Faulty motherboard?

I have been using this MSI K9A2GM-FIH for a little over a month now and lately I have been having problems with it. For starters, the other day, the onboard LAN just simply quit working. It doesn't show up in Windows and I have tried everything I can think of. The activity light flashes yellow consistently. For a while I was running my CPU (Athlon x2 5400 black edition) at 3300 (275x12) @ 1.47vcore. It was running very stable and I saw no errors from P95 and Orthos after 12 hours of testing with both.
Now, it sometimes doesn't want to POST at all. It will hang when booting and after I reset the system, the clock settings will revert back to default and I will get an error message on POST (in red letters) saying that the previous overclock had failed. It even does this when I try to boot from stock clock settings. I do have the latest BIOS.
Any thoughts?

I fixed this issue. Apparently I went overboard with cable management and unknowingly crippled my system. I had tucked the 4 pin CPU power cable in the space between my video card and motherboard. The LAN chip is right there in that area and this cable was basically laying against the chip. Apparently interference from the voltage within the cable had crippled the chip and was also causing issues with the mobo in general.
I discovered this problem when I got the notion to take the motherboard out and inspect it for swollen capacitors and also wanted to reapply the AS5 properly. When rebuilding, I decided to leave the CPU power cable loose and not tuck it away as I had done previously. Viola, upon rebooting the Realtek LAN chip is back to life and I am no longer getting errors from BIOS during POST (kept telling me that my overclock had failed). I cannot reproduce that problem after booting 10 times repeatedly.

Or maybe something had just gotten loose. Who knows what speck of debris might be causing a problem?

Glad you're fixed up, though ...

It's interesting though. This same cable was also in very close proximity to the PCI-E slot and vid card, if not touching. And sometimes my BIOS screen would get the 'jitters' like a CRT would do after being degaussed. But like you said, it could have been a piece of dust or anything.

Now it's back to the same problem again. Only this time, it started acting up after I attempted to take the CPU OC further. I was trying to see if I could squeeze 3400 out of it, but it wouldn't. Now, when I go back to previous, known stable OC it starts this series of problems all over again....dead onboard LAN and constantly saying 'Previous overclock has failed blah blah blah....' in red letters during POST. Hell, it even does this when everything is back to default and on stock settings.
So here's what I figure for a solution. I'm going to reflash the BIOS to the previous version (the version that the board shipped with) and if this doesn't solve anything, I'm going to buy a new mobo from the egg, then have this one RAM'd and take it from there. Any suggestions on a cheap motherboard to replace this one?

Sounds like maybe the chipset it heating up or you're getting short on voltage somewhere. Video jitters are typical of video faults, sometimes from lack of voltage, sometimes from hot chips. If your chipset and MOSFET sinks are plenty cool then the PSU would be the next place to look. Have you checked the voltages yet ...?

QuietIce said:

Sounds like maybe the chipset it heating up or you're getting short on voltage somewhere. Video jitters are typical of video faults, sometimes from lack of voltage, sometimes from hot chips. If your chipset and MOSFET sinks are plenty cool then the PSU would be the next place to look. Have you checked the voltages yet ...?

Click to expand...

Hmm....the mobo has onboard video but I am using a Geforce 9800 GT PCI-E card. According to Everest the temps are fine on the north and south bridges. I have the voltages on the HT and Northbridge set to auto in BIOS...perhaps they are undervolted?
The PSU is a 500w Antec with dual 12v rails @ 18a. This should be suitable for my system since I'm not running SLI.
How should I go about testing the voltage on the PSU without a multimeter?

You might try manual settings on the chipset voltages or even up them one setting just to see if that resolves the issue. With that many things going wild I'd think it's something more general (like the chipsets) that's involved.

That's a good PSU and plenty for your system. There's no good way to test a PSU w/out a meter but even a cheap < $10 one would work ...