Sign in to follow this  
sunama

so many people's hard disks fail, yet none of mine have. why

Recommended Posts

ive seen many posts of people reporting many hard disks failures. in fact, some people get through as many as 3 hard disks in 3 months. is it something perhaps that these people are doing (or not doing)?

ive owned the following hard disks:

maxtor 6.4gb - 5400rpm (no cooling)

quantum 9.1gb - 7200rpm (air cooled)

quantum 60gb - 7200rpm (air cooled)

seagate bara IV 60gb (watercooled)

seagate bara IV 80gb (watercooled)

WD1200JB (watercooled)

the only problem i had was with the seagates. however, once i changed the side of the hard disks that was being watercooled, no further problems.

ive also built 2 pcs for my friends. using the following hard disks:

ibm 75gxp - 30gb (air cooled)

2 x ibm 60gxp RAID - 60gb (air cooled)

no problems with the above hard disks.

now from the above, 9 hard disks, none have caused me any problems. the key to them all (except the 5400rpm maxtor) is that they have all had active cooling on them.

so im hypothesising that perhaps, the main cause of hard disk failure could be heat related. do u guys agree or disagree?

its just that i find it really strange that ive not had a single hard disk failure from 9 hard disks, all from different manufacturers (and please dont say its luck, 9 hard disks working perfectly isnt luck).

perhaps if people who have hard disks fail on them could report the cause that they believe led to their hard disk failures. also perhaps if they could mention what sort of cooling they had, if any.

Share this post


Link to post
Share on other sites

I have a feeling it's more to do with handling and shipping by the distributor/supplier.

I buy almost all my hard-drives from the same supplier, and never have any failures. The few drives I got off other suppliers are the only ones I've had problems with.

My preferred supplier always has a good stock turnover (I saw them shift 16,000 Barracuda IV 40Gb drives in a week), and ships the drives in sturdy boxes with over 1" of foam rubber surrounding the drive.

Share this post


Link to post
Share on other sites
I have a feeling it's more to do with handling and shipping by the distributor/supplier.

I buy almost all my hard-drives from the same supplier, and never have any failures.  The few drives I got off other suppliers are the only ones I've had problems with.

My preferred supplier always has a good stock turnover (I saw them shift 16,000 Barracuda IV 40Gb drives in a week), and ships the drives in sturdy boxes with over 1" of foam rubber surrounding the drive.

I agree... it's handling... the distributor.... the retialer... and the consumer all have to handle the drive properly, if one of them can't manage that task then the drive is toast.

I've had more HDD's than you, and I've only had 1 fail in service(80gb 120gxp). I don't use active HDD cooling.

there are exceptions... some people have crappy power supplies or inadaquate case cooling, or some other factor that probably contributes more than anything else to *their* falures.

Seagate U series 4gb

Seagate ATA IV 40gb

2 x IBM 75GXP 15gb

IBM 120gxp 80gb

WD 1200bb

WD 800JB

Fujistsu 10GB

3 x quantum prodrive 200-500MB

2 x WD caviar 200-400MB

2 x Conner 1GB

I've had a couple DOA's from poor shipping(100 gb WD drives), and the 120GXP RMA that was bad when I got it. The original 120GXP failed after 1 month of use.. so you be the judge whether it was in-service or DOA.

Share this post


Link to post
Share on other sites

have u guys ever felt the surface of your hard disks, while in operation? cuz my wd1200jb was so hot (when no cooling was used and i was intalling win2k) that i couldnt keep my finger on the pcb side of the hard disk for longer than 5-10 secs. now i cant believe that without cooling, this hard disk would still be functioning today. Now alot of people dont use any cooling and surely its because of this that alot of hard disks (that get as hot as my WD1200jb) may actually be failing.

might i suggest that before people leave their hard disks running (and closing the lid of their computer case) that they at least check for excess heat buildup on the surface of the hard disk casing.

btw my seagate baracudas are alot cooler than the wd1200jb. they get warm, but not hot.

