Double Failure WD5000AAKS drives

[doughonea]

I have two WD5000AAKS drives that were Windows RAID-1 in a Windows 2003 server system. After a power failure, neither drive is recognized by the BIOS in that system, nor in another system I tried. In the original system, the BIOS can tell that *something* is plugged into the SATA channel, but it cannot ideintify what it is.

I tracked down a drive with a PCB that matches one of the drives (waiting for delivery), but one PCB vendor I spoke with said it might me corrupted firmware rather than a PCB issue.

I have been a computer pro for about 20 years, but when it comes to hard drives, swapping PCBs is as far as I have ever gone...in short, I am clueless about checking/updating firmware, but I would LOVE to recover one of these drives long enough to clone it.

I'd appreciate any thoughts on what the real issue might be and how to go about testing/recovering, etc. I apologize that I am going to need some fairly detailed assistance since this is new territory for me. I can follow instructions, though!

Thank you in advance for any and all assistance!

[Alexii]

post full model here , also need symptoms the drives are exhibiting upon power up. Are they spinin at all ? Clicking ?

[jono-ats]

If the drives aren't spinning up, check for threads about the TVS.

If the drives are spinning up but clicking, then you likely have preamp failure . . . definitely not a DIY fix.

If they are spinning up but aren't clicking and aren't recognized either, then you probably have a firmware issue.

There are some relatively inexpensive software tools that folks on the forum have mentioned that I haven't tried personally. It's easy to brick a drive when you are manipulating FW. But most important is the knowledge of how the FW works and which modules are critical, etc. That info you won't find on the forum.

If it's not a simple TVS issue, then I'd recommend pro assistance if the data is important to you.

[doughonea]

Since I am not sure just what info on the drives to provide, I am posting a pic of each drive's label.

The drive initially spin when plugged in, but stop after about 15 seconds. There is no clicking. The drives were not subjected to any drops, etc.

Attachments

[doughonea]

And for what it is worth, I would really like to retrieve the data on either drive, but it's not worth spending big bucks to get it back. More of an issue of conveinence vs. necessity. I can recreate everything in time.

[unknown]

he drive initially spin when plugged in, but stop after about 15 seconds. There is no clicking. The drives were not subjected to any drops, etc.

Most likely SA problem.

[lcoughey]

Both drives in the picture are of a different family...so, I don't think you can get the same PCB to work for both drives. As you say the drives spin, then spin down, you have eliminated the blown TVS out of the picture and are left with:
1. PCB issues common with ROYL drives
2. Head issues
3. Firmware issues
4. Surface issues

No matter how you look at it, you really need the assistance of a pro with the right tools to help you properly diagnose the actual problem and quote you a price for the recovery.

If your issue is PCB only and you supply the donor PCB, you might be able to convince a pro to reprogram it for you at a lower price.

[doughonea]

I understand that the PCB has to match the specifc drive. The replacement that is coming matches one of the drives.

In the meantime, I did find a firmware module for the WD5000AAKS-00UU3A0. Would anyone please offer advice as to what software is available to load this to one of the drives?

Like I said before, I am not willing to spend much money on recovery, so I don't have much to lose trying it myself, if possible.

[jono-ats]

There are many, many firmware modules, and many different versions.

I'm 100% sure you won't improve matters experimenting with the firmware -- you will only make things worse.

I think you've already received good counsel from a number of contributors. Of course, you are free to ignore it.

Good luck!

[Alexii]

Sry D. I am all for helping DIY guys but u are way in over ur head. The cheapest tools to work with WD SA are ~ 500$ i think. Someone will correct me if i am wrong. So even if someone agrees to walk u through every step of the way u still get no way of doing it. Its a good thing data is re constructable by u , i congratulate u on the backup. At his point i think its done. Get 2 new drives and best of luck in ur future endeavors.

[doughonea]

Thanks Alexii...I really hate to drop $500 for this; I was hoping it could be a learning experience for me, and acutally, it has been...Thanks everyone for your help and advice.

[fzabkar]

AFAIK, the cheapest SA tools are WDR-UDMA and WD HD Pro (~US$150). You should be able to use this software to read the "ROM" on the patient PCB and transfer its contents to a donor PCB. These tools will also allow you to read and write the firmware modules in the System Area (SA) on the platters.

Alternatively, if your WD5000AAKS-00UU3A0 drive uses the following PCB ...

http://ledisquedur.com/526-thickbox/206 ... rev-p2.jpg

... then a ROM transfer involves moving the serial flash memory chip at location U12 to your donor. Some board suppliers (eg hdd-parts.com) include such a service for US$10, otherwise your local TV/AV repairer should be able to handle it.

