ST4000LM024 head swap

[epeace2009]

Hi

I have this drive from 2020. Client reported that the drive had failed and stop working. After examining I found out that PCB was dead with few comoponets shorted and dead. I swapped PCB with donor same 100% and I have read ROM with external programmer XGECU.Put the ROM on donor and then I found out that heads are dead too. So I swapped head and disk initilze now. I have read SA even it reports me that head 4 is bad. Drive catch ready state but no UA access. I also unlocked ROM and blocked SA writings but still I get abort when access user area. Also I have read 3 files 3f.40,41 SED files and ther was a lot FFFF fields which I tried to fill with zeros and write back to drive but still cannot read user data.

Can you help me with this case?

Thanks

 ST4000LM024-2AN17V

FW: 0001

PReamp 8430 as on donor ad on patient.

Patient 2020 year, Donor 2018.

[pbzcbf...@gmail.com]

Which components? The failure mode may provide a clue.

https://recuperodatos.com/sites/default/files/webform/donantes/02826-PCB.jpg

[epeace2009]

What you mean?  I cannot access UA I get abort every time. In log file there is no error except standart SMP in Terminal I get this:

(P) SATA Reset

 SIM Error 1002 LBA 0000000000024366 FD FCFFF3FF
 RW Error 44090081
[MCMTW_DSW] Written to Primary
 RECOV Servo Op=00A5 Resp=0005IDLE1
 RECOV Servo Op=01A5 Resp=0005
 RECOV Servo Op=00A5 Resp=0005
 RECOV Servo Op=01A5 Resp=0005
 RECOV Servo Op=00A5 Resp=0005IDLE1
 RECOV Servo Op=0195 Resp=0005
 RECOV Servo Op=0295 Resp=0005
 RECOV Servo Op=0395 Resp=0005
 RECOV Servo Op=0495 Resp=0005
 RECOV Servo Op=0595 Resp=0005
 RECOV Servo Op=0695 Resp=0005
 RECOV Servo Op=0795 Resp=0005IDLE1IDLE1IDLE1IDLE1
 RECOV Servo Op=01A5 Resp=0005
 RECOV Servo Op=00A5 Resp=0005
 RECOV Servo Op=01A5 Resp=0005
 RECOV Servo Op=00A5 Resp=0005IDLE1IDLE1IDLE2
ASCII Diag mode

F3 T>
ASCII Diag mode

F3 T>
Input_Command_Error

F3 T>
TCC:0017 EMCDataErr:00000013

[epeace2009]

Here is marked on pcb which components are  shorted. I desoldered them 2 of them and they are dead short out of PCB.

[Alandata Recovery]

My Basic Seagate flow

Does anybody do different here? 

1 tech mode patch

2 try to get terminal

   Edit ROM for just h0 or H1

3 if have terminal

   Back up critical system files

3a ^L print preamp info

4 if terminal, then edit smp flags

   Turn off most everything

5 test ability to write sys files

   No write - no fixy... 

I'm probably forgetting some steps.... 

Any steroids you guys suggest?

If you can't access sa 0 or 1 then try head swap

If head swap doesn't help

Then it's platter and not recoverable. 

Seagate are POS

Cheap-ass thin protective platter coating fails quickly.

Moronic firmware retries drive to death instead of detecting damage and stopping to protect itself. 

--
Data Recovery Certification Group / for issue with google group please email sc...@myharddrivedied.com
---
You received this message because you are subscribed to the Google Groups "DataRecoveryCertification" group.
To unsubscribe from this group and stop receiving emails from it, send an email to datarecoverycertif...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/datarecoverycertification/f8edce53-c35e-4fe8-92a0-69a8a7c9bdc4n%40googlegroups.com.

[Alandata Recovery]

Why mess with the board at all? 

Just write ROM to good board. 

You have to be real careful when reading ROM that you have a good read. 

Sometimes pc3k reads wrong size. 

You only get half of the ROM. 

If you then rewrite ROM you corrupt it and lose critical adaptive.. . No recover at that point. 

[Alandata Recovery]

Looks like platter damage. 

Heads might help. 

Most likely it will die before recover

[epeace2009]

Hi Alandata

Heads are already replaced but now h4 terminal said that is bad. But disk initilize it self SA backed up no errors, ROM unlocked, Terminal available, write blocked, read and write SED sys files successful.