Share this post


Link to post
Share on other sites

I have a 1200JB in my home computer and do not notice any excessive heat. Inside case doesn't go beyond about 35 C or so, but it's a mid-size tower with a P3-550 and Matrox G450, 2 HDD's (1200JB and 200BB), CDROM, and CDRW. Maybe I just get enough airflow or something.

Share this post


Link to post
Share on other sites
I have a 1200JB in my home computer and do not notice any excessive heat. Inside case doesn't go beyond about 35 C or so, but it's a mid-size tower with a P3-550 and Matrox G450, 2 HDD's (1200JB and 200BB), CDROM, and CDRW. Maybe I just get enough airflow or something.

but have u actually tried touching the hard disk while it is in full flow? eg try transferring a bunch of files (say 2gb worth), transfer from one folder to another, then try touching the pcb side of the hard disk. it was only when i started installing win2k that i noticed this "extreme" heat.

although your case temp might be 35C (or lower) this doesnt mean that the casing of the hard disk isnt way way way more higher than that. in fact, when i touched the hard disk, i was running it outside the computer case (i was installing win2k on it, before i installed the hard disk into the computer case, with its waterblock. i just wanted to make sure the hard disk was in full working order before i went through the hassle of waterblock attaching). so while the hard disk was too hot to touch, the ambient temp that the hard disk was operating in was well below 20C!

Share this post


Link to post
Share on other sites
have u guys ever felt the surface of your hard disks, while in operation?
If you've actually done that and your disks are still running, I'm quite amazed! :)

Yes, Im also amazed that he has a finger left... But I think he ment the outside of the disk, not the platter...

Yes some drives can be warm to the touch while they're in operation, but Ive never felt the impulse to provide any active cooling solution for them...

Share this post


Link to post
Share on other sites

I don't know why we have such an obsession with air-cooling these devices... They're designed to run hot, aren't they? I feel if you want a cooler drive, buy a 7,200 or a 5,400 rpm instead of a 10,000 one, or don't buy a hot drive like the IBM 36Z15.

I believe a properly mounted drive will convey heat through the metal screws and metal mounting bracket it's attached to. Additional cooling fans near the drive only impart more vibration and noise and maintenance headaches for the chassis, don't you think?

I know I'm going to get flamed for this--let me have it.

Share this post


Link to post
Share on other sites
I believe a properly mounted drive will convey heat through the metal screws and metal mounting bracket it's attached to.  Additional cooling fans near the drive only impart more vibration and noise and maintenance headaches for the chassis, don't you think?

aaah. but i watercool my hard disks, further more watercooling the hard disks enables me to wrap up the hard disks in denim. this allows me to completely isolate the disks from any part of the computer case, hence no vibration, but also no passive cooling.

when i did air cool my hard disk, i used to cool it cuz i heard many other people recommending that this be done. ive had 0 hard disk failures. whether this was as a result of cooling the hard disks, i know not.

where are the people who had lots of hard disk failures? they dont seem to be coming on here and explaining the circumstances under which their hard disks failed. i read some people who would go through 3 hard disks in the space of a few weeks. could someone who had this experience please post and give us the setup under which the failure occurred.

thanx.

Share this post


Link to post
Share on other sites
but have u actually tried touching the hard disk while it is in full flow? eg try transferring a bunch of files (say 2gb worth), transfer from one folder to another, then try touching the pcb side of the hard disk. it was only when i started installing win2k that i noticed this "extreme" heat. 

I have touched the drive after powering down to see how hot it got after leaving it running for a while. I have not tried touching it while it's running, since klutz that I am, I'd likely bump the drive too hard and either damage it, or cause a short. ;)

Hard drives are designed to operate with temperatures up around 50-60 C. Of course, that temperature is just about as hot as I can comfortably handle with my bare hand (I've tried fishing test drives out of ovens without using mitts and anything above 60 C I can't do :D).

