Joined: 11 Jun 2007
|Posted: Mon Jun 11, 2007 11:42 pm Post subject: WD2500JB-00REA0 Recovery Success Story
I just wanted to post a little success story as a way of encouragement. This post is a near clone of one I posted on ActionFront's data recovery forum, which, like this forum, is a good resource.
Some folks (like me) may have lost data that they consider valuable -- but not valuable enough to pay the (often) high prices of data recovery for hardware failure.
Unfortunately, my 250gb WD drive, model WD2500JB-00REA0 failed not too long ago. The drive would spin up, the heads would click there times (the sound they make when they park) and then the drive would spin down again. On second spin up, the heads would click fewer times, then it would spin down again. This whole process would repeat over and over. Except for one time, the drive was completely unreadable (and the one time it only started to boot before death).
Anyway, I was unhappy. I looked online and bought a new WD2500JB-00REA0 from ebay. There were slight shipping issues, but no biggie. Then the new drive shows up, it ended up not being 00REA0, and it ended up being "Recertified." I'm already sad about my data loss, so it just added insult to injury to get a refurbished drive of a different model whose PCB was clearly different (in shape and size). I lost patience, opened the cover, looked inside, jiggled the heads by hand, everything looked kosher, so I closed the cover again. I do _NOT_ recommend that anyone does this. I kept the cover open only for a very short period of time. Anyway, all this has little to do with this story because I didn't use that drive from eBay in the end, nor did I need to open up the drive. I just wanted to build up the sympathy and suspense.
It also occurred to me that I bought another two of those drives at the same time -- one other EIDE and an SATA. Looking at the EIDE drive, sure enough, the model was identical. I swapped PCB's, plugged it in. The drive now spins up with no clicking sound. I was pleased. Plugging it into my PC, the bios detected that there was a drive, but all information (drive model, and geometry) was completely blank. I thought I was screwed.
Then after a fair amount of googling, I did find out about the U12 chip. Thanks to the folks on this forum and www.deadharddrive.com. I have no experience swapping SMT chips, but figured I'd try anyway. Rather than going all out like Amit (referring to how someone called Amit on ActionFront's forum removed and replaced his chip), I figured I'd google removing and replacing SMT components. After finding a useful site (link & summary to follow), I performed the following steps:
1. Removed U12 ("old U12") from the board of the busted HDD "bad board".
2. Removed U12 ("new U12") from the board of the working HDD "good board."
3. Placed "new U12" in "bad board." This step is optional, but I did it because I never tried soldering an SMT chip before (just caps and resistors in the past). I recommend this as a practice run. In fact, I recommend a practice chip removal as well. Unfortunately, you don't want to break "old U12" because you'll need it. Also, you don't want to break "good board," cause you'll need that. In other words, you'll have to practice your chip removal on something else.
4. Plugged "bad board" with "new U12" into the working HDD. Powered on the HDD, and heard clicking sounds. At this point, I rejoiced. The defect seemed to be in (at least) the "bad board," and my U12 removal/insertion appears to have at least somewhat succeeded.
5. Now I placed "old U12" into "good board." I put this into the busted HDD, and powered it on. No clicking sounds! I plugged it into my PC and it was recognized at boot time. I promptly started taking off the data, prioritizing in order of most precious data first. I didn't know if I was on borrowed time because my experiment could fail, and I also didn't know how much junk I introduced into the drive when I opened it. In the latter case, I've only myself to blame. In the former case, for folks who think "just replace U12 and you're gold," I ask you this: These drive seem to have a fair amount of problems. Do you _really_ want to give them a second chance with your data?
The question that I don't have the answer to, which maybe someone can enlighten us with, is: do these drives fail as much as posts online suggest? Or is it that the WD drives are just so popular that the problems are proportional to their popularity?
