage be set to?
m
0
l
October 31, 2011 9:52:47 PM

So received the new memory in the mail today. I decided to install Windows 7 64 since I know this memory is compatible. I set the CPU to 1.14V and set memory voltage to 1.492V and set the timing to 9-9-9-24.

I have not had any hickups or crashes.
m
0
l

Best solution

a b B Homebuilt system
November 1, 2011 9:22:32 AM

Set it to 1.14v. this should be good, and not too much.
Share
November 1, 2011 5:21:31 PM

If your system is crashing when your voltage is on "auto" you either have a bad bios, bad motherboard, or bad CPU. Unless of course, you messed with other settings. I had suggested clearing your CMOS...
m
0
l
November 2, 2011 4:02:44 PM

Thanks casualbuilder for giving me some confidence in the setting at 1.14V. It runs great at the settings mentioned in my last post. No issues so far.

JonnyDough
I updated the bios; it was very easy.

I tend to agree with you that something is wrong if the computer will not run on AUTO. But I'm not sure how to figure this out. And I never messed with any other settings

I will try clearing the CMOS.
m
0
l
November 3, 2011 6:49:12 PM

Let us know if it works!
m
0
l
May 5, 2013 10:27:05 PM

I cleared the CMOS, it still crashed on AUTO. So I still don't know why it does this, but it's been a year and a half still going strong using the manual voltage setting. Just wanted to let people know that in my case, the manual settings work fine long term and temps are normal on stock CPU heatsink.

I use this website so much for research that I felt I should reply and choose "best answer".

Thank all.
m
0
l
!