I really wouldn't worry about it too much. Over on Ars, many users have reported failures with the 75GXP line in spite of plenty of cooling (taken on the drive itself). While adequate cooling isn't likely to hurt longevity, I think it'd be hard to prove it significantly increases it (I'm not talking letting the poor drive exceed its operating specs). (Mis)Handling plays a much bigger role.

Share this post


Link to post
Share on other sites

If your IDE drives are getting too hot to touch than you have major case airflow problems.

I once worked on a crapaq mini tower POS... the only fan was a single 60mm PSU fan blowing into the case... and practically no ventilation anywhere else... this caused the ATA IV in it to become dangerously hot... too hot to touch... however, in a normal case this drive runs cooler than body temp.

In my current enlight case ( Here )

I had 2 75GXPs and 1 bulk storage drive (went from ata IV->120GXP->1200bb)

All the drives ran slightly warm to the touch, never hot. This case has an ~1800RPM PSU fan and a single 80mm medium speed(~2000rpm) panflo case fan.

Share this post


Link to post
Share on other sites

9 is hardly a large #. I alone have 10 running in my house right now. I also build a lot of computers, i'd say out of 100 of them, only 9 or so have died, 6 were IBM and 3 were maxtors. Most of them were actually IBMs, i haven't really had any problems with the IBMs up until recently, when they started dropping like flies left and right. I also put in about some 75 WD 200BB 20GB drives in comps at school close to 2.5 years ago. So far only 4 have died, and 2 were DOA.

Share this post


Link to post
Share on other sites

In the last 12 months I've bought 16 hard disks for myself. None are actively cooled in any way. Out of those 16, 3 have failed. All were Western Digital drives spanning the whole range of the product lineup.

Out of the 60-odd operating disks in my apartment, only two others have failed, an old Quantum AS and a 75GXP. Both had head crashes.

All the drives in all the computers in my apartment have the same handling by me, and close to the same environmental conditions. My drives come from, well, everywhere. The WDs mostly came from retail superstores like Circuit City, and they're the only guys I seem to be having problems with.

I *never* turn my PCs off. Maybe that has something to do with it. In general I'm pretty philosophical about things. I have way more drives than any "home user", so even if I don't see as many as a retailer, I still have a big enough sample size to make generalizations. So here's my generalizetion: It's luck of the draw, man. I've got a 5MB drive that still spins up. I had a four month old 120GB drive die. It's disappointing but for a "normal" drive, it isn't like you can point to one thing and say "this is why this drive failed."

Share this post


Link to post
Share on other sites

is there anyone here then who was actively cooling their hard disk and had a failure. cuz people who have reported failures (above) have all said that they havent used any sort of cooling.

im just trying to find some sort of pattern, cuz im not the sort of person who believes in "luck". you make your own luck.

btw i recd my wd1200jb wrapped in some bubble wrap (no 1" thick sponge). it wasnt shipped in ideal conditions. the drive has served me perfectly for the last 9 months...and counting.

the way i see it is that if a drive was shipped/handled badly, it will either be DOA or it will run ok for a few days then fail. it wont fail after a year or so due to bad handling. the 75gxp/60gxp drives were failing after a whole range of durations (from DOA to a couple of years).

also did anyone notice in the reliability survey that the 75gxp are actually more reliiable than the 60gxp. so ibm released the 75gxp, then they went and released an even worse drive. this must've cost them an arm and a leg to deal with all those RMA's!

Share this post


Link to post
Share on other sites

perhaps the reliability survey should have an extra field where the users can fill in whether or not the drives are/were actively cooled or not.

this would give us a much better idea of whether or not active cooling can actually increase the life of a drive or not. perhaps Eugene could implement this?

Share this post


Link to post
Share on other sites
the way i see it is that if a drive was shipped/handled badly, it will either be DOA or it will run ok for a few days then fail. it wont fail after a year or so due to bad handling.  

They can and do fail a lot later due to handling damage. Mishandling also includes ESD damage. This kind of damage can take months to show up.

also did anyone notice in the reliability survey that the 75gxp are actually more reliiable than the 60gxp. so ibm released the 75gxp, then they went  and released an even worse drive. this must've cost them an arm and a leg to deal with all those RMA's!

