Hi folks, looking for advice on recovery options for a Western Digital My Passport (2TB) USB drive. A colleague sent this drive to me.

Drive details:

Model: WD20NMVW-11W68S0
Date: 28 SEP 2012
PCB: 2060-771801-002 REV A

Symptoms and behaviour, from start of drive being plugged in (behaviour is the same, tested on both Windows 7 as well as FreeBSD):

1. Access LED is steady for about half a second
2. Drive spins up
3. Access LED blinks (LED on/off interval ~0.25s)
4. After about 10-12 seconds, drive emits a single click + audible "chirp" (sounds like actuator being forcefully reset or set to landing zone)
5. Access LED goes steady again
6. Access LED blinks (same interval as above)
7. After about 45 full seconds or so, #4 happens, rinse lather repeat #5-7. The access LED interval also sometimes varies.

However: between steps #3 and #4, I/O to the drive does appear to work. On Windows the drive pops up a warning about wanting to run CHKDSK (no way in hell am I doing that!) and a drive letter is assigned (not going to dare access that). I'm able to get full SMART attributes (provided below) during this time as well; ATA IDENTIFY commands as well work just fine (drive capacity, capabilities, etc.). But after #4, all I/O to the drive essentially times out, and eventually begins returning full SCSI sense data indicating a hardware issue (vendor-specific ASC/ASCQ).

The SMART data I can understand fully (trust me -- been doing this part of it for about 15 years now, quite familiar with ATA protocol as well) and see absolutely no anomalies, so I'm left thinking there is something more nefarious going on. (BTW, the SMART selective LBA self-test that didn't complete was due to me attempting to induce said SMART test; eventually something somewhere went into power save mode and hence the test aborted after several minutes. Also, the "SMART Status command failed" error is normal for smartmontools 6.4 via a USB/SAT layer; it's not indicative of a problem).

My gut feeling is that the drive locks up or begins misbehaving when trying to access some particular LBA (possibly outside the normal accessible region, i.e. internally used by WD for whatever) -- the fact IDENTIFY and SMART data can be obtained briefly, to me, is proof of that -- but then again I'm used to doing data recovery on PATA, SATA, and SCSI drives; this USB-only interface nonsense is making my troubleshooting efforts very painful.

I understand it's possible to wire up SATA XMT/RCV +/- to solder points E71 / E72 / E73 / E75, but this drive uses a JMicron JMS538S so odds are the data is AES-encrypted thus wiring that up wouldn't do me any good. And yes, I've read the full PDF documenting circumvention and decryption, but there really aren't any tools out there for getting at any of the relevant data that don't require $3500-$10000 data recovery hardware/equipment.

I'm not sure if a PCB replacement (or SATA PCB replacement) + ROM swap would be at all useful.

Looking for advice on what to do.

If the advice is "get a DR company to deal with it", I'm fine with that, but after reviewing several of them the costs tend to be in the mid-to-high 3-digit US$ range, which is outside of said colleague's acceptable range (hence why he sent the drive to me). If there are folks here who do DR on these drives, are located in the US or Canada, and are willing to cut me a serious discount (I'm talking US$150-200 or so -- I can provide a target drive with same or larger LBA count), I'd love to hear from you.

Thank you!

Sounds like a directory problem, or bad sector 0, or possibly a firmware issue.

Clone the drive and then run a file recovery program.

_________________
http://www.datasaversllc.com

Try cloning the drive with ddrescue. Test it first by reading a few sectors at random places -- start, middle, end.

_________________
A backup a day keeps DR away.

Couple things:

1. LBA 0 can't be bad -- Windows is able to detect a filesystem, which means it was able to read the MBR or GPT (not sure which scheme is used -- see #2 below).

2. Any program trying to do I/O to the drive past the point of it locking up will not work. As described, any I/O past that point times out, then shortly after returns in an immediate I/O error with SCSI sense code data indicating hardware failure (I'd have to pull down the exact ASC/ASCQ but they're vendor-specific). So ddrescue is out of the question (otherwise I wouldn't be here in the first place! ).

So drive cloning or similar things (dd, etc.) aren't possible given what the drive does. Again: I/O to the drive is only possible for about 10-12 seconds, up until it makes a click + "chirp" noise -- after that it goes catatonic and the only way to bring it back is to power-cycle it (a bus reset doesn't help). I don't like power-cycling drives repeatedly (I shouldn't have to explain why, you guys understand).

Questions that are lingering:

1. If the problem is with firmware, are we talking about what's on the U12 or U14 EEPROM (and which one)? Are these flashable using some kind of serial or TTL interface to the drive itself? Is this procedure documented? Where would I find a replacement firmware (other than a donor PCB)?

