Why does my HDD like to shut itself off?

EvilNick

Limp Gawd
Joined
Jun 14, 2004
Messages
149
Here's my problem: My HDD just likes to shut itself off from time to time while I'm using my com. Obviously, turning itself off results in a hang/crash/BSOD. I know it's the HDD because I can hear it spin down and shut off. Sometimes, it even cause an UNMOUNTABLE_BOOT_VOLUME error and I have to run Recovery Console just to get back. What I want to know is, is this caused by a HDD that's facing emminent death? Or a bad PSU? I'm using a WD 800JB 80GB PATA drive and my PSU is an Enlight 420W unit.

Would really appreciate some insight into this. In the mean time, I think I should go burn all my stuff to backup, hehe. :eek:
 
the only reason it would be spinning down in the middle of a session is power loss

Im pretty busy weekends but if you list the rest of your config (everything drawing power including all fans) and include your room temperature the fine folks in the PSU forum will give you a hand ;)
 
Hrm ok, off the top of my head, here's the stuff in my rig.

P4 2.4C @ 3.0GHz
Asus P4P800-E Deluxe
2 x 512MB Adata Vitesta DDR500
128MB 9800 Pro
Audigy 2 ZS
BenQ DW1640 DVD+RW
WD800JB 80GB PATA HDD
Fans:
120mm High Speed Panaflo @ 5V
2 x 80mm Low Speed Panaflo @ 12V
45mm Delta @ 5V
120mm Adda @ 12V
There's also a little temp readout on the front which has 2 small backlit digital readouts. Not LCDs, just backlit.

As you can see, I only have one HDD and one optical drive, so it's not at all a heavy load. Can heat cause a HDD to spin down as well? Cause I know it gets pretty hot, but I can't check cause it doesn't support SMART monitoring.
 
Heat can cause the PSU to drop capacity
its rated amperage was likely measured somewhere between 21C and 30C
whereas its operating temprature is likely around 40C to 50C (or worse)
its exhausting the case and the CPU heatsink
thus its likely only producing 2\3rd of its rated capacity (depending)

Im still busy writing the frontpage and so will leave the number crunching to others
http://www.xbitlabs.com/misc/picture/?src=/images/video/ati-powercons/t4.gif&1=1
http://takaman.jp/D/index.html?english
http://hardforum.com/showthread.php?t=844691

but another important question would be are you running a UPS \ AVR or surge protector?
and if so what kind and how old is it.

Most are MOV technology and they wear out, thus between the capacity dropping becoause its hot and the AC current from your utility being unstable because its hot or Air Conditioners or other heavy draw motors being on the same circuit as the PSU you might be having brownout severe enough that the PSU is shutting down to protect equipment
 
Hard Drives can also be shut down from too much heat on the drive. They all have temperature sensors and most of their firmware probably has a shutdown point to prevent damage. I have a WD as a second drive that I have had forced shutdowns in the summer months (room ambient temp 90+ degrees, no air conditioner =( ) when I do multi-gigabyte transfers to it while the rest of the computer continues to operate fine. If you don't have good ventilation over the drive try getting some and seeing if that helps.
 
I do use a UPS, I'm not sure what kind it is, but it's about 2-3 years old already. I do have an air conditioner in my room which is probably off the same circuit, but I've never had problems with running everything on at the same time before.

Btw, today I opened up my case and I noticed something. The LAST corner pin on my 20pin ATX connector looks like it may be a lil burnt. The plastic is still white, but I can see thru it and the metal pin looks a lil different. It's the orange wire btw.

Geshtar: I read on the WD firmware site and they said that drives spinning down during use can be cause by overheating or a bad power/IDE cable. It's possible that it was overheating because my drive doesn't support SMART monitoring, so I checked my air filters in the front of my case and they were pretty thick with dust. So I cleaned them out now, but the problem still persisted. And I doubt cablegami with IDE cables will create folds bad enough to cause cable problems right? :)
 
EvilNick said:
Geshtar: I read on the WD firmware site and they said that drives spinning down during use can be cause by overheating or a bad power/IDE cable. It's possible that it was overheating because my drive doesn't support SMART monitoring, so I checked my air filters in the front of my case and they were pretty thick with dust. So I cleaned them out now, but the problem still persisted. And I doubt cablegami with IDE cables will create folds bad enough to cause cable problems right? :)


Well, I personally have never done cablegami (too lazy), but lots of other people do and if it caused lots of problems I'm sure we would have heard about it by now....so you can probably rule cablegami out.

Just for giggles, you might want to try running WD's drive test programs and see if anything turns up about the drive. If not, I dunno, I suppose your power supply could be getting marginal. Not very familiar with how good Enlight's are.
 
