HOT HOT HOT.. hard drive :eek:

S

SudburyOntario

Gawd
Joined
May 5, 2008
Messages
575
:eek: I just had a 2001 40gb western digital drive inside a computer with the panel off, just to format and test how the sectors are. It was burning hot to touch !

I have never felt a hot drive besides the Maxtor slim profile drives ( which was cooler than this WD ) !!!

Question is.. would this be an overspin problem causeing the extra heat, or just a bad drive model they produced - Enhanced IDE Hard Drive WD400BB-00AUA1
 
No, just some drives run hot... if you check temperatures with HW Monitor another program they're probably 35-40c which is over 100 degrees Fahrenheit so it'd feel hot.
 
I left a 1TB WD black running overnight in a HD dock and the paint got tacky. :eek:

I shoulda used my infrared thermometer to read the temp but it was a little too early in the morning for me to be thinking correctly.
 
I put my server in a cabinet with no airflow (just to see how it'd run) and by the next couple days temps had reached pretty high... low 60's for idle on the processor, and the HDD was in low 50's for c.

Those were the "dreaded" 7200.11 Seagates but they're still kicking, now running at a cooler 30c in a well ventilated case/area :)
 
SudburyOntario said:
Question is.. would this be an overspin problem causeing the extra heat, or just a bad drive model they produced - Enhanced IDE Hard Drive WD400BB-00AUA1
Click to expand...

The problem is you need to have some airflow near the drive.
Stick a fan near it or put the drive near an intake fan.
You have undoubtedly caused excess wear to the drive running it that hot, if you want the drive to last, get the temp down.

You can use the free HDTune to check the drives temperature.
My drive temps are around 30 to 32C when the drives are running (ie not in power save mode).
 
Nenu said:
You have undoubtedly caused excess wear to the drive running it that hot, if you want the drive to last, get the temp down.
Click to expand...

No, this is untrue. Google did a study several years ago about hard drive failure rates in their server farms and they debunked two commonly-held myths about hard drives.

1. high disk usage does not make drives fail faster
2. high temperatures (within reason) do not make drives fail faster.

Here's an excerpt (from a synopsis of the study):
they did find a few interesting patterns in failing hard drives. One of those we thought was most intriguing was that drives often needed replacement for issues that SMART drive status polling didn't or couldn't determine, and 56% of failed drives did not raise any significant SMART flags (and that's interesting, of course, because SMART exists solely to survey hard drive health); other notable patterns showed that failure rates are indeed definitely correlated to drive manufacturer, model, and age; failure rates did not correspond to drive usage except in very young and old drives (i.e. heavy data "grinding" is not a significant factor in failure); and there is less correlation between drive temperature and failure rates than might have been expected, and drives that are cooled excessively actually fail more often than those running a little hot.
Click to expand...

http://labs.google.com/papers/disk_failures.pdf Here is the full study if anyone is interested
 
From their chart, it looks like once you get past 45C you start getting a higher chance of failure.
 
Eva_Unit_0 said:
No, this is untrue. Google did a study several years ago about hard drive failure rates in their server farms and they debunked two commonly-held myths about hard drives.

1. high disk usage does not make drives fail faster
2. high temperatures (within reason) do not make drives fail faster.

Here's an excerpt (from a synopsis of the study):


http://labs.google.com/papers/disk_failures.pdf Here is the full study if anyone is interested
Click to expand...

Luckily I have my own experience and that of others around me to help make my own conclusions on this as the google report does not reflect the usage patterns of home machines.

Heat does increase wear, as does power cycling.
The higher temp the drive gets over ambient, the more wear power cycling does.

You talk about debunking myths but you ignored the ops first post!
In your point 2. you said high temps "within reason" do not make drives fail faster.
Read the first post again then you will see that what you posted does not apply.
 
I had this drive running in an open case as did all the other drives I tested, but this one takes the cake.. I'm going to try my hdtune which is what I was running.. to see the temps this time.
 
Nenu said:
Luckily I have my own experience and that of others around me to help make my own conclusions on this as the google report does not reflect the usage patterns of home machines.

Heat does increase wear, as does power cycling.
The higher temp the drive gets over ambient, the more wear power cycling does.

You talk about debunking myths but you ignored the ops first post!
In your point 2. you said high temps "within reason" do not make drives fail faster.
Read the first post again then you will see that what you posted does not apply.
Click to expand...

