PC freezing at random

[Wally's World]

Non-CM related problem if you don't mind.

Three days ago, my PC all of a sudden started freezing at random times. Could happen anytime, during a game of CM, cruising the web or viewing the desktop. Can happen 10 minutes after logging on or after 3-4 hours of no problems. Everything will just freeze and ALT-DEL-CTRL will not work.

The freezing will also occur while I am in just in BIOS or even while Windows XP is loading (during the loading screen).

Only solution is to hit the Reset button on the case to reboot the machine. It usually reboots fine, but once in a while after hitting the Reset button and trying to reboot, the motherboard's (Epox 8RDA+) alarm will go off with a post code of 25. At this point, I have to turn off the PC with the power button at the back of the machine and then restart.

Post code 25 for this motherboard shows up as 'Reserved' in the Epox manual. However research on the web indicates possible issues:

1) BIOS is corrupted at which I should clear CMOS

2) Video card problems (GeForce 6600GT)

3) PSU issues

I don't think it's an overheating issue as the CPU temp hangs around 54-60C and the video card around 64C. Now I do overclock my video card by about 7% and when I upgraded my PC last year with a new CPU (Athlon 2800+ Barton) and the 6600GT, the tech told me that the PSU is almost at borderline with the hardware. I have no idea what the PSU other than it's one of those generic PSU's that was installed with the original build of the machine 3.5 years ago.

I did add another RAM module a couple of weeks ago and the machine has been running fine until last Friday. And it was running fine with the new CPU and video card since last fall.

I did clear CMOS last night and the machine booted and ran fine for about 30 minutes while I changed some of the default BIOS settings to what I want. And then I shut it down for the night. Maybe clearing CMOS will solve the problem.

Or is it the overclocking of the video card is putting pressure on the PSU? But I have overclocked the card since I installed it last fall.

Could the PSU be failing or be too small for the hardware?

Any other suggestions or ideas would be truly appreciated.

Many thanks to any and all replies.

Specs:

Epox 8RDA+

Athlon 2800+ Barton

1 gig PC2700 Ram (1x512, 2x256)

60G hard drive

GeForce 6600GT (overclocked manually by 7%)

1 50x CD-RW drive

1 floppy drive

1 Vantec 80mm case Stealth fan

[ April 04, 2006, 10:24 AM: Message edited by: Wally's World ]

[Schrullenhaft]

The PSU would sound like the most likely culprit. With it being 3.5 years old it might be having little "hiccups" every-so-often that might cause your problems. Either problems providing the requested amperage or filtering problems due to aged capacitors that allow for too much 'noise' on the voltage that could cause the motherboard/videocard problems.

Newer power supplies are supplying much more amperage for the +12V line since that is what newer CPUs and videocards use heavily. However the Epox 8RDA+ and the Athlon XP 2800+ Barton still use the +3.3V line for power (there's no need for the 4-pin 12V line you see on Pentium 4 motherboards). The 6600GT on the otherhand does put more strain on the +12V line from the power supply. Overclocking it will also draw more amperage on the +12V line too, particularly overclocks of the core (how much more amperage, I'm not sure).

So a distinct possibility is that the power supply might not be providing the voltages/amps as necessary and a voltage drop may cause the lockup (of the videocard or motherboard). However this is only a guess. I do doubt that the CMOS would be the culprit. Typically it either has an incorrect setting or it doesn't and the failure should be fairly immediate (though there are always those stange cases).

Will you get a lockup consistently while using the computer (however long it may take) or do the lockups occur even more random than that (a random interval of days, etc.) ? What are the 'DC Output amperages' of your power supply for the +5V, +12V and +3.3V (they'll typically be printed on a sticker on the power supply) ?

Additional things to try: 1) Reduce your videocard's clock to it's standard setting 2) Swap your 6600GT for the previous videocard (if you still have it) temporarily 3) A more expensive option, get a new, preferably name-brand power supply with more wattage.

If reducing/eliminating the overclock on the videocard makes the system stable, then upgrading the power supply should (generally speaking) allow you to overclock again. If you tend to upgrade your computer incrementally (new motherboard and CPU now, videocard later, drives later, etc.), then a new power supply could enable the next upgrade (24-pin connector, SATA connectors, SLI/PCI-E power connectors, higher +12V amperages, etc.). However a good power supply can be fairly expensive, spending near or over $100 is quite easy.

