Is suspension a good idea for extended use?

OmegaZero · Post by **OmegaZero** » Fri Oct 07, 2005 11:50 am

Here is the situation:

I have a server with three drives in it (see the "SPACE" system in my signature). It is much louder than I would like, mostly due to features of the case; hard mounted drives, 80mm fans, poor airflow/turbulence, etc. I am intrigued by the new P150 case Antec has announced as a possible replacement. However I am a bit apprehensive about the HDD suspension; particularly its effect on the life of the drives, since they are on and busy 24x7x365.

My fear comes from a recent conversation where I was told that suspending a HDD in elastic or any other method that decouples the drive completely (not including foam pads and/or soft grommets) will significantly reduce it's the lifespan. I specifically asked if this was thought to be true due to thermal issues and he answered:

WD Tech. Support Rep. wrote:"No. The extra strain on the spindle and motor assembly caused by the lack of physical mounting will inevitably shorten the unit's operational life. These types of configurations are not supported by WD and will void your drive's warranty."

What I need is a solid technical explanation of this issue. Please don't post opinion - I am searching for an authoritative answer here (MikeC, Ralf, Rusty, etc. help me please!). Is HDD suspension detrimental to the lifespan of HDDs?

Thank you for you help in advance.

NOTE: I know I am going to get razzed for this post, but before you start in on me let me say: Yes, I did search before I posted this. Using several different combinations of keywords I got over 1000 results each time, none of which seemed to answer my question.

shades_of_blue · Post by **shades_of_blue** » Fri Oct 07, 2005 12:17 pm

The WD rep was probably referring to those who use rubber bands wrapped physically around their hdd. The extra pressure it causes to the top plate could technically put additional stress on the drive motor assembly.

OmegaZero · Post by **OmegaZero** » Fri Oct 07, 2005 12:22 pm

I got the feeling he meant the extra work the motor has to do to get the platters moving without a firm surface to push off of. I know this causes higher temps, but is this really bad for the drive (assuming it is adequately cooled)?

Post by **Rusty075** » Fri Oct 07, 2005 12:37 pm

Sounds like an overly cautious tech support email-reader blowing smoke. (oops, that was opinion.

)

Torque loading, the effect of the motor having to work harder because of the reaction from the suspension, only applies when first spinning up. Once at speed it's in equilibrium. (the higher temps are from the lack of heatsinking to the case, not the motor working harder. Suspended drives don't pull any more watts). I could see where it might effect the arm movements, but considering that the actuator is engineered to work just fine when mounted vertically where it has to fight gravity, the torque loading shouldn't have an effect.

I bet that the engineering issue is that being softmounted allows the movements of the heads to setup reactionary movements of the entire HDD. From a performance standpoint this effect is pretty well documented; the read/write tolerances are so tight that the micro-movements of the drive causes the head to "miss" its track occasionally and have to be readjusted. The aggregate effect of these misses is tiny, but real.

But from a lifespan issue, I have trouble seeing the impact. HDD's are rated for a specific in-operation load and vibration G-force maximums. As long as the acceleration on the drive stays below that threshold, there shouldn't be a reliability or warranty issue. Considering the tiny weight of the heads compared to the big mass of the rest of the drive, plus the counteracting effect of the platter's gyroscopic stability, I can't imagine that the G-loads are much of anything. (we could actually calculate it if we had accurate numbers for the parts involved) The G-loading from a truck driving by your building is probably 100 times bigger.

Anecdotal evidence doesn't point to a correlation either. There haven't been really any reports of suspended drives suffering bearing failures.

frankgehry · Post by **frankgehry** » Fri Oct 07, 2005 2:06 pm

this link has some information about suspending drives - http://www.arcade-tech.com/arcade/shock_mounting.html

another -
http://forums.silentpcreview.com/viewto ... ard+drives

more -
http://www.seagate.com/support/kb/disc/ ... ators.html

OmegaZero · Post by **OmegaZero** » Fri Oct 07, 2005 2:13 pm

Thanks for the info - you guys are awesome.

I figured it was a bunch of crap. I doubt there would be so many suspension advocates if it was actually bad for the drive...

Devonavar · Post by **Devonavar** » Fri Oct 07, 2005 2:23 pm

Read through this thread for a heated debate about the effects of vertical mounting on hard drives. You should be able to get a good idea of the issues involved, even though the thread doesn't relate directly to soft-mounting. I won't try to resolve the debate, but my opinion is similar to Rusty's: An overly cautious tech support rep blowing smoke.

vertigo · Post by **vertigo** » Fri Oct 07, 2005 5:20 pm

Remember, tech support people are usually rather cautious to advise anything risky. Suspending drives is risky if you allow the temps to get too high (because less heat conducts away), or if it is loose and the drive moves and shorts something, etc.

Also, tech support reps usually must conform to company policy, so expect to get mainstream advice from them. They can't control how securely you suspend the drive, so they won't condone it.

Ultimately it's your computer, so the decision is yours as to what you do with it. You have the responsibility to educate yourself to the risks involved. They will give you advice that works, but beyond that it's up to you.

OmegaZero · Post by **OmegaZero** » Fri Oct 07, 2005 5:23 pm

Well if I were to suspend the drives, it would be in the hardware provided with the P150. Still waiting on a review of course, but I'm sure it'll be secure enough and such.

Tough topic, this is. There seems to be just as much feeling either way...

ozdoc · Post by **ozdoc** » Fri Oct 07, 2005 7:41 pm

"No. The extra strain on the spindle and motor assembly caused by the lack of physical mounting will inevitably shorten the unit's operational life. These types of configurations are not supported by WD and will void your drive's warranty."

Hmm. I'd be suprised if they would know if the drive had been suspended, when returning an individual drive to be RMA'd. Sounds like hot air.

However, if you need be be cautious, how about suspending the entire drive cage? That way the drives are physically supported by mounting, but vibrations are decoupled from the case. My hard drive temps in such a suspension have never exceeded the low 30's, usually in upper20's. There are examples in the gallery forum, particularly for a 3000B case. (Kudos to Antec for the P150 design also.)

chirs · Post by **chirs** » Fri Oct 07, 2005 8:06 pm

ozdoc wrote:However, if you need be be cautious, how about suspending the entire drive cage?

Exactly what I was going to suggest. If you have a large mass being suspended, then the additional weight of the other drives will tend to dampen the vibration of the drive(s) doing the seeking. The probability that you will get a worst-case timing event is minimal, so on average you should see a benefit.

If you wanted to be really paranoid, you could add inert mass to the suspended drive(s) to further dampen the vibrations.