Slashdot Mirror
Reviving A Dead Hard Drive The Hard Way
An anonymous reader writes "This guy went to the trouble of swapping logic boards on a dead hard drive to get his NeverWinter Nights save games back and took photos." I would have just used a character editor to get my stuff back, but clearly, I lack the dedication this gentleman has. Regardless of reason, nice work!
It should be done early and often. Hard drives do fail and can do so without warning. Therefore it is very important to back up that valuable data.
It's interesting how he found that the same brand and model of hard drive can have a vast array of different firmware configurations. This seems like it is a bit dishonest to the consumer who assumes he/she is purchasing the same thing that was recommended to them.
Visualize the world of wine
RPGs: They kill. They ruin lives. Just say no.
To make laws that man cannot, and will not obey, serves to bring all law into contempt.
--E.C. Stanton
The Plist and Glist are stored on some hidden track on the HDD platter. As long as the firmware is the same the drive should work. Although I believe drive companies change firmware without changing the "Official" firmware number. This is done because the changes are only "manufacturing" related. (-;
I was doing this stuff in the early 80's.
I even replaced platters on 10 gig drives..
Ummm... CN: the drive was -dead-. Ain't nothin' short of a new board that would've fixed it. (Okay -- sending the platters out for oodles of money would have, too.) Also, I don't know why this is labeled "the hard way." I've done it three times, en-toto, and it takes about ten minutes so long as you've got the correct Torx/Phillips/whatever. [Note: DON'T try doing it with the wrong tools; you'll probably just strip the head, and then it gets more fun.]
$.02...
Now, I wonder if I can make use of the warranty on the original drive.........
In other news: how long before he's swapping logic boards on the webserver?
My mom, a teacher, made a banner with this quote and posted it in a faculty lounge:
Blessed are the pessemists, for they have made backups.
To make laws that man cannot, and will not obey, serves to bring all law into contempt.
--E.C. Stanton
I had to do something similar with some wet floppy disks back in the day. (backups, I hear you say? Those *were* my backups!)
pb Reply or e-mail; don't vaguely moderate.
What's the deal with this? More people I know have lost new IDE drives than I ever recall in the past. Are my friends just unlucky, or do drive just not have the quality anymore? I know this assumes that drives used to be better, and that may well not be true, just this is the trend I've noticed. Is it worth buying a new drive (I do need one...), or is it just going to die on me in a few months?
As far as the article goes: What a waste! It must be damn nice to be able to buy TWO new drives to replace the logic board on one! Sure, one of the new drives is usable, but the other is shot.
"These people look deep within my soul and assign me a number based on the order in which I joined" --Homer re:
hmm... so he switched the whole logic board?
I did the same thing with a bunch of 1,6GB western digital hard-drives a few years back, I got a pile of broken ones for free and was able to salvage 4 into working condition by changing the logic boards from those that made funny noises to those that sounded fine but the BIOS did not detect.
But it totally kills the warantee..;)
But my 60 gig recently bit the dust, and the first thing people told me to do was stick it in the freezer... (just like he did in the article) Of course I naturall say "But that'll kill it."
theirs? "It's dead already, idiot"
Nine times out of ten, a hard drive dies because of media defects -- then you're (pretty) screwed. Sometimes, the stepper motor dies. Then, you're screwed. But, if you give it juice, and either -nothing- happens (no LEDs, etc.), or the BIOS doesn't see it, it's likeley the board. As always, troubleshoot starting with the obvious, and work toward the unlikely.
He seems somewhat surprised that the price of repairing a hard drive is more than buying a couple of new ones. You are paying to get the data salvaged, not the physical disk back.
Having worked in technical support with a database company, I can tell you how upset people can get when you tell them it's going to cost almost $400/hr to salvage their database. Sometimes it could take upwards of 16 hrs to do it depending on the size and extent of the damage.
How far a little proactiveness and an occasional backup of important data will go.
In C++, friends can touch each others private parts.
