WDC WD10S21X-24R1BT0-SSHD-8GB

[MasterT]

I have WDC WD10S21X-24R1BT0-SSHD-8GB drive, Big BearH 2.5" family, basically it's a WD SSHD, taken out of a PS3 console.

It's the first one that I receive this model for recovery and has other problem than Slow Responding or heads.

Drive is always BSY. I don't hear motor spinning.

I blocked SA access in ROM.

I noticed drive gets ready if i disconnect pcb with NAND otherwise always BSY. So i guess NAND or it's content is bad.

I uploaded dir into HDD RAM from SA and I get this:

SA dir reading (ID)..................... : Ok
Module 02 uploading..................... : Ok

Drive identification reading ........... : Ok

SA dir reading (ID)..................... : Ok
Servo overlay........................... : LDR already loaded

ID is not ok, as it does not show serial, capacity etc.

Drive starts spinning and stops short while after loading ldr.

I discovered that I can read all modules by ID and made full backup. Motor starts when i try read modules and stops after read is done.

Unfortunately I cannot write modules. I get "Unknown error code: 3745". So can't change anything.

If I allow SA access in ROM and NAND pcb disconnected, drive never gets ready, and motor never spins.

Anyone has experience with this model?

Ace TS said they don't know of any way to clear nand like in seagate or turn it off like in toshiba.

Anyone tried to swap NAND pcb only? How the drive behaved? (considering that the donor's pcb NAND will have donor data inside).

I don't have a donor for this model so any tips would be much appreciated prior to buying one.

Any ideeas are welcome. Thanks.

[MasterT]

View ROM information
ROM Data size........................... : 512 Kb
Firmware version........................ : 0003002T
Heads configuration..................... : by map
Heads number............................ : 4
Heads number in use..................... : 4
Switched off heads...................... : No
Heads map............................... : 0,1,2,3

ROM Modules
------------------------------
ID : Description
------------------------------
0A : Heads map
0B : Flash ROM dir
20B : Flash ROM dir (ext)
30 : SA Translator
47 : SA Adaptives
0D : Flash configuration
4F : Microprogram version

SA Regions
+---+-----+-----------+---------+
| N | Blk | ABA | Size |
+---+-----+-----------+---------+
| 0 | | 0 | 300 288 |
| 1 | | 400 384 | 300 288 |
| 2 | | 800 768 | 300 288 |
| 3 | | 1 201 152 | 300 288 |
+---+-----+-----------+---------+

Micro Jogs
------------------------------
LH : PH : Micro Jog (Hex/Dec)
------------------------------
0 : 0 : 00/0
1 : 1 : 00/0
2 : 2 : 00/0
3 : 3 : 00/0

Attachments

ROM-2021-08-20-22-57-24.rar

(266.06 KiB) Downloaded 1011 times

[MasterT]

And some pictures:

[fzabkar]

FWIW, I would think that an enhanced secure erase, or maybe just a secure erase, would clear the donor's NAND cache.

[fzabkar]

If U1 on the daughter board is a NAND flash IC, then why does it need a crystal (Y1)? Why does the board have 4 (or 5?) power supplies?

The large array of capacitors would suggest that these are for power loss data protection.

Is the same MCU used in non-SSHD applications?

[MasterT]

Close-up picture:

[MasterT]

Cannot find any info about these codes, but as it states and considering the power supplies I guess this an SSD not just a NAND.

[MasterT]

More photos:

[unknown]

Anyone tried to swap NAND pcb only?

Yes I did, and it worked.

[pepe]

I guess this an SSD not just a NAND.

yes, it interfaces to the other pcb through 2 differential lanes, which turn out to be a SATA interface. It connects to a 88SM9642, which appears to be a port multiplier. So the drive should work perfectly without the SSD board, perhaps with some modifications...

pepe

[fzabkar]

This dual drive needs a "driver" so that both drives (SSD and HDD) can be seen by the OS. Otherwise only the SSD is visible since it occupies the first position.

https://community.wd.com/t/disappointed/16880/3

Aside from modifying the ROM to disable the SSD, I suspect that one could simply tap into the HDD's SATA lanes to access the HDD directly.

[fzabkar]

Aside from modifying the ROM to disable the SSD ...

Sorry, this doesn't make sense. The ROM wouldn't see the SSD since it exists on a port multiplier. I think only a hardware modification would make sense. To this end we would need to see the other side of 2060-771980.

[fzabkar]

ISTM that a tool such as PC3K or MRT should be able to select any drive behind a SATA port multiplier simply by setting the PM Port bits in the first dword of each FIS. These words are reserved in the original SATA spec and default to 0. This means that port 0 is the power-on default. The SATA Port Multiplier spec assigns 4 of these reserved bits to a PM port number (0 - 15).

See page 10 of the first document, and page 183 of the second.

Serial ATA II Specification - Port Multiplier:
https://www.seagate.com/support/disc/manuals/sata/sata_im.pdf

Serial ATA II Specification Port Multiplier:
https://www.t10.org/ftp/t10/document.03/03-157r1.pdf

In the present case the SSD is on SATA port 0 and the HDD is on port 1.

[MasterT]

Other side of 2060-771980.

[MasterT]

Full Firmware Backup:

WDC WD10S21X-24R1BT0-SSHD-8GB.7z

(5.85 MiB) Downloaded 1245 times

[pepe]

Sorry, this doesn't make sense. The ROM wouldn't see the SSD since it exists on a port multiplier. I think only a hardware modification would make sense. To this end we would need to see the other side of 2060-771980.

yeah, i wanted to write the same just didn't click on the submit btn.

The drive starts but gets stuck in bsy state if i removed the SSD board. However, attaching it live makes it get ready.
Doing some more tests...
pepe

[pepe]

Now this is pretty odd. The above pic shows a 88SE9171, which is a PCIe-SATA controller, with one sata port.
Just noticed this is a BigBearH, what i have is a BigBear, with a different design, having a 88SM9642 on it, which is a port multiplier indeed.
So this drive must have considerably different design than mine and the guy at https://community.wd.com/t/disappointed/16880/3 .

I can't read the part number of the main controller on the pic, it must be something like 88i944x-xxx
mine is 88i9446-BRL2.
The SSD board is almost the same on mine, 771925-504, layout is similar but the CAPs are not populated. The SSD chip is the same, so it is a SATA intf that goes to the PCIe-SATA controller. Which is probably interfaced to the MCU with the PCIe lane and more integrated into the HDD, probably not accessible as a separate device from the outside.
And as such, the FW must be aware of the SSD part.
One could attach a prefilled sata drive on the SSD port as a cheat, i suppose...

pepe

pepe

[MasterT]

Controller is Marvell 88i9441-BRL2:

WhatsApp Image 2021-08-26 at 19.08.14.jpeg (69.23 KiB) Viewed 33149 times

[pepe]

yeah, so it is different, as i expected.
can you check where the main sata port is routed? I suppose it is to the 88i9441.
(it is routed to the 88SM9642 as a main difference)

pls also check routing of the SSD sata lines, i suppose it goes to pins 21-22 and 24-25 of 88SE9171
and i would also make sure that the PCIe signals of 88SE9171 are routed to 88i9441 but i think it will be hard to check due to the burried traces.

https://www.marvell.com/content/dam/marvell/en/public-collateral/storage/marvell-storage-88se9171-datasheet-2015-11.pdf

Anyway, the cheating can work on this one too...

pepe

[fzabkar]

FYI, the 771984 PCB is shown in this review:

[url]https://www.eteknix.com/western-digital-black²-dual-drive-review/2/[/url]