Hello every one,
I was tinkering around a JMVW. Following is the problem definition:-

Problem Definition:-

Take backup of a working Fblite of the entire SA and ROM by every way. Trying to use a donor rom of the same microcode and trying to ID the drive normally

Steps-Round 1

1) Fblite with normal operation was connected and SA+Rom was backed up nicely and zipped to preserve.
2) Donor rom code of same uC was flashed over the existing hdd.
3) Load ATA.



Observation-Round 1:-

Drive spins normally, DRD DSC. The moment after DRD DSC when auto get detailed button is clicked, the drive starts making high frequency scratching like noise (sssssssss)
No capacity and no ID, no mod access.

None of the loaders upload. Shows failed to upload.
Loading 01(original) in RAM and list fails to give SA access as the light status changes from DRD DSC to DRD DSC DRQ.

Steps-Round 2
1)Get DRD DSC, go in kernel so SA is not sought. Bypass SA in ROM by manipulating mod 30 and reboot.
2)Faster DRD DSC, Auto get details.
3)Load ATA.

Observation-Round 2:-
1)DRD DSC status is achieved faster.
2)ATA LOADED nicely..all mods 11 etc etc.

Loaded 01(original) in ram and tried to access SA but same status flags are lit.
Tried to read tracks. No luck..

Steps-Round 3
1)Flash Donor ROM code again to neutralize all changes done previously.
2)One by one patch each and every original rom module to see where the drive IDs.

Observation-Round 3:-

It was Module 4F that is solely required to get access to SA. Drive IDed normally after 4F is put. SA access is independent of every other module (depending unless you wish to bypass 30).
The moment original 4F is patched, there is all christmas but when donor 4F is used, the drive loses its ID.
With original 4F drive is not at all noisy. Non original 4F causes noise in such drives SYMPTOM TO BE NOTED.


Steps-Round 4
1)Get drive to ID normally by using original modules.
2)Analyze original 4F and eliminate blocks of data by zero filling and generating checksum and writing it back to locate important portions which enable the heads to read the data.

Observation-Round 4:-
The required region is located in the displayed offset as highlighted in the pictures.
Now the noises are a little different as each bit is zero filled from right to left one by one and the ID is checked. As it tends to the threshold digit, the SA access becomes a slower process as the adaptive are lost. If the threshold digit is changed, ID is lost.

Conclusion:-:-
The highlighted area in the edited 4F photograph displays the threshold point. Any change made beyond this point has less effect on the SA. Any changes made before this point or at this point can disallow SA access.



Thanks.

Could you upload the BIN file for module 0x4F? ISTM that, rather than deleting bytes at random, we should try to make sense of any groups of data. For example, I can see that there are several sections consisting of 4 records, where each record is associated with a particular head.

_________________
A backup a day keeps DR away.

Here is the original 4F and donor 4F.

I had already started analysing a 4F module (see attachment). ICBW, but these are the groups as I see them.






I don't know what the following data mean. 0x55AA and 0x80AA, 0x90AA appear to be signatures of some kind.

_________________
A backup a day keeps DR away.

@sin, I would compare your 0x4F modules against those from models with different numbers of heads. That should shed more light on the module's structure.

_________________
A backup a day keeps DR away.

4F module from WD20NMVW-11AV3S0

_________________
A backup a day keeps DR away.

4F module from WD2002FAEX-00MJRA0

_________________
A backup a day keeps DR away.

This observations seems contrary to expectations. AIUI, this last section of the module contains the adaptives for the dual stage actuator. :?

https://forum.acelaboratory.com/viewtopic.php?t=9854
viewtopic.php?f=1&t=35523

_________________
A backup a day keeps DR away.

... or this ?

_________________
A backup a day keeps DR away.

Have been playing a bit with this today as I had a spare half hour.

Two drives of same fw version and 1 day apart in manufacture date, non-native 4F prevents drive from working.

When I get some more time, I’ll try and figure out which are the most important sections of it. But it may take a lot of trial and error!

_________________
PC Image Data Recovery
http://www.pcimage.co.uk

New!! HDD-PCB.COM for all your PCB and donor HDD requirements!

Assuming my observations are correct, I would limit any edits to the primary head. I would select one particular adaptive value and then adjust it up and down until the drive fails to ID. This will tell you the sensitivity (tolerance) of each adaptive. There may be better approaches, though.

_________________
A backup a day keeps DR away.

If the dual-stage actuator is the reason for the extra adaptives, then could the problem of a missing ROM be solved with donor heads and donor head adaptives?

_________________
A backup a day keeps DR away.

Hi there sorry for the delay in reply. You have done amazing job of dissection of 4F. Great amount of my love and respect for that.



Possible, the conclusion can go wrong. Hence I need more people to try out chopping all the things in 4F but just staying consistent with the last portion and further comment on functionality.




Great info thanks.
--

I have exactly the same thing in my mind. If the 4F is locked to one set of physical heads. May be swapping heads with an appropriately matched drive and using the donor 4F maybe could locate the servo tracks and start to read the SA?

Feels like buying two Fblites and trying out swapping heads.....
The major problem is, when some one delibrately overwrites the ROM code with a donor without having original patient rom backup.
Has any one got success cracking such cases? It will be inspirational personally.


--

Very well said.

--

So what I will try tomorrow is, I shall bypass the SA in rom and just keeping head 0 or head 1 engaged while only manipulating the values as marked by you.
Thanks for the idea.

--

It might be interesting to see how a drive behaves if we could disable the piezo actuator, assuming this is the problem. I'm wondering whether disabling this second stage actuator would still allow the drive's primary actuator to follow the track. Perhaps this would be a better approach than feeding incorrect adaptives to the secondary actuator.

That said, the HDA connector doesn't appear to have any aditional pins (18), so I am at a loss to understand how the piezoelectric transducers are driven. I would expect that the motor controller would generate a driving voltage, and that this single voltage source would be multiplexed to the selected head via an IC on the headstack. Alternatively, it could be that all PZTs are activated at the same time.

Perhaps a photo of the preamp would help?

_________________
A backup a day keeps DR away.

I think my suggestion may have been ambiguous. I'm suggesting that you confine your edits to head #0, leaving all other heads unchanged.

BTW, could you identify the SA module that MRT uses to rebuild 4F?

_________________
A backup a day keeps DR away.

.




Aye Aye Captain. I shall try this and post my findings here.It has to be multiplexing, they can multiplex a voltage level surely, but to drive a whole set of coil to move something around, they may require some amount of current too as just voltage won't just do enough. So may be at the same time,when the voltage gets multiplexed, the current is feed by the servo IC(voltage controlled current source which is taking its voltage input feed from the multiplexed voltage. Once voltage detected, due current is pushed from the servo IC to move the coils??? Who knows?
I think a great deal of multiplexing is done to read from so many heads one by one while keeping the pin count constant to company mfg standards. However, multiplexing of a current line instead? sounds quite out of my experience tbh.
Or is it that very small current is required to drive this coil and the required current can be pulled from this multiplexed voltage source itself.
no idea.

Was just caught up with the conclusion of 7th semester exams and hence could not reply in time.

I have a friends NMVW which he just give away for a small amount...I haven't tested it for uniqueness as we are currently imaging the drive so that he can move his data to his bigger drive.
Let me see if the uniqueness applies here too.

Let me see if I can send you a snap of heads etc.

But why do I have a feeling (or lack of experience possibly) that there are no secondary set of heads.


--
Arctic Monkeys - Arabella

Maybe I'm mangling the terminology, but I'm referring to head #0 as the primary head while all others are secondary.

_________________
A backup a day keeps DR away.