[ April 04, 2006, 08:14 AM: Message edited by: Schrullenhaft ]

[Redwolf]

PSU most likely, or mainboard.

60 C is very much the upper level for an AMD64, you might want to look into that.

Anyway, you want to replace the PSU about right now. Overclocking CPUs or GPUs does indeed increase power consumption by a lot. I can drive my 120 watts full-load-at-stock-speed PC to take 280 Watts just by overclocking the CPU (with overvolting). Granted that is pretty extreme. But the point is, a 2000 MHz CPU overclocked to 2400 MHz usually takes a lot more power than a native 2400 MHz CPU.

Given the symptoms you describe, given that you overclock, given that you almost certainly have a complete piece of junk as a PSU It is likely that your PSU is mortally wounded and ready to kill all components connected to it (possibly including you) at the most inappropriate moment.

[Wally's World]

Thanks Schrullenhaft and Redwolf both for your posts. I did try running the video card at the stock speeds and I still had random freezes even in just XP's loading screen. Therefore based upon your advice, I need a new PSU.

I am either looking at the Enermax 420W EG425P-VE SFMA Noisetaker:

Enermax PSU

Or the Antec TruePower 2.0 430W:

Antec PSU

I don't think I need a more powerful PSU, but can I get away with a smaller unit like a 350W or a 380W based upon my config?

Sorry, but PSU info is foreign to me although I am learning.

[Schrullenhaft]

Either one of those power supplies should work for your current setup. The Antec is a bit more future-proof in that it offers the 20+4 power connector that can power future motherboard upgrades (Intel or AMD). The wattage and amperages that both offer should work fine.

[Redwolf]

Some people bash Antec. Antec resells back and good manufacturers, I'm not sure what the Truepower 2.0 line is coming from.

In any case I bet the Enermax has more real power anyway despite the lower wattage rating.

[Wally's World]

Many thanks to both of you for your help. This weekend, I'll go buy a PSU and install it.

[Hpt. Lisse]

Should do it - though I must say, 60C for a Barton is climbing up there (esp. if that's an idle reading.) My OC'ed Barton 2500 (to 2.1 GHz, and 360 FSB) may hit 50C after CMing for hours (w/Thermaltake Silent Boost HSF, running toward the slower/quieter end of the fan control.) I also have the same Mobo as you do.

You may wanna throw down for a decent 500W PS, just in case you upgrade your rig anytime soon...

[Wally's World]

Hpt. Lisse, my Barton usually stays around 49-50C at idle and climbs as high as 54C at heavy load. I've only seen it climb once to 60C and that was when my old case fan was going. I've replaced it since. Last night, the CPU was hovering around 50C after playing CM.

[Hpt. Lisse]

Gotcha.

Have you run MemTest since the new memory module?

Have you physically checked the mobo for bulging capacitors?

[Wally's World]

Have you run MemTest since the new memory module?

Yes, MemTest ran fine with the new module.

Have you physically checked the mobo for bulging capacitors?

Motherboard looks good. Can't physically see any problems.

Actually my PC has been running just fine the last three days. No problems or issues whatsoever. I even overclocked my 6600GT again by about 5%.

Will continue to keep an eye on it.

[Wally's World]

After replacing the PSU with a more powerful on, I'm still having problems with random freezes and then sometimes a C1 post code on my 8RDA+ motherboard. A close inspection showed about 5 capacitors with crude on the top of them. Does this mean my 8RDA+ is toast?

*** Update ***

Weird thing here. I decided to experiment with the voltage settings in the BIOS and by bumping up the DIMM voltage to 2.90V (from first the default voltage and then 2.77V), my system seems to have stabilized. Before I couldn't run it for 5-10 minutes without a freeze, now it's been running for almost an hour with no problems. Just hoping it will stay that way.

I also changed the memory timings to 3-3-3-10 to stabilize the memory. From what I've read on the internet, 10 or 11 seem to be the best timing for my motherboard (nForce2) and it appears that the memory timings don't have much if any affect on system speed anyway.

Does bumping up the DIMM voltage and upping the memory settings make sense? Admittedly I am just grasping in the dark.

Many thanks for any replies.

[ April 22, 2006, 11:40 AM: Message edited by: Wally's World ]