ST4000NM0165 ROM problem

[n3dv3d]

I didn't expect that you would find any voltages. Instead, I was expecting you to measure resistances.

Both e-fuses have latched into a fault state and switched off their outputs. This suggests that there are shorts downstream. There appear to be two MOSFETs between the e-fuses and the motor controller, so I was wondering if there was a short after the MOSFETs.

Sorry, I had quickly read your previously post.

Below the resistances between ground and each of the voltage test points.

Thanks for your time

Attachments

[SWM]

I understand that the MOSFET? component marking is needed. Maybe fzabkar can identify it.

[fzabkar]

I don't know how those MOSFETs are connected or what they do. Fortunately (?) I'm involved in a Reddit thread with exactly the same PCB. Perhaps I can ask the redditor to provide the same measurements for comparison purposes.

[fzabkar]

This is that Reddit thread:

https://old.reddit.com/r/datarecovery/comments/1mukcvu/hdd_wont_spin_up/

The user measures "OL" at each of the MOSFETs, presumably on the 200 ohms range. This suggests that your 0.3 ohm reading is a problem.

[fzabkar]

I'm wondering whether the MOSFETs are configured as ideal diodes, and I wonder if they are OR-ing the 5V and 12V rails for the purpose of charging the backup capacitors. I think the answer to the problem will involve determining where those 0.3 ohm pins go. Do they end up at the motor controller or the backup capacitors? Do they connect to the e-fuses or the 5V or 12V SATA inputs?

[SWM]

If you look at the photo, then 0.3 Ohm goes to the component "mosfet?" SOP8. I asked about its marking - it is not visible from the photo. I would have already unsoldered it and checked if 0.3 Ohm remained. But for this it is desirable to know what it is.

Maybe I am wrong - it is not possible to determine from the photo where the tracks go.

[n3dv3d]

If you look at the photo, then 0.3 Ohm goes to the component "mosfet?" SOP8. I asked about its marking - it is not visible from the photo. I would have already unsoldered it and checked if 0.3 Ohm remained. But for this it is desirable to know what it is.

Maybe I am wrong - it is not possible to determine from the photo where the tracks go.

The component is marked P2040US.
Could it be this one?
https://www.alldatasheet.com/datasheet- ... SS-13.html

Leaving the PCB fault aside for now – why doesn’t the drive spin up after moving the ROM to a donor PCB? Could it be because of different RAM?

Thanks

[SWM]

Quoting my previous answer:
"If the ROM was re-installed from 2 Mbit to 1 Mbit - most likely bb."
After all, I still don't know which of the boards has 2Mbit RAM...

[fzabkar]

Can you measure the voltage test points on your donor? I feel that the problem is a simple one, if we can understand what the MOSFETs are doing. The DMP2040USS datasheet seems to be the correct one.

[n3dv3d]

Can you measure the voltage test points on your donor? I feel that the problem is a simple one, if we can understand what the MOSFETs are doing. The DMP2040USS datasheet seems to be the correct one.

Hi,

all values seem to be pretty similar.
The big differences seems to:

Attachments

[fzabkar]

Can you measure the voltages?

[n3dv3d]

Can you measure the voltages?

Sure:

Attachments

[fzabkar]

I think the second MOSFET is acting as an ideal diode between the 5V and 12V supplies. I suspect that the 12V output goes to the backup capacitors. During a power loss event, if the 12V rail decays faster than the 5V rail, the 5V rail then powers the backup capacitors. I still don't understand how the smaller MOSFETs fit in, though. Nor do I understand how they find a path to ground.

[fzabkar]

Can you measure the resistances between the GDS pins of each of the 3 MOSFETs?

[n3dv3d]

Can you measure the resistances between the GDS pins of each of the 3 MOSFETs?

Hi, below the results ( highlighted differences with donor PCB ):

P2040US

G <-> S = 3K
D <-> S = 133K donor pcb= 264k
D <-> G = 128K donor pcb= 267k

Si1413DH

G <-> S = 0,6M
D <-> S = 6,3K donor pcb= 387k
D <-> G = 0,6M

Si1413DH ( near sata connector )

G <-> S = 375k
D <-> S = 326k
D <-> G = 387k

[fzabkar]

Sorry, I have no idea what is going on.

[SWM]

Can I see a photo of the patient's board from the other side?

[n3dv3d]

Sorry, I have no idea what is going on.

Thanks again for your time

Quoting my previous answer:
"If the ROM was re-installed from 2 Mbit to 1 Mbit - most likely bb."
After all, I still don't know which of the boards has 2Mbit RAM...

OK, but in your opinion, why does the initialization fail when import the patient adaptives into the donor ROM? Is the RAM check integrated into adaptives?

Can I see a photo of the patient's board from the other side?

Sure, in attached.

Attachments

[SWM]

The back side of the board is a ground plane and control points. Nothing can be tracked...
The donor's FW ROM version does not match the FW disk version. The main program is on the disk.

[SWM]

There are no easy ways. Either supply from a laboratory power supply 5V with a limit of 1A at 0.3 Ohm and watch what heats up or smokes... Or reball the RAM. Or an identical donor board.