WD1002FAEX overvoltage, TVS ok

[cymerio]

Hi. First of all, sorry about my English.

I have 3 hard disks that died in the same incident (overvoltage in the 5v line, I guess).
One was an old Maxtor drive and had the 5 V TVS diode burnt out. I fixed it myself just replacing the diode. So far so good.

The other 2 hard disks however show no visible damage, so I'm a bit confused here.
Neither of them get detected by the BIOS and they make no noise when powered.


One of them is a WD1002FAEX:

• I tested both diodes and 0 Ω resistors with a multimeter and they are all fine (i.e. resistors measure 0 Ω, diodes measure about 260 Ω and 260 kΩ when tested in place IIRC).
  I even tested the 5 V TVS diode outside the circuit (high resistance in one direction, low on the other, ~0.6 V drop IIRC). I soldered it again in place.
  
• I tested the resistance between the 4 motor contacts at the center of the drive's chassis, with the PCB removed. They measured 1.3 Ω and 2.3 Ω depending on which contact pairs were measured. So I guess the motor is fine, right?
  
• I cleaned the contacts at J1 with a soft rubber. No result.
  
• I measured voltage at the SATA power pads on the PCB, and they are all OK according to specs (i.e. 3.3, 5, 12 V plus ground and something else at pin 11).
  
• I measured voltage at some random points of the PCB, and couldn't find any place with 5 V or 12 V except near the power pads. Most places measured around 0.5 V.

Any tips to identify the problem? What else could I measure?
Is it beyond DIY repair?

Here are some photographs of the drive:

• Top label (model, serial number, etc)
• Back side
• Back side, PCB removed
• PCB closeup
• PCB closeup, component side

Thank you for your time.

[fzabkar]

Measure the voltages at the indicated test points.

Voltage_test_points.jpg (59.01 KiB) Viewed 15284 times

BTW, your English is better than 99% of native speakers.

[HaQue]

the first post should go sticky on the newbies faq page as a prime example of how to ask for help, say what has been done and show obvious effort on part of the victim of HDD Failure.

fzabkar, are they the TVS diodes?

cheers

[fzabkar]

are they the TVS diodes?

No, here they are:
http://www.users.on.net/~fzabkar/HDD/bi ... diodes.jpg

[pcimage]

I must say that is an excellent first post, all the detail required and in excellent. English

The good news is that the PCB has a ROM at U12, so it should be straightforward to exchange the PCB and transfer the 8-pin chip at U12.

Assuming there's no internal (preamp) issues, then this should fix the drive.

For €50 I'll provide a new PCB and transfer the ROM for you, as I'm so impressed with your post

Of course, try fzabkars suggestions first. Maybe he can help you fix the PCB for free!

[cymerio]

Thanks everybody for your kind responses!

Measure the voltages at the indicated test points.

They measured 0.1 V, 0 V and 0 V respectively, from left to right. Kind of sad.
While I was at it I also measured the voltage around C76, which was 5 V and 0 V on each side, if it helps.

For €50 I'll provide a new PCB and transfer the ROM for you, as I'm so impressed with your post .

Thank you very much! A very good price indeed. I'll definitely accept your offer if I can't fix it myself.

[fzabkar]

ISTM that the SMOOTH motor controller IC is dead. This chip generates the onboard Vcore, Vio, and V- supplies. All three supplies are at 0V. I suspect that C6, C9, and C14 will have either 5V or 12V across them. If so, then that would be added confirmation that the chip is dead.

[cymerio]

ISTM that the SMOOTH motor controller IC is dead. This chip generates the onboard Vcore, Vio, and V- supplies. All three supplies are at 0V. I suspect that C6, C9, and C14 will have either 5V or 12V across them. If so, then that would be added confirmation that the chip is dead.

Yes, C6 has ~12 V and C9, C14, C13, etc. have 5 V. The SMOOTH chip must be dead.
Thank you for your help.

