Logical step to troubleshoot a WD5000

[northyork]

I am new to this board and am interested in learning about HD repair. I have started to read through several of the beginner posts but there's a lot. I assure you that I respect the experts opinions here and will try to give back as I learn. So in short, thank you for any knowledge you share with me.

So on to my question:

I have a WD5000AAKS that isn't recognized by my PC. I believe that I caused the short when I attached the graphics power to the HD (doh!). The PCB shows several burn marks around the screws that hold the PCB to the drive. These screw holes also appear to be part of the circuit - maybe ground points?

I am not certain if the drive, pcb, or chips on the pcb are damaged.

I am looking for a set of logical steps to follow to try and determine where the failure may be.

I have been told that I should start with finding a replacement pcb but that the bios will need to be removed and re-soldered. Is this the right place to start? Are there other things I can/should try first? Is it possible to find a pcb with the correct BIOS already together rather than having to remove the chip and re-solder?

Thank you very much for your input.

[fzabkar]

Can you show us a photo of the damaged area? Does it involve TVS diodes D4 or D3, and resistors R64 and R67?

[Vulcan]

I have a WD5000AAKS that isn't recognized by my PC.

More detail please What exactly do you mean by "isn't recognized"? There are several possible interpretations...

Also:

• I'm assuming that the drive is powered from the PC's internal PSU. Does the PC power-on with that drive attached to its PSU or not?
• If the PC does power-on with that drive attached, then listen to the drive carefully at power-on. Does the drive attempt to spin-up when you listen to it, or not?
• If the drive does attempt to spin-up, what else do you hear e.g. drive stays spinning, or drive stops spinning, or drive makes click-click noises for a few seconds, or ... ?

[northyork]

Attached is the requested picture of the PCB. I highlighted in red where I visably see discoloration.

The R64 & 67 look intact but I can't be sure I'd recognize what bad looks like. They both have a readable white box on them

To answer Vulcan - Isn't recognized I mean, I plug it in and nothing. No sounds, no action from the PC, just nothing.

Attachments

[fzabkar]

The discolouration is just normal oxidisation. I can't recognise any burn marks.

Did you remove diodes D4 and D3, or were those locations already vacant?

Unfortunately the flash data are internal to the Marvell MCU, so a DIY board swap is not feasible. U12 is wired for an external serial flash memory, but this location is vacant. Your only real hope of recovering your data, without resorting to professional data recovery, is to attempt to repair the board. That may not be as difficult as it sounds, if you have a digital multimeter (~US$10).

Some board suppliers offer a firmware transfer service (US$10 - $20), but they will probably need to know whether the MCU is still alive.

[harddrivespecialist]

Did you buy that drive in used condition?

It seems like motor IC was replaced before and both TVS chips are gone.....

[northyork]

I did not buy the drive used.

[northyork]

What should the PCB look like?

[northyork]

The discolouration is just normal oxidisation. I can't recognise any burn marks.

Did you remove diodes D4 and D3, or were those locations already vacant?

Unfortunately the flash data are internal to the Marvell MCU, so a DIY board swap is not feasible. U12 is wired for an external serial flash memory, but this location is vacant. Your only real hope of recovering your data, without resorting to professional data recovery, is to attempt to repair the board. That may not be as difficult as it sounds, if you have a digital multimeter (~US$10).

Some board suppliers offer a firmware transfer service (US$10 - $20), but they will probably need to know whether the MCU is still alive.

No I didn't remove anything from the pcb, this is how it came out.

How do I go about repairing the board?

Thank you for your assistance.

[fzabkar]

Normally the board would be protected by a 12V TVS diode at D4 and a 5V TVS diode at D3. I believe harddrivespecialist is alluding to the discolouration of the SMOOTH chip's pins. However, its YWW (Year / Week) date code, 726, matches the date code, 0726, on the Marvell MCU, which suggests both may be original components.

The tests I would do would be to first to confirm that there are no shorts on any of the supplies, and then power up the board on its own and measure the actual voltages.

Set your multimeter to the 200 ohms range and measure the resistances of R64 and R67. They should be close to 0.

Then measure the resistance between ground and each of the +5V and +12V pins in the SATA power connector.

http://pinouts.ru/Power/sata-power_pinout.shtml

Measure the resistance between the corner pins, 1 (Vdd) and 54 (Vss), of the hynix SDRAM. Pin #1 is identified by a dot on the PCB and is at the top right corner of the chip. Also confirm the resistance between pins 4 (top left) and 8 (bottom right) of the vacant location at U12. Both devices appear to require a +3.3V supply.

HY57V281620ETP-6, Hynix, 128Mb SDRAM, 2M x 4Bank x16, 3.3V:
http://www.hynix.com/datasheet/pdf/dram/HY57V281620E(L)T(P)Series(Rev.1.1).pdf

Measure the resistance between the metal tab of the transistor at Q4 (below the SDRAM), and each of its end pins. Also measure the resistance between ground and each of the end pins. Q4 appears to be the pass transistor for a linear regulator.

