WD5002AAEX Doesn't spin up

[icon]

The computer's power supply needed to be replaced after a surge. Now the drive does not spin up. BIOS shows blank, instead of no drive connected or reading the drive. There are no signs of damage on the PCB. If this is not a DIY, I would like to team with a guru who would be local to my area.

[fzabkar]

Does the BIOS respond differently if you install the PM2 jumper on the drive?

Try this with the board removed from the drive.

[icon]

With PM2 enabled, the BIOS shows there is nothing connected.

[fzabkar]

The PM2 jumper enables Power Up In Standby (PUIS). This results in the drive identifying itself using the information stored in onboard flash memory. The flash memory also contains the boot firmware.

If PUIS is disabled, then the drive spins up and attempts to access the reserved System Area where the full complement of firmware is stored. The drive also loads the defect lists, translator, and other modules necessary for its operation. If successful, it identifies itself with a complete Identify Device information block.

The fact that there was nothing in BIOS (with the board disconnected from the drive?) would suggest to me that the board is dead. However, I believe that some models may report a blank model number as normal behaviour, although I haven't seen this for myself.

Assuming that the board is indeed at fault, then you would need to transfer the "adaptive" data to your donor board. If location U12 on the PCB is vacant, then these data would be internal to the MCU, in which case it would not be a DIY proposition.

That said, before spending any money, I would check the 12V and 5V TVS diodes, plus the zero-ohm resistors nearby. If these are damaged, then there may be an easy, zero-cost DIY repair.

To this end, could you upload a photo of the component side of the board?

[icon]

Gladly.

Attachments

[fzabkar]

The good news is that your board has a discrete serial flash memory chip at U12. If you are unable to fix your board, then this chip will need to be transferred to your donor. Some board suppliers offer such a service for US$10.

That said, I would first set your multimeter on the 200 ohms range and measure the resistances of components D4, D3, R67, and R64 near the SATA power connector:
http://www.users.on.net/~fzabkar/HDD/bi ... diodes.jpg

[icon]

D4 177k
D3 227k
R67 .7
R64 .6

The measurements seemed to vary slightly over a series of tests, however they were all very close to the numbers I provided.

[fzabkar]

They're all OK.

Assuming that a replacement board spins up the drive, then transfer U12 and you should hopefully be up and running again. Best of luck.

BTW, I'm assuming that the drive doesn't vibrate or make any sound as if it is trying to spin up.

[icon]

The drive never makes a noise or vibrates.

This drive was provided to me with an extra PCB. The other PCB does not spin up the drive either. Also, the extra PCB always shows there is nothing connected in the BIOS, under any configuration previously discussed.

[fzabkar]

The drive never makes a noise or vibrates.

This drive was provided to me with an extra PCB. The other PCB does not spin up the drive either. Also, the extra PCB always shows there is nothing connected in the BIOS, under any configuration previously discussed.

Then it doesn't look like a PCB problem. :-(

If you isolate the 20-pin preamp contacts at J1 with insulation tape or a business card, does the drive then spin up? If so, then this would suggest that the preamp inside the HDA is shorting the PCB.

[icon]

I covered the contacts at J1 on both boards. Neither spun the drive up.

I did notice a change with the extra board in the BIOS, instead of reporting nothing connected, it now showed blank. This is the same result I received initially with the original board.

[icon]

Is there anything more I can try with this drive before asking which professional to send it to?

[fzabkar]

In the worst case you could buy a replacement board for US$40 and transfer the serial flash memory at U12 to your donor. Some suppliers include such a transfer service for $10. Therefore your total DIY cost should be less than $50, if there is no internal damage.

Howver, I would first take a few more measurements to narrow down the problem.

You say that there are no sounds or vibrations coming from the drive, so this would suggest that you don't have a seized motor or a stiction fault. I would measure the resistances of the motor windings. Typically you should expect to see 2 ohms between each of the 3 phases and common, and 4 ohms phase-to-phase.