The OPs post stated only that the drives were "burning hot to touch". Where on the graph does that fall?

The linked PDF is an awesome source of information and I'm glad to see it. Anything that makes a hard drive's spin more difficult is obviously going to increase wear, I'd quess the salient temperature then would be the lubricating grease's liquidity/viscosity limits (when it gets too hot and liquid, too cold and chunky) and the metal warp of the platters.

Wonder if the google report takes into account the fact that server HD's don't power cycle much.
 
I've had some 4 and 5 platter drives that got pretty damn hot, even what I might call "burning to the touch". I would add some active airflow and if the temps still seem out of control, then start to worry. Some drives really need active cooling to not get extremely hot.
 
I don't think "it felt hot to the touch" is the most accurate measure of a hard drive's temperature or whether or not it's running too hot. ;)
 
NoEcho said:
The OPs post stated only that the drives were "burning hot to touch". Where on the graph does that fall?

The linked PDF is an awesome source of information and I'm glad to see it. Anything that makes a hard drive's spin more difficult is obviously going to increase wear, I'd quess the salient temperature then would be the lubricating grease's liquidity/viscosity limits (when it gets too hot and liquid, too cold and chunky) and the metal warp of the platters.

Wonder if the google report takes into account the fact that server HD's don't power cycle much.
Click to expand...

The key part is how hot the ops drive got.
Burning to the touch suggests temps in excess of 70C, google only tested to 50C so the graph does not reflect the ops situation.

70C+ (Burning hot!) is a massive change from ambient, power cycling the drive with such hot temps can increase many failures that do not show their effect statistically on the google report.
ie solder joints breaking, connections moving enough to let particles diminish electrical contact, increased corrosion rate, electrical component failure, electrical components values changing, etc...
The google report does consider power cycling but it does not demonstrate the large increase in failure rate that occurs when drives are run very hot and cooled to ambient.
Note that use of power saving makes the drive power cycle much much more so the damaging effects of excess heat are amplified by this unless power saving is turned off.
 
Nenu said:
The key part is how hot the ops drive got.
Burning to the touch suggests temps in excess of 70C, google only tested to 50C so the graph does not reflect the ops situation.

70C+ (Burning hot!) is a massive change from ambient, power cycling the drive with such hot temps can increase many failures that do not show their effect statistically on the google report.
ie solder joints breaking, connections moving enough to let particles diminish electrical contact, increased corrosion rate, electrical component failure, electrical components values changing, etc...
Click to expand...

"Burning hot" is just a phrase, I've handled drives that were 50C (122F) right out of a rack and I wouldn't want to be be holding onto those drives for very long, either.

Point being, the OP stated "burning hot" and that's not enough information to go with. 70C is pretty hot, but I would gander that at 70C, the drive would pretty much be toast. The OP needs to be more specific and get some numbers before we can recommend him any advice. If anything, extra airflow to cool it down wouldn't hurt, but if "burning hot" to the OP meant the drive was in reality at 45-50C, the drive is certainly within tolerances.
 
I shall clarify as best I can since I have nothing available to tell the correct temps.. yes 40gb hard drive harr harr.. ( someone might need one lol ),. Burning hot to the touch meaning it was difinitly hotter than 50c. Also meaning I could have cooked an egg on it :D
 
40 gb hard drives make nice targets for the range.
 
No, this is untrue. Google did a study several years ago about hard drive failure rates in their server farms and they debunked two commonly-held myths about hard drives.
Click to expand...

I've read that study a coupla times because I couldn't believe the conclusions.

The overall statement that "high temps don't make HDs fail faster" refutes everything I've experienced or read.

A computer HD is electro-mechanical and most mechanical devices use cooling to extend their life because heat decreases their lifespan.

Like most stats, you can interpret them a coupla different ways.

Personally, I think IBM's study that concluded.....
As the temperature exceeds the recommended level, the failure rate increases two to three percent for every one degree rise above it. For example, a hard disk drive running for an extended period of time at five degrees above the recommended temperature can experience an increase in failure rate of 10 to 15 percent. Likewise, operating a drive below the recommended temperature can extend drive life
is a little more believeable.
 
A guy I worked with years ago (1998-2001) had a drive that had been running hot so long that it wouldn't work until it spent an hour getting good and hot after a shutdown. His theory was the tracks shifted over time from running too hot.

