PC Hang/Hard drive failures...

[majmera]

Hi Guys,
CPU : Athlon XP 1600+
Motherboard: K7 420 nVidia nforce (MSI 6373)
RAM : 2 x 256 MB 2100 DDR
HDD : WD 100GB and Seagate 200GB (ATA 100)
OS : WinXP Pro.

Guys my system was running fine for a year and half, but then suddenly I started seeing blue screens randomly and then few days later, the PC wouldn't boot up at all. So I decided to format the hard drive, re-install XP and see what happens. After re-installation, the system worked fine for a few days and then again back to the same problem. So I figured, maybe my hard drive is having problems (Norton disk doctor showed some problems, but would not fix them since XP had the exclusive access to the drive).
I went ahead and bought a Seagate drive (200 GB), made that my primary drive, re-installed everything. The system worked fine for a month until recently the I started seeing random hangs. The system would be idle and suddenly freeze up (no mouse movement or reaction to Ctrl+Alt+Del). One strange thing during all this was the hard drive light stayed on continuosly.
A few days later, now my system won't even boot, (Blue screens every single time), I again re-installed XP, but started seeing the hangs immediately. Today I ran the Seagate Disc Utility program and it found numerous bad sectors on my new hard drive.
Now I am completely baffled as to why there would be bad sectors on a new drive (only a month old), what other components of my system could be bad, that could cause this? I can't believe that it is just a coincdence that both my drives went bad on their own.

Thanks
Mayank

 

[majmera]

Update...
I ran the Seagate Disc Utils yesterday and it ran for more than 20 hrs and found numerous bad sectors.
Can a bad power supply or bad RAM cause such failures?

 

[Zee]

bad sectors generally mean the drive has... bad sectors. bad disk clusters i think it used to be called? But anyway, it's just your harddrive getting old.

 

[majmera]

Well my Seagate is barely a month old.

 

[DetroitSportsFan]

Update...
[I ran the Seagate Disc Utils yesterday and it ran for more than 20 hrs and found numerous bad sectors.
Can a bad power supply or bad RAM cause such failures?
/Q]

To answer your question ..... YES! An under performing Power Supply can cause read/write errors and file corruptions. Try monitoring your system voltage using either your motherboard's utility or something along the lines of mother board monitor (mbm5). Be sure to set it up correctly or your readings may be "out to lunch." To be honest though, the "verdict" is still out on the accuracy of software monitors like this from within windows. Still, its worth looking at first before you pull your case off and start checking voltages with a multimeter. If you suspect a problem, proceed with that multimeter to check your voltages.

As for ram .... bad ram usually causes other problems. Although I suppose its possible and should still be considered. Memtest does well for ram testing, however I'm wondering if you would even make it through ONE round of testing on memtest with the problems you've described.

One other thing ..... don't rule out heat. Is your heatsink/fan dirty or plugged up? Is there a layer of dust on the mother board? Crashes caused by heat can cause file corruption .... and heat itself can cause premature hardware failures ..... including your brand new hard drive.

 

[majmera]

To answer your question ..... YES! An under performing Power Supply can cause read/write errors and file corruptions. Try monitoring your system voltage using either your motherboard's utility or something along the lines of mother board monitor (mbm5). Be sure to set it up correctly or your readings may be "out to lunch." To be honest though, the "verdict" is still out on the accuracy of software monitors like this from within windows. Still, its worth looking at first before you pull your case off and start checking voltages with a multimeter. If you suspect a problem, proceed with that multimeter to check your voltages.

As for ram .... bad ram usually causes other problems. Although I suppose its possible and should still be considered. Memtest does well for ram testing, however I'm wondering if you would even make it through ONE round of testing on memtest with the problems you've described.

One other thing ..... don't rule out heat. Is your heatsink/fan dirty or plugged up? Is there a layer of dust on the mother board? Crashes caused by heat can cause file corruption .... and heat itself can cause premature hardware failures ..... including your brand new hard drive.