Rule 1. Always have a backup.
Rule 2. If you changed data, see rule 1.
But, what people forget is to test their backup to see if it can be restored from.
Fight Spammers!
I've also had good luck pulling data off 2.5" drives by pulling the covers and simply running them through a hardware cloning box (about $120 now). The fact that you're reducing their MTBF to something like 10 hours is irrelevant if you get the job done in 20 minutes.
Oh, act lawyerish: only charge for successful recoveries. That way, the clients even sympathise with you if you don't succeed.
Suppose your drive dies and it has personal information on it, and you can't recover the drive. What's the simplest and most effective way to wipe the data on the drive so you can throw it out?
Programs like character editors allow you to make a new saved game (on a new hard drive) and then do all the hex editing required to change the character's name, level, experience, skills, equipment, etc. No need to get at the old save game.
Kind of an afterthought to an earlier comment of mine paraphrased as "Doesn't MTBF mean anything anymore?"
Hard drives have warranties. Sure, these warrenty periods are shortening, but that's neither here nor there. Given that a drive is going to fail eventually, would it be beneficial for drive makers to offer 'data insurance'? Data recovery is expensive because it's not a common practice. If you paid some reasonable, optional $x when you buy a drive, and the drive goes down, and you could send it back to the maker for recovery (having paid 'insurance' on it), the practice would be more common and the price would decrease. The idea being, like most forms of insurance, you are paying less than what the recovery would cost because the rest is subsidized by the other people who pay but never need it. A third party recovery service could offer this as well.
There are a number of issues I can see with this arrangement (privacy, confidentiality of data, what happens when the drive can't be recovered, what if they just SAY it can't be done, etc), but it's something to think about.
"These people look deep within my soul and assign me a number based on the order in which I joined" --Homer re:
That's not even close to "the hard way". Every bench tech worth their minimum wage has done this same thing more times than they can count. Execpt they usually know that you need the same firmware before they start.
I'll be impressed when someone gets fed up enough to build a clean room in their guest bathroom and recovers a drive with crashed heads.
Do not fold, spindle or mutilate.
I think that this kind of hardware swashbuckling is pretty neat. I think I would probably just have accepted defeat and called it a day.
But what's even cooler is that the guy went and got his own domain for his dead hard drive. Nice.
This approach seemed expensive, but as far as bringing a dead drive back to life through surgery, this seemed pretty easy.
"The hard way" would have been buying a new drive, taking it to a cleanroom and transplanting the platters! You'd more than likely lose the use of the 'donor' drive, and there's a higher chance of failure in this much more invasive procedure, but that would be much more article-worthy.
"These people look deep within my soul and assign me a number based on the order in which I joined" --Homer re:
I had a physically dead drive...you know, the dreaded click of death.
Being pissed as I was, I opened up the damn thing and got ready to wreak havoc on the platters.
But I chickened out, (what kinda chemicals might that thing spew out?) and put the drive back together.
To my surprise, the drive worked again!
My room is was a nasty, dusty place too...so I bought a new drive, mirrored the old, and never used the fixed drive again.
I still have it in my house...an old Quantum 6 Gig drive.
Any ideas what was wrong, and how opening the sealed platter compartment might fix anything?
Agreed, this is a bit silly to post as a "wow, this is just sooooo amazing!" idea. We got in a batch of those crappy little micro dells, the ones that don't even have a CD-ROM drive, and they all came with the same model of Western Digital Caviar (YAAACK!) drives. One by one almost 50% of them failed, onboard controller card just stopped working. Everytime I swapped a card out to salvage the data, I had people ooohing and ahhhing my efforts like it was magic or something. This is not rocket science, anyone reading this article should be capable of doing it themselves.
