Toshiba DT01ACA100 won't spin after power surge

[gameboybin]

I have this Toshiba Drive that came in with 19V power surge (Someone plugged in wrong 19V power to 12V rail).
The drive came in with fried PCB ( after swapping the PCB i realised the drive is knocking also)

- So 1st step - I changed the PCB, swapping the ROM from the patient over. After PCB swap, the heads starts knocking after powering on. So i guessed the heads are gone too

- Step 2: i proceeded to change the heads
HERE WHERE STRANGE THINGS STARTED TO HAPPENED

I swapped totall of 5 pair of heads (believing it is the donor compatibility problem).
- 1st pair has bad head 1. The HDD initialized fine. And been able to read all head 0 data, SA, and backup
- i proceed to swap the both working heads donor. However after swapping. The drive starts knocking twice then spins down. I thought it is platter damage or not compatible donor.
- but after 3 sets of new heads. New donors. ( I did check the surface - it is very clean without any damages) i excluded that the problem is with the compatibility
- I started to suspect it is problem with ROM, RAM, FW of some sorts
- I swapped to the 5th pair of heads. Now with same exact match as the patient. The drive still have old symptopms. Knocked twice then stopped.
- I proceed to Turn off head 1 in NV-RAM. And the drive started up spinning fine with full ID ( head 0 is readable through the whole drive)

I now suspect that is there is something wrong with ROM, RAM, translator, FW problem

If someone has encountered this problem before. Please help me out!
Many thanks!

Patient MODULES+RAM
https://drive.google.com/file/d/1sDCwnxghPQNlr_lIoRPoQv8_ycsQIpIT/view?usp=sharing

PATIENT drive

2 DONORS with only head 0 working heads

USED HEAD DONORS

PC3000 after disabling head 1 in NV-RAM

[fzabkar]

It seems strange to me that an overvoltage on the 12V rail is now showing up as a problem with head #1.

Normally, the only damage would be a 12V TVS diode and the associated fuse.

https://recuperodatos.com/sites/default/files/webform/donantes/04107-PCB.jpg

What was damaged on the patient PCB? Is it possible that the drive was connected internally when it initially failed, and was it then connected externally with the wrong adaptor?

[gameboybin]

It seems strange to me that an overvoltage on the 12V rail is now showing up as a problem with head #1.

Normally, the only damage would be a 12V TVS diode and the associated fuse.

https://recuperodatos.com/sites/default/files/webform/donantes/04107-PCB.jpg

What was damaged on the patient PCB? Is it possible that the drive was connected internally when it initially failed, and was it then connected externally with the wrong adaptor?

It was connected externally many times with the wrong adapter. up until it fried these 2 diodes i marked them in white. I haven't checked more on the PCB

[fzabkar]

Presumably the larger diode was the damaged one. The smaller diode would have been OK. There would also have been an open fuse.

It seems to me that the solution was to remove the large diode and replace, or bridge, the fuse, assuming that the short was cleared.

FWIW, I notice this model came with several different PCBs -- OA90352, OA90377, OA90381.

[gameboybin]

Presumably the larger diode was the damaged one. The smaller diode would have been OK. There would also have been an open fuse.

It seems to me that the solution was to remove the large diode and replace, or bridge, the fuse, assuming that the short was cleared.

FWIW, I notice this model came with several different PCBs -- OA90352, OA90377, OA90381.

I already got through PCB problem. By swapping ROM to the donor board.
I am now having problems with the head 1 not wanting to read or ready the drive.

The head 0 already got all the data out, and backed up all modules

[fzabkar]

Have you tried the patient's heads in the donor that had the "same exact match"? BTW, I only mentioned the PCB because I'm wondering whether you introduced some kind of weird fault or incompatibility issue.

[gameboybin]

Have you tried the patient's heads in the donor that had the "same exact match"? BTW, I only mentioned the PCB because I'm wondering whether you introduced some kind of weird fault or incompatibility issue.

Yeah! I havent tried that

[pepe]

i vote for servo corruption on side 1 due to preamp damage...

[gameboybin]

i vote for servo corruption on side 1 due to preamp damage...

Anything that can be done with it??

[fzabkar]

What kind of pathway would result in damage to the preamp that affected head #1 but not head #0? Moreover, since the preamp's supply voltages are derived from the 5V rail, how would an overvoltage on the 12V rail find its way into the 5V side of the motor controller?

[gameboybin]

What kind of pathway would result in damage to the preamp that affected head #1 but not head #0? Moreover, since the preamp's supply voltages are derived from the 5V rail, how would an overvoltage on the 12V rail find its way into the 5V side of the motor controller?

What i think this is some kind of RAM damage or module damage over time. Because the patient drive was not just plugged in with 19V once, but many times.

[pepe]

What kind of pathway would result in damage to the preamp that affected head #1 but not head #0? Moreover, since the preamp's supply voltages are derived from the 5V rail, how would an overvoltage on the 12V rail find its way into the 5V side of the motor controller?

drive was knocking after pcb replacement, so i assume preamp was hit. How it happened exactly is another question.
My suspect is that it was used as an external drive and someone plugged a laptop supply in (this is the most trivial way of giving a hdd 19V). In this case the 5V rail is also regulated from the incoming power (12V normally) on the bridge board. OP wrote it worked a few times before things took a bad turn. I suspect the bridge's regulator was also burnt which resulted an OV on the 5V line straight to the preamp. I saw quite a few cases where a burnt preamp caused uncontrolled write current to some of the heads, resulting in similar damage. It is usually easy to confirm with a scope.

could OP confirm my suspects (bridge board damage)?

pepe

[fzabkar]

The OP didn't find a shorted 5V TVS diode or open fuse on the 5V rail. I would repair the patient PCB and measure the supply voltages. That wouldn't take very long.

[pepe]

these are not sensitive to neither pcb nor heads, so it should not click like that. But clicking after pcb swap clearly points towards preamp damage. and with a damaged preamp anything could happen, this is why i always advised checking preamp after a suspected OV, without trying to power the drive with a replacement pcb.
There's one thing worth noting, namely that there are 2 types of pcbs these drives came with: OA90377 and OA90381, the main difference being the motor controller. However, i still don't think this is the issue, as the drive cloned just fine with H0.

another possibility is that the overvoltage did not kill the preamp, but only triggered uncontrolled write current on channel 1, damaging the content on that surface. This could be checked by putting the orig head stack back and disabling H1. I would also check the patient MHA in a donor, where it cannot do significant harm (disabling one head at a time).

[pepe]

... and this pcb is another wonderful example of genius engineering where the fuse protects the tvs instead of the circuit...
they seem to have corrected it on the newer pcb (90381) though.

[fzabkar]

It seems to me that all is OK.

This is OA90377:

TVS_diodes_fuses.jpg (136.41 KiB) Viewed 67 times

OA90377:
https://s.turbifycdn.com/aah/yhst-14437584971410/91711039-19.gif

OA90381:
https://s.turbifycdn.com/aah/yhst-14437584971410/91711049-18.gif

OA90352:
https://s.turbifycdn.com/aah/yhst-44463692235217/90352-13.gif

[pepe]

nope, the circuit is connected to the 3 vias next to the connector terminals on both the 5V and the 12V rail. i have just removed the fuses on one of these and the tvs remained unconnected while those 2x3 vias were still connected to the motor controller.

[fzabkar]

OK, I see. In earlier versions it was the grounded end of the diode that was fused. That's so stupid.