Thread: Full size HDD...reliability concerns?

Originally Posted by DigitalVampire

I should have been more specific.
Sorry...I was not talking about home brewed shock mounting schemes...It may work too, but as you've pointed it out, it may have negative effects since a badly designed shock mount with poor damping can amplify the shock/vibration to a catastrophic oscillation.

Over the years, we have received engineering samples of various shock mounting options from various HDD manufacturers ranging from simple rubber grommets, Teflon Posts, damping foam pads... to complex weighed wire-rope isolators. During the shock & vibration testing, it was observed that any form of professionally designed shock mounting mechanism have a positive impact on the drive life and also on the data r/w speeds..Some better than the other...

I'll see with the concerned staff if I can post the information of some products we have tested...Most of the time they are samples shared under an NDA. I hope some of them might have matured to reach public domain by now, but I'm not sure on what is available in the market.

I dont have field data for hard mounted HDD's since we dont use that scheme in any of our products. We have come to this conclusion after consultation with the Engineering Dept of atleast 2 major HDD manufacturers. We would have loved to hard mount the HDD's and pocket the 30% of storage costs we currently spend on shock and vibration damping.

The 5 years that I've mentioned was just an approximation...I have also seen hard mounted HDD's fail in less than 48 hrs under continous vibration levels within the manufacturer recommended levels...Another important parameter is data R/W frequency. A typical OS/Data drive will not see reads/writes in a contionous fashion, but our test application writes to the HDD continously to mimic the real world HD video recorder usage. An automobile may never see such vibrations nor a CarPC see sustained data rates and the drive could theoretically last longer. YMMV

Cool, thanks for the clarification.
It sure seems like an expensive solution for a $2 problem but that would be my view.

I am well aware of potential for test execution methods potentially changing the results I do software performance testing. If the test is set up wrong it is possible to show either very good or very bad results even when the software should perform perfectly.

Based on 'common' uses for a car PC (music, GPS, and video playback or capture) the read/write cycles will be fairly low and a system like you describe not likely to change the already low failure rate. In fact the elevated read/write cycles in your tests may contribute to some of your failures even without motion.

Very cool though.

Originally Posted by TruckinMP3

It sure seems like an expensive solution for a $2 problem but that would be my view.

I personally do agree it is bit of an over-design. But at the product level which are sold in tens of thousands, we cannot take such risks and it is always better to be safe than sorry...No HDD manufacturer would recommend us hard mounting the HDD's even though we wish we could inorder to increase our profit margins. Btw..It is not just us. I have studied almost every competitor model and they ALL have some sort of shock mounting.

On a CarPC hobby level, we can always take the risk of Hard Mounting the HDD...Like I said earlier, even if it lasts only 5 yrs, do we really care? The technology would have bettered so much by then warranting an upgrade anyways.

Originally Posted by DigitalVampire

SNIP

On a CarPC hobby level, we can always take the risk of Hard Mounting the HDD...Like I said earlier, even if it lasts only 5 yrs, do we really care?

SNIP

Sure, but many are lasting more than 5 years and the ones that do not.... it is hard to say what actually caused the failure.

Originally Posted by TruckinMP3

I would suggest finding a drive that does not use Fluid Dynamic Bearings (FDB)
It is a feature to reduce audible noise in HDs but has trouble in cold temps.

Yeah mine was mainly cold I think. Not totally sure though. I haven't looked into non FDB, I thought everything had FDB nowdays.

I am thinking I will get a SSD though they really have become quite reasonable and are so much faster than a 2.5" it is ridiculous. And they are faster than a 3.5" as well in many cases unless you get a velociraptor or something (which is a 2.5" now anyway )

seagate and maybe others manufacturer have 5 years warranty.
I don't care if it's dead, just send it back .
By the end of the warranty, SSD would be dirt cheap.

Originally Posted by nobb

Strange, I have also killed several regular hard drives after exposing them to the winter chill. They were free though, so oh well. But my automotive grade drive has been extremely resilient to both shock and temperature. I just bought the new Seagate ee25.2 80gb automotive grade drive, so if anyone wants my old 20gb ee25 one, send me a pm. Its still under warranty until 2011.

Guess i've been lucky then, I have just a cheapo 2.5" drive that has FDB and it's survived 3 years now thru new england winters with the only real issue being that when the temp dropped down about 7-8deg F the computer would not boot till it warmed up.

I think down to -10C, most drives should be fine. But being in Canada, my carpc is exposed to temperatures as low as -45C during the winter. Personally, I would use a normal drive if it only gets down to -10C max. But for temperatures lower than that, I would say an automotive grade drive, or ssd is a necessity.

Originally Posted by nobb

I think down to -10C, most drives should be fine. But being in Canada, my carpc is exposed to temperatures as low as -45C during the winter. Personally, I would use a normal drive if it only gets down to -10C max. But for temperatures lower than that, I would say an automotive grade drive, or ssd is a necessity.

Did the temp get that cold (Damn) or was it wind chill?

Wind chill temps affect living things is the reason for the question.

No, that is the temp. It gets worse with windchill =(. Very tough to be able to build a working carpc in that kind of environment.