WD Caviar Black slowly failing

[Epstein]

I have am in the midst of a hard drive failure. It still works, but very slowly. Hopefully, this is the sort of thing that can be fixed with a PCB swap, but I'd like to get some opinions on the nature of this failure.

Western Digital Caviar Black, WD1001FAES-22W7A0, PCB:2060-701640-002 RevA.

What's going on:
- I've successfully copied some files off of it, but extremely slowly. Large files (several gigs) are problematic.
- Read speeds are highly variable.
- WD Diagnostics didn't complete (drive disappeared), but there weren't any SMART errors.
- Drive spins up normally.
- Makes seeking noises in bursts when transferring files.

Again, I'd like to PCB swap this drive. I have experience reworking PCBs, so moving the BIOS isn't a big deal. But, I just want some opinions on whether this is even the culprit, or if maybe the preamp is to blame. I'd also be curious to hear about which component might be going out on the PCB. I do hardware design for a living.

[Cleanroom]

If you want to try a board swap, then find a board that match 2060-701640-002 and swap the U12 chip that is located on the flip side of the board near the top center. It is the small 8-pin chip.

If that does not help that you have a heads issue and you will require professional assistance.

Good luck

[Epstein]

If you want to try a board swap, then find a board that match 2060-701640-002 and swap the U12 chip that is located on the flip side of the board near the top center. It is the small 8-pin chip.

If that does not help that you have a heads issue and you will require professional assistance.

Good luck

That's the plan. I just wanted to know, based on the symptoms, if anyone had any insight into whether the PCB swap will actually fix the issue or whether it's the heads/preamp/whatever. I realize at the end of the day a PCB swap will either fix it or it won't. I'd just rather not throw parts at it unnecessarily.


Also, I had the PCB off and under the microscope. I didn't find any damage, signs of overheating, or failed solder joints. WD needs to wash their PCBs better, though.

[Cleanroom]

there is no way of really knowing until a professional looks at it as to what is exactly wrong, but chip/board swap is prolly something you can easily try on your own. It should not makes things any worse.

[northwind]

Well allow me to disagree on this. If things go wrong during soldering/desoldering, then....

[mr_spokk]

I think the problem is elsewere, and it's not DIY...sorry to say.

[Epstein]

Well we will find out shortly. I have a matching PCB arriving today or tomorrow. Soldering shouldn't be a problem. I have to rework SO8's all the time at work.

Still, has anyone seen symptoms like this and can point to a specific problem?

And one final thing that I thought I'd post in case someone has a similar question... I've found that the "use the same PCB number" may not be completely accurate in all cases. The 2060-701640-002 RevA that I needed apparently comes populated in 2 different flavors: 32MB and 64MB cache. Tracking down the larger one that matches my drive was a bit trickier. In theory, the firmware could detect the size of the cache on init, but I have no idea if that's actually happening. I have the proper PCB so that shouldn't be an issue for me. I thought I would post my finding so far, though.

[alchemy]

sounds similar to my problem in the thread I just posted "live PCB swap..... ?" except I may have compounded the problem.