Hard Drive Overheating?

  • Thread starter Thread starter AdenOne
  • Start date Start date
A

AdenOne

In my second home PC I have a 250GB Seagate 7200.10 SATA drive, CPU is
Pentium D 925 3GHz.

Now what I don't understand is, my hard drive feels hotter than the
CPU heatsink!

Is this right? Using a test utility, it shows hard drive at between 44
and 49 degrees C depending on what im'e doing. Have 2x 80mm case fans.

Otherwise there are no issues, get a good 70-80MB\second read rate and
no clicking or other funny business, but was just wondering if it
would shorten the drive's life at all? Drive is about 5 months old but
was only installed about 3 months ago.
 
AdenOne said:
In my second home PC I have a 250GB Seagate 7200.10 SATA drive, CPU is
Pentium D 925 3GHz.

Now what I don't understand is, my hard drive feels hotter than the
CPU heatsink!

A fan will do that.
Is this right? Using a test utility, it shows hard drive at between 44
and 49 degrees C depending on what im'e doing. Have 2x 80mm case fans.

That's fine.
 
Somewhere on teh intarweb "AdenOne" typed:
In my second home PC I have a 250GB Seagate 7200.10 SATA drive, CPU is
Pentium D 925 3GHz.

Now what I don't understand is, my hard drive feels hotter than the
CPU heatsink!

Is this right? Using a test utility, it shows hard drive at between 44
and 49 degrees C depending on what im'e doing.

Way too hot.
Have 2x 80mm case fans.

Is either of them blowing past the HDD?
Otherwise there are no issues, get a good 70-80MB\second read rate and
no clicking or other funny business, but was just wondering if it
would shorten the drive's life at all?

Almost certainly, by quite a bit I'd say.
Drive is about 5 months old but
was only installed about 3 months ago.

I don't even consider running a 7,200rpm (or faster) drive unless it's
_directly_ cooled by at least one 80mm fan. (I prefer a 120mm fan.)

You *do* have a lot to learn don't you?

Word; it's far less expensive to learn about something *before* you do it.
 
Is either of them blowing past the HDD?

No.
I don't even consider running a 7,200rpm (or faster) drive unless it's
_directly_ cooled by at least one 80mm fan. (I prefer a 120mm fan.)

And how do you propose I mount such a fan, as my case has no mounting
holes for a fan near the HDD area? I cannot buy a 5.25" HDD cooler as
both of my 5" bays are used.
You *do* have a lot to learn don't you?

Word; it's far less expensive to learn about something *before* you do it.

I admit I might not know everything about everything like some people,
thats why I ask. Luckily I have the cash to replace anything that
breaks in my computers. And I have never once yet had to replace
something I broke through stupidity, in over 8 years of working with
computers every day. HDD companies should say "heatsink and\or fan
required" - nowhere do they provide information on maximum heat levels
and so on.
 
AdenOne said:
And how do you propose I mount such a fan, as my case has no mounting
holes for a fan near the HDD area? I cannot buy a 5.25" HDD cooler as
both of my 5" bays are used.

I install my HDs vertical, either hanging from the drive rack or on
the side of it or sitting on the floor of the case, Then a fan at
the lower front of the case can blow over the HDs.
 
~misfit~ said:
Somewhere on teh intarweb "AdenOne" typed:
.... snip ...


I don't even consider running a 7,200rpm (or faster) drive
unless it's _directly_ cooled by at least one 80mm fan. (I
prefer a 120mm fan.)

I am running a 30 GB and an 80 GB 7200 rpm drive. They don't even
feel warm. No extra fans on them.
 
AdenOne said:
And how do you propose I mount such a fan, as my case has no mounting
holes for a fan near the HDD area? I cannot buy a 5.25" HDD cooler as
both of my 5" bays are used.


I admit I might not know everything about everything like some people,
thats why I ask. Luckily I have the cash to replace anything that
breaks in my computers. And I have never once yet had to replace
something I broke through stupidity, in over 8 years of working with
computers every day. HDD companies should say "heatsink and\or fan
required" - nowhere do they provide information on maximum heat levels
and so on.

I've used fan such as this in the past with moderate success:
http://www.newegg.com/Product/Product.aspx?Item=N82E16835888105

Those small fans are a bit of a crap shoot, as some of them are noisy
right out of the box. It took 15C right off a pair of hard drives I had
in a situation similar to what you've described.
 
In my second home PC I have a 250GB Seagate 7200.10 SATA drive, CPU is
Pentium D 925 3GHz.

Now what I don't understand is, my hard drive feels hotter than the
CPU heatsink!

