Wanted: Dead Hard Drive

[Zvuqniki]

Just tried it. No change.

[Zvuqniki]

Here are the two boards.

Attachments

Both Boards

[fzabkar]

Have you tried cleaning the connections to the heads?

Just FYI, I can now confirm that Brasso metal polish, applied with a cotton bud, does an excellent job of shining up oxidised PCB pads.

I have recently tried this on a WD1600BB PCB. The preamp contacts were OK, but I tested this procedure on the JTAG conector, and on the parallel set of SDRAM pads.

I suspect that any metal polish would be suitable. You may like to wash off any residue with isopropyl alcohol.

[Zvuqniki]

Ok, I tried the hdd again today. It spun up, and started working. Managed to get ~600MB out of it. Then it stopped.

What do you guys think it is?

[fzabkar]

Your problem report is confusing. In your first post you stated that your drive "does not spin up". This would have suggested a PCB failure.

Now you are saying that the drive spins up, works for a while, then spins down. This could still point to a PCB fault, but it could also indicate a head or media fault.

Furthermore, your PCBs have the 88i6745-TFJ1 Marvell MCU that is affected by a "head mimic" problem. I have no firsthand experience with this problem, so I can't say whether your drive's symptoms are consistent with it. In any case, the location at U12 is vacant, which means that your drive's unique "adaptive" data are stored within the MCU. In this case a "ROM" transfer won't be a simple DIY proposition.

BTW, I suspect that your drive may be one of those that does not spin up if it cannot detect a preamp. If so, then this may explain the initial no-spin symptom. However, you have since cleaned the preamp contacts (J1), so this rules out a connection issue. Nevertheless you may like to see how the drive behaves if you place some insulation tape over the preamp contacts on the PCB. Do this with your spare PCB.

[northwind]

is affected by a "head mimic" problem. I have no firsthand experience with this problem, so I can't say whether your drive's symptoms are consistent with it.

No they're not.

[Zvuqniki]

Ok, fzabkar, tried using insulation tape over J1. It spun up.

So, you think that my only solution would be swapping the MCU from the Old board to the new board?
I can make that happen. Although I don't have the equipment to do that myself, I have people that have that, and know how to change chips on boards.

In your opinion, should I go and do that tomorrow?

[fzabkar]

Your initial no-spin symptom is even more puzzling now.

northwind is saying that your drive is not behaving as if it has a "mimic" fault, so one might think that the MCU is OK.

Before doing anything drastic, I suggest that you at least examine the drive's SMART report. Look for reallocated, pending, or uncorrectable sectors.

HD Sentinel (DOS / Windows / Linux):
http://www.hdsentinel.com/

HDDScan for Windows:
http://hddscan.com/

See this article for SMART info:
http://en.wikipedia.org/wiki/S.M.A.R.T.

If you do decide to swap the MCU, you should try to backup its ROM's contents before doing so. One way to do this is to via an MHDD script. You'll need version 4.5 of MHDD, and you'll need to configure your SATA controller for IDE legacy or compatibility mode in your BIOS setup. I can give you more information if you can get to this point.

BTW, have you confirmed that your power supply is OK?

[Zvuqniki]

First of all, I'm on a laptop, and this was an external HDD.
I have one working Seagate 1TB HDD, and one non-working WesternDigital 1TB HDD, which I'm trying to get the data from. I have only one external housing.
If I use the Seagate HDD on the external case, it works perfectly. When I use the WesternDigital HDD it doesn't work. Therefore does that mean that the external case is not at fault?
I dismantled and threw away the original Western Digital external case that housed the hdd.

Now, when I connect the Western Digital HDD to the external case, and turn the case ON, nothing happens. Not a beep from the HDD. The case isn't even recognized by windows as an external device.
If I use the spare circuit board that I ordered on the Western Digital HDD, and plug it to the same external case, when I turn it on, the drive starts working and starts spinning, BUT it is not recognized by Windows as an external HDD.

So what are my next steps, before changing the MCU?

I don't have a desktop computer here at home, but I have one at work, if needed for future steps.

[fzabkar]

It sounds like you may be describing a Power Up In Standby (PUIS) issue.