2. If I was to get a SATA-based donor PCB, does anyone know if the AES encryption used by these drives is decodable (by software) at this point given the insecure nature of the implementation, or is the "get a MyBook PCB and hook that up in-line" process still considered the go to? I'm referring to the process described here: http://www.hddrecovery.ca/blog/slow-pas ... -recovery/

3. A continuation of #2: does anyone have a list of the MyBook PCBs which can be used for this? I keep finding people talking about using them, but nobody seems to mention where to get them or what exact products are compatible. Yes I'm aware WD probably uses different ICs across different MyBook products and revisions, so I'd need to find one that uses the JMS538S, but it'd be good to have PCB model strings listed somewhere... was hoping someone might know where.

I found a post here -- http://mybroadband.co.za/vb/showthread. ... ners/page2 -- that outlines basically the PCB in the WD5000H1U might be a compatible PCB using hte JMS538S, but I can't really find anything useful on Google/etc. trying to track one of these down. :/ (This is very frustrating)

4. If the answer to #2 is "no, nothing's available right now", does anyone have documentation further explaining the problem described here, section "Managing Firmware Problems"? http://www.hddrecovery.ca/blog/slow-pas ... -recovery/ -- this is the stuff that I haven't spent any time with. The document references "using tools from {companies}" but said software is accompanied by hardware that ranges from something like $3500 and up -- way, way outside of my budget, obviously.

Ok, quick reply.

For what you stated i would try a PCB swap to start with (with ROM chip transference) just to be sure that the problem is not caused by some sort of power issue at the PCB level or any failing component on the PCB.

To check the firmware you will need WD firmware tool and a SATA compatible PCB. Firmware tools might be expensive. Cheapest available option might be WDMarvell. In any case you need to convert PCB from USB only to SATA so that you can bypass the USB bridge and send the vendor specific commands to the drive (by the use of the firmware tool).

Maybe a hardware based cloning tool might help as it will send soft-reset, hard-reset and power off-on the drive as needed. If this issue only happen when one particular sector / LBA is read you might get lucky with expensive hardware cloning tool like DDI4.

1 - Firmware is on the PLATTERS mainly as modules and there is some unique to the drive adaptives on ROM U12 that you need to copy to the new SATA/USB PCB if you do a pcb swap. ROM U12 is unique to the drive, contains adaptives and ROM code that must match overlay - module 11 on platters so the drive will properly work. U14 deals with USB bridge aspects as encryption.... You can't write to the firmware by TTL. You need proper tools.

2 - Perfectly decoded ! No problem ! You just need a PC-3000 Unit with Data Extractor. Can handle the decryption on those drives without any problem at all. Cost of the tool is more or less $6000 USD so i guess a little bit too much for a single drive. MRT Pro tool can also deal with it and decrypt on the fly and it's half the price of PC-3000.

3 - List of PCBs that are SATA and compatible with USB only : - http://www.hddoracle.com/viewtopic.php?f=7&t=1080 - you will still need to swap ROM to access USER DATA (encrypted). You can buy PCBs out of ebay.

4 - Slow response problem is a firmware issue that causes the WD drive to respond slowly with very low access speed and transference rate. Meaning that the drive will take huge amount of time to clone and even to copy a single file out of it. This is because of some issues on some firmware of WD drives not dealing with sectors that are pending relocation properly. It's solved by deleting the re-lo list (list of sectors pending relocation) and disabling relocation. It can be done with firmware tools and even with some free scripts and tools :

http://www.hddoracle.com/viewtopic.php?f=86&t=848

But on your case the drive just looses I/O instead of been slow. So it's a different issue.

Now i would advise you to start by reading this :

http://www.hddoracle.com/viewtopic.php?f=56&t=176

If you want to understand firmware better read this :

Newbie info, from and for newbies. About firmware, SA, etc

For advantages of hardware cloning read here :

Hardware assisted cloning of a drive with Bad Sectors :

For firmware tools that will work on WD like WD Marvell go here :

http://www.hddoracle.com/viewforum.php?f=146

To understand more about modules of firmware on WD drives go here :

http://www.hddoracle.com/viewtopic.php?f=86&t=1071

some examples :

PLATTER MODULES :





ROM PCB Modules :



Regards and good luck !

_________________
1Q9xrDTzTddUXeJAFRn37aqh1Yr6buDCdw - (Bitcoin Donations)
paypal.me/Spildit - (PayPal Donations)
The HDD Oracle - Platform for OPEN research on Data Recovery.

^^ Spildits definition of "quick" differs from most

lol

_________________
1Q9xrDTzTddUXeJAFRn37aqh1Yr6buDCdw - (Bitcoin Donations)
paypal.me/Spildit - (PayPal Donations)
The HDD Oracle - Platform for OPEN research on Data Recovery.

Hi folks, thought I'd post a current status update on this recovery project.

* I have two PCBs coming in from Hong Kong and China respectively:

2060-771801-002 PCB (USB)
2060-771960-000 PCB (SATA)

The first is a replacement PCB for the drive itself (I verified all the silkscreenings on the chips match so there won't be any spindle motor oddities etc.). I'll need to swap U12 (easy) and U14 (hard; damn surface mount stuff is so tiny!).

The reason I bought this board is because I don't want to risk damaging/tinkering with the original, and because there's a possibility the original board is misbehaving/bad in some way. Worth $25 to try.