[cymerio]

Post update:

In the end PCImage offered me the same deal for two drives (notice how in my first porst I mentioned 2 hard disks with no visible damage, although in this thread I only talked about one).
That was unbelievable kind from their part. Big thumbs up.

Well, the fixed PCBs came back today (super quick job ).
Unfortunately it only worked for one of the drives:

• WD20EARS:
  • Spins up (normal noise), the BIOS detects it.
  • I din't dare to access the data right away, but I guess it's fine. Should I take additional steps before trying to access the data?
• WD1002FAEX:
  • Doesn't spin up (no noise), BIOS doesn't detect it.
  • I checked the PCB voltages near the SMOOTH chip and this time I got sensible voltages (-5V, etc).
  • The 4 motor contacts still measure ~1.3 Ω and ~2.3 Ω
  • So I guess it's the preamp? Can I check it with the multimeter (without opening the case)?

[fzabkar]

If you install the PUIS (Power Up In Standby / PM2) jumper on the WD1002FAEX drive, does it then appear in BIOS, albeit with a blank model number?

Jumper Settings Info Sheet - Western Digital:
http://www.wdc.com/wdproducts/library/o ... 001037.pdf

Can you measure the voltages at the HDA connector (J1) with the PCB on and off the drive? It may be difficult to find the test points on the rear of the PCB.

[cymerio]

If you install the PUIS (Power Up In Standby / PM2) jumper on the WD1002FAEX drive, does it then appear in BIOS, albeit with a blank model number?

No, it doesn't appear.
My motherboard is an Asus Crosshair V Formula-Z with BIOS 2002 (latest).

Can you measure the voltages at the HDA connector (J1) with the PCB on and off the drive? It may be difficult to find the test points on the rear of the PCB.

Do I need to measure all 20 pins?
I measured 4 and 6 using these probe points (guessed):
Probe points I used

With the PCB on the drive:

• 4: ~0V
• 6: 5V

With the PCB off the drive:

• 4: -5V
• 6: 5V

[fzabkar]

It could be that the preamp is shorted and is pulling down the -5V supply. If you place a business card between the HDA contacts and the PCB, does the drive spin up? If so, then this would suggest an internal fault.

[cymerio]

It could be that the preamp is shorted and is pulling down the -5V supply. If you place a business card between the HDA contacts and the PCB, does the drive spin up? If so, then this would suggest an internal fault.

Yes, it does spin up when I do that.
Also the drive is detected by the BIOS with a blank name.

[pcimage]

Seems you have a bad preamp then

[deftrue]

Seems you have a bad preamp then

I have to agree

This one's not gonna be that easy!

[cymerio]

Well then... I won't be fixing this one. Too expensive for me.
Thankfully I have the other drive which has a backup of my data, although several months old.
So I lost some data, but not all.

Let this be a warning for all of the readers that come to this thread:

• Backup your data frequently. Don't be lazy.
• Don't put all of your drives in the same computer (i.e. backup to external drives / cloud).
• Even with a UPS and a good PSU your drives are not safe. A faulty motherboard still can still screw you.
• If you find the most minimal fault with new hardware, return it ASAP.

What did happen to me was that I kept a new motherboard that had some issues.
It was slightly bent and had a dead SATA channel. Nevertheless I still kept it.
Apart from that everything seemed to work... until I installed a new graphic card (2 cards total, extra power lines to the mobo).
When I pushed the start button I saw a flash and smelled smoke.
The motherboard, CPU, a PS2 keyboard and a card reader died in the incident.
There were 4 drives in the system ATM:

• A 256GB OCZ SSD. Still working fine. 0 problems. All data intact.
• A 1TB WD Caviar Black (WD1002FAEX). PCB's Smooth chip dead. Preamp dead.
• A 2TB WD Caviar Green (WD20EARS). PCB's Smooth chip dead.
• An old 300GB Maxtor drive. TVS diode dead.

So in the case of the HDDs, the newer they were, the most damaged they resulted due to overvoltage.

Draw your own conclusions. Learn from my mistakes.

Thanks to everyone on this forum. You've been the most helpful.
Oh, and PC Image rules!