I was moving my pc innards to a new case. Everything was fitted, powered up: smoke. From the front, so I suspected a hard drive. Checked I hadn't made any shorts in the case etc. and tried again. This time Win 7 started up but THREE of the drives (five total) were not detected. Nor detected in BIOS. After switching around sata and power cables, and trying the drives in another PC and external HDD case, discovered the three drives had been damaged – none of these spin up nor are they detected.

After discussion and consideration discovered I had put a 4 pin molex to sata power adapter on in reverse! Ack!!

Dead HDDs:
1.
WD 750GB Caviar Green WD7500AACS
WD7500AACS-00D6B0
14 Aug 2008
PCB code 2060-701537-004 REV A

2.
Seagate 1TB ST31000333AS
PCB code: 100512588 Rev A

3.
Samsung 250GB SP2504C

The Samsung was a scratch disk and Im not too worried about it, but the other two had months worth of work on them (ie since last I backed up. Ack!)

Below are photos of the damaged PCBs:

Green: http://www.flickr.com/photos/45441303@N07/5391359361/

Segate: http://www.flickr.com/photos/45441303@N07/5392007956/

From what I've gathered I'll need to change the PCBs then switch back the ROM chips. Discovered the ROM chip location on the segate but not on the Green. Any help, suggestions much appreciated. Also where best to get a replacement Caviar Green PCB (have already ordered one for the segate)? Located in Australia (Melbourne)

Much appreciated!

Cheers,
Richard

Here are direct links to your photos (no ads or banners):
http://farm6.static.flickr.com/5212/539 ... b575_b.jpg
http://farm6.static.flickr.com/5219/539 ... 9c22_b.jpg

On the WD7500AACS PCB there is a vacant location at U12. This is set aside for an external flash memory. Unfortunately this means that the ROM data are internal to the Marvell MCU.

The Seagate drive has a burnt 5V TVS diode. This suggests that there was an overvoltage on the +5V supply. I can't see it clearly in your photo, but the 8-pin chip nearest pin 2 of the preamp connector (J4) appears to have a blister near its top right pin. This chip is probably the MOSFET chopper and Schottky flywheel diode for the preamp's negative supply, in which case there could be preamp damage. Measure the resistances between ground and each of pins 6 and 8 of the preamp connector, both on the board and on the HDA itself. Use the 200 ohms range of your multimeter. If you don't have one, a digital multimeter can be had for AU$10 from Jaycar or Dick Smith.

_________________
A backup a day keeps DR away.

Thanks so much for your reply!


With the rom on the MCU of the WD caviar green, what are my options to retrieve data? I suspect none other than forking out large sums of money?

The seagate: Ah, in the meantime I have retrieved another PCB (same 100512588 Rev A), and swapped the ROM chips. I'm getting power now, but a 'chk...chk...chk..." sound and no detection. Does this also mean the drive itself has been damaged? Should I check the pins on the connector of the preamp on the drive itself as suggested?
I'm learning as I go. I've determined the pins 6 and 8, but where do I connect for ground (I've a strong suspicion that's an embarassing question)?

AISI, the WD drive requires professional DR.

As for the Seagate, the preamp looks like it may be damaged, but that's for someone else to tell you.

When taking measurements, I would use SATA power ground as your reference.

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

AIUI, the fact that the Seagate spins up means that all the onboard supplies on your replacement PCB are OK. This in turn means that none are being shorted by the preamp. I'd still take some measurements, though, just in case I'm wrong.

I suspect that preamp pins 2 and 4, and possibly 10, may be grounds. Pins 5 and 7 could be the two ends of the voice coil. One of these pins would be connected to a parallel array of current sense resistors, probably the set of 4 x 1R00 resistors near the SMOOTH motor controller. The bank of 5 x 1R00 resistors would constitute the current sensing for the spindle motor. I believe they would be grounded at one end.