There are no manufacturers left for me to choose from. I've had a failure from every manufacturer that currently sells HDD's, and many that used to in the past. Plus, the landscape of HDD manufacturers has changed vastly over the years. When I got my first HDD, Seagate was cheap crap, and Control Data Corp (CDC) and Micropolis made the best drives out there. Later on, Seagate purchased CDC's drive business, so go figure. So Seagate went from crap to gold depending on the model you bought, at least IMHO. It looks like a stroll through the local Best Buy shows Seagates (which burned me a long time ago), Western Digital (which seem troublesome right now), and Maxtor, which I think failed me a long time ago too, and I hear that some folks have problems with them. As for DeskStar drives... Well, I bought one of those ages ago too, and it did live up to the DeathStar nickname. I'm not suggesting that I have praticularly bad luck, but I've just bought lots of drives and at one point or another any brand failed for me. And I still have not learned to back up religiously. Drives continue to get bigger, data continues to amass, backing it all up is a growing challenge. Whenever I bought a tape drive, I always found myself behind the curve -- a full backup requiring 5-10 tapes, and then cycling through incremental tapes. I've had pretty shoddy success with a tape's ability to retain data too...
Anyway, setting down the whiny violin for a minute, I wanted to share my soldering experience. I got advice from this link:
First off, I must thank Tim for the reassuring tone of his article. His article suggests that desoldering SMT chips isn't the most impossible thing in the world, and I felt comfortable trying it after reading his advice.
What he refers to as "Solder wick," is, to my understanding, what Radio Shack calls a desoldering braid. So between a decent soldering iron, some solder, a desoldering braid, and a little patience, you can do this. I used a small blade to pry the chip out. It was the small knife that came with one of radio shack's soldering kits. You want something very flat that will fit under the chip, but not needle thin. An exacto knife would work well too, but be careful what you're scraping under the chip!
Please don't flame me for using Radio Shack tools. When you're anxious, and it's a Sunday, they're the only ones that are open! I work 10 minutes from Jameco electronics, but they are closed during the weekends.
So, to desolder U12:
1. I put enough solder on both sides of the chip so that I had a large bead stretching across the 4 pins on each side. Make sure it's nicely melted on there and that the bead covers all the pins on one side.
2. I held the knife on one edge of the chip (one of the two ends with no pins).
3. I melted one side with my sodlering iron and twisted the knife slightly to raise the chip up a bit from that side.
4. I switched to the other side, melting the solder and lifting it up a little bit by twiting the knife the other way.
5. I repeated steps 3-4 a couple of more times, and the chip came right off, with virtually no sodler stuck to the pins.
6. Since I had it, I used a desoldering gun from Radio Shack (like a soldering iron with a built-in suction bulb) to remove the big beads of solder left on the board. You could do that with a desoldering braid: Place the braid on the blob of solder, heat the braid up, and it will suck the solder for you.
To place U12 back:
1. Line it up. If you're an idiot like me, don't use tweezers. At least the burning sensation it made me work faster
2. Heat up the pins. If you're lucky, the chip will already sink into the solder. If not, you'll have to add a small bead of solder.
3. Just try to solder the pins at this point. First, I pressed the iron to the pin and saw if it sank into place.
4. If it didn't, I heated both sides up to make sure it was well positioned, and added solder as needed.
5. If solder bridges two pins, don't panic. At least they're holding the chip in place. Continue soldering the rest of the chip, place some desoldering braid on the shorted pins, heat it up, and suck up the excess solder. There may even be enough solder left over to fix those pins to the board. If not, then don't dispair, try adding a little solder again.
It's not that hard. I'm just being verbose because I want to be clear, not to make it sound more complicated than it is.
If the thought soldering or desoldering still really, really intimidates you, then don't do it. First, decide how important your data is to you. If the whole experience is a science fair project to you, then go for it anyway and see if you figure it out. If your data is a bit more valuable, go to a friend with good soldering experience. If your data is a lot more valuable, then you shouldn't be reading this forum, but instead you should be comparison shopping for the data recovery service that offers you the most "warm and fuzzy feeling" per dollar that you're willing to spend on recovering your data.
Anyway, thanks to this thread and other links that helped.
I hope my story helps encourage people that, for those of who have lost a drive but don't think it's worth the money of (or cannot afford) professional data recovery, there is hope.
If this experiment didn't work for me, the next thing I would have tried doing was actually transplanting the platters or the heads. One thing I didn't know for sure was, is the angle of each platter relative to each other important? Or is it just important that the platters are placed in the order in which they were on the old drive?