Is this my HDD failing?

[wWooDd]

I have not had this problem long but its truely scarey. This has happened only one other time during the summer and now. It only started happening recently again after I started to overclock.

Last summer I started having a strange issue where it sounded like my system was shutting off. I could hear the distinct ker-chunk + whine when the system shuts down followed by the spin-up sound a second or two later; nothing except the noise happens (no errors/shut down/flashes/etc.). This happened several times over the span of a couple days and just never happened again. I did ask a local computer specialist about it and had not a clue.

Recently, a couple months afterwards, I started to overclock. Originally a couple months ago with no problems; I ran into other issues and had to leave the overclocking for later. I had it set to the most stable variables I had last time with great success. With my Radeon, I did artifact scanning with ATI Tool just in case when I started getting that "shut-down" issue again. Only when I used ATI Tool. Happened twice, no more. While playing FEAR tonight, several days later, I had a complete system failure. My computer crashed followed by that distinct "ker-chunk + whine". The system reboots and hangs until I hear [The HDD?] spin up and every thing runs like nothing happened except a Windows "Your system had recovered from a major crash" error.

Im wondering if this is my HDD failing on me or if it some thing I am actually doing wrong. I am really not in a position to go out and grab a new 120 gig drive so if my drive is dieing I am going to start backing up my vitals to CD until the drive actually *dies*. Thanks much!

 

[SuperSilicon]

It could be your powersupply, overclocking can lead to problems. Post your specs

 

[RebateMonger]

I had something similar once. A WD 120GB drive that I figured was on the verge of failure. It turned out that the 4-pin power cable was slightly lose on the drive and would disconnect for a split second every once in a while.

 

[ziplocpicker]

I have the exact same problem. I'm not sure why it happens, but i suspect it is my PSU (only 13A on the 12v). It is also one of the main factors driving me to upgrade the PSU. If anyone could think of another solution/possible problem it would help me as well.

 

[Captante]

Another important question is what your system is, in particular the motherboard... if its a current chipset, even if your overclocking via the fsb, your hd won't be effected, but if its an older Athlon XP or P-4 theres a good chance the hd (& all IDE devices) are running out of spec which can easily cause problems like you describe.
Post your system specs for more info.

 

[ziplocpicker]

ASUS N6600/TD/256 Geforce 6600 256MB AGP
AMD Athlon 64 3000+ ClawHammer 800MHz FSB
MSI K8N Neo
CORSAIR ValueSelect 1GB (2 x 512MB) 184-Pin DDR SDRAM DDR 400 (PC 3200)
Western Digital Caviar SE WD1600JD 160GB
Raidmax 420W PSU (can't find the exact specs. I remember 13A on the 12v though)

Yes, i know i skimped a lot on the PSU. It was a bad move on my part.

 

[FlyingPenguin]

Very similar problem with my own workstation and a couple of clients rigs. Bad capacitors in the PSU. System ran fine as long as you didn't shut it down. On shutdown or power up the PSU would spike the drive (PATA drives would make a clunk - never a good sound to hear, SATA drive in my rig would make a loud BOING sound). In all cases this eventually led to drive failure.

I went through 3 RMAs on my own rig before I figured it out. And this was a quality Antec PSU.

Unusual noises from the HDD are usually bad omens.

I'd look into it ASAP. Shine a flashlight into the fan opennings on the PSU and look for bulging/leaking caps. Examples here:
http://www.pcstats.com/articleview.cfm?articleID=195

I bad PSU will almost always show itself in the voltage 12 volt or 5 volt rail. Use a digital multimeter ($20 at Home Depot) and measure the voltages at a spare HDD connector. If either the 12 or 5 volt rail are outside of 5% tolerance, then ditch the PSU.

If you are overclocking, DON'T. You need to eliminate that as a possibility. It's not likely since O/Cing doesn't affect the PCI bus in modern mobos (in Pentium 1 & 2 days o/cing could damage a HDD if the mobo didn't support seperate PCI and CPU clock dividers). Don't overclock again until you resolve the issue.

Hope this helps...

 

[RealTime]

Originally posted by: ziplocpicker
ASUS N6600/TD/256 Geforce 6600 256MB AGP
AMD Athlon 64 3000+ ClawHammer 800MHz FSB
MSI K8N Neo
CORSAIR ValueSelect 1GB (2 x 512MB) 184-Pin DDR SDRAM DDR 400 (PC 3200)
Western Digital Caviar SE WD1600JD 160GB
Raidmax 420W PSU (can't find the exact specs. I remember 13A on the 12v though)

Yes, i know i skimped a lot on the PSU. It was a bad move on my part.