User data still cannot be read. When I try I get abort message!

So my asumption is that some module is damaged or not read good?

[epeace2009]

I have tried to read Translator file but it cannot read it in terminal there is error: LED:0x000000BD FAddr:0x0000900

[Alandata Recovery]

6 test defect lists

    v10, etc

    Read mcmt

Pc3k has a solution for led-cc

You can write modules? 

To view this discussion visit https://groups.google.com/d/msgid/datarecoverycertification/f30dc58e-4612-43e4-b264-7121f8f87d0dn%40googlegroups.com.

[pbzcbf...@gmail.com]

Usually when you have a head fault after a PCB failure, you would expect that the drive had been subjected to an overvoltage. Otherwise it would be strange to have a PCB failure at the same time as an unrelated head fault.

The components you have marked as shorted are mostly OK.

The large component nearest the SATA connector is an inductor. These normally measure close to 0 ohms.

The component marked 30Y is a 2 ohm resistor (EIA-96 marking code).

The capacitors adjacent to the MCU could be filtering the Vcore supply, in which case a resistance of less than 100 ohms would be normal.

[epeace2009]

I have tested defect list and it seams ok. Also SMART reading in terminal ok. I disabled in sys93 everything but still no user area access!?

What else I should try? Can I do translator regenaration?

Thanks

[Alandata Recovery]

Pc3k has read sector using utility mode. 

It disabled some error checks. 

Maybe it uses PBA.

It may give garbage. 

Does that work? 

--
Data Recovery Certification Group / for issue with google group please email sc...@myharddrivedied.com
---
You received this message because you are subscribed to the Google Groups "DataRecoveryCertification" group.
To unsubscribe from this group and stop receiving emails from it, send an email to datarecoverycertif...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/datarecoverycertification/1548e3ca-f8d0-4e28-b17c-d02ac3668762n%40googlegroups.com.

[epeace2009]

Yes it works using utility mode in sector edit. What should I do next?

[Alandata Recovery]

This sounds like the situation where the translator is bad. The symptom is that you can read fine up to a certain address, after that you can't read. What's happening is that when a sector is read the drive seeks to where the sector should be. Then it reads the sector and the meta data with that sector and verifies that the sector read is really the sector it expected. 

Utility ignores the address and ECC errors. You could get real data, partial Data, or wrong sector. 

Ace has manual and automatic repairs for this situation. 

Is sector 0 correct? 

To view this discussion visit https://groups.google.com/d/msgid/datarecoverycertification/733d71f4-166f-4432-9917-2086296895bfn%40googlegroups.com.

[Alandata Recovery]

This situation often occurs when amateurs follow Internet advice and use the serial port. 

They manually clear the alt, glist and Regen the translator. 

M0, 6,22..... 

I4, 22.... 

Maybe that's what happened before you got it? 

[epeace2009]

Hi Alandata

Noone before me open the disk or do something with the disk. Client said that the disk has folen on floor and stopped to work. But also USB connector was broken but luckly this disk is SATA and has small USB pcb as addon and SATA connector was ok. PCB of disk is dead so I read ROM and put it on sam donor PCB then when I powered the disk HSA was dead and I swapped the heads with donor.

The disk initilze now but no UA access. I did not regen translator or cleared P, G lists. I just unlock ROM and applied write protect on SA every time i start the disk .

When I tried to sector read I get "Abort" message and I tried to read using utility and I get reading of data but not useful. I tried in DE to read using utility but no still no data just rubish.

Please if you could help me to solve this problem and get data for client? I do not have to much experience with these drives.

Thank you

[epeace2009]

Yes I forgot to say that I cen read Translator modules 348 both copies but it seams for me that they are ok in hex view but not sure.

[Alandata Recovery]

Can you read the critical modules 

1b

28

35

93

348

Some others I can't recall. 

Try to get ready on just h0 or H1

When you read sector 0 with utility

It's it zeros or random

To view this discussion visit https://groups.google.com/d/msgid/datarecoverycertification/aa371645-304e-4ea1-8b5e-cccab70d7112n%40googlegroups.com.

[epeace2009]

Hi

Yes I can read all these sys files except 348 copy 1, copy 0 reading ok. 

Sector 0 in sector edit with using utility reads random?

[Data Recovery Guru]

This is a model that encrypts data through the native PCB, with keys residing in the MCU. This is why you get ABR errors, because the keys from the donor PCB don't match.