Does your original WD PCB have a PM2 (PUIS) jumper? If not, then perhaps the drive is configured for PUIS in the firmware. If PUIS is enabled, then you could attach the drive to a SATA port on your desktop PC, and then use HDAT2 (http://www.hdat2.com) to disable PUIS. To wake up the drive, you would need to start HDAT2 with the /w switch, as follows:

hdat2 /w

If the drive then spins up in your external enclosure, this would suggest that the USB-SATA bridge board in your original WD enclosure was responsible for spinning up the drive.

[harddriverecovery]

That is some pretty freaky action on this drive. I would say that you're probably pretty lucky you got that first 2.5 gig.

[Keatah]

Ok, this is what happened. I got the replacement circuit board, and put it on the hdd. Used the external case, powered it up, and voila, the drive spins up perfectly.
HOWEVER, it is not being recognized! I guess I have to swap some chip on the circuit boards.

But listen to this. I thought, what the hell, I'll try the old non-functioning circuit board again. To my surprise, the drive spun up, and started working perfectly. I started to copy the data, but then it died at 2.5GB. (I still need some 38~ to go, to get the important stuff (family pictures) out).
Then I did the switch again. Used the replacement board, spun up the drive, then powered it down, and then used the original. I got another ~800MB out.

What do you guys think is wrong? Any steps that I should take?

EDIT: Ok, I tried it a third time, using the replacement board to spin it up, and then trying the original board to get it to work, but now it won't power up at all.

I have a feeling the problem is a lot simpler than what we're discussing. And I think we'll get lucky with this.

I repaired 4 drives with this exact same problem (at least the symptoms match) by reflowing ALL the connections on the board with a heat pencil. In my case the tipoff was a slight dull appearance of the solder.

The replacing and swapping of the boards is making intermittent connections on the faulty PCB sometimes work. I had observed these precise exact identical symptoms on the drives I repaired. I had blindly reflowed 3 PCB's and these 3 drives are in the field working just fine. On number 4 I got curious and took my time. I had used a microscope and dental pick to look at the solder connections. And far too many of them were "soft". The solder was almost like a thick paste that was never heated fully. All it takes is ONE intermittent connection and your drive is out of commission!

Now, I'm taking a guess here, based on what I read, that this is the issue. And if the reflow job is done carefully and gently, there's no harm done to future recovery operations. Unless you screw up the reflow and make all kinds of shorts. Then it's game over!

[Keatah]

I would suggest a competent tech do this for you. If you've never soldered these exact chips before, then I wouldn't make this your first try.

Be aware that soldering is a skill, that, while, anyone can do it, only the best of the best techs have a feel for it and do it correctly. And then there's standards to follow if you want the job to last in the face of bumps and vibrations and temperature fluctuations. All of which are present in the HDD environment.

Again, I'm recommending a re-flow only, no chip changes no other complicated and risky stuff. If this doesn't correct the issue, then we'd look into the problem further.

Good luck!

[Keatah]

Sorry to make so many posts ladies and gents. But I feel compelled in this case, I looked at your PCB numbers and your drive label, and your drive was seemingly made at the same time my other 4 drives were! The date code is less than a week away! And the PCB numbers are far too close for it to be anything else.

Your drive's PCB was probably even made on the same exact assembly line, by the same equipment, with the same solder paste lot.

Most telling of all, the serial number is a few hundred counts away from my other 4 drives.

[Keatah]

.
.
.
.

[JWCC]

LOL, that explains a lot. Probably even the same person doing 'quality control' on your 6 drive's boards?

In my cases, the drives had corrupted firmwares. It wasn't the software as testing SE on MHDD worked first thing on a test drive on the exact same controller chipsets. GIGO - the d*** drives themselves were borked, likely by someone trying to do something they're not supposed to (messed with firmware).

So, what's the progress? Did the original poster ever get all their data off? It certainly does sound like a case of the 'flakeys' with those intermittent issues. "Wiggle the cable... OMG it works!" "Press against the board... OMG it works!" "Detach and reattach all harnesses... OMG the car runs!" (see, it works not just for HDD's. ) Bad solder joints and all that jazz.