I would then measure the voltages with the board disconnected from the drive. There will be a Vcore supply for the MCU (~1.2V), a Vio supply for the RAM (+3.3V), and a negative supply (-5V) for the preamp. To this end I would measure the voltages at pins 2, 4, and 6 of the preamp connector at J1. Also measure the voltage at pin 8 of U12 (Vio), taking care not to short any adjacent pins, otherwise you will do major damage to the flash IC. Finally, measure the voltages at L1 and L2 near the SMOOTH chip. One of them will be the Vcore supply.

The final measurement will be to check for shorts at the preamp connector in the HDA. If the preamp is shorted, then it may have killed your replacement board, which in turn would explain why neither board was able to spin up your drive, even when the preamp was isolated.

[icon]

I received a tone from a couple of pins on the HDA. Most of the measurements on both boards were not consistent with what you specified.

[fzabkar]

I received a tone from a couple of pins on the HDA. Most of the measurements on both boards were not consistent with what you specified.

Sorry, but that tells me nothing.

Excuse me if I'm stating the obvious, but resistance measurements should be performed in the absence of power, and voltage measurements must be made with respect to ground, ie the black probe should connect to SATA ground, or to a mounting screw hole.

[jono-ats]

With this model, if a head is bad, it won't try to spin.

For DR, check out DRC on the forum -- I think he is located in CT.

Jon

[icon]

My apologies. Here are my findings.

Original Board
L1 1.1v
L2 2.5v
J1
Pin 2 nil
Pin 4 .1mv
Pin 6 5.2v
U12 .5mv
J6 2.2k and 4.4k

Extra Board
L1 1.1v
L2 2.5v
J1
Pin 2 385mv
Pin 4 .1mv
Pin 6 5.2v
U12 .3mv
J6 2.2k, 3.3k, 1k

Additionally, I visually inspected the pins on the HDA which are ok.

If you think there is more we can try, I would be more than happy to continue. Otherwise, I will take jono-ats's advice and contact DRC.

[fzabkar]

Thanks very much for the feedback. I'm not a data recovery professional, and I don't have access to failed boards, so I have to work from photographs and feedback from end users.

L1 appears to be part of the Vcore supply (= 1.1V) for the MCU.

L2 is part of a second Vio voltage regulator, let's call it Vio2 (= 2.5V). It supplies the SDRAM chip.

Your result for U12 make me wonder whether you measured the correct pin. Pin 8 is at the top right corner of the chip in your photo. Pin 1 is adjacent to the dot, and the pins are numbered counterclockwise, down the left side and then up the right side. This chip requires a 3.3V supply, let's call it Vio1. Vio1 is supplied by transistor Q1 below the SATA power connector. I would test for the presence of 3.3V at Q1. AIUI, the tab should measure +5V, and the two end pins should measure 3.9V and 3.3V.

Your results for J6 don't actually test the motor. You need to test the resistances at the terminals on the motor itself.

I don't understand why -5V is missing from J1. IIRC, this is sometimes not present even on a good board, so I'm wondering whether the MCU disables this voltage if it detects a fault. Also, I previously theorised in another thread that some models may test for the presence of a working preamp. From jono's response, it appears that this is indeed what is happening, and this would also explain why neither board was able to spin up the drive when J1 was isolated.

In short, my approach would be to test for the presence of +3.3V at U12, and confirm the resistances at the motor terminals. If these are OK, then I can't see that there would be anything more you could do. Check my PM.

HY5DU121622DTP-D43, Hynix, 2.5V, 512Mbit DDR SDRAM:
http://www.hynix.com/datasheet/pdf/dram/HY5DU124(8,16)22D(L)TP(Rev0.1).pdf

[icon]

The measurements from the motor were 1.7 and 2.7. I double checked U12 at the correct pin 8 and received 2.5v. I checked Q1 and the voltages were exactly what you stated. 3.3v, 3.9v, and 5v.

[fzabkar]

Thanks again. Clearly U12 is a 2.5V part, like the SDRAM, not 3.3V, so that's OK. I couldn't see the markings on the chip, so I couldn't be certain.

Vio1 (= +3.3V) must be used internally by the Marvell MCU, so that's OK, too.

The motor windings also check out, so AFAICS your DIY efforts have come to an end.

Hopefully it wasn't a waste of time, though. At least I learned something.