_________________
A backup a day keeps DR away.

Seagate:

pins:
Current (150 ohms) on 6
no current on 8
-- this applies equally to both old and new PCBs

and current on pins 6 and 8 on the 'HDA' connectors

Let me double check I've transferred the correct chip!! From my research this is the ROM chip, but perhaps I was wrong: 25P40VP See attached image (strangely this is not the same numbers/brand on both PCBs, though both PCBs are coded the same, i.e. 100512588 Rev A.) I, perhaps wrongly, used the details from this discussion which was about same drive with different PCB: http://www.techsupportforum.com/forums/ ... 68684.html

The changed chip:


As for the WD drive, though that large chip looks burnt out, I've discovered one of the (protection?) diodes has shorted, so there's a slim chance I could save it anyway. However, where would I find a replacement diode? code reads "LG 87A" but google offering nothing I can see. See pic

"25P40VP" is correct. The Winbond part in the other thread is electrically equivalent.

M25P40VP, ST Microelectronics, 4 Mbit (512 K × Serial Flash memory:
http://www.numonyx.com/Documents/Datasheets/M25P40.pdf
http://www.datasheetcatalog.org/datashe ... Xyuxsz.pdf
http://www.farnell.com/datasheets/40267.pdf

The "LG" part on the WD board is the 12V TVS diode (D4).

SMBJ13A, General Semiconductor (Vishay), marking code LG, 600W, SURFACE MOUNT TRANSZORB TRANSIENT VOLTAGE SUPPRESSOR:
http://www.datasheetarchive.com/pdf/get ... .pdf&scan=
http://www.vishay.com/docs/88392/88392.pdf

That said, when the SMOOTH chip is damaged, replacing the diode would be pointless.

See my notes:
http://www.users.on.net/~fzabkar/HDD/HDD_ICs.txt
http://www.users.on.net/~fzabkar/HDD/TVS_diodes.txt

_________________
A backup a day keeps DR away.

Preamp can sometimes be damaged, but not short the new board in my experience.

To continue diagnosis on Seagate I would say it is necessary to create RS-232 adapter for Seagate to view terminal output, but this is a lot of work just to get diagnosis for something that is not *likely* fixable without access to a proper laminar flow bench, and necessary tools and practice to safely replace heads on Seagate desktop drive.


For the WD, it is tough because the ROM is embedded into MCU. This will require pro tools such as PC3K to repair, as you will need to recreate the ROM using some "fancy tricks".

There are a few members who likely have PC3K from Australia who patrol these boards, seek them out and let them know your from HDDguru, they will probably give you a much more honest price then going directly to a big company. Also other members of the board will be able to vouch for their credibility.

I am of course not purposely trying to discourage DIY; just simply letting you know that both of these drives are damaged beyond the level that typical DIY procedures will be able to help.

Best of luck!!!

Update:

the WD caviar:
Checked R64 and R67. R67 was not near zero ohms, so removed and bridged. This overheats D3. Managed to pull the power before D3 melted entirely. Suspect removal of R67 was BAD IDEA. R64 measured near 0 ohms. So... still stuck on this one.



Seagate. Only thing I can think of is I may have soldered in the rom chip badly, so attempting resoldering now.

Thanks Russwinters, will follow up your suggestions if this final DIY attempt fails on the seagate.

Gah. Have been told I've been testing diodes incorrectly. So..

back to the caviar green.
The D4 is fine, it's the D3 that is shorted. Can anyone confirm this is a protection diode and can be removed?
R67 (200 ohm) needs to be put back in. Should definitely not have been removed. And I can put the D4 back in too. So hopefully D3 can be removed, and hopefully on removal Im not going to just find out that the Smooth is damaged anyway.

As for the seagate... still getting the same after re-soldering. Drive (very slowly) detected by bios, get the "chk...chk...chk..." sound, and then no OS detects the drive. Can only assume the ROM chip (which has been swapped to the new PCB) was damaged (unlikely?) or (likely) the drive itself has been damaged. Seems like DIY on the seagate is impossible.

R67 and R64 should both measure close to 0 ohms. A shorted D3 (5V TVS diode) would suggest that there was an overvoltage on the +5V supply. This is consistent with the damage on the Seagate drive.

As for DIY on the Seagate, you could at least build a serial terminal interface and observe the drive's terminal report at power-on. This may shed some light on the problem, although it will probably only be a post mortem exercise.

_________________
A backup a day keeps DR away.

Alas alas oh.

Time to put the DIY attempt to rest. Thanks to all who've offered assistance. Really appreciated!

Now to chasing (semi-)professionals...

Cheers
Richard