I can't say that I noticed that in the reliability survey, but I do have a possible explanation for what looks like a stupid decision on IBM's part. Quite often, due to the short development cycles of disk drives, a product that is currently in development is too far along to incorporate design changes that seem like a good idea (based on how an earlier product fared).

For example, let's say there is a fundamental design flaw in the 75GXP line (I don't know if there is, but let's pretend). The team working on its successor, the 60GXP line, sees this, but maybe they don't have the time/resources to devise a suitable fix. They release the 60GXP as-is, potentially with the same flaw the 75GXP has. Instead, they focus on the next generation, the 120GXP, and work on putting a fix into that product, since there is more time to develop the fix.

Share this post


Link to post
Share on other sites
btw i recd my wd1200jb wrapped in some bubble wrap (no 1" thick sponge). it wasnt shipped in ideal conditions. the drive has served me perfectly for the last 9 months...and counting.

the way i see it is that if a drive was shipped/handled badly, it will either be DOA or it will run ok for a few days then fail. it wont fail after a year or so due to bad handling. the 75gxp/60gxp drives were failing after a whole range of durations (from DOA to a couple of years).

All the drives I've purchased, except 1, have been recieved this way or worse... My retail 800JB is the only retail drive i have purchased directly, it came in plastic isolators... my replacement 120GXP arrived in plastic isolators as well... but it was basically DOA.

AFAIK most problems occur from handling either by the consumer or the retailer. If the retailer doesn't specialize in hard drives and they are handling the drives, chances are they're doing it wrong.

Also, i think most people who buy HDD upgrades and install the drive themselves are probably doing something wrong.

I've seen workers at Circuit City drop drives on the counter, I've seen customers at microcenter spinning drives around and catching them (these are new $100+ drives)... I've seen unpadded new drives in computer shops stacked 10 more more high on shelves...

This behavior really makes me not want to buy a hard drive.. i wish I could buy direct from a distributor...

Share this post


Link to post
Share on other sites
ive seen many posts of people reporting many hard disks failures. in fact, some people get through as many as 3 hard disks in 3 months. is it something perhaps that these people are doing (or not doing)?

How many times have you heard:

"Hey, my hard drive has been running great for the last 6 months! Thanks WDC!!"

?

Share this post


Link to post
Share on other sites
i read some people who would go through 3 hard disks in the space of a few weeks....

I have nine drives in four computer, the oldest of which is getting to about 5 years. None of my personal drives have ever failed. Of the drives I sell, I occasionally (2-3 per year) fail within a couple of weeks of leaving the shop, I normally put these down to either manufacturing defect or unlucky shipping. All the computers I sell use high quality powersupplies so that may be a reason other people have problems. Does anyone regularly sell cheap and nasty cases/powersupplies and have a high rate of returned hard-drives?? :wink:

I had one customer who went through four hard-drives in rapid succession, after the first two died I though it may have been something to do with the electricity supply (gut feeling) in her area (country town), so I loaned her an old 600va UPS. Two more then died, and when I came to replace those she had unplugged the PC from the UPS "because it kept going flat". This time I gave her my big UPS, 1000va with buck and boost circuit. I got a call from her a week later, the UPS was going flat... I convinced her to complain to the electricity company, who humoured her and installed a monitor in her building. A week later she again called me, the electricity company was installing a new transformer on her block. In Australia the voltage is meant to be 230v +-15%. Over the week, in her building the highest voltage was 210 and the lowest was 150. Since then she purchased my old 600va UPS which isn't going flat and has not gone through anymore hard-drives (18 months and counting)...

My guess is, many drives die because of cheap powersupplies and/or low quality electricity supply. As it is, every computer I own runs on a UPS and hopefully I'll keep my perfect record :wink: touch wood...

Share this post


Link to post
Share on other sites

sunama, what method did you use to water cool your drives?

I'd be interested in anything that is NOT aluminum, as I can't risk my swiftecs or heatercore (which is also copper).