You need to either fix the native PCB and put back native ROM, or swap MCU to donor PCB (which is not as easy).

To view this discussion visit https://groups.google.com/d/msgid/datarecoverycertification/b4fbea55-36bc-498a-b06e-22e9db7142een%40googlegroups.com.

[epeace2009]

OK thank you labtech. I will try to fix patient pcb.

Thanks

[Data Recovery Guru]

Another potential problem at this point is that by testing with the donor PCB as you did, the keys from the donor MCU may have overwritten the keys in morules in the SA. Therefore, say you put back the native PCB with native ROM back on the drive, the drive may still return ABR errors.

If you can accomplish to get native PCB back in working condition, you will need help from somebody really smart in understanding how to fix this problem (like AceLab, DFL, MRT support, or a really good technician).

Document this model number and PCB code for future reference, so that testing with donor PCB is not a good idea. It is also important to check the drive arrived with native PCB, as somebody else may have already worked on the drive swapping PCBs, etc. 

To view this discussion visit https://groups.google.com/d/msgid/datarecoverycertification/221a8a96-df6c-4731-91de-66dc6bf36e80n%40googlegroups.com.

[epeace2009]

Hi

I finaly replaced components near power connector and PCB change light to Bsy but there is no info in terminal and when I measured current it starts 35mA and in a second turn to 0. I have tested working PCB and PCB without disk has 136mA current continous.

Can you help me what to check more on PCB why current drops to 0 in a second?

[epeace2009]

Also I check voltages on other components on PCB and compared to working they are same in ange 0.8V 1,7V and 4.9V.

[pbzcbf...@gmail.com]

Can you measure the marked voltages? Start with your donor and compare with the patient.

[pbzcbf...@gmail.com]

Which components? There was nothing wrong with those two components that you wrongly diagnosed as being "shorted".

On Tuesday, January 7, 2025 at 6:23:50 AM UTC+11 data...@ymail.com wrote:

I finaly replaced components near power connector ...

[epeace2009]

Here are measurments.

[pbzcbf...@gmail.com]

V2 seems shorted to ground. What do you measure at this point on the donor PCB?

The serial flash memory and SDRAM both appear to be 1.8V devices.

https://datasheet.ciiva.com/26786/k4t1g164qf-bce7-26786841.pdf (K4T1G164QF-BCE7)

https://www.onsemi.com/pdf/datasheet/le25s161-d.pdf (LE25S161)

[t...@desertdatarecovery.com]

Is that a cold solder joint the other side of V2 marking?

 

Tim Homer - Lead Engineer

Desert Data Recovery

t...@desertdatarecovery.com

www.desertdatarecovery.com

--

Data Recovery Certification Group / for issue with google group please email sc...@myharddrivedied.com
---
You received this message because you are subscribed to the Google Groups "DataRecoveryCertification" group.
To unsubscribe from this group and stop receiving emails from it, send an email to datarecoverycertif...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/datarecoverycertification/adab1737-03ab-4336-828f-7b71fae568d1n%40googlegroups.com.

[epeace2009]

On working pcb v2 VOLTAGE is 0.8V

[pbzcbf...@gmail.com]

There's your problem. If reflowing the solder doesn't fix it, then measure the resistance between the V2 test point and ground. If it is close to 0 ohms, then there is a short on that rail. This could correspond to the shorted capacitors adjacent to the MCU. There appear to be two Vcore supplies (0.8V and 0.88V).

If you confirm the existence of a short, then you need to check all those capacitors that are connected to this test point. You can locate them most easily on your donor by performing point-to-point resistance tests between V2 and each of the MCU capacitors.

[epeace2009]

I am sorry bad measure V2 has 0.8V on both points as on good board.

[pbzcbf...@gmail.com]

That photo is one I found on the Internet. Perhaps the OP should upload his own.

[pbzcbf...@gmail.com]

Is there any difference between the donor and patient in any of the voltages? I can't understand why the supply current drops to 0 if the voltages are present. :-?

[pbzcbf...@gmail.com]

If there is no difference in the voltages between donor and patient (other than the preamp supply), then check the 40MHz crystal oscillator near the MCU. If you can hear rattling when you shake the PCB near your ear, then the crystal has shattered.

[epeace2009]

Hi

I don know why current drops. I tried few times on PCB without disk and current drops to 0 in a second on correct working PCB it stays at 136mA.

