Pleonasm wrote on Apr 30^th, 2007 at 12:20pm:

Pleonasm,

That would be an interesting study. Whether cooling fans actually increase the life of a HD or paradoxically decrease the life of a HD. The Google study indicates low temperatures may be a problem.

Quote:

Brian, you are right:  based on the Google research, there is an increase in the AFR (annualized failure rate) of hard disk drives from about 2% at 30C to 9% at 17C (see Figure 4 in the Google research paper).  Note, however, that the “low temperature effect” is generally present only during the first three months of the hard disk drive lifecycle, after which it is ameliorated quite a bit (see Figure 5).  The temperature range of 30C to 40C appears to have the lowest associated failure rate across the entire lifecycle (although that AFR is not significantly different from the 40C+ range, as judged by the standard error bars in Figure 5, in general).

Rama, “actual temperatures” are in fact reported in the Google research paper (see especially Figures 4 and 5).  Your idea about correlating the AFR with geography is interesting, but it wouldn’t account for whether or not the unit is in an air conditioned environment, for example, where the ambient temperature is controlled.

As a practical matter, I do not see any reason for a person to use a hard disk drive fan in the absence of experiencing temperatures which are outside of the operating limits established by the manufacturer (e.g., 0C to 60C by Seagate, in general).  In the spirit of “don’t fix it if it is not broken,” the first step – in my opinion – would be for a PC user to monitor the hard disk drive temperature, and only then make a decision to install a cooling fan if circumstances justify it.

Baseem, if Western Digital says that the acceptable operating temperature for the hard disk drive is between 0C and 60C, then anything in that range is – by definition – acceptable.  I doubt that anyone on this forum has engineering credentials that exceed those of the professionals at the manufacturer, so why not simply accept the assertion as it is stated?

Baseem, Seagate references the hard disk case temperature (provided by an internal disk sensor) when discussing “operating temperature” (usually 0C to 60C).  Seagate also discusses ambient temperature (usually 5C to 55C), which refers to the temperature of the room in which the PC is located.  I suspect that the same terminology applies to Western Digital, too.

The downside risk of using a hard disk cooling fan would appear to be minimal.  However, based on your prior post, it doesn’t seem that you need to take any action.

Rama, according to the Seagate website, the operating temperature of the ST328040A hard disk drive is 0C to 55C.  If, for example, your drive was at 50C (122F) it would be both “hot to touch” and operating within its design specifications.  In other words, your observation that the drive was “hot” doesn’t necessarily imply that it was running outside of its design tolerances - and, your success in cooling the drive doesn’t imply that that it will last longer or run more reliably.

Now, if the temperature of the drive was in fact above 55C, then clearly corrective action would be required.  Did you actually measure the temperature of the drive (e.g., using SpeedFan)?

When you touch your running HD and have to immediately pull your blistered finger away with a resounding "Ouch!" (or much worse verbiage) it's obviously too hot.
Nothing in a computer should ever run that hot.

Any electronics designer who has to use "Hobby Grade" components, like our home computers use, will tell you quite frankly that "".

Why else would we have CPU coolers, Video chip coolers, PSU coolers, case fans, on and on.
HD coolers are just one more step in the effort to keep our PC's alive just a little bit longer.

Electronics are damaged when they heat up from room temps to time after time.
The expanding and contracting literally them apart.

Google? They are not electro/mechanical engineers and I sure wouldn't put much stock in anything they say.

Hard drive coolers are not Refrigeration devices.  They don't CHILL a drive, but just blow away excess heat. (it can't go below room temperature)
Keeping a drive at room temps has NO downside.  They will last longer and your valuable data will be better protected.

If room temperature was damaging to a drive, they would come with built in HEATERS. 

The last Deathstar drive that squalled, balled and stopped on me, (sans cooler) was so hot when I reached in to remove it, that it literally my fingers.
I had to wrap a rag around it to take it out. ( I couldn't find my Oven Mits)

Since that day, I've had two-fan coolers mounted under every drive I've owned.
That same technique is now being used (with the 1/4" standoffs) by hundreds of PC users world wide. 
One guy wrote me to tell me that his 10,000 rpm Raptor drive that used to run scalding hot, now runs cool with the installation of the little cooler. 

Here's one guy that took my cooling tips to the extreme!



The inlet and exhaust fans are obvious. What's not so obvious is that the RED fan is actually mounted in a clear plastic Left-Side case panel.  And the two blue fans in the base are sitting on top of the HD, taped to the base plate with duct tape. 

When he sent me this picture, my only comment was, "COOL!"

Y'all have a great day now, Y'hear?
Shadow 

Rama, it is true that the Google research is based upon consumer-grade hard disk drives used in a server environment.  But, doesn't that make their findings even more solid?

If the temperature of a consumer-grade hard disk drive (while operating within the manufacturer's design tolerances) isn't highly correlated with failure when used in a server environment, then one would expect the finding to be all the more descriptive of a less demanding home environment, no?

* * * * * * * * * *
TheShadow, if the temperature of your own hard disk drive exceeded the manufacturer's design tolerances, then certainly it was proper for you to take corrective action to cool the hardware.  It doesn't logically follow, however, that installing a cooling fan when a hard disk drive is operating within its tolerances has any benefit.

By the way, the primary author of the Google research is Eduardo Pinheiro, who holds both a PhD in Computer Science from Rutgers University as well as a BS in Computer Engineering.  The second author of the paper, Wolf-Dietrich Weber, holds a PhD in Electrical Engineering from Stanford University.  The third author, Luiz André Barroso, has earned a PhD degree in Computer Engineering from the University of Southern California and holds both B.S. and M.S. degrees in Electrical Engineering.  Therefore, you may wish to reconsider your assertion that Google personnel "are not electro/mechanical engineers and I sure wouldn't put much stock in anything they say."