Share this post


Link to post
Share on other sites
Hard drives are designed to operate with temperatures up around 50-60 C. Of course, that temperature is just about as hot as I can comfortably handle with my bare hand (I've tried fishing test drives out of ovens without using mitts and anything above 60 C I can't do :D).

I may be showing some chutzpah by saying this, but I totally disagree.

Perhaps 60C is OK for SCSI disks, but no freaking way for IDE. I would say 45C average, maybe 50C max, should be allowed. Anything higher... prepare the "drive coffin".

The only times that I've had reallocated sectors, on my IBM 14GXP and 75GXP, were when I let my drives overheat, due to poor case cooling. (Common PC cases stack HD mounts right on top of each other, and without active cooling (a fan) nearby, this isn't enough ambient cooling for 7200RPM HDs, unless you have so many case fans that your case screams like a wind tunnel.). The temps weren't even around 60C, they were peaking over 50C, and I started to experience slowdown and eventual lockups of W2K, and upon restart (after a cooldown), there were also some reallocated sectors. This happened slowly over an overnight period while doing something with the machine (that I couldn't interrupt), and you could literally "feel" the slowdown of the machine as the HDs started to fail. I was running DTemp at the time, and could watch the error count fields climb up. It was kind of scary, actually.

I've also known other people who's WD IDE HDs overheated, and took out the filesystem boot sector. (Statistically the most likely point of failure of filesystems in W2K, since it keeps re-writing this sector to update a timestamp.. very poor FS design.)

I think that the "too hot to hold your finger to the side" test is probably a pretty good indicator of "too hot", if that does indeed correspond to ~60C.

With active cooling in my new case, a fan blowing air inside the case directly on the HDs from a front intake vent, with a Maxtor 60GBDM+9 mounted directly on top of an IBM 30GB 75GXP, using MBM 5 to measure SMART temps, I get something like 5C above case abient for the Maxtor, and a couple degrees below abient for the IBM, probably due to the case abient temp sensor being near the middle of the mobo, and the HDs closer to the "fresh" cool intake air. (Roughly 32C ambient, 43C CPU, 38C Maxtor, 30C IBM.)

I did just have the IBM 30GB 75GXP fail on me though, this afternoon, after bringing my computer back from where it was for a LAN party all week. When I got it home and turned it on, I got the infamous IBM "scritch-scritch" of doom. Remembering what one of my friends did with his flaky 40GB WD drive, turning it upside-down, well, I tried that, and it worked, the drive spun up. (Before, it would attempt to spin-up, and fail, and all of the optical drives in the machine would reset too, like there was a brief PSU short or something.. wierd. I think maybe the PCB-to-drive-assembly connector on the IBM is going, that was a known problem with them, among all the others.)

I had also just moved my brand-new 60GB Maxtor drive out of my "LAN party" box into my fileserver, because I thought that the rampload IBM drive would be able to take the shock of transportation better.. guess that theory didn't work. Now I'm not quite sure what to do about a HD for a "luggable desktop" machine, or how to properly transport it. (I also had someone else load the machine into my car today, because I was busy. Perhaps they just tossed it into the back seat instead of laying it down carefully like I do, and that damaged it. Who knows.)

Share this post


Link to post
Share on other sites

unfortunately im using an alumunium waterblock to cool my 2 hard drives that i have at the moment. however, you can buy (in the usa at least) copper hard disk waterblocks. i believe they cost around US$75.

you can read about my hard disk watercooling in the link below:

http://forum.oc-forums.com/vb/showthread.p...&threadid=81923

btw im using copper cpu, gpu, chipset blocks. AL hard disk block and radiator. im using water whetter to slow down corrosion. ive had this setup for a year now and all is fine so far.

Share this post


Link to post
Share on other sites

so to summarise, we are saying that apart from luck, the following can have an effect on hard disk life

heat

handling/shipping/installation technique

psu quality

wildly fluctuating electricity supplies

is this about right?

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this