Seagate SSHD 2.5 ST2000LX001 over-voltage repair questions

[jEnoch]

Hello all

I got a new power supply (EVGA Supernova) installed in my system and got my power cords confused and ended up using my old PSU (Corsair HX) SATA cords to power my hard drive with all my data on it. This definitely screwed it up. Unfortunately I hadn't figured out a backup solution and my data is all on that disk. I do know more or less what happened though. I checked with a multimeter and checked some pinout diagrams from pc-mods (which may or may not be completely accurate) on the power supplies and found the following:

This is a pc-mods pinout diagram for the voltages of the SATA plug in the Corsair HX PSU. I verified that the SATA power connector end matches the voltages on the diagram with a multimeter.

Corsair SATA pinout voltages

Here is a pinout diagram for the voltages coming out of the EVGA PSU SATA plug. This made me question the accuracy as my no pin is in the opposite corner but the voltages are right.

EVGA SATA pinout voltages

In essence what ended up happening was this voltage was applied to the HDD:

I have been doing some research although I am pretty much a noob at this stuff. I heard that sometimes there is over-voltage protection in place in the form of a TVS Diode or other circuitry. So I took it apart and looked at the PCB and at first glance there are no obvious scorch marks but some "test points" look a little dark, and then there is what looks like a burn on a chip I am unfamiliar with.

PCB:

Perhaps a TVS Diode or something?:

Burn mark that looks suspect...

I haven't contacted Seagate because I know for a fact this drive did not come with DR and I am sure that my bad installation falls out of warranty anyway. I just wanna try to get my data off of it without spending $500+ if I can. Thank you if you read this and double thank you if you spend the time to reply. Any help is greatly appreciated!

[fzabkar]

It looks like the overvoltage got past the TVS diode and went on to damage the SMOOTH motor controller. Usually when this happens the preamp on the headstack is clobbered as well.

[Lardman]

It doesn't look promising. You could try a board & rom swap just in case but chances are you're going to need a professional to also swap out the heads. Kudos for your fault finding but the burn out the smooth chip is not a good sign.

[jEnoch]

Thanks for the replies!

It looks like the overvoltage got past the TVS diode and went on to damage the SMOOTH motor controller. Usually when this happens the preamp on the headstack is clobbered as well.

Is there any way to check for this?

It doesn't look promising. You could try a board & rom swap just in case but chances are you're going to need a professional to also swap out the heads. Kudos for your fault finding but the burn out the smooth chip is not a good sign.

Is swapping out the heads an expensive service to hire?

Thanks again you guys I'm grateful for help

[fzabkar]

I think this job will be more complicated than most.

Firstly, the "ROM" (serial flash memory) chip is a BGA type, so transferring it to a donor PCB would be a delicate operation.

The NAND flash (Toshiba TC58TEG6DDLBA0M) on the donor PCB would need to be initialised.

If you want me to help you with the preamp testing, you will need a multimeter and some patience. That's because I haven't yet analysed this particular PCB.

The SMOOTH chip generates the onboard power supplies, so I would testing for shorts to ground at the relevant test points.

Can you measure the resistance of the two capacitors adjacent to the 3.3V regulator near the NAND flash?

Can you measure the resistance between ground and each of the Vx, Vy, V1, V2, V3 test points adjacent to the SMOOTH controller?

Can you identify the markings on the two mystery ICs which I have labelled as "IC ??"?

Attachments

ROM.jpg (39.76 KiB) Viewed 11207 times

HDA_3V3_reg.jpg (59.7 KiB) Viewed 11207 times

Regs.jpg (99.43 KiB) Viewed 11207 times

[sin]

I doubt a board swap with BGA rom swap would work.

I guess you would also need to swap the " Solid State " part of the hard drive.

In fact, i have seen this drive would spin with a DONOR rom of a non-sshd drive (rosewood with same firmware)
+ Unique rom mods of patient patched into the donor. Yet, it would yield subfile errors.
Also, the length of RAP is different for a SSHD and non-sshd counterpart..


--
PS: Apprently there was some seminar today by ACE which suggested they have a solution for LX series. I Couldn't attend it as I was keeping busy.

[jEnoch]

This is sounding like it might not be feasible for someone like me, but since this is so far within my power to do, might as well make the measurements!

Can you measure the resistance of the two capacitors adjacent to the 3.3V regulator near the NAND flash?

Can you measure the resistance between ground and each of the Vx, Vy, V1, V2, V3 test points adjacent to the SMOOTH controller?

Can you identify the markings on the two mystery ICs which I have labelled as "IC ??"?

Here are the measurements:

HDA_3V3_reg.jpg (63.66 KiB) Viewed 11126 times

Regs.jpg (105.66 KiB) Viewed 11126 times

Thanks again, does this look like I measured correctly?

I doubt a board swap with BGA rom swap would work.

I guess you would also need to swap the " Solid State " part of the hard drive.

Yes I was wondering about the the solid state part. Thanks for the info!

[fzabkar]

V1 appears to be shorted to ground. Hopefully that's not the supply for the ROM.

The 10F20 component is actually a Schottky diode. I believe it is part of the circuit which generates a negative supply voltage for the preamp.

Did you confirm whether the 5V TVS diode was shorted? If so, just snip its pins with flush cutters. That will prevent it from influencing the other resistance measurements. I suspect that the shorted capacitor adjacent to the 3.3V LDO is probably connected to the 5V supply.

Your next step is to identify which of the pins on the HDA connector are ground and supply pins. To this end you should test for continuity between the HDA connector pins and the Ground and +5V pins at the SATA power connector. You should also test for continuity with the V1, V2, V3 test points, plus the anode (non-striped end of the Schottky diode). In this way you will identify which power supplies are required by the preamp.

Once you have done this, you will need to measure the resistance between the corresponding pins on the HDA. This will tell us whether the preamp is shorted or otherwise damaged.


NVLUS4C12N, ON Semiconductor, MOSFET, N-Channel, 30V, 10.7A, marking AGx, UDFN6:
https://www.onsemi.com/pub/Collateral/NVLUS4C12N-D.PDF

NSR10F20, ON Semiconductor, Schottky Barrier Diode, 20V, 1A, marking 10F20, 152AD:
https://www.onsemi.com/pub/Collateral/NSR10F20-D.PDF