Measure the resistance between ground and each of pins 2 and 4 at the 20-pin preamp connection points at J1. These are the +5V and -5V supplies for the preamp.

Below the SMOOTH chip you will find two inductors. These are the black components above Q3 and to the left of D1. Measure the resistance between each inductor and ground. Also measure the resistance of R45 (R050) to the left of the inductors, and R48 (0) at the top right corner of the MCU.

The next step depends on what you find after completing the above measurements.

[northyork]

I am trying to think back to my college days and just to confirm, all resistance testing should be done under no load - correct?

[fzabkar]

All resistance testing should be done without power applied.

[northyork]

R64 is showing .04
R67 is showing 1 or open loop
All the +5 and +12's to ground are showing open loop

I was confused with the hynix instructions - You refer to a dot in the top right corner to reference pin 1 but my dot is in lower left. And unfortunately the 2nd link is no longer active so I couldn't look at the pin out.

Normally the board would be protected by a 12V TVS diode at D4 and a 5V TVS diode at D3. I believe harddrivespecialist is alluding to the discolouration of the SMOOTH chip's pins. However, its YWW (Year / Week) date code, 726, matches the date code, 0726, on the Marvell MCU, which suggests both may be original components.

The tests I would do would be to first to confirm that there are no shorts on any of the supplies, and then power up the board on its own and measure the actual voltages.

Set your multimeter to the 200 ohms range and measure the resistances of R64 and R67. They should be close to 0.

Then measure the resistance between ground and each of the +5V and +12V pins in the SATA power connector.

http://pinouts.ru/Power/sata-power_pinout.shtml

Measure the resistance between the corner pins, 1 (Vdd) and 54 (Vss), of the hynix SDRAM. Pin #1 is identified by a dot on the PCB and is at the top right corner of the chip. Also confirm the resistance between pins 4 (top left) and 8 (bottom right) of the vacant location at U12. Both devices appear to require a +3.3V supply.

HY57V281620ETP-6, Hynix, 128Mb SDRAM, 2M x 4Bank x16, 3.3V:
http://www.hynix.com/datasheet/pdf/dram/HY57V281620E(L)T(P)Series(Rev.1.1).pdf

Measure the resistance between the metal tab of the transistor at Q4 (below the SDRAM), and each of its end pins. Also measure the resistance between ground and each of the end pins. Q4 appears to be the pass transistor for a linear regulator.

Measure the resistance between ground and each of pins 2 and 4 at the 20-pin preamp connection points at J1. These are the +5V and -5V supplies for the preamp.

Below the SMOOTH chip you will find two inductors. These are the black components above Q3 and to the left of D1. Measure the resistance between each inductor and ground. Also measure the resistance of R45 (R050) to the left of the inductors, and R48 (0) at the top right corner of the MCU.

The next step depends on what you find after completing the above measurements.

[fzabkar]

R67 is open circuit. Bridge it with a short wire link, or flow a blob of solder between its ends. Then retest the +5V supply for a short to ground.

The hynix link still works. It's the brackets in the URL that are confusing the forum's "linkifier". Just manually copy and paste the full link.

Sorry, I would have responded earlier, but it appears that automatic email notifications have been suspended on my account.

[northyork]

Unfortunately, every one of those tests show an open loop.

I bridged R67 like you said and that didn't change the results of the +5 or +12's. Also, when I tested 1 (Vdd) and 54 (Vss), of the hynix SDRAM, it also showed an open loop.

Are you sure that my multimeter should be set to 200 ohms? I ask only because I do get readings at higher ohm settings.

Again, thank you for your assistance.

[fzabkar]

Those measurements are OK. You have confirmed that there are no shorts on any of the supplies.

That said, the resistances of R45 (R050) and R48 (0) should both be 0 ohms (not their resistances to ground).

If the resistors test OK, then power up the drive and good luck. If it still doesn't work, then we can do some voltage measurements to narrow down the fault.

[Amarbir[CDR-Labs]]

Those measurements are OK. You have confirmed that there are no shorts on any of the supplies.

That said, the resistances of R45 (R050) and R48 (0) should both be 0 ohms (not their resistances to ground).

If the resistors test OK, then power up the drive and good luck. If it still doesn't work, then we can do some voltage measurements to narrow down the fault.

As Usual ,
You always do a great job with newbies .Good and Great Going Mate

[northyork]

R48 is .05
R45 is showing OL

I have tried powering it before going through these steps with you and have had no luck and since we haven't changed anything (only tested), I'm not sure it's worth trying to power it up again...

[fzabkar]

Just to be sure, R45 is the black component immediately above "R45", not below it. It should have a resistance of 0.05 ohms. I expect that it is the current sense resistor for one of the DC-DC converters. If I'm right, and if it is open (OL), then the associated supply has a problem. It's load would probably be damaged, too.

I can't be sure from the photo, but I suspect that R45 is part of the Vcore supply for the MCU. Could you test the resistances of the 4 nearby capacitors (C24, C25, C37, Cxx)?