Is this right? Using a test utility, it shows hard drive at between 44
and 49 degrees C depending on what im'e doing. Have 2x 80mm case fans.

Otherwise there are no issues, get a good 70-80MB\second read rate and
no clicking or other funny business, but was just wondering if it
would shorten the drive's life at all? Drive is about 5 months old but
was only installed about 3 months ago.

You are vague about what this "test utility" is, but
generally it would be reporting a Smart value for the drive
temp. While the Smart value is certainly better than
nothing, it is only the spot temp of one component on the
drive and does not tell the whole story about whether any
individual parts are cooler or hotter than this temp.

Generally speaking, the drive you listed above should not be
feeling particularly warm when installed in a case with
adequate drive bay airflow. What case are you using and
where is the drive mounted in it?

Certainly one of the more common ways to keep a drive cool
is when the case has a front fan mounted, a case intake fan
that blows through the bay. Providing the case does not use
a mostly obstructed stamped-out set of slits or holes, this
can be a very effective way to cool HDDs, how many can be
cooled this way depends on the specifics of the room ambient
temp, position of the drives relative to the fan, fan
diameter and placement, whether there's a filter installed
in front and the case front bezel intake area.

A well designed case can cool a few drives in a rack
adequately without a front case fan if care was taken in
it's design such that most if not all of the intake airflow
comes in through the HDD rack. Unfortunately these days
cases are being made with several aux. intake areas that
countermine this cooling strategy, such as side vent holes,
rear vent holes. Covering these side and rear passive vent
holes will improve the front intake rate, though in some
cases the intake area is insufficiently low such that either
the metal or (typically plastic) front bezel area may need
to be increased... it varies quite a bit based on the
specific case.

Another concern as significant is that if there is low
enough intake flow rate that the drive is running warmer
than usual, it may also result in other parts running warmer
than usual, not just those we often have temp reports on
like a chipst or CPU.

Therefore without more info the only suggestion I have is to
look over all areas of the case cooling subsystem to assess
where there is a weakness as the aforementioned model of
drive is not difficult to keep cool enough that it only
feels mildly warm if that, providing the case airflow is
good and the ambient (room) temp is not excessively high.
 
And how do you propose I mount such a fan, as my case has no mounting
holes for a fan near the HDD area? I cannot buy a 5.25" HDD cooler as
both of my 5" bays are used.

Start by telling us what case you have, and providing a link
to a picture of it that shows the front and inside angle
shot of the HDD bay where we can see in at the front metal
wall of the case.

Often I *refurbish* old OEM cases, and one of the things I
always do is pop the front bezel off and put a large fan
hole in front of the HDD bay. Sometimes I don't even put a
fan in it, just the big hole allows more airflow through the
bay, but often I'll go ahead and put the fan in and if the
need for airflow isn't high it will be a very low speed fan,
perhaps even throttled back to lower than default RPM.

Cutting out a hole is best done when the case is empty of
course, otherwise bits of metal may cause harm as they're
harder to keep from flying everywhere unless you use a metal
nibbler which may work fine on a cheap thin walled case but
not so well on older or more expensive cases with 0.8mm or
thicker sheeting (at least not with the hand, manual
nibbler, an electric one will cut through thicker metal with
ease but also increase the risk of throwing bits of metal in
random directions). I prefer using a sabre saw before the
case is populated with parts, but that's not a good option
now unless you are willing to empty the case and clean it
afterwards.

In some situations the better option is to buy a new case,
fully prep the new case for great airflow, then transfer the
parts all at once when you're done. This minimizes system
downtime and then you have the original case empty to mod
towards better airflow, and a platform upon which to build
your next/replacement system without taking apart the
current build yet.


I admit I might not know everything about everything like some people,
thats why I ask. Luckily I have the cash to replace anything that
breaks in my computers. And I have never once yet had to replace
something I broke through stupidity, in over 8 years of working with
computers every day. HDD companies should say "heatsink and\or fan
required" - nowhere do they provide information on maximum heat levels
and so on.

Sometimes there is an unwritten assumption that the system
integrator takes it upon him/herself to ensure the design
keeps parts cool enough. There are too many variables to
simply insist a fan is needed, as OEMs do often carefully
design such that they can keep one HDD cool enough with only
a rear exhaust fan that also cools the CPU (2nd exhaust fan
in PSU). Unfortunately older OEM cases were designed for
an era when even HDDs ran cooler, and many aftermarket cases
weren't designed with much cooling in mind at all.
Fortunately the average branded case today is designed much
better than 4 years ago, but the generics are often still as
bad as 4 years ago.
 
