Sign in with
Sign up | Sign in
Your question

Dead pc - bad motherboard?

Tags:
  • Homebuilt
  • CPUs
  • Systems
Last response: in Systems
Share
June 10, 2011 1:42:13 AM

Hi,

I built a PC three months back. I mostly use it for software development and media encoding. Around 2 weeks back, it started freezing randomly. It would freeze within 2 minutes of booting into windows 7, sometimes in BIOS, and other times nothing would come up on screen but all components would be on. The only common thing was high cpu fan speed. It's around 6k rpm and temperature hovering between 40 and 50.

It then started working 3 days back and cpu fan speed was normal but Today it finally gave up ghost. Now nothing comes on; not even POST works. I can see the cpu and other lights come on but nothing on monitor. I disconnected everything leaving just cpu but still no POST. All I get is high pitch CPU fan sound. Also, CPU fan speed is not constant. No beeps because I don't have motherboard speaker, which I am planning to get tomorrow

Specs:
AMD Phenom II X4 965
Stock cooler
Gigabyte GA-880GM-UD2H matx MB
PNY 8GB DDR3 memory
WD 1 TB hard drive (blue) 7200 rpm
Radeon HD 5770 XFX
Asus PCE-NI3 internal PCI-x wireless adapter
Corsair CX600 PSU
Windows 7 Professional
Samsung 12x blu ray drive
Roswill card reader

I have never overclocked it and have left most of the BIOS on default. I think it's the motherboard but I would really like to confirm it. Any help is appreciated.

Thanks!

More about : dead bad motherboard

a c 113 B Homebuilt system
a b à CPUs
June 10, 2011 2:42:39 AM

Was it cooler in the room suddenly three days ago?

What were you using to test the temps, just the BIOS readings? Those are often very low, and it's better to use a program like HWMonitor to get the temps.

Do get the speaker.

If there was a problem with the CPU cooler, say a defect or foreign particle between the heatsink and CPU, that might cause an issue. You may want to inspect both surfaces for flatness. A single-edged razor blade is good for that. Use denatured alcohol and a coffee filter to clean up the surfaces before applying some new thermal paste.
m
0
l
a c 136 B Homebuilt system
a c 87 à CPUs
June 10, 2011 2:49:58 AM

The case/motherboard speaker is a great next move .

Google "bios beep codes"
m
0
l
Related resources
June 10, 2011 11:29:47 AM

Room temperature has been more or less constant. I don't think cooler is the culprit because I encoded a movie successfully 2 days back, which is really the ultimate stress test.

I'll report my findings after installing BIOS speaker.

Thanks.
m
0
l
June 17, 2011 1:46:05 AM

I finally got the speaker but situation has not really improved. I get repeating 4 long beeps. It's AWARD bios and apparently there is no such code. The closest I see is "repeating beeps", which means bad memory. I have had run memtest before POST went dark and it reported success. I ran it for one pass only, however. It is PNY Optima 8GB DDR3 1333 memory.

Any ideas?
m
0
l
a c 113 B Homebuilt system
a b à CPUs
June 17, 2011 1:54:43 AM

Try each stick of RAM individually.
m
0
l
June 17, 2011 2:34:35 AM

Proximon said:
Try each stick of RAM individually.


I have tried that too, all combinations of slots and ram. If I completely remove all RAM, I get 4 long and one 1 short beep.
m
0
l
a c 113 B Homebuilt system
a b à CPUs
June 17, 2011 3:28:34 AM

So the beeps are different with and without RAM.

I would want to try switching out the PSU at this point, if possible, to rule that out. It's not too likely that both sticks failed at once so RAM can't be too likely any more. That leaves board, CPU, PSU.

m
0
l
June 17, 2011 11:19:32 AM

Unfortunately, that's not an option as I don't have spares. I have opened a ticket with Gigabyte, lets see what they have to say.
m
0
l
June 23, 2011 2:14:38 PM

Finally heard from Gigabyte and they suggested that long beeps are usually sign of memory fault, and requested me to try with a single 2 GB memory module. I reluctantly bought one yesterday and ... nothing changed. Same long 4 beeps. Now it is down to cpu, motherboard and power supply, and I have a hunch that it is MB. I have reported my finding to Gibabyte; waiting for reply.

This is my first build and I have to say that it is astonishing that methods for locating faults are as primitive as shooting in the dark.
m
0
l
a c 113 B Homebuilt system
a b à CPUs
June 23, 2011 8:57:02 PM

Usually it's easier for someone with experience, but occasionally it does come down to that.

Mostly any issues in "newbie builds" come down to user error, but there will always be the 5 to 10% that encounter an actual hardware failure up front. Just don't assume that it's common. I've built 6 machines from newegg this year and not one of them had a serious hardware issue... there is just one that will not sleep properly.
m
0
l
June 24, 2011 1:48:43 AM

This should be down to exact science. I have the exact error code beeping out of the heart of the motherboard and yet the manufacturer is playing a guessing game. I think being a programmer has exacerbated my frustration.
m
0
l
a c 113 B Homebuilt system
a b à CPUs
June 24, 2011 2:09:47 AM

Any sort of error code generation is going to rely on a certain amount of processing ability. If that ability is compromised then the error code generation is also compromised.

You can train people and buy equipment that will help you sort out these kinds of troubles, but it's not cost effective. Anyone in business will just swap out parts because it's easier. Any private builder will either need spare parts or friends with spare parts. ;) 
m
0
l
June 24, 2011 2:31:00 AM

Well, I had set out to build a machine with most bang for buck within a budget. Considering the time, and now spare parts involved, I think I feel defeated.
m
0
l
!