I just realized the hot spot on the bottom of my laptop was the 7200 rpm hard drive rather than the battery. After 3 years, you'd think I would have figured that out sooner. :rolleyes: At least it wasn't "ouch" hot after running all day. A couple days ago I got a drive failure message when I tried to wake the machine up from sleep mode. Hitachi's disk test failed maybe an hour in with a drive overheated message, so I ordered an SSD. I'm not taking a chance on my disk going belly up at an inconvenient time.

Other than the laptop, I keep my disks nice and cool. I make sure to buy cases that have fan mounts positioned to cool the drives. What can I say, I'm a recovering SCSI junkie.
 
Syntax Error said:
Get some actual temperatures rather than going off something empirical, please. :)
Click to expand...

agreed. Everyone in this thread saying "it's BURNING HOT so it must be more than 50C!" are most likely overestimating the temperature. I think you'd be surprised how hot 50C can feel, especially in something with a large thermal mass like a hard drive. If I stick my hand onto my 74G raptor (which is reporting 45C) it feels pretty freaking hot.

Nemu said:
Luckily I have my own experience and that of others around me to help make my own conclusions on this as the google report does not reflect the usage patterns of home machines
Click to expand...

Care to elaborate? Come on, the hard drive doesn't care what programs are using it and it what patterns. Disk access is disk access. The only difference would be that home drives power cycle more, which very well may contribute to disk failure, but neither the OP's post nor the google study had anything to do with power cycles. The OP was talking about temperature.
 
I'm getting some temps since I just bought something here.. and what's the deal with power cycles ? ( haven't read into that yet ) Thanks
 
Eva_Unit_0 said:
agreed. Everyone in this thread saying "it's BURNING HOT so it must be more than 50C!" are most likely overestimating the temperature. I think you'd be surprised how hot 50C can feel, especially in something with a large thermal mass like a hard drive.



Care to elaborate? Come on, the hard drive doesn't care what programs are using it and it what patterns. Disk access is disk access. The only difference would be that home drives power cycle more, which very well may contribute to disk failure, but neither the OP's post nor the google study had anything to do with power cycles. The OP was talking about temperature.
Click to expand...

The op has already qualified that he could fry an egg on it.

I did elaborate in my reply and the 2 subsequent replies.
Why are you talking about programs and drive access?
This is about drive failures caused by excess heat. No mention of programs and disk access has been made by me.

I'm glad you agree that power cycling increases drive failure.
It matters because if the op has not prevented his hard drive going into a power saving mode or uses power saving intentionally, it will power cycle.
With the temps his drive runs at, that will age the drive much faster.
 
I see.. that should have been obvious.. and no I don't allow my hard drives to power down. That would just get me mad.

Syntax Error, I already wrote that hdtune won't report the temps of this wd hdd.
 
SudburyOntario said:
I'm getting some temps since I just bought something here.. and what's the deal with power cycles ? ( haven't read into that yet ) Thanks
Click to expand...

An electronic device has an offline resting temperature and running temperature(s).
The larger the difference between the two, the greater the stress on components.

Here is an example for soldered joints:
The circuit board on the hard drive is made of many different components, metallic, plastic, fibrous etc.
All of these expand with heat at different rates, the higher the heat, the greater the expansion and thus the greater the difference between strongly attached objects.
This difference in expansion rates creates extra stress in any joints.

Soldered joints are immovable and thus massive stresses build up in these joints due to the difference in expansion of the circuit board, metal tracks, the component soldered and the solder itself.
Being the weakest point, after enough power cycles soldered joints start to develop cracks that end up running through the whole joint.
Its when this happens that is of concern.

The greater the temperature difference between a drives hot and cold states, the faster the cracks will appear in soldered joint.
The more power cycles the drive goes through the faster the failure will occur.
Failure is inevitable if a drive is stressed enough, its better to reduce these stresses so failure does not occur in the expected lifetime of the drive.

There are many more modes of failure due to heat, this is one of them.
 
Syntax Error said:
Are you able to boot into it or access it? I'm surprised it hasn't gone up in flames yet! :eek:
Click to expand...

It's loading and unloading files just fine. It came out of an old Compaq 500mhz pc. Although when I ranhdtune the second time, I noticed the single bad sector was now gone :p

I really am bothered to use the drive for a dummy pc, maybe add two fans for it. This temp came after one hour check in time using hdtune sector check utility to run the drive in.

Regular house hold thermometor told me the temps :D
 
You must log in or register to reply here.
Back
Top