Ok, I don't know the case make or model, all of its drive bay's are
used up, and I back up every 4 hours due to this being a critical data
storage PC for the work I do.

So if the HDD fails I will just buy a new one, which would work out
about the same as a new case these days. Ime not too phased about the
current setup as its been fine ever since setup, but I was just
wondering, as I opened the case to snap in a new graphics card and
felt the HDD and wondered if it should be so hot. I have since managed
to temporarily place an 80mm fan blowing onto the HDD from below, and
its dropped by about 10 degrees, but i cant leave the fan here as its
balanced on the cables, precariously..

Thanks for the help though.
 
"AdenOne"
And how do you propose I mount such a fan, as my case has no mounting
holes for a fan near the HDD area? I cannot buy a 5.25" HDD cooler as
both of my 5" bays are used.

I cut two pieces of scrap sheep metal to attach the fan with, one for the
fan's top two holes, one for the fan's bottom two holes. The drives stick
about an inch closer into the case, which of course necessitates drilling
new mounting holes for the drives.

If you are able to locate the fan farther forward you wouldn't need to drill
new drive mounting holes.

Jon
 
~misfit~ said:
Somewhere on teh intarweb "AdenOne" typed:

That is OK - the heatsink is cooled by a fan, so it shouldn't feel too hot.
Way too hot.

Don't let misfit scare you - his name indicates his status in this group!

I had reason to contact Western Digital regarding hard drive temperatures
recently and they confirmed by email to me that the operating temperature of
the high performance SE16 drives is up to 60 degrees C. Asuming that Seagate
drives have similar tolerances, then 44-49 degrees is not way too hot, that
temperature is possibly a few degrees warmer than average, but well within
normal tolerances.

I see from your other replies that you have placed a fan near the drive to
lower its temperature by 10 degrees - nice going. So I don't think you need
to worry any more.
 
That is OK - the heatsink is cooled by a fan, so it shouldn't feel too hot.


Don't let misfit scare you - his name indicates his status in this group!

I had reason to contact Western Digital regarding hard drive temperatures
recently and they confirmed by email to me that the operating temperature of
the high performance SE16 drives is up to 60 degrees C. Asuming that Seagate
drives have similar tolerances, then 44-49 degrees is not way too hot, that
temperature is possibly a few degrees warmer than average, but well within
normal tolerances.

I see from your other replies that you have placed a fan near the drive to
lower its temperature by 10 degrees - nice going. So I don't think you need
to worry any more.


http://209.85.163.132/papers/disk_failures.pdf
 
Jon Danniken said:
"AdenOne"

I cut two pieces of scrap sheep metal to attach the fan with, one for the
fan's top two holes, one for the fan's bottom two holes. The drives stick
about an inch closer into the case, which of course necessitates drilling
new mounting holes for the drives.

If you are able to locate the fan farther forward you wouldn't need to drill
new drive mounting holes.

Jon

Yep...
I do the same thing...just make my own fan brackets.


The HD should not run that hot. I had one in an enclosure than did not allow
for proper cooling.
I was getting occasional errors until I figured out the drive was almost too
hot to touch.
When I removed it from the enclosure...the temp came down to "just slightly"
warm.
 
Jon Danniken said:
"AdenOne"

I cut two pieces of scrap sheep metal to attach the fan with, one for the
fan's top two holes, one for the fan's bottom two holes. The drives stick
about an inch closer into the case, which of course necessitates drilling
new mounting holes for the drives.

If you are able to locate the fan farther forward you wouldn't need to drill
new drive mounting holes.

Jon

Yep...
I do the same thing...just make my own fan brackets.


The HD should not run that hot. I had one in an enclosure than did not allow
for proper cooling.
I was getting occasional errors until I figured out the drive was almost too
hot to touch.
When I removed it from the enclosure...the temp came down to "just slightly"
warm.
 
Ken said:

Interesting, but an inconclusive paper. It presents a lot of data and facts,
but draws very little in the way of conclusion or recommendations. The say,
"One of our key findings has been the lack of a consistent pattern of higher
failure rates for higher temperature drives". In other words, temperature
(within the operating range) makes no difference to drive failure rates!

The charts in this study appears to show that a drive temperature of 40-50
actually gives a significantly lower chance of drive failure, but this is
probably due to the higher volumes of drives in the other bands. Therefore
the lines will tend to balance out and result in temperature in the normal
operating range of a drive to have no effect on failure rate. Temperatures
outwith operating temperatures are likely to damage the drive and cause
failure.

