Odd Problems unsure what to fix

I have a homebuilt system specs below:

Gigabyte ep45-ud3 rev1.0 mainboard
seating an intel core 4 duo q6600 slacor 2.4hz chip over clocked to 3.4
8 gb cosair ddr 1066 mhz
Radeon Sapphire 5770 hd grahics card
2 Western Digital Blue Cavair 500 gb disks
a seasonic 600w psu
a couple of optical drives

it's running win7, the disks are in ide mode. I built it in April this year.

I've had a few odd problems, which I probably should not have ignored,

When first setting up the system I tried to use a raid, but this somehow killed one of the drive, I suspected a faulty drive and replaced it but never tried to raid again.

Firstly the system very occasionally blue screens with a kernal power failure as the only error in the logs. When the system restarts I get no information about the problem even after sending the dumps to microsoft.

This was an annouance but rare. So rare that I dismissed the idea it was a problem with the psu.

Today things got worse.

The system freezes every few seconds for a second or two. The logs show a disk error a bad block. this is throwing up every few seconds. Now normally i'd just accept it and get ready to buy a replacement and reinstall (any adivce to save me time or money welcome!) but I restarted and the disk error switched from disk dr0 to dr1... Does windows re-assign these randomly each time the pc starts? Also when the error was on dr0 I was getting errors reading and writing to the registry, no such error occurs when it says problems with disk dr1.

Also the error first occured on dr1 on the 25 Aug. twice, and then stopped until this morning.

Like I said normally I'd assume it was the drive alone, but two failed drives so quickly sounds like some other problem maybe with the main board. I also have a huge problem, in that my master's dissertation is due in on Wednesday and I need a working pc fast.

Any suggestions as to what to try or which component is failing. I fear it could be the main board.
14 answers Last reply
More about problems unsure
  1. To troubleshoot the system, loose the OC first!

    Yes, your PSU could be the culprit, since 2 HDD were dead.
    Check also the grouding of your PC.
  2. How do I check the grounding? And a psu can kill hard drives?
  3. 1. PSU can kill everything connected into it (including you :D )
    2. Use a multitester and see if there is some voltage between your PC and ground, more than let say 1.5 V is actually not recommended. Sometime, touching the case helps, if got electrified...the grounding is really bad.

    BTW : Returning to stock speed is really important during trouble shooting.

    Almost forgoten, bad grounding has killed 2 of my HDDs (Maxtor 20GB and Seagate 80GB) about 3 years ago, within the same week.
  4. How do I check the grounding?

    ground is where? the pc case housing? so I place one pin on that, and the other on what?

    Sorry i'm not a whiz,

    I'm supprised if the PSU is the culprit, to kill two drives over a 4-5 month period? if there was a fault wouldn't it have done it closer together? Also I stress tested the system maxed out the cpu adn gpu for over 24 hrs (while overclocked) with no errors at all. Of course that doesn't mean the psu couldn't have developed one, but it's an expensive part to replace without confidence it's the culprit.

    mind you it's the only lead I have on the random bluescreen kerenal problems. (i'm still not sure BSOD and a note saying the kernal stopped, maybe power was turned off etc message is very helpful..)

    I've tested the hard drives in WD digital life software, running in dos, so the cpu and psu would be doing very little, seems one of the drives has a real problem, it fails, when I change power cable, data cable, sata port, on it's own with the other drive running. It is also the primary system drive (the older of the two the twin of the other failed drive). I'm going to try to get as much of my data off as I can... sigh..

    not what I need right now.

    Later, once mt wife is at work I'll put the drive in her pc and test it in dos (bootable cd-rom) see there if anything happens.

    but at the moment we are thinking PSU is the culprit?
  5. PC housing is usually connected to grounding.

    Dead components could be caused by your electrical system, either the PSU, grounding, or simply bad quality electric supply (not clean, voltage unstable).
    Defect components are also possible, please keep that also in mind.
  6. Yeah this is why I'm supprised i'm looking at the psu, it's a Seasonic S12 Series 600W it was rated as an awesome psu and it's powered two systems in its time, until now flawlessly.

    However it has been running since Sept 06 but an expensive psu should last longer than that no?

    Other than that I either got unlucky with the hard drives,
    Or the sata controller on the main board is really bad.

    Also the system runs much better if i disconnect the good drive, which is odd and makes me wonder if it's the sata controller.. I really want to identify the right part before ordering new parts! Is there any software that can moniter the voltages from a psu?

    I don't have a multimeter, but I'm still not sure, even if i borrow one, where to put the nodes, one on the case, the other? on the hard drive casing? or the psu casing or where? bit of a newbie at that sort of thing.

    my bsod errors were:
  7. would a psu tester find problems which could cause hard drive failure?
  8. Ok so I think I've figured out the BSOD happening occasionally after start up,

    With all my components and the over clock, my system comes in under the wattage of my psu, even when 100% loaded, slap on some capatior aging, and blam the thing wants to draw more than that.

    I guess a psu tester would show me if the voltages were steady.

    If the psu is killing my hard drives would the psu tester show me that error, or could it only be happening when the system draws too much wattage. is there anyway to test the maximum wattage the psu can draw?
  9. Hmmmm...this is difficult...
    the normal PSU tester, which you can easily find, can only measure the voltages of the PSU under no load a.k.a not connected to the system.
    This one for example :
    But if your PSU went wacky and delivered higher/lower voltage as it should be even under no load, then yes, you can find it out using that thing above or using a multitester.
    It is also a bit slow, you can not see voltage spikes, etc.

    Hmmm...27 bucks...If I were you, I would simply buy a new PSU...I know Seasonic is agood brand, but PSUs are also mass produced, defect products are normal....just make sure the grounding is also in tip top condition.
  10. should i try to fix the drive, or rma it do you think?
  11. antec signature 850 should do the trick... wow another £120 on a psu...sigh..
  12. Ok then,

    So i've removed any devices I don't need, so both opticals gone, one of the hard drives, gone, any usb devices gone, the fans turned down (using fan controller so I'm not sure this actually saves power) and the graphics card throttled to the lowest settings I can (I have no spare low power pci express card)

    Naturally the over clock is off too...

    The disk has been repaired using western digitals tools and I have new psu on order...

    So I guess either power to the hard drive or to the whole system must have failed while vital data was half written from the hard drives cache to the drive.

    I'll wait for the new psu to turn up before I throtle the machine again...

    at least a new psu doesn't require a reinstall of windows...
  13. I wish you luck with the new PSU!
  14. Ok so the psu didn't fix it...

    I replaced it, with a brand new spanking antec signature 850, which should be able to cope easly.

    I did not reapply the overclock.

    I fixed the disk using western digitals utilities.

    Everything ran fine for a couple of weeks now, it's got the same disk error.

    So The PSU wasn't the problem.

    What then is?

    I have a kernal warning about the registry but I've had that since this problem frist started. I also have a Circular Kernal Context logger stopped message.

    I've no idea what's wrong.

    Is the drive crappy? or the sata controller on the main board (not sure how it could that that as I'm running in ide mode) or the processor?

    Any advice please?
Ask a new question

Read More

Homebuilt Systems Product