Sounds like the PSU to me. Whether it is loose somewhere or it is just dying...you should pick yourself up a good PSU. If you are overclocking then that makes it even more important to have a good PSU. I like Antec PSUs.
www.antec.com

 

[wWooDd]

This issue on my rig started/ended months before I even had the thought of overclocking. I checked for a loose connection and found that while every thing was secure, the power for my HDD came directly after my video card; assuming this would mean I wouldn't have to over clock to have my power hungry card take too much juice. I switched cables around now so the HDD and vid card are separate. I'll post my specs any ways:

Radeon 9800 Pro 256Mb
P4 3.0Ghz
MSI 875P Neo
Corsair XMS series 1024Mb
Western Digital 120Gb
Altec True 480 PSU

 

[SuperSilicon]

It really does sound like a psu issue, I would consider if your still having problems to aquire another powersupply and switch the current one out and see if that helps. You can then go from there, if you take the machine to a shop your looking at spending what the powersupply would cost + a grip more for labor and shop cost's so its a better idea to start there.

 

[ziplocpicker]

Thanks for the info. Nice to know my suspicions were pretty accurate (a little ego boost for me). I'm working on a new build right now and this PC will cycle out the crappy one my dad is running. The new PSU is already on its way.

 

[wWooDd]

Relating to over-clocking, though, that does not necessarily mean a bad PSU? I run fine with out over clocking my processor. Ive been playing FEAR all night with my Radeon tweaked with no failure or hint of it. Im doing a new build this summer so if I can get away with what I got right now it would be great.

 

[FlyingPenguin]

wWooDd: Now that you've posted your system specs I would IMMEDIATELY examine that PSU for bad caps. All PSU manufacturers were hit with a bad batch of caps 2 - 3 years ago and they're showing up now. Antec has better quality control than most of them, but they got hit too for a while and every single case of bad caps in an Antec PSU I've come across has been in Antec True 430s and 480s. You can shine a light into the back fan vent and look left at the four largest caps grouped together (see photo in link below) and if you see brown (not white) crud or bulging tops, stop using it IMMEDIATELY.

Photos of an Antec 430 with the brown crud here: http://www.dslreports.com/forum/remark,14116577?hilite=bad+caps

The brown crud is the dialectric gel inside the cap that's leaked out. Normally it looks like brown liguid but inside the hot confines of the PSU it bakes into a brown crud.

Ignore any white goo you see around the caps - that's just an epoxy used to glue them together.

You can also open the PSU case to examine it (make sure it's unplugged from the wall electric) but that will void the warranty and Antec has a 3 year warranty although they (unlike others) require proof of purchase (a receipt) which sucks.

The overclocking is just clouding the issue. Turn it off for now. The PSU works HARDER when you overclock (CPU and memory draws more current). This will amplify any voltage instabilities in the PSU. If the PSU is putting out bad voltages, overclocking will make it even worse.

Additionally, the True 480 is an older design that does not have a dual 12 volt rail (seperate 12v bus for the drives and the mobo). On modern systems with power hungry video cards you really need a modern design dual rail PSU for voltage stability.

I'm very fond of the Enermax Noisetaker EG495 which NewEgg sells for a very reasonable $83 shipped. Nice rock stable PSU. I install it in all my own rigs, and I use the Noisetaker 425 in client rigs. Also any modern Antec PSU is an excellent choice.


Hope this helps...

 

[wWooDd]

Interesting. Ok, well, I checked for the caps & any brown material emerging. With out having to take apart my rig to really examine it (Im actually quite uncomfortable to take apart the PSU its self), I saw several caps but no brown or any material coming out of the caps. I ask why I should immediatly stop using the PSU if this is happening, assuming its most likely system damage? If I can avoid spending any ammount of money it would be best considering the holidays took its toll on me. As with the situation, I will be getting a PSU whether or not this one is bad. Is Antenc and Enermax still suggested for the performance of a top end (power house) rig? Couple years ago when I built this one Antec was highly suggested and a bad batch would not deter me.

Thank a lot!

 

[ziplocpicker]

Antec is still a good brand, but is having some quality control issues from what i hear. people seem to be getting a lot of Antecs with bad caps, which could fry your motherboard/cpu/vga and cost you a lot of money. I wouldn't take that chance.

The best brands for value seem to be Enermax, Fortron, and Sparkle. OCZ, Seasonic, and PC P&C are better but pricier.