• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Heatsinks on hard drives?

Joined
Jun 14, 2004
Messages
57
Hey peeps,

I've got a pretty nice storage setup at the moment with 4x147GB 10k SCSI drives. Very fast and plenty of room for pr0n. ;)

Naturally, these drives get pretty warm. Not crazy hot, but warm nonetheless. At the moment, I have a single 120mm fan (enermax adjustable, I believe) blowing over these four drives and while they're definitely not overheating, they're still warmer than I'd like. This is a typical setup where the 120mm fan is pulling cool air directly through the front of the case and blowing it over the drives.

In an effort to cool these things down a bit more (and hopefully lower my fan speeds for less noise) I'm going to remove the rubber noise-isolation nubs that prevent them from making metal-to-metal contact the the drive cage, but I'm also thinking about picking up some of those generic stick-on heatsinks (like the type you'd use on memory) to increase the surface area a bit and enhance the cooling provided by the 120mm fan.

Whaddya think? Will it make a difference? Anyone done this before? Am I wasting my time?
 
Most of the heat is dissipated out the sides of the drive. You could get some 5.25->3.5 convertors; they usually have good heat transfer properties.

 
A 120 mm fan is probably enough to cool the drives. I wouldn't bother spending more money on cooling solutions that really will have little or no effect on the drive lifespan.

These drives are built to take reasonable amounts of heat for a very long time.
 
unhappy_mage said:
Most of the heat is dissipated out the sides of the drive. You could get some 5.25->3.5 convertors; they usually have good heat transfer properties.

True. I've got a pair of 36gb 10k drives in a home-built server and I have them sitting in hefty aluminum 3.5->5.25 converters.. even with pure passive cooling, they stay remarkably cool. The problem here is that would require 4 5.25 slots, which I dont have free. ;)
 
There are convertors that take 4 drives and put them in 3 5.25 bays - is that good enough? What case do you have, and what's in it?

 
unhappy_mage said:
There are convertors that take 4 drives and put them in 3 5.25 bays - is that good enough? What case do you have, and what's in it?

Interesting, I haven't seen those before. Who carries them? Gotta link? :)

I'm running a slightly modified D8000 from www.coolcases.com. Inside the case you'll find:

- 3.2ghz P4 (stock) on a supermicro P4SCT+ motherboard
- 4x147gb 10k Seagate SCSI drives
- LSI 320-2x raid contoller
- eVGA 6800GT
- M-Audio sound card
- Seasonic S12 600w PSU
- 2x DVD drives

I'm planning to pick up a quiet cooler for the 6800GT and rip out one of the DVD drives, leaving me with 3 available 5.25" slots. I suppose my only worry at that point would be whether the 10k SCSI drives would be noisier in the 5.25" bays than they would in the standard drive cage... although that would allow me to ditch the existing 120mm front fan (currently the loudest component in the system) for a low-rpm "silent" model.

I tell you, the balance between performance and noise is a real bitch sometimes. I use this machine for audio production work and gaming, generally.
 
Try these. There are also 5-disk versions, if you might be expanding in the future. They're also available in 4-port sata/sas and 5-port sata/sas. I don't own any of these enclosures yet, but I've heard good things about them.

 
cut and paste

Ice Czar said:
there are three main variables to any heat transfer solution

the amount of medium being transfered to in a given amount of time
(in this case increase airflow)

the size of the transfer interface
(employ a finned heatsink to increase interface area)

the temperature differential
(decrease the air temperature your transfering to, twice the difference in temperature between the drive and the air twice the heat transfered)

in this case an additional factor or two could well be at play
radiation is another means of heat transfer, although a very small factor most of the time
two drives close to one another can be radiating into each other, still just a drop in the bucket comparatively, generally more to the point is the lack of airflow close proximity creates

Each 10°C (18°F) temperature rise reduces component life by 50%
Conversely, each 10°C (18°F) temperature reduction increases component life by 100%.
which applies roughly to both mechanical as well as silicon devices (there is obviously a lower limit)

excess HDD heat can be a sign of a bearing going bad, but unless you have seen a rise in temperature at the same ambient I wouldnt jump to conclusions

consider what a few simple air deflectors or ducts might do to force more air past the drives, as well as a more powerful fan, better location or accessory might offer


Id caution you about drive caddies that employ an added interface, sometimes they are fine sometimes they are not
in an external RAID device you have an EE carefully looking at impedence and noise on the dataline, in cobbled together aftermarket systems that isnt the case, I had to learn that by corrupting a RAID array

as mentioned its the drive frame that is the major thermal mass not the covers
the temperature differential plays a big role as does the airflow pattern in convection, youd be suprised how much of a difference it can make (dead spots\ better turbulance to swapout boundary layers) experiment with some deflectors and ducts but you really need to employ a stable ambient for comparision purposes
 
Back
Top