Better HDD cooling by contact with case or open to air?

"Rod Speed" wrote:
In practice the cage is pulled up against the body
of the drive when the screws are tightened.

I've only inspected Maxtor hard drives closely, but all
the Maxtor drives that I've seen have raised bosses
around the screw holes. They are so low that one has
feel them with a finger nail to be sure they're there, but
they *are* there, and they make a small air gap between
the body of the drive and the HD cage. That means
that by metal-to-metal conduction, the bosses and the
mounting screws transfer more heat to the cage than the
sides of the HD.

*TimDaniels*

Timothy Daniels wrote
Rod Speed wrote

In practice the cage is pulled up against the body
of the drive when the screws are tightened.

I've only inspected Maxtor hard drives closely, but all
the Maxtor drives that I've seen have raised bosses
around the screw holes. They are so low that one has
feel them with a finger nail to be sure they're there, but
they *are* there, and they make a small air gap between
the body of the drive and the HD cage. That means
that by metal-to-metal conduction, the bosses and the
mounting screws transfer more heat to the cage than the
sides of the HD.

They still use the cage as a heatsink, otherwise
the cage wouldnt feel warm to the touch.

"Rod Speed" wrote:

Timothy Daniels wrote
Rod Speed wrote

In practice the cage is pulled up against the body
of the drive when the screws are tightened.

I've only inspected Maxtor hard drives closely, but all
the Maxtor drives that I've seen have raised bosses
around the screw holes. They are so low that one has
feel them with a finger nail to be sure they're there, but
they *are* there, and they make a small air gap between
the body of the drive and the HD cage. That means
that by metal-to-metal conduction, the bosses and the
mounting screws transfer more heat to the cage than the
sides of the HD.

They still use the cage as a heatsink, otherwise
the cage wouldnt feel warm to the touch.

My hard drive rack screws have rubber grommets to help reduce
vibration. The only heat conduction from the hard drive to the hard
drive rack is through the four small screws. The intake fan blowing
over the hard drives cools them.

On Thu, 23 Jul 2009 08:05:48 +1000, "Rod Speed"
wrote:


The case does not act like a large heatsink.

Wrong.

You can see this easily enough by measuring the large
difference between the hard drive frame and the metal
wall of the drive cage to which it is attached.

What matters is that the drive cage metal is warmer
than the rest of the case. It always is, and that means
that its acting as a heatsink for the drive.


For it to effectively heatsink it has to have a large area
of conductance, meaning two nearly perfectly flat areas and
an intermediary layer like a silicon pad or grease which is
not present.

Surely you know these very very basic facts about
heatsinking components? If not, let me introduce you to
Google...

Or you could do it the dumb way, unplug all fans and put the
drive in the case, get it's smart temp then shove two pair
of isolating washers onbetween the rack and drive frame then
take the temp again (being sure you let the drive run for
same period and that ambient temperature remains constant).



Since that drive cage is quite thin metal relative to
the case, if anything the temperature differential
would be even less than with normal heatsinks...

Meaningless crap.

Yes Rod, you write meaningless crap a lot.


but it is not because to effectively transfer the heat you
must have two mostly flush surfaces that conduct well.

Have fun explaining why its at a higher temperature than the rest of the case.

??

That proves what I wrote, if it were a good heatsink then it
would not be at a substantially higher temp on the rack
metal adjacent to the drive frame, remembering that the
drive frame itself is just a heatsink too, not the portion
of the drive that creates the heat.

Now contrast something more relevant. The metal heat
spreader on top of a CPU, the corner of it that's several
millimeters from the die, compared to the bottom of the
heatsink which has a much flatter finish than a drive frame,
and grease or a thermal pad. MUCH lower temp difference
between these two mating surfaces because they were MEANT to
be a thermal interface.

