i think my rig is killing my HDDs

I upgraded my rig's hard drives in March, SSD Corsair 60Gb 50k iops and 3 (striped) Western Digital WD5000AADS 500GB SATAII 32MB Cache old school drives. beautiful load times all round. then last week failed to boot (flashing underscore before all the bios screens) unless i unplugged the solid state, so i tried a number of things including unplugging all sata cables (power and data) from all the drives and only pluging in the SSD - this time it got past the flashing underscore but failed to detect the drive. retailer agreed to an exchange. . . but while waiting for the replacement I was running windows off my old HDD and now this morning steam freezes and i lose my striped drive - exactly the same symptoms upon reset, plugging the drive in post boot and rescanning disk in disk manager is unable to find it. so 2 completely different drives both dying in seemingly the same way. any ideas?

Power supply OCZ700SXS 700W
MoBo MSI 770-C45 (MS-7599)
Processor AMD Phenom II X4 965
RAM 4.00GB DDR3 1333
GFX Nvidia GeForce 9800 GX2

not sure if it's entirely relevant but my case doesn't fully seat the GFX card, it's a clip in case and the card isn't clip in friendly so it occasionally unseats itself from the MoBo (power connections are fairly secure) but that was since i had the trouble with the SSD and messed around with the iguts of the machine

any input would be appreciated . . as you can imagine, reporting the same fault with a different drive may ring warning bells with the retailer so I'm hoping to avoid that.

 

Hi token,

I'd suspect the PSU.

 

so ditch the PSU and save the other/new hard drives?
I'd have thought 700w was ample so am I wrong or is it likely a fault with the PSU?


or are there any other guesses :tired

 

It's likely to be a fault that's developed in the PSU. Greater claimed output doesn't mean greater quality components.

Some makers have a better reputation than others but even within one manufacturers range, quality can vary significantly. See the latest returns rates for PSU's from a French retailer.

 

Thanks for the input, I've gone for
Antec TruePower Quattro 850W Modular PSU
and
Startech 4 Port PCI Serial ATA (SATA) Storage controller Raid 0 1 - Card
just to be safe, as well as 3 clip in sata cables, just to treat myself - now there is nothing (bar the scenario of malware causing the problem) that could cause the problem to repeat. . am I right? :confused

 

Hmm, difficult to prove anything with 100% certainty when it comes to troubleshooting PC's but given your signs and symptoms, I'd say that a new PSU should fix all the problems you've related here, except the gfx card seating problem (and any potential issues that may yet arise due to the old PSU having been out of spec. and causing other damage).

 

i managed to force an 8 pin mobo connection into my gfx card, I'm sure it can take anything a faulty PSU can throw at it

 

That's never a good idea, forcing a connection. This may help.

 

wont be malware related. Malware cant stop your PC from posting.

 

right - update: with a new SSD, new PCI Sata controller and a lovely 850W power supply i tried again, fresh install of windows on the SSD and new raid array of striped 500Gb WD caviar green x 2 both are the old ones from march this year.
everything hunkey dorey then a steam game freezes and on reboot failed to detect one of the 500Gb, on re-reboot it found it as a 0Gb drive in the PCI Sata Boot UI and fails to find it in the windows pci sata gui.
so i wonder if i mixed up the drives and accidentaly put the faulty 500Gb in and to my suprise the other 500Gb was detected and could be set up into an array!!! so thinking I'd made a silly error i initialized and formatted then started copying some files into it - with a speed of <1 Mbps usually <100 Kbps so obviously i hadn't mixed the disks up, I've tried the WD diskchecker through dos but it freezes, guess it doesn't like working through the PCI card.

I'm going to try and test the drives through the on board satas now. wish me luck

anyone have any further ideas? or could it just be that the old power supply messed up 3 hard drives and the fault just didn't surface on one of them until now:confused

 

Wait, these are Green drives in the RAID? That strikes me as being kinda counter-intuitive, green drives being about spinning down to save power, variable spin speeds etc., I think I'd be looking for the WD feature tool to override the default settings in the drives.

Just found this from WDC.

 

oh well, shows what i know - the retailer's feedback had every post talking about raiding the drive so i didn't really think twice . . .

If i summerise
drive x,y,z are the green caviar WD drives
drive s is the solid state

all purchased in march with a fresh install, drive S fails (by all accounts it's non existant as far as the PC is concerned) in april so replace with my old system drive while it's being replaced, while running off the old drive drive x fails - DLGDIAG (WD hard drive tester) says 're-allocated sector count 42 : threshold 140' cannot test. so i order a PCI sata controller and power supply.

A PCI sata controller can read drive x with VERY limited access speeds but the motherboard just kicks out a SMART status=bad error and wont run windows if connected and crashes windows if hot plugged.

drive S's replacement comes back, fresh install of windows, new raid with drive Y & Z - filled it back up with family files + steam. drive y fails 36 / 48 hours later same symptoms as drive S.

My question is: if the power supply and the sata controllers are new is there anything else in the system that could be causing these drives to fail or could it be that the old parts (PSU & MoBo sata controller) created a redundant fault that surfaced 2 weeks later?

Also I did the WD drive test on the 1 remaining Caviar green drive all tests were fine but writting 0 to all sectors kicked out error/status code 0112 = IRQ timeout. should i be worried?