The second PCB is obviously the SATA interface version of the drive PCB, and I will need to transfer the U12 chip from the original PCB.

I'm thankful the U12 chip is easier to transfer; lots of space around the PCB makes the desoldering/soldering easier. But that U14 chip... ugh.

* I purchased a WD My Book Essential 2.0 (P/N WD5000H1U-00) off eBay. I planned to use this in combination with the SATA interface PCB to do decryption, as a post I found on a South African forum stated it used a JMS538S chip. I received the product today and found it contain an Initio INIC-1608L chip. Total waste of $35...

Does anyone know where I could reliably get my hands on one of these SATA-USB adapter boards with a JMS538S chip? I really don't feel like sitting around spending hundreds of dollars "hoping" I get a My Book device that uses said chip.

All of this is so incredibly frustrating. Needless to say I won't be buying (or recommending) any of WD's USB 3.0 products ever again; better choice is just to get a 2.5" SATA-USB enclosure that uses a good/reliable chip w/out encryption (ex. ASM1153E) and buy your own drive.

Anyway...

If replacing the original (USB-based) PCB doesn't work, I'll try the SATA interface version instead and then give WDMarvel a try. Thank you SO MUCH for the links and utilities!

Really bad advice since this is a USB drive with all the issues that a USB drive brings to the table.

_________________
Buy your friends Toshiba\Hitachi and your enemies Seagate.

More bad advice, the PCB is working but the drive has either a weak/bad head, classic firmware issue, or bad sectors.

_________________
Buy your friends Toshiba\Hitachi and your enemies Seagate.

Total waste of money. Your more likely going to kill the drive completely in your DIY attempts.

_________________
Buy your friends Toshiba\Hitachi and your enemies Seagate.

agreed



It is very unlikely a PCB problem, haven't seen a pcb fault in passport series so far.

What your colleague said happens to drive ?
was it dropped ?

_________________
Data Recovery Pakistan

The drive was not dropped or mishandled. If the drive has a physical problem then it's a loss -- there's nothing I can do about it (I do not open drives under any circumstance) and my colleague can't afford a data recovery company (I've gotten several quotes from different companies, all estimate US$650 or higher). So I'm his best bet, barring physical damage.

So, let's summarize your options
in case of:
1) bad/weak heads, you think you gonna pull off a head replacement with no experience ?
2) of hardware cloning needed, you gonna invest in that kind of things ? (cost a lot more than 650$)
3) firmware problems, do you have such tools in your disposal ?

bottom line, the time and efforts you gonna invest, is gonna be a lot more than 650$ how do the americans say time is money, not to mention if you'll need firmware tools and hardware tools Etc. .

Hard drives are not lawnmowers, washing machines, or coffee makers. They are incredibly complicated, precision devices.

You simply don't have the knowledge, equipment, or experience to pull this off. It's nothing personal . . .

_________________
http://www.datasaversllc.com

proper diagnosis is required before proceed for any DIY.
if drive has physical problem than DIY is not possible, sorry.

i can assist remotely to do some basic checks to verify if problem is with heads or not. let me know if you are interested.

_________________
Data Recovery Pakistan

Any update on this ?

Did you recieve the PCBs and test them out ?

_________________
1Q9xrDTzTddUXeJAFRn37aqh1Yr6buDCdw - (Bitcoin Donations)
paypal.me/Spildit - (PayPal Donations)
The HDD Oracle - Platform for OPEN research on Data Recovery.

The PCBs arrived this afternoon.

The 2060-771960-000 (SATA) board I received has me concerned. It uses a slightly different CPU model (Marvell 88i9446-NDB2) than the original board (Marvell 88i9346-TFJ2).

The 2060-771801-002 (USB) board I received uses an Initio INIC-3608PN USB controller, not JMicron JMS538S, so I cannot use this board.

Instead of MOVING the ROM i would COPY the ROM content with something like this from the USB only to the SATA PCB :

http://www.hddoracle.com/viewtopic.php?f=117&t=865

Then i would check if the drive can be cloned. Of course it would be ENCRYPTED... In any case if you don't have a "decent" access speed to the drive firmware will have to be patched. Assuming the drive is detected as it should to start with.

_________________
1Q9xrDTzTddUXeJAFRn37aqh1Yr6buDCdw - (Bitcoin Donations)
paypal.me/Spildit - (PayPal Donations)
The HDD Oracle - Platform for OPEN research on Data Recovery.

UPS ......

Your original board is ...

PCB: 2060-771801-002 REV A

That one ins compatible with 771823 and NOT with 2060-771960-000

http://www.hddoracle.com/viewtopic.php?f=7&t=1080

2060-771960-000 could have been used for instance to replace an USB ONLY 771961.

YOU CAN'T USE NEITHER OF THE PCBs...


Sorry.

_________________
1Q9xrDTzTddUXeJAFRn37aqh1Yr6buDCdw - (Bitcoin Donations)
paypal.me/Spildit - (PayPal Donations)
The HDD Oracle - Platform for OPEN research on Data Recovery.