If the side of the drive frame were meant to be an interface
it would be obvious that they had machined it as flat as
reasonably possible. You obviously have never looked
closely at a hard drive as this is very easy to see lacking
on any modern drive. I've several makes in front of me at
the moment, none of which have had any attempt at all made
to machine them as flat as the bottom of the cheapest of
heatsinks... nor any instruction to use thermal interface
material as you'd find required on any proper heatsink.



The drive has is fairly flat but the side of the drive isn't
even close, typically only the screw hole area is machined
flat which is a very tiny % of the drive sides.

In practice the cage is pulled up against the body
of the drive when the screws are tightened.

In practice it is not making but a tiny % contact. We could
claim any heatsink is pulled up against a part with a clip
or screw/etc, but still they are machined flat first and
thermal interface material used.



A normal / average computer case does not act
as a significant heatsink for any part in the system.

Wrong, as always.

Basic science Rod, something sorely misunderstood. A
thermal path is only as effective as it is designed to be.
There was no effort put towards it being designed for
heatsinking and it is terrible at it.

Even with your denial of the obvious you can easily see for
yourself. Turn off system fans and touch the side of a
then-hot running drive. Next touch the side of the frame
next to the drive. With anything effectively heatsinking an
adjacent part you would not, with touch alone, even be able
to discriminate a difference in temperature between the two,
but obviously we can all easily tell the drive frame is
barely, if at all, feeling warmer than ambient temp unless
it was only because you'd left the sides of the case on so
the entire interior of the case was signficantly elevated
above room temp already.

If you were/are really anal about it you could design a case
to be a better heatsink. First you'd need to lap the sides
of the hard drive which will remove all the (typically
black) paint applied. Next you'd choose a good drive cage
material, sheet aluminum thicker than 0.06". Next you'd
measure that sheet to be sure it was flat.

Next you'd apply heatsink grease to the side of the drive
and flex the drive cage outward so it wasn't all scraped off
when you put the drive in.

Next you'd screw it on, and be discontent at what minor
difference even such an optimal cage to drive-frame
interface made because it is a small % of total drive frame
area, and a distance away from the heat producing portion of
the drive (chips on PCB, motor, platter air resistance w/air
as an insulator).

The rest of us, including the hard drive industry, know it's
not worth the bother so they don't. Take a look at a hard
drive someday, then look at the insides of a drive rack and
tell us if you see big creases in the metal where it bent to
conform to the irregular sides of the drive. Most of us
already know you won't see that happening. Some of us even
know that if you did see it, it still won't help a lot
because the coefficient of expansion of the drive frame is
different from the typical steel drive cage.

Funny thing is, I hadn't even mentioned yet that a lot of
cases these days don't even have the drive cage designed as
plain flat sheets of metal anymore, most better cases now
have rails or slide-lock tabs for quick mounting which keeps
the drive from having ANY direct contact with the drive
cage, 0%... ok maybe 0.03% if the screw threads happen to
brush up against it instead of having a plastic sliding rail
of some sort.

I sometimes ask myself why I bother replying to a troll.
The answer is that some who are new to computer cooling are
better off learning it right than repeating urban myths.

On Fri, 24 Jul 2009 07:49:05 +1000, "Rod Speed"
wrote:

Timothy Daniels wrote
Rod Speed wrote

In practice the cage is pulled up against the body
of the drive when the screws are tightened.

I've only inspected Maxtor hard drives closely, but all
the Maxtor drives that I've seen have raised bosses
around the screw holes. They are so low that one has
feel them with a finger nail to be sure they're there, but
they *are* there, and they make a small air gap between
the body of the drive and the HD cage. That means
that by metal-to-metal conduction, the bosses and the
mounting screws transfer more heat to the cage than the
sides of the HD.

They still use the cage as a heatsink, otherwise
the cage wouldnt feel warm to the touch.


It doesn't feel warm to the touch unless there's another
problem like already grossly (still doing so) overheating
drive or whole system at high interior temp due to
inadequate airflow (and case side panels left on until right
before comparing cage temp to hold in heated air rather than
the heat being 'sunk from the drive).

