Somewhat unusual question - Hard Disk "Air Bearing" failure 2

[monkeysolder]

I hope this is in the right forum -

We are having an unusual problem with a piece of equipment that uses a 2.5" fixed hard disk drive housed in a small, water and airtight sealed enclosure. Prior to sealing, we purge the entire enclosure of air using tetrafluoroethane. (Commonly available in 'canned air' electronics dusting products). This is a common process on our other equipment, but the first time we have done this for a device housing a disk drive. A problem has just emerged where we are seeing a number of failures for the hard drive and I was wondering if this may have to do with the "air bearing" that the read/write head on the disk platter rides on during operation. Could there not be enough gas density (or perhaps too much), which is causing the air bearing to fail?

Thanks,

 

[MikeHalloran]

Air bearings work on air. Everything in that tiny hard drive is optimized for the 'flying height' of the head, which will be different for a different gas. I'm guessing the gas you use is denser than air, which would increase the flying height and greatly weaken the already miniscule magnetic signal being read. If the drives work okay when returned to a normal atmosphere, that would tend to confirm the hypothesis.

Try purging with nitrogen.

Or using flash drives.



Mike Halloran
Pembroke Pines, FL, USA

 

[monkeysolder]

Thanks for the sanity check - We will give that a shot.

 

[IRstuff]

But what are the failure modes? How can you come up with "solutions" without having detailed knowledge of the failures?

TTFN

 

[MikeHalloran]

Given the reliability of hard drives, it seems reasonable to conjecture that the cause of multiple failures clustered within a single product, but not reported as significant in the total drive population, is likely to be some factor external to the drives themselves.

We know that hard drives don't work in a vacuum, hence the filtered vent ports you find on all of them. Given that, a little information about the unusual environment, and (okay, dangerously little) knowledge about how air bearings and magnetic recording work, the posted hypothesis seemed plausible.

So I proposed a simple test that can be executed at little cost, to test the hypothesis. If recovered drives work in air, than a more air-like purge gas is indicated.


Perhaps it was arrogant to go further and propose a possible strategy for restoring production, if the test supports the hypothesis, and a strategy for making an end-run around the problem, whatever its root cause may be. For that, I apologize.


Mike Halloran
Pembroke Pines, FL, USA

 

[ivymike]

What temperatures have you measured within the enclosure? Hard drives seem a bit finnicky about excessive heat.

 

[tlee123]

This company makes sealed hard drives. They look expensive.

http://www.mt-optech.com/

Also, it's a minor point, but I think the problem with the other gas (if it is the problem) is the viscosity difference with air, not density difference.

Tom

 

[monkeysolder]

Thanks for the advice and suggestions everyone. I vacuum purged one of the returned defective units and the hard drive started working immediately, so the gas density was the issue.

Now the question is... was there any permanent damage done to the drives that ran with the denser gas? Did the read/write head fail because it was spinning above the platter or because it was making contact with the platter? On booting the defective unit, I heard the 'click of death' from the hard drive. I suppose more long term testing is in order.

 

[monkeysolder]

Oops- gas density or more likely, gas viscosity is the problem, I should say. Thanks tlee.

 

[IRstuff]

Vacuum-purged followed by dry air or nitrogen?

TTFN

 

[monkeysolder]

Nope - vaccuum pump followed by release of vaccuum in normal room air fixed the problem immediately. We plan to do a dry air fill with the units from now on and do some experimentation with N2 to see if this has any adverse effects on the drive.

 

[IRstuff]

What were the failure symptoms, aside from the click of death?

The most likely scenario is head crash, so there's probably some latent damage. You should run CHKDSK and see how many bad sectors there are now.

TTFN

 

[monkeysolder]

No other failure symptoms except a high pitched intermittent whining intermixed with the aforementioned click of death. I will run chkdsk to see what the damage is.

 

[monkeysolder]

Woohoo.. no bad sectors. They were only running a total of about 3 minutes under the adverse conditions so hopefully they are ok.

 

[monkeysolder]

Oops, nevermind. Bad sectors!

 

[unclesyd]

Do you use any type of lubricant on any of the drive components? Reaction with miniscule amounts TFE might be a contributor to some of the noise you are hearing.