If your other PCB (WD5000AAKS-75A7B0) looks like this ...

http://pcb-hdd.com/images/WD10EAVS-00D7 ... %20PCB.jpg

... then location U12 is vacant, in which case a ROM transfer will require a different approach. This is because the ROM data are internal to the Marvell MCU.

That said, it may be that neither MCU (88i9045-TFJ2 and 88i8845C-BHG2) is affected by the "mimic" problem which affects MCUs such as the 88i6745-TFJ1. The latter MCU often fails with symptoms similar to a head or media fault. I'm not a data recovery professional, so others will be able to confirm whether this is the case.

BTW, late model WD drives have oxidisation issues, probably as a consequence of materials changes mandated by RoHS. Therefore the first thing I would try is to gently clean the 20-pin preamp connection points on the PCB with a soft white pencil eraser.

There are ways to read and write the ROM and SA modules using freeware, but I don't know how to go about it when the drive isn't spinning.

See wdc-wd5000abys-01tna0-re2-needs-recovered-t21632.html

[jono-ats]

There are ways to read and write the ROM and SA modules using freeware, but I don't know how to go about it when the drive isn't spinning.

It's not so much HOW but WHAT and WHEN.

You are very helpful with PCB-related issues, but you are really doing end users a disservice by encouraging them to muck about with the firmware, especially the SA modules. There is a lot to know, and random manipulations are ALWAYS a bad idea. Why do you persist?

[fzabkar]

... but you are really doing end users a disservice by encouraging them to muck about with the firmware, especially the SA modules. There is a lot to know, and random manipulations are ALWAYS a bad idea. Why do you persist?

AIUI, the OP has already declined the price of professional recovery. He is now looking for an alternative to throwing his drives in the trash bin. Why would you want to stop him from experimenting when he has nothing to lose? In any case, the first thing I propose to do is to READ the firmware, if at all possble. That's perfectly safe to do. It is certainly not "random manipulation", as you put it. It is standard practice.

In fact, if you really care about the OP's predicament, perhaps you could advise us as to whether the abovementioned MCUs are affected by mimic faults. That would at least save him the cost and bother of replacing his boards.

[jono-ats]

AIUI, the OP has already declined the price of professional recovery. He is now looking for an alternative to throwing his drives in the trash bin. Why would you want to stop him from experimenting when he has nothing to lose? In any case, the first thing I propose to do is to READ the firmware, if at all possble. That's perfectly safe to do. It is certainly not "random manipulation", as you put it. It is standard practice.

In fact, if you really care about the OP's predicament, perhaps you could advise us as to whether the abovementioned MCUs are affected by mimic faults. That would at least save him the cost and bother of replacing his boards.

I'm sorry. It's just that the notion of safe practices is so ingrained, it's hard for me to recommend something to anyone that might brick the drive for good . . . IMHO, it's always good data recovery practice to try to maximize options -- not to close doors.

Sometimes people change their minds. Or they get "experimenter's remorse," and wish they hadn't done step A or B. Or, they later realize that there was something very valuable on the drive that they wanted after they ruin it.

Perhaps this drive has nothing invaluable -- that's the impression I get. The OP stated that he thinks he has a firmware fix that he wanted to try. However, there is no universal firmware fix, and no single patch that will help him. I know this because I work with WD drives every day, as one of their Recovery Partners.

You have to be able to determine which modules are corrupted, and of those, which ones matter, and which ones are unique to a drive and cannot be replaced; which can be regenerated and how, etc. etc.

I haven't bought the software you recommended, but I'm skeptical that it can do all of those tasks. Even if it can, it will not teach the user what he or she needs to know.

Reading FW does not put the drive at risk unless there is head / media damage. But you didn't initially suggest that the user stop at that point. Nor did you reveal that writing of firmware should not be done casually; that there are significant risks and that knowledge and experience are prerequisites.

Anytime you mention working with firmware, you owe it to the community at large to FULLY disclose those risks. Someone else Googling the topic may come across your post months or years later and have no context and no clue. They might just brick their drive thinking they had a solution to their problem.

If you fully inform a person when you dispense advice, then the reader is able to weigh the risks vs. rewards and decide for themselves. Like I said, anything less is a disservice.

You claim that there is "no harm" if the OP doesn't care about the data. But what you are recommending is like suggesting that someone drop a playing card from a table to see if it will land standing straight up on its edge. That's not going to happen, and neither is the poster going to learn anything of real value by playing with the SA firmware in this case --- except not to do it.

The most common PCBs for the "mimic" problem are 2061-701444-xxx and 2061-701477-xxx, but other PCBs are subject to failure for various reasons, so it is part of the diagnostic process to rule them out.