Funny thing is, since I had a running system open in front
of me I reached over and touched the rack. Guess how it
felt...?

It felt cold, touching the metal 'sunk a little heat away
from my finger because unlike a hard drive my finger
conforms to the metal and has a slight (fingerprint level)
of moisture to aid as a thermal interface grease). So not
only did the rack not feel hot, it felt colder than ambient
room air.

@

"kony" wrote:
[....] some who are new to computer cooling are
better off learning it right than repeating urban myths.

OK, I guess your bottom line is that most HD cooling
is done by air flowing over the HD's surfaces, and that
adequate air flow is forced by a fan, and not by convection,
right?

*TimDaniels*

kony wrote
Rod Speed wrote
Timothy Daniels wrote
Rod Speed wrote

In practice the cage is pulled up against the body
of the drive when the screws are tightened.

I've only inspected Maxtor hard drives closely, but all
the Maxtor drives that I've seen have raised bosses
around the screw holes. They are so low that one has
feel them with a finger nail to be sure they're there, but
they *are* there, and they make a small air gap between
the body of the drive and the HD cage. That means
that by metal-to-metal conduction, the bosses and the
mounting screws transfer more heat to the cage than the
sides of the HD.

They still use the cage as a heatsink, otherwise
the cage wouldnt feel warm to the touch.

It doesn't feel warm to the touch unless there's another
problem like already grossly (still doing so) overheating
drive or whole system at high interior temp due to
inadequate airflow (and case side panels left on until
right before comparing cage temp to hold in heated
air rather than the heat being 'sunk from the drive).

Wrong, as always. It feel warm to the touch in any system
that doesnt have a fan that specifically moves air over the
drives and feels warm to the touch even if there is a fan
that specifically moves air over the drives too.

Funny thing is, since I had a running system open in front of
me I reached over and touched the rack. Guess how it felt...?

It felt cold,

How odd that mine dont.

touching the metal 'sunk a little heat away from my finger
because unlike a hard drive my finger conforms to the
metal and has a slight (fingerprint level) of moisture to
aid as a thermal interface grease). So not only did the
rack not feel hot, it felt colder than ambient room air.

You're lying now. It cant possibly be at a lower temperature
than the ambient room air. VERY basic physics.

kony wrote
Rod Speed wrote

The case does not act like a large heatsink.

Wrong.

You can see this easily enough by measuring the large
difference between the hard drive frame and the metal
wall of the drive cage to which it is attached.

What matters is that the drive cage metal is warmer
than the rest of the case. It always is, and that means
that its acting as a heatsink for the drive.

For it to effectively heatsink it has to have a large area of conductance,

Wrong, as always. For some heat to be removed from the drive that way,
ALL it needs is a path for the heat to flow from the drive to the cage.

And it doesnt even need that when the drive can radiate to the cage as well.

reams of your mindless pig ignorant raving flushed where it belongs

John Doe wrote:
"Rod Speed" wrote:

Timothy Daniels wrote
Rod Speed wrote

In practice the cage is pulled up against the body
of the drive when the screws are tightened.

I've only inspected Maxtor hard drives closely, but all
the Maxtor drives that I've seen have raised bosses
around the screw holes. They are so low that one has
feel them with a finger nail to be sure they're there, but
they *are* there, and they make a small air gap between
the body of the drive and the HD cage. That means
that by metal-to-metal conduction, the bosses and the
mounting screws transfer more heat to the cage than the
sides of the HD.

They still use the cage as a heatsink, otherwise
the cage wouldnt feel warm to the touch.

My hard drive rack screws have rubber grommets to help reduce
vibration. The only heat conduction from the hard drive to the hard
drive rack is through the four small screws. The intake fan blowing
over the hard drives cools them.

Thats wrong too. The drive also radiates to the cage.