Today I made some rearranging of IDE cables and power connectors, the drive seems abit better now, hasn't spun down today... yet. :eek:

Anyway, I've run Western Digital Diagnostics and the Quick Test which I used to be able to pass last time now give me this:

Test Result: FAIL
Test Error Code: 06-Quick Test on drive 1 did not complete! Status code = 07 (Failed read test element), Failure Checkpoint = 65 (Error Log Test) SMART self-test did not complete on drive 1!

Does this mean it's my HDD and not my PSU? Then why does my PSU have a lil burnt orange 3.3v pin on its 20pin ATX connector? :confused:
 
It's reasonable to assume that if an electronic device is not receiving a reliable supply of electricity, it could cause it not to operate properly...just because your HDD fails doesn't mean it's not the PSU. ;)
 
WD800JB.

'nuff said, I've had the exact problem with this drive.
 
Great. I hate it when something's wrong and I can't pinpoint for sure what it is.

However, seeing that there's a lil burnt pin on the orange wire of my PSU's 20pin connector, I think it's time to be careful with that PSU. Any ideas what could have caused it to overheat/burn? It's the 3.3V orange wire.
 
If the female pin in the connector coming from the PSU has opened up a bit it will cause that wire to have a poor connection which will cause the pins to get hot, cause burning and discoloration.

Warm temps can aggrivate this since resistence to electrical flow is a function of heat AND the warmer a pin type connection is the looser it becomes due to expansion of the metal, during the winter it might not have been loose enough to cause a problem but with the added heat of summer it got *just* loose enough to cause issues and from there it began to snowball.

What kind of PSU are we talking about here? I never saw it listed but I might have missed it while skimming the thread. If it's just some cheapy-came-with-the-case-P.O.S. I'd suggest ditching it and even if it's not with it burning pins on the mobo it might not be a bad idea to look into a new one.
 
madmat said:
If the female pin in the connector coming from the PSU has opened up a bit it will cause that wire to have a poor connection which will cause the pins to get hot, cause burning and discoloration.


agreed likely poor contact

gee said:
Bad connector. If the burnt pins were caused by a power supply failure, then chances are your computer wouldn't be working right now.

Make sure the pins in your motherboard's ATX connector don't have any carbon on them - if there is any, scrape it off with a jewler's screwdriver or something.

The connector ended up having a high resistance, and since your computer pulled a fair bit of current through this resistance, this overheated and burned up the connector.

;) make sure the pin is properly seated in the connector and not loose, and make sure its clean and not misshappen
 
It's an Enlight 420W, it is a decent PSU although it's not as solid as something from Antec, OCZ or Coolermaster. Definitely not a POS-that-came-with-the-case item. :D

One other symptom: Sometimes in the morning when I want to turn on my computer, turning on the PSU switch round back results in the front case power and HDD LEDs blinking nonstop and the fans moving just a lil. I usually resolve this problem by removing and reconnecting the 20pin ATX connector and then it works fine.

The pin might have widened abit since my sleeving job last time, so that and the problem I just mentioned above might show that it's a bad connection. Can I just squeeze the pins back together or should I get a new PSU? I'm short on cash, so a repair is definitely preferred over a replace. :) Since the orange wire is the 3.3v one, I also noticed that MBM reports my 3.3 line to be at 2.98 most of the time. :eek: 5v and 12v lines still look OK though.

*UPDATE*
After awhile it got worse. BSOD's before I can even get into Windows, before I can even start a Windows installation from the boot CD... so I popped open my case, took the 20pin PSU connector and took a look. Sure enough, almost all the pins were abit wide, as in wedged open. And I reckon none of them made proper contact because under close scrutiny, all of them looked alil worse for wear.

So I used a lil jewellery screwdriver and mashed each together so that they're back to a tighter fit and now, the com is back up and fine, and my 3.3 line is back to 3.41v instead of the lowly 2.98.

But obviously, this is only a temporary solution, leaving it like this will only mean it'll happen again eventually. Short of replacing my PSU, I want to just snip off all the 20pin wires, solder on new pins and put them into a new connector plug.
What I'm wondering now is whether normal cheapo solder will suffice for the job? As in, the kind I use for fan/LED wires. If it is, I'll just try this instead of buying a new PSU since I have a bunch of new pins around.

Thanks for all the help so far, really helped me out here guys. ;)
 
Yes, regular rosin-core solder will work fine but really, as long as the pins are tight now they shouldn't loosen up unless you're plugging and un-plugging the 20 pin connector a bunch.

I can understand wanting to swap the pins out but honestly, if it works I wouldn't mess with it until I had something I needed to do (like replacing the mobo) that involves unpluging that connector.
 
Hrm, I see. Well yeah, the though of crimping/soldering 20 of those tiny pins is not a fun one, but I'm worried if the connection worsens again over time and cause the connectors to burn/melt. That wouldn't be so bad for the PSU, but I'd hate for it to take my mobo or anything else along with it.

Maybe the next time it acts up, I'll replace the pins straight away. Maybe crimping will be much faster and easier than soldering. :D
 
Back
Top