Click of Death from a WD5000AAKS - any chance of rescuing data?

I had a WD5000AAKS-00YGA0 mounted in my Windows Home Server and it recently started to click. I had a new drive ready to be put in, but before I managed to disable the clicking one from WHS the clicking got worse and now it fails completely to show, both in WHS and in Win7. 

It spins up, clicks rapidly for a few times, then stops clicking and spins down after like 30 seconds. Hello Click of Death, I didn’t want to hear from you…

Since it stops by itself, can I assume that the arms is functioning properly, ie makes their way home to base instead of just floating around over the discs?

I checked the PCB and it looks okay to me, except for some of the connectors along the biggest thing of the board (it says “hynix”?), it seems a bit burned on the longer golden things facing the power supply, just off C64 and C63. That’s the only thing I can see that looks odd.

Is there any way to rescue the data off the drive? I have a friend that’s got two drives of the same model, but not of what I assume is the same series, 00TMA0 and 00A7B0 instead of my 00YGA0.

I’ve read some threads here that makes me wonder if I’ll ever pull this one off, but swapping the PCB’s, would it “maybe work, in very rare occasions” or is it “forget it, it’ll never work”…? 

Hey, I’d even try to swap the physical discs between chassis if that would work… ;) 

I think you will need a professional data recovery service.

lol dont try any of those things if data is important to you

if you try to swap the pcb and the phisycal device is bad, it may kill the good pcb

if you try to move the platters the data will get corrupted and lost forever

those drives are finely tuned. changing the boards most likely wouldn’t work.  don’t even think about opening the drive case.  you’d ruin the disks inside doing so.  data recovery is your best bet.


haha, men of great minds think a like.  sorry david I didn’t see your post.

Your WD5000AAKS model is a member of the “Tornado” family. Normally clicking would be associated with an internal fault. However the Marvell MCU (88i6745-TFJ1) in this family often fails in a way that mimics head or media problems.


If the above is your problem, then a board swap plus firmware transfer should fix it. Unfortunately the “adaptive” data are internal to the MCU, so a “ROM” transfer is not a straightforward DIY proposition. Some board suppliers include a firmware or ROM transfer for US$10 - $20.

BTW, the hynix part is an SDRAM. The following article should help you identify the other components on the PCB.

One thing you could try is to gently shine up the 20-pin preamp pads on the board. Use a soft white pencil eraser.

Okay, I’m with you all the way, but I lost you at “20-pin preamp pads”… Somewhere on the board itself or is it the SATA-connectors?

I think we all lost you, fzabkar.  but I am glad you have a lot more knowledge about this than we do.

I’m sorry, I’m not a very good communicator. Let me try to explain it a little better.

Inside the drive’s head/disc assembly is a head stack, and on this head stack is a preamplifier IC. The signal from the read heads is extremely weak and needs to be boosted by the preamp. The preamp connects to the main circuit board via a flexible cable and a 20-pin connector. The PCB mates with this connector via 20 copper pads. The amplified read signals are fed to the Read Channel logic, either in a separate IC as in earlier drives, or incorporated within the microcontroller (MCU), as is the case with the WD5000AAKS.

embedded servo data <–> read head <–> preamp <–> preamp connections <–> read channel

Each track on the platters has embedded servo information. That is, the user data are surrounded by prerecorded information that is required for positioning the head over the track. If a drive is unable to read its servo information, then it will sweep the heads forwards and backwards, resulting in a clicking sound.

The usual culprits are platter damage, head damage, or preamp failure. “Weak” heads appear to be a common problem in modern HDDs, probably because the heads fly lower and faster than in earlier models. When a head encounters an obstruction (thermal asperity) or sustains a shock induced head slap, then its temperature increases dramatically, often resulting in permanent damage to its read element. The read signal amplitude is then significantly reduced.

Early drives sometimes had problems with the preamp becoming unglued, or bad solder joints on the flex PCB. Preamps can also be damaged by overvoltages due to failures in the external power supply

Sometimes the preamp contact pads on the PCB develop oxidisation, resulting in intermittent contact. In such cases you could use a soft white pencil eraser to gently clean the pads. An alternative method, which I routinely use in VCRs but have not tried in HDDs, is to clean the pads with a cotton bud and a metal polish such as Brasso. Modern WD (and Seagate) models are afflicted with oxidisation issues, probably due to materials changes mandated by RoHS.

On rare occasions the clicking can be due to a failure in the read channel logic. This is a particularly common occurrence in the Tornado family. AIUI, the culprit is the Marvell 88i6745-TFJ1 MCU chip. When such a read channel failure occurs, the data recovery profession refer to it as mimicking a head or media fault.

In the following article, the 20-pin preamp connector, and the connection points on the PCB, are referred to as “heads contacts”:

This article explains the servo structure:

For anyone who requires more information on the ICs, here is my collection of datasheets:

anyone have some advice on a replacement for a wd5000aaks? 

thanks, fzabkar.  very enlightning and appreciated. 

another question.  why is it that sometimes there’s clicking when the drive doesn’t get enough power?