Another point on these results is that their temperature range results don't
extend enough at the 'hot' end. They include all temperature reports above
49 degrees in one band. The operating temperature for the SE16 extends to 60
degrees, so their 49+ band counts for a significant number of drives
operating within 'normal' temperatures as well as those operating outwith
recommended temperatures. This explaines the large variation on standard
deviation observed in this banding.
 
AdenOne said:
Ok, I don't know the case make or model, all of its drive bay's are
used up, and I back up every 4 hours due to this being a critical data
storage PC for the work I do.

So if the HDD fails I will just buy a new one, which would work out
about the same as a new case these days. Ime not too phased about the
current setup as its been fine ever since setup, but I was just
wondering, as I opened the case to snap in a new graphics card and
felt the HDD and wondered if it should be so hot. I have since managed
to temporarily place an 80mm fan blowing onto the HDD from below, and
its dropped by about 10 degrees, but i cant leave the fan here as its
balanced on the cables, precariously..

Do you have small children, or can you make one? They are excellent for
tasks such as holding cooling fans up until the age of about 12-13, when
they start getting resentful.
 
With some cases the adjacent drive bays don't have enough
space between them in a HDD rack to allow airflow between
the drives. If this is the situation you could leave an
empty bay between each drive, use a different case, or try
to put a fan or another fan in somewhere. Not knowing the
exact case it is difficult to say if adding another fan
would be effective enough.


Why not just have a decent case you can reuse? A case that
causes overheating risks more than one drive, and a case
does tend to cost less than a drive if you keep an eye out
for sales and rebates (at least in the US) but we dont' know
how many HDD bays you need.


That may help, though you wrote you had all drive bays used
so is this effective for the other drives? In situations
where it's difficult to find a place to mount a fan what I
do is take a piece of cardboard and make a template for a
fan bracket. Bend the edge to make a lip that will screw to
some portion of the case and trace out a curve on the other
end against the inside of the fan frame, then use the
template to cut the part out of sheet metal (I prefer
aluminum, at least .06" or thicker though thicker is harder
to bend the first edge.

They do make HDD fans on plates that screw to the bottom of
a HDD, my main gripe with those is that often the fans are
cheap failure-prone and noisey high RPM junk.
Do you have small children, or can you make one? They are excellent for
tasks such as holding cooling fans up until the age of about 12-13, when
they start getting resentful.


LOL
 
I had reason to contact Western Digital regarding hard drive temperatures
recently and they confirmed by email to me that the operating temperature of
the high performance SE16 drives is up to 60 degrees C. Asuming that Seagate
drives have similar tolerances, then 44-49 degrees is not way too hot, that
temperature is possibly a few degrees warmer than average, but well within
normal tolerances.

Smart temp is the temp of one point on the drive where the
sensor is, not a comprehensive way to know if the entire
drive including individual components are cool enough.

Unfortunately since a hard drive is not a candle, it doesn't
just melt at an exact temp but rather individual components
create different heat levels. For example motor controllers
or their powered transistors often get quite hot, or the
logic chip/*processor* being hot can cause logical lockups
which aren't fatal but require cooling down and possible
corruption or loss of data.

I see from your other replies that you have placed a fan near the drive to
lower its temperature by 10 degrees - nice going. So I don't think you need
to worry any more.

This may be true, it would be easy enough to touch test the
drive to get an idea if it is relatively cooler than it was.
 
Interesting, but an inconclusive paper. It presents a lot of data and facts,
but draws very little in the way of conclusion or recommendations. The say,
"One of our key findings has been the lack of a consistent pattern of higher
failure rates for higher temperature drives". In other words, temperature
(within the operating range) makes no difference to drive failure rates!

The charts in this study appears to show that a drive temperature of 40-50
actually gives a significantly lower chance of drive failure, but this is
probably due to the higher volumes of drives in the other bands. Therefore
the lines will tend to balance out and result in temperature in the normal
operating range of a drive to have no effect on failure rate. Temperatures
outwith operating temperatures are likely to damage the drive and cause
failure.

Another point on these results is that their temperature range results don't
extend enough at the 'hot' end. They include all temperature reports above
49 degrees in one band. The operating temperature for the SE16 extends to 60
degrees, so their 49+ band counts for a significant number of drives
operating within 'normal' temperatures as well as those operating outwith
recommended temperatures. This explaines the large variation on standard
deviation observed in this banding.


Yes, the problem with that study is it was done on drives
they *needed* to use, that were implemented with the idea
they would be cool enough in a properly designed system, not
drives allowed to get overly hot beyond a threshold.

The other problem is what I mentioned in another post, a
drive is not just a blob of evenly conducting metal so one
temp report at one spot on a drive is not telling the whole
story, particularly in a situation like the OP had where
drives may be close together.
 
Back
Top