• Please review our updated Terms and Rules here

Catastrophic HDD failure (large pics)

mR_Slug

mR_Slug

Veteran Member
Joined
Nov 28, 2006
Messages
941
Location
UK
I've been going thru some old hard drives and chucking out the knackered ones. Came across two WD drives that failed in an odd way.

#1 - The studs holding the magnet in place have somehow broken:
hdd-1-broken-stud.jpg



#2 - Total carnage!
hdd-2-close.jpg


there's a few more of the 2nd here: http://108.59.254.117/~mR_Slug/projects/hdd_fail/

I pulled up the top platter and it came off like you'd pop those components out of the plastic for an air-fix model. It seems each head held on to the head assembly until it had dug nearly halfway through the platter.

I also found some IBM re-branded 36GB drives that state their 15K. IBM PN: 06P5332, actual PN: Seagate ST336605LC. According to Seagate's datasheet these are 10K drives. Not sure if this is a mistake or some under-handed practice.
 
Holy crap shoot batperson! That platen look like the rotors on 2006 my Ford Ranger beater.
 
Last edited:
Back in the mid 90's I had a service call from a hardware store. The owner had gone on vacation and his computer was making 'a strange noise'. As soon as I walked in, I could hear the drive screaming in pain. Apparently the owner had told his employees not to shut down the computer for any reason. Opened up the drive and it looked pretty much like your second picture, one head came off, a spiral scratch to the park zone, and a groove half way through the platter. They had left it running for two days.
 
Yeesh, that's some carnage. I suspect the bolt was either overtorqued or had a defect in it that caused it to snap.

The worst failure I ever had was a 15k RPM SCSI drive where the top platter sheared off the spindle and sliced through the lid seal before exploding to bits inside the server case.
 
GiGaBiTe said:
Yeesh, that's some carnage. I suspect the bolt was either overtorqued or had a defect in it that caused it to snap.

The worst failure I ever had was a 15k RPM SCSI drive where the top platter sheared off the spindle and sliced through the lid seal before exploding to bits inside the server case.
Click to expand...

Was it a Deathstar drive?
 
No, it was a Seagate.

IBM never used the Deskstar moniker for their enterprise SCSI drives, think they called them Ultrastar or something. I have a small pile of IBM SCSI drives and they're built like tanks, never had any issues with them.

Deathstar drives on the other hand, I had 6 of them fail back to back in the early 2000s.
 
IBM hard drives (except for that one batch of deskstars) have been rock solid for ages, kind of a shame they sold the line off but then again IBM does not play in the commodity field which is what most PC hardware is these days.
 
VERAULT said:
so the rare-earth magnet came free at full RPM and shattered to bits breaking everything? NEAT!
Click to expand...

No, there are two different drives of the same make and model, that failed in different ways. They are "COMPAQ" but I think they are re-branded WD drives. The link to more images shows one of the drives top covers. The one with the broken magnet would have made a great door-knock sound.

GiGaBiTe, I didn't know the 15k drives could do this. I had a ~150GB 15K drive die at night time a few months ago. The sound it made I can only describe as terrifying. A real guttural sound. It let out a scream like a cat that was having its very soul ripped out of it straight into hell. I though I was having a nightmare until I heard it again in the morning. I'll have to get to that one sometime.

lyonadmiral, Funny you should mention IBM drives. I've been trying to get some 0662 drives to work this evening. Two of them were formatted to 528 byte sectors, which threw me a bit. sg_utils sorted that out. One seems to be dead-dead. These were kinda proto-Ultrastars. Their the odd ones that have a copper tape band through the middle, to connect the two halves. There is no top. I pulled apart the dead-dead one, interesting design.
 
I've been going through some IBM DRVS drives. I had quite a few failures of these drives. It turns out the PCB connects to the motor and HDA with contact pads instead of pins. I scratched them up and most of the drives are working again. If the drive gives a start unit failure this may be the issue. Thought I'd pass this along.

Images of drive, pads etc: http://108.59.254.117/~mR_Slug/projects/hdd_fail/ibm/

I've got a few PCB's that still wont power up a good drive though. Anyone know what the most likely component that's failed in this area: (The arrow points to a 470uF cap on the other side that checks out ok either in-circuit or out of circuit.)
close-up-circuit.jpg



I don't know SMD stuff that well. I'm not going to spend a huge amount of time on these, its just I've got more disks that work than PCBs:)
 
Unknown_K said:
IBM hard drives (except for that one batch of deskstars) have been rock solid for ages, kind of a shame they sold the line off but then again IBM does not play in the commodity field which is what most PC hardware is these days.
Click to expand...

Around 1993 I was building a dual boot Win 3.1 and OS/2 machine based on a DX2-80 cpu and I was lucky enough to trade some stuff for an IBM 0664 2GB SCSI drive to use with it. Once Adaptec sent me a ROM update for my A1522 it was a fast and rock solid drive in that system. If I recall correctly, it was about a $2k drive at that time, so I was happy I wasn't actually buying it.
 
CDC Wren drives have also held up well. I'm still running a 386 with a FH 300MB SCSI Wren IV drive. Driven with an Ultrastor 14N controller (a fairly rare bird).
 
You must log in or register to reply here.
Back
Top