Yes I check all voltages and compare them with working one and they are same.

[epeace2009]

I will check crystal later today and will report to you.

Regarding the photo of PCB that you put fzabkar, patient PCB is 100% same. There is no visible burnouts or solder melt or something else. I have examine PCB under microscope and could not see any difference with working one.

[pbzcbf...@gmail.com]

If your donor PCB's ROM is desoldered, examine how the terminal reports the missing ROM. If it reports something like "Seagate Boot ROM", then that proves that the MCU is not brain dead. Then do the same for the patient PCB.

[epeace2009]

ROM chip i did not desoldered at all niether on patient or donor PCB. Working PCB report Boot:xxxxxxxxx  

Patient PCB does not show anything and BSY flag is on but dimmer than normal. Before all flags were on.

[pbzcbf...@gmail.com]

I am trying to determine if the MCU is alive. If the ROM is missing, the MCU should report a "SEA Boot ROM" message in terminal. If it doesn't, then either the MCU is dead, or the crystal is not ticking, or the reset pin is being held low by the motor controller.

[pbzcbf...@gmail.com]

Something like this (for a different model):

SEA-3 Yeti Boot ROM 2.0 (12/06/2007)
Copyright Seagate 2007

[epeace2009]

This is message on normal working PCB.

Second pic patient PCB BSY dimmer and no message.

[pbzcbf...@gmail.com]

I'm confused. Are you seeing these messages when the ROM is removed or disabled? If yes, then the patient MCU must be bad, or the 40MHz crystal is bad, or the MCU is being held in the reset state.

This is a normal terminal:

https://groups.google.com/g/datarecoverycertification/c/kubMb2A6Of8

Boot 0x80MMD00440001    
Boot 0x80MMD00440001    
SpinUp
TCC:0023

[epeace2009]

Yes this is normal terminal on working pcb, but on patient i dont get anything nor disk sipn up when  i put it on disk.

Here is the picture of patient PCB

[pbzcbf...@gmail.com]

You don't appear to understand what I'm asking or why I'm asking it.

[epeace2009]

Sorry did not understand than. Please simplify me :)  Sorry<

[pbzcbf...@gmail.com]

It appears that you don't want to remove the ROM. Don't you have a backup? 

If you are prepared to accept a small level of risk, short the Data Out pin of your donor ROM on your donor PCB to ground and tell us what you see in terminal. This will prevent the MCU from trying to boot from the external ROM. Instead it will boot from its own internal SEA boot ROM. By doing this we eliminate a lot of external variables and narrow the scope of the problem.

https://ae01.alicdn.com/kf/Sfd909140f8fe44ff8f663490cc5de04ew.jpg

[epeace2009]

OK ROM chip removed no messages

[pbzcbf...@gmail.com]

If there are no messages on the donor PCB, and if you have selected the correct baud rate, then I don't know how to narrow down the problem without an oscilloscope. Sorry.

[epeace2009]

When shorted on donor PCB here is message:

Flash boot code Diital Signature Verification failure!
0x0020
0x0020
0x0020
0x0020
0x0020
SEA5 Boot ROM 1.0 (7/31/2015)
Copyright Seagate 2015

[pbzcbf...@gmail.com]

At last!  I think that your patient MCU, or one of its support components is bad. Or it could be that the motor controller is sending a power-on-reset signal to the MCU.

The easiest thing to try would be to swap the crystal. Or you could measure the DC voltage at each pin of the crystal and compare with your donor. I expect you'll see 0.9V (= 1.8V / 2).

[epeace2009]

Thank you. Please if you could be a litttle bit more polite and mark on picture crystal because I m not very good with smd electronic :)

[pbzcbf...@gmail.com]

Attached.

Measure Vx1 and Vx2 (voltages).

https://www.google.com/search?q=crystal+40MHz&tbs=isz%3Am%2Citp%3Aphoto&udm=2

[epeace2009]

Thank you again. I will try tomorrow  because it is late here. I will report my progress. :)

[epeace2009]

HI

U swapped crystak and both Vx1 and Vx2 and still the same.

Thanks

[pbzcbf...@gmail.com]

It looks like the MCU might be dead. An oscilloscope would be helpful, but I guess you don't have one.

[epeace2009]

Hi

No I don't have oscilloscope. The is returned to the client.

Thank you anyway.