I plan to run memtest as soon as I can get my system to come up. When my system was up, I used my motherboard utility (MSI PC Alert IV) to track the voltage and heat, but that seemed to be fine. The CPU was running at about 40-45C. The voltage was about 0.1 to 0.2 volts below the required levels. Another think is I had left my case open since the last time I opened it (about a month ago) would that cause any problems?

I can understand the RAM or PSU causing file system corruption, but actual bad sectors surprises me.

 

[DetroitSportsFan]

I can understand the RAM or PSU causing file system corruption, but actual bad sectors surprises me.

Fluctuating voltage can damage your hardware. I'm not sure if its the cause of your problem, but its certainly something to be considered. Have you stripped down to as bare bones as you can get .... and then add your components one by one .... testing heat and voltage between each round of new hardware? Without having your system in front of me so I can actually see whats happening ..... I'm still questioning your PSU as the source of your problem. Especially considering software monitoring tools really aren't that reliable.

 

[majmera]

Update...
I was able to bring the system up by removing one of the memory banks (I have 2 x 256 MB). Also I ran memtest+ on the system with both the banks and it showed some failures. Currently I am running with just 1 bank (removed the bank that I think is faulty). Is there some way to verify if this is a memory problem or a memory controller problem. I think my motherboard is one of the early implementations of the Twinbank Memory Architecture.

 

[mechBgon]

What brand, model and wattage is the power supply? If it's old, cheap/generic or underpowered, any of those three, then replace it with a new, high-quality, beefy one. Antec's SL400 is easy to find, and Fortron or Enermax are a couple other high-quality brands.

 

[vskagentsmith]

I'd say clean the system inside and out, make sure all the fans are clear of dust and that the PSU doesn't have any major dust build up inside of it (don't open it!). As far as your harddrive issue, the drive could just be bad, it happens sometimes. Typically bad clusters can be cleared up with a FORMAT, the detection and marking of bad clusters is done by the FORMAT program that basically writes F0 (I might not be totally accurate on that) on all the storage locations on the drive, then checks them by reading it back, any locations that don't read back F0 are marked as bad.

If you format and still have problems then there could be other problems with the drive. The oxide coating of the harddisk can sometimes flake off, which pretty much nurfs the drive right then and there.

If you've run the provided disk diagnostic software and it's coming up with errors on a freshly formatted (nothing installed) harddrive, then you may need to contact tech support for the drive manufacturer. Typically most companies will gladly replace the drive if it's defective, and carry a decent warranty, and no matter what's wrong they'll usually replace it. Or could provide more indepth information about what will be needed to fix it anyways.


Best of luck to you, hope to see you get this worked out!

 

[majmera]

The Power supply came with the Antec case. It is about 3 yrs old. When the system is up, I can using the MSI PC Alert tool to monitor the voltages and they seem to be dip to 0.1 - 0.2 volts below the required levels.
Is that good enough?
Also my CPU temp seems to be at around 44-46 C. Is that ok?
Currently the system seems to be running fine (I have removed 1 bank of memory and all peripheral other than my HDD and CDR).
I think over this week I am going to test the memory modules out using the memtest and then decide where to go from there.

Thanks for all your help
Mayank

 

[majmera]

memtest86 failed on 1 memory module. It ran successfully on the other module for almost 10 hrs. So I think I have atleast 1 bad memory module. The system's been stable without that module plugged in.

 

[Double Trouble]

Sounds like you probably have a bad memory stick, and that perhaps your power supply isn't working as it should anymore either. If you take out the 'bad' memory stick and run memtest again and everything comes out "oK", then odds are that it's the memory stick causing the problem. The temps you are seeing are OK, so don't worry about those. The voltage seems a little low, but nothing to be concerned about.......

Just my 2 cents.

 

[Sideswipe001]

Even if the memory was the cause of the problem, I'd still RMA that hard drive. After a month, you shouldn't see ANY bad sectors. If you're getting them already, get it replaced. Sometimes they do die that fast. HDs are one of the worst components for failure, if not THE worst.