An unusual case. Already submitted to Ace TS, but we are all drawing blanks. If there are any suggestions from others, then they are welcome.

Patient HDD: IC35L040AVVN07-0
Patient PCB Sticker: 07N8979 H69144A
When powered on HDD enters BSY. No spin up. No short on 12V or 5V.
The customer reported the HDD failed after a power outage.

I found a matching donor:
Donor HDD: IC35L040AVVN07-0
Donor PCB Sticker: 07N8979 H69144A


1) The patient NV-RAM was transferred to the Donor PCB. Patient HDD was then powered on with Donor PCB + Patient NV-RAM. HDD does not spin up. HDD reports RDY. It is possible to read the (patient) NV-RAM in PC3000. The contents looks normal.
2) Using the donor PCB with donor ROM and donor NV-RAM, mounted on the patient HDD, then the HDD does spin up enters RDY. But then the normal error is reported when drive ID or other command is sent to the HDD.
3) Finally, I transferred the patient ROM and the patient NV-RAM to the donor PCB. When this happens the original fault (no spin up and BSY) is replicated.
4) I then repeated the process with a second donor PCB just in case the first donor PCB was faulty. The results were replicated.
5) Inspection inside the HDD has taken place. No signs of contamination of damage to the platter.

I have spoken to someone else at a long time data recovery company today. Curiously he has been working an identical case, but a "PLAT" series IBM/Hitachi, which again came in after the customer reported a power outage.

Any thoughts?

Thanks,
John

_________________
CDR - Manchester Data Recovery Services
0161 408 4857
http://www.cheadledatarecovery.co.uk/

Read the full ROM chip (not just NVRAM) and post a copy here. I had a case a while back where the ROM was corrupted, but with a couple days of analysis in HEX I was able to rebuild it. I can probably do it again. There's usually two copies in the ROM and with some analyzis comparing the two copies as well as comparing other similar ROMs, it may be possible to figure out the flipped bits and get it correct again.

_________________
Data Medics - Hard Drive, SSD, and RAID Data Recovery Service Company

These have ROM and Nvram in separate chips. I miss the nvram image of the patient you read, did you? Rom is one thing, the worse thing is nvram which contains sa cyl.
However, you may run nvram recovery in pc3k, i think it should be supported.

pepe

_________________
Adatmentés - Data recovery

Patient ROM is corrupt.

Donor ROMs are incompatible with patient NVRAM.

Could that be the answer?

_________________
A backup a day keeps DR away.

What about patient PCB + donor ROM + patient NVRAM? Would there be any point to that???

_________________
A backup a day keeps DR away.

When the donor ROM (U6) + patient NVRAM (U5) are on the patient PCB I get the same result as when they are mounted on the donor PCB. That is the HDD will enter RDY with no spin up. I can then read the (donor) ROM and Patient NVRAM within PC3000. This indicates that the patient PCB is likely to have good functionality with the exception of the ROM and potentially the NV-RAM. Likewise, if I mount donor ROM and donor NVRAM on patient PCB, the patient PCB can be used normally with the donor HDD.

At this stage whilst it has been possible to read the patient NVRAM (see attached in .zip) via PC3000, it has not been possible to read the patient ROM chip. Whenever this is mounted on a PCB it causes the PCB to enter BSY even when jumpers are set to safe mode. I have dismounted the ROM (U6) and I am trying to read with an external reader. However, the chip used is a LE25FV051T (datasheet attached) and is not supported by either external reader I have available (RevelProg or MiniPro). According to the pinout in the datasheet this chip uses quite a different ordering to most "25" chips. Trying to re-order the wiring to match the LE25FV051T pinout, using a 3.3V supply does not allow reading this chip. I have a support request in with RevelProg to see if they can add support or make a suggestion.

As Fzabkar suggests, perhaps there is donor ROM incompatibility. For reference, I have taken a photograph of the patient and donor HDDs and PCBs. The results are replicated using a second donor HDD, which only has a match on the first line of the PCB sticker.

@Pepe - To use the NV-RAM recovery it is necessary to get the HDD to spin up. The only method to do this is using the Donor ROM and Donor NVRAM chips mounted. The HDD spins, enters RDY, but gives an Abort error on any command - as would be expected given donor ROM & NV-RAM. Using the NVRAM recovery option the process runs (for hours) with 'lights' flashing on the Status Register suggesting the SA is being read. But by the end it reports failure and no adjustments are made to donor NV-RAM mounted on the PCB.

I'm still waiting on Ace TS to reply. Allegedly this has gone to the 'developers'. I'm not overly optimistic about getting a response from them.

John

_________________
CDR - Manchester Data Recovery Services
0161 408 4857
http://www.cheadledatarecovery.co.uk/

Just a small nitpick ...

When you withdraw the flex cable from the connector, you should retract the two locking tabs on each side.

_________________
A backup a day keeps DR away.

AFAICT, these text strings relate to the ROM/NVRAM versions:

Patient NVRAM


Donor NVRAM


Donor ROM

_________________
A backup a day keeps DR away.

Can you piggy-back the patient ROM on top of the donor ROM, and switch the data pins between one and the other with a suitable SPDT or DPDT switch?

You could POST with the donor ROM, wait for the donor to come ready, then switch to the patient ROM and use PC3000 to dump it. That's assuming that PC3K dumps the actual ROM rather than its image in RAM.


Edit: Maybe it would be sufficient to isolate each /Reset pin and switch one of them to ground to disable the inactive ROM?

_________________
A backup a day keeps DR away.

There's a couple of things that needs to match, check in pc3k manual, i won't write it here.
what does patient and donor pcb do with patient rom and no nvram?

no point in mixing rom and nvram of different fw version coz they won't work together, at least not just as they are.

btw, just noticed patient did not get ready, which suggests to me that the rom might be corrupt, as Franc said.
And starting to see your problem around reading and writing that rom
Sorry, i am too slow already.

sent pm...

pepe

_________________
Adatmentés - Data recovery

After a lot of attempts, it has been possible to resolve the issue. Big thanks to @pepe for his assistance on this. I cannot give too much detail, but in summary, there was a fault with the ROM. The patient ROM was unreadable and external chip readers did not support it.

It was necessary to make alterations to the patient NV-RAM, and using donor ROM and modules to write to patient SA. However, it was a final 'surprise' step which allowed access to user data areas & drive ID. The surprise being, that the change made in theory should not have made a difference to LBA access, but it did. This HDD is definitely being put in the "unusual" cases, and a good example of having to be methodical in the approach to recovery.

Once the disk was reading it was possible to read 100% LBA - not a single bad sector.

Thank you for everyone's input and suggestions.

Regards,
John

_________________
CDR - Manchester Data Recovery Services
0161 408 4857
http://www.cheadledatarecovery.co.uk/

I don't understand why you didn't just flash the ROM with a working copy. Can PC3000 flash the ROM?

_________________
A backup a day keeps DR away.

no, it can't write rom on old ibm arch.
Your idea seemed viable at first but since pc3k cannot write it we have had to find some other solution.

pepe

_________________
Adatmentés - Data recovery