Tell me he replaced the platter head amp board inside the drive, ok, then I'll be a little impressed. Actually I'm still a bit surprised people can open up the drives and get away with it... more than once I've given people the advice to open the drive and gently spin the platters (by the edge please!) in cases where the motor was going out and wouldn't spin it up and they needed the data NOW. Sure it voids the warranty and probably will tear up the drive, but when the data is more important than the drive, it's a worthy one-shot. One fellow I told that to got his data off, and used a can of compressed air to blow out the drive thoroughly while replacing the lid, and to my knowledge, the drive is still working. (tho I sure wouldn't trust it)
I work for the Department of Redundancy Department.
Who on earth would spend that type of effort just to recover a drive with XP on it...
Got Code?
An improvement can be designed to make the product better.
It can also be designed to make the product cheaper to produce, even if there is some kind of trade off.
At the end of the day, some executive is going to look at a suggested change and think: "will this help us make more money?"
So the latest version is always the best for the company, but is it the best for you? You can't be sure of that.
The Internet is full. Go Away!!!
I had a guy at work who had been saving his work to the HD instead of to the servers where it would have been backed up. The day before he was to present the results of his project to the president of the company, the drive failed.
It would spin up, and apparantly work for a few minutes, then spin down.
Suspecting heat-related problem, I stuck it in the freezer for a few hours, tried it again, got it to run long enough for the PC to finish booting & to copy the data, then it failed again.
Like your people said though, I wouldn't necessarily try it on a drive where less dramatic measure might work.
Many years ago I used to buy large quantities of dead harddrives from Gateway Computers. I took the logic boards off every one of them and using a known good logic board and a known good drive I'd quickly figure out which logic boards were good and which drives were good. Combine good with good and I'd usually end up with a nice pile of working drives which I resold on Usenet for a nice profit. The dead drives I would either RMA back to the manufacturer or sell as dead drives. That was back when a good drive was worth $1/MByte and I was buying dead ones for 10 cents/MByte. As a side note, all those dead drives used to be someone's good drive and naturally all their files and data were still on the drives.
I can't belive this guy spent money registering "deadharddrive.com" for one page on how he got his saved games back.
I would have thought that name would have been snapped up by a data recovery service years ago!
What post? The one you're carrying inside your rusty innards!
I beg to differ. I have two of the same Seagate drives mentioned in the article. One of them has been running over 2 years. The Barracuda IV series are some of the most reliable I've seen.
This guy's logic board was fried. It was not a mechanical failure like most HD failures are. He could have gotten a power surge and fried the electronics.
BTW, this is the easy way of reviving a HD, not "The Hard Way". Boards are designed to be easily replaced in most HD's. Now, if he had opened the other side. It would have been a total loss, unless he was in a clean room free of any dust or debris in the air.
It's lazy kids like you that have ruined this industry. Back in my day, we didn't have operating systems.
Hell, I had to write a WYSIWYG word processor on an abacus.
And it made me a better person, I can tell you!
Thanks for the new definition of "idiot." Never knew its that simple.
The first drive he bought had a different part number, as you can see by looking at the close-up pics he took of the labels.
.bin file, just like flashing the BIOS on your motherboard (except you can usually do it in Windows). I would contact tech support first, though, and make sure the firmware you need is compatable with the different hardware. You'll probably have to contact them anyway to get the .bin.
Also, firmware can be changed. All it takes is a utility and a
Lastly, if you plan on trying this at home you need to know that Seagate and IBM/Hitachi (and, I presume, other vendors, but those are the only ones I deal with in a professional capacity) classify drives in catagories, like "generic 80GB 7200RPM ATA100" and there are often several part numbers that satisfy that description which the vendor considers interchangable for the purposes of warranty replacement (sometimes that isn't really true, which is the only reason I know this, but usually it is). Basically that just means that there's no real guarantee that you'll get the exact same drive. I've been able to get around that by making enough noise, but I also was acting as an agent of a $25billion company, so YMMV.
Under capitalism man exploits man. Under communism it's the other way around.
Backing up is like voting--most people don't do it but they still think they have the right to complain about the results of their laziness.
