Freezing HDs ?

Sorry if this is is wrong place (or not worth posting at all, even)

I just wanted to get some opinions on this - freezing (literally) a dying a hard drive to get it working for enough time to perform some data retrieval...
see http://lifehacker.com/5515337/save-a-failed-hard-drive-in-your-freezer-redux.

(I saw some things about oven-baking dead boards, eg video cards, motherboards which have defective solder joints to get a chance of resurrecting them)

I am not recommending anyone try this !!

Dumb_Question
30.August.2013

 

I can say I have successfully put a drive "on ice", but you need to take precautions and keep it cold too.

Also, a low humidity environment is important. You do not want to do this on a rainy day...

 

Interesting to hear a first hand report that this can be successful.

You wouldn't want condensation to appear, which might be a risk in even moderate humidity (40-50%)

Dumb_Question
30.August.2013

 

The % of times when 'freezing' is the correct method for data recovery is very small; there are both mechanical and electronic faults that it works for; prior diagnosis is key; it's looked on as the last resort.

My success rate is 100%, only had to use it 2x in 10+ years, probably 10% of 'difficult' recoveries.

 

It is extremely rare for this to work and I am surprise satrow saw it work successfully twice in his lifetime - you should apply to Guinness World Records!

I have seen it tried dozens of times over the years, but successful only once, and that was over 20 years ago - when a 120Mb drive cost nearly $350.

Understand this ONLY works when the problem is due to worn bearings seizing. The idea behind it is extreme cold causes "matter" to contract. By freezing the drive, the hope is the seized motor bearings will contract enough to allow the motor to spin (or step) again (without hardening the lubricant and thus bearings in place).

The problem is (besides the aforementioned moisture problems) as soon as the drive motor starts spinning, the lubricant and worn bearings begin to warm up again. And when matter heats up, it expands. And when bearings or the tracks they ride in expand, the bearings tend to seize in their tracks again.

So my point is this, IF you are successful at getting the drive to spin again, you literally have seconds (a minute or two if lucky) to locate the file (as in one, maybe a few small files) you need before the drive seizes again (and condensation develops and creates a shorting hazard).

Note with the one successful attempt I was a part of, we first put the drive in a refrigerator for two reasons. First was to see if it would work without freezing (it didn't) but second (and most importantly) is less damaging condensation and fewer, smaller ice crystals form if the "matter" is chilled to near freezing first (same is true for meat and other foods, BTW, so you should always chill your food before freezing if you want to preserve the textures and flavors).

Before putting the drive in the fridge, we connected long flat ribbon (it was an old EIDE drive) data and power cables to the drive. Then when we moved the drive into the freezer, we routed the cables to a computer on top of the freezer, waited several hours for the freeze to set in, then powered up with the drive still in the freezer. We got about 3 minutes before the drive froze (seized again) but we were able to retrieve the spreadsheet the boss so desperately wanted.

Common mistakes sure to increase chances of failure:

• Sealing the drive in plastic. It seals in moisture, and takes time to unwrap allowing more time for condensation to form.
• Putting warm (or even room temperature) drive directly into freezer without chilling to near freezing first.
• Attempting to clone entire drive rather than prioritizing and going after the top priority files first - before the drive heats again.

And of all the claims where it did work, I have never heard of any case where re-freezing allowed for successful recovery of more files. In other words, it is a one-shot deal. All the more reason to know exactly what file you want to recover and where it is located. You need a plan of attack.

Of course, it is the same old story here - having a current backup of your important files is crucial, and negates the need for such desperate and likely unfruitful attempts as freezing a drive.

Oh, one more thing - Mythbusters, the TV show did this once and it worked - but they emphasize they only got a couple minutes of run time - enough to recover a couple files only.

 

NO no no, Bill - note that I wrote "and electronic faults" (but I reckon those that this works for are very rare, as well!).

Picture this: a PC is brought in that crashes to a BSOD with ~2-3 minutes of boot, client wants access to his valuable data ASAP.

On connecting it up and booting it, you get a similar crash/BSOD - you also smell something acrid; on investigating the source of the smell, you find a tiny blackish crater on one of the ICs on the HDD controller board. A quick online search suggests the make/model drive is pretty rare and not in stock at any of your usual trusted suppliers, so swapping boards is out, for several days, at least.

Connecting the drive to the laptop via USB whilst on the bench showed a tiny wisp of smoke rising from the crater ~1 minute after it started up and a fast shutdown was made. My partner in crime was sent to collect plastic baggies, rubber bands and to check that there was a supply of beer cooler freezer packs ready frozen. Wrapped up tight, the drive was banished to the freezer for ~6 hours.

Apart from Explorer being it's usual dumb self, crashing out when it hit a bad block/unreadable error, we got all the data off in the first recovery session, probably 12-15 minutes uptime before I was satisfied we had all the useful/valuable/personal data off it. Client had his data back the same evening.

 

I think for sure, your case is an exception to the rule. Nevertheless, a happy customer is a returning customers so kudos to you for making it happen.

 

I have recommended it when someone is desperate but I can't recall ever trying it let alone having it work.

One trick I have tried on components is a solder reflow, the whole component is heated in an oven which melts the solder fixing any dry solder connections, I've fixed multiple PS3's this way so I can confirm it works. It's just an upscaled version a of standard solder reflow where each dry joint is manually soldered by hand.

If someone is desperate enough I will try putting their drive in the freezer simply because there's nothing to lose.

 

As an added note I have come across few hard drives that totally don't work, most of the drives I fix have started throwing up errors and causing crashes long before they are no longer readable at all.

I often wonder how many hard drive problems are misdiagnosed because the hard drive seems to work.

 

Yes, it's all down to the diagnosis, Rikky, getting the full gen from the user/client and digesting that fully before making any decision on the next step. When it comes to saving data, you might only have one shot at it.

 

Right - but sadly, too often those errors are ignored instead of taken seriously as warnings to make a backup NOW!