Crashing and freezing on my homebuilt PC

[ttbwpg]

Hi all,

I built this system last fall as a second pc and have recently been having problems.

MOBO: MSI RD480 NEO2
CPU: AMD64 4000+
RAM: 3X 1gb
OS: xp pro sp2
PW: 450w
VC: evga 8600GT

Recently, my pc started acting up. Out of nowhere, the lcd monitor would go black a read "no signal". At first, I had trouble getting it to boot.

My first thoughts were virus or heat issue. Performed a full virus scan with avg, nothing there. I opened an cleaned things up inside. I blew our all the dust and checked connections, removed and reseated my vid card. The problem remained. At first I thought it was flash related as it seemed to crash when I clicked on a flash page. I don't think that was the case as after a few days of trying to figure it out, the freezes seemed to be random.

I could do a reformat, but I haven't determined whether it's a hardware or software issue.

My question is where do I go from here?

 

[Shadow703793]

What exact PSU do you have?

This looks like a hardware issue and not a software issue. Download and run Memtest86+ to see if RAM could be a cause.

Tbh, this looks like either the card or the LCD is dying or could be PSU if it's low quality.

 

[ttbwpg]

OKIA 450axt
generic type psu

 

[shortstuff_mt]

+1 to running memtest86+ for a few hours to test for RAM errors. I've seen dead/dying RAM cause this type of problem a few times. I would take out all but one stick of RAM and try running the system like that. If it doesn't work, swap out the stick for another one. You can also try each stick of RAM in each RAM slot to try to eliminate a dead RAM slot on the motherboard.

This checklist has some more troubleshooting ideas:

http://www.tomshardware.com/forum/261145-31-build-post-checklist

 

[ttbwpg]

I'll run memtest a little later. I checked to cpu temp and it was running hot at 54c right after start up. I have a big fan on it right now and the temp is down to 26c. I have a decent cpu cooler. Is it possible that this machine has always run at this temp? If it doesn't crash over night, then cpu temps would seem to be the issue, yes?

 

[Onus]

That PSU is offal, although your rig shouldn't be drawing even 200W from it.
Did you reset your BIOS to defaults? Try that if you didn't, otherwise go through the checklist that shortstuff posted.

 

[shortstuff_mt]

Yup, 54c is way to hot for an idle temp. You need to remove the HSF, clean off the old thermal compound, and reinstall the HSF with fresh thermal compound. Make sure to carefully read the instructions on how to correctly install the HSF.

 

[ttbwpg]

I pulled one stick of ram. The remaining matched pair seem to be doing better. The one stick felt hotter than the others when I pulled it. I didn't get a crash until I ran FEAR2. I put a room fan on the job and FEAR2 ran without crashing. Where to from here?

As for temps, here is where is is now with the sides off. The gfx card looks a little hot.

Field Value
Sensor Properties
Sensor Type Winbond W83647HF (ISA 290h)
GPU Sensor Type Driver (NV-DRV)

Temperatures
Motherboard 36 °C (97 °F)
CPU 44 °C (111 °F)
Aux 42 °C (108 °F)
GPU 47 °C (117 °F)
WDC WD5000AACS-00ZUB0 30 °C (86 °F)

Voltage Values
CPU Core 1.28 V
Aux 3.22 V
+3.3 V 3.30 V
+5 V 5.56 V
+12 V 13.25 V
-12 V 3.02 V
-5 V 3.30 V
+5 V Standby 4.95 V
VBAT Battery 3.10 V
Debug Info F FF FF FF
Debug Info T 36 44 42
Debug Info V BF C9 CE CF DA DA DB (03)

 

[Onus]

Hopefully the sensor is not accurate, because a +12V reading of 13.25V is way too high, over +10%.

 

[ttbwpg]

I checked it again. Here are the latest temps and such from Everest:

Field Value
Sensor Properties
Sensor Type Winbond W83647HF (ISA 290h)
GPU Sensor Type Driver (NV-DRV)

Temperatures
Motherboard 32 °C (90 °F)
CPU 39 °C (102 °F)
Aux 42 °C (108 °F)
GPU 43 °C (109 °F)
WDC WD5000AACS-00ZUB0 31 °C (88 °F)

Voltage Values
CPU Core 1.28 V
Aux 3.22 V
+3.3 V 3.30 V
+5 V 5.54 V
+12 V 13.56 V
-12 V 3.52 V
-5 V 3.60 V
+5 V Standby 4.95 V
VBAT Battery 3.10 V
Debug Info F FF FF FF
Debug Info T 32 39 42
Debug Info V BE C9 CE CE DF E0 E1 (03)

Everest might be off on the temp as my MSI software has the =12v at 11.93.

It looks like the machine crashed again over night.

 

[Onus]

I think you should replace your PSU. You don't need a big one (450W is fine), but you need better quality. Choose one from Antec, Corsair, PC Power & Cooling, Seasonic, or Enermax.