I bought 4 Maxtor 80 GB drives and had one seize up on me. I was fairly certain that the logic board had fried itself (the screws anchoring the drive came out and the drive started floating free in the metal chassis).
Since I had 4 identical Maxtor 80 GB, I waited until Maxtor sent me a replacement, swapped the logic boards, brought the drive up immediately, and dumped everything over. I sent the drive with the bad logic board back and resumed work.
I doubt I would have gone to the trouble of asking vendors to look up their firmware versions had I not bought several identical drives!
Jory
Try doing it at 3:00am on a detacenter floor with a leatherman on a pair of quantum atlas 10k u3 160 drives. Somehow 2 drives in a raid5 array failed within 6 hours of each other and the customer needed to get back data from changes they made that day.
I did it, it worked, but I never expected to see a headline about someone doing it.
01:36AM up 426 days, 2:46, 1 user, load average: 0.14, 0.11, 0.05
First:
....(this drive doesnt fix it)....
I look at some businesses that do hard drive recovery - the prices are exhorbitant! I could buy 2 replacement drives for those prices.
then...
So I go get a replacement hard drive
So I ring around some places and besides having to deal with some hopelessly non-tech sales people I actually find a shop that goes to the effort of looking on the drive for me and it's the right firmware! Cool! I go and buy this one.
So he doesn't want to have the data recovered cuz it costs the same as 2 new drives...
but he buys 2 new drives to recover this hard drive?
no comment
One of my favourites - put the hard drive in a plastic bag and put it in the freezer - cooling it down shrinks the parts and may enable the drive to spin up properly. I actually try this and get lots of funny looks from my wife. Still, it doesn't work.
This trick can work on some IBM hard drives. IBM had a problem where you would hear a clicking sound. The reason for the clicking was sometimes that the disk had increased in size due to the heat, and the heads were unable to compensate. Putting the drive in the freezer made the disk shrink getting the heads correctly aligned again.
Obviously, the drive did the same thing after 10 min, but atleast you got the most important data off the drive.
Um, he did buy two replacement drives in the process of fixing the dead one. (He said he was going to try to return one of them.) The DIY approach was probably a lot faster, though.
I had to revive a drive that failed after a power failure. The machine had been on for a few years straight and the old scsi drives it had used oil bearings. These bearings seize up sometimes if they are allowed to cool.
So i took the drive out of the computer and did everything you would normally do to a drive that was not spinning up, Shaking it, trying different power connectors, etc. Nothing worked. I figured there was not much damage that could be done with a little brute force, so i took a screw driver and started hammering on the side of the disk while it was plugged in. That didn't work either, so i figured it was time to use some REAL brute force. I took the drive and lifted it up about 3 feet off of the ground (still plugged in and powered up) and let it drop. That drive spun up and worked fine for another 6 months until the whole system was scrapped.
Your mileage may vary, but when it comes down to a broken drive, if it's not spinning, there's not much more damage you can do to it.
Why read the article when I can just make up a snap judgement?
I've done the "controller swap" to recover data off a dead drive at least 6 times in the last year. One of me co-workers just recently had to do the same on another drive.
It's not generally something you want to do as you could end up with two dead drives instead of one. But in certain situations it is the only way to recover a system that HAS to be up and running and contains critical data that may not have been backed up recently.
Someone has already mentioned cars, but in the context of a change that happens during different model years. In fact, cars change during a single year as well! It's not uncommon for people to consciously wait for a few months after the latest car model has arrived in dealerships before making an order. This gives the manufacturer time to "debug" the current model. Little things get fixed or changed here and there. So, on average, the later cars of the same model year are a little bit more reliable.
I knew a guy (I'll be nice enough not to name him) who discovered a dead drive and took a multimeter to it. Found the power wasn't making it past the power connector. There was a tiny surface-mount resistor that was serving as a fuse. He replaced that, and got the data back. Much cheaper to pay for a $.01 part than a replacement drive.