Switch to full style
Discussions related to DFL tools: general chat, firmware requests
Post a reply


July 24th, 2018, 18:09

Hi, I had a drive WD5000LPVT-24G33T1 and it was not detected normally
I fixed it and did a slow read fix.

It gave sector access to user area, i could also do a media test on DDI and image but slowly.

Suddenly it stopped giving data. Full detection but all sectors logically bad.

04 was corrupt and not readable or writeable.

Tried to move it around the SA but no luck
again it would read just a 1% on head 0 and varied percentages on head1

I tried to read module partially by changing its length in the Hex view of dfl and i could see the header and module

So I edited the 01 and made its length to 400 and saved it.

Now I can image the hdd.

I hope some one finds this useful.


Re: WD5000LPVT-24G33T1

July 24th, 2018, 22:28

Nice job, @sin. Thanks for sharing!

Re: WD5000LPVT-24G33T1

September 13th, 2018, 16:40

Welcome Larry. I love your pic...here and on DFL propriety forum too! :D

Re: WD5000LPVT-24G33T1

September 13th, 2018, 17:14

@sin, I confess that I don't understand why your method worked. I also don't understand the function of module 04h.

BTW, does a length of "400" correspond to 0x400 bytes, ie 2 sectors?

Re: WD5000LPVT-24G33T1

September 15th, 2018, 18:37

Hey there. I believe it is in sectors...there was one drive with translator module not reading at all...
Luckily it read tracks and i got the P-list.
Using this P-list in a donor, i regenerated the translator and took a copy of it in a folder.
Now where ever i moved the translator on the patient drive, it made changes to mod01 about re locating the module however the translator never opened up and read till a few percentage and dfl displayed module read error.
So now in the hex tab of DFL, i tried to punch in the aba of this newly shifted module and tried to read only 10 sectors off it, it opened up gracefully. Now I tried to write something there in the same length.
It wrote. Power off the drive, switch in on and tried to read the same aba of the relocated translator. It showed the data (junk) that i wrote before powering off.

So now this meant the module read partially.

So i read the module to the maximum extent it could..Like increasing from 10 sectors to some % of the over all length of module, it read gracefully.
So I wrote the newly regenerated translator and it wrote it...

Power off and power on....
now the mod didnt read by ID in the module checking tab. Neither it help check the copy on 0 or 1

So to reconfirm, in the hex view i punched in the translator aba number only to see the royl header existed with perfect checksum and then I accessed the module in the hex view by punching in the same aba and length that i used to write this translator as discussed before...
It read.

So now it reads only a few % of the whole module length....hmmm...what can i do?

So I edited mod 01 and edited the length of the translator mod till where it could possibly read...

saved the changes and power off, power on the drive.

Now the translator simply read by double clicking the drive and data was readable and showed read ok and checksum ok while i made the utility check it

Lol Squeeze theorem in calculus ? may be on similar lines...

this means the mod checksum actually may not depend on the overall readable/writeable length of the module but just depends on the content of mod...0 filled area fails to help in checksum calculation possibly

hope the case is explained to. If you find any ambiguity in language, please highlight that area.

Last edited by sin on September 15th, 2018, 18:44, edited 1 time in total.

Re: WD5000LPVT-24G33T1

September 15th, 2018, 18:42

so as to the question why mod 04 caused all bad sectors...i dont have a clue...isnt it backup of zone table ie backup of 03 in 04?

Re: WD5000LPVT-24G33T1

September 16th, 2018, 0:25

Here is a discussion about the importance (or non-importance) of module 04h:

WD Passport issue with module 04-family models configuration:

The differences of opinion don't inspire much confidence. :-(

I examined mod 04h in the following resource dump:

http://files.hddguru.com/download/PC-3000-UDMA%20Support/WDC%20Marvell%20family%20utility/Hubble%20LT/WDC%20WD5000LPVT-24G33T1-02-01A02-WD-WXJ1AA3P1527.ZIP (45.43 MB)

I have worked out the structure of this module, but I confess that I don't understand the meaning of the data. For example, there is a large section where each word has odd parity, the MS bit being the parity bit. FWIW, Toshiba's Glists also use word parity.

Based on your experiments, I suspect that the above example could be shrunk to a size of 0x5600 bytes (0x2B sectors) without affecting data accessibility.

Re: WD5000LPVT-24G33T1

September 16th, 2018, 12:46

Hi there my lovely friend.
I also do not have the ability to understand the true nature of 04...eventually i believe when my experience grows more....
However, this method works confirmed when one gets a drive with poor translator readability ie when the mod does not fully read but can only be read partially.

It was done on one JPVX recently....However let me try to find a similar lpvx and destroy 04 to see what happens....may be it turns all sectors logically bad may be not....
However, the translator which fails to be fully read can be shrunk beautifully and the drive will give data confirmed.


Re: WD5000LPVT-24G33T1

September 16th, 2018, 17:12

sin wrote:let me try to find a similar lpvx and destroy 04 to see what happens....

I would change the module's ID in mod 01, eg from 04 to FF04. This will tell you whether the drive can work without it.

Re: WD5000LPVT-24G33T1

September 16th, 2018, 21:30

Module 04 appears to be structured as follows:

offset          description
0 - 2F          ROYL header
100 - 5FF       2 blocks of 60 x 4-byte records, 1 block per head
600 - 55FF      data block #1 (0x5000 bytes, several sections)
5600 - A5FF     data block #2 (0x5000 bytes, several sections)
A600 - F5FF     next 0x5000-byte data block

The 0x5000-byte data blocks all appear to have the same structure, with similar data, suggesting that only the first block may be required. I have examined other examples of mod 04 where the majority of the space is "noise" (eg remnants of other firmware modules), so this would support the idea that the full module contents are not important. That said, there is no information in the header which would tell us the number of valid data blocks.

Re: WD5000LPVT-24G33T1

September 18th, 2018, 12:34

Thankyou Sir, Great info. So the technique worked as it read header + some blocks ahead of it...


Re: WD5000LPVT-24G33T1

September 30th, 2018, 9:23

Module 04 is the format DCM and you shouldn't need that module for normal data access / working of the drive.

Module 04 is only important for HEAD CUT, etc .... Just try it out ...

Get a Marvell based drive, delete 04, power off/on and check if data is still accessible.


Unless of course your drive is "special" ....

Re: WD5000LPVT-24G33T1

October 16th, 2018, 16:39

IDk how special...but it showed all sectors as logically bad without having functional 04....will try it again and report if I get one more such drive

Re: WD5000LPVT-24G33T1

October 16th, 2018, 17:25

Ok, thanks.
Post a reply