WD5000AAKX-001CA0 slow response and now corrupted firmware

[facuarmo]

Hello there, a couple days ago I decided I could be the hero of the day and started to poke around this forum for sofware and examples to get the drives I had laying around with software issues back to life through both the serial terminal and ATA commands which worked out just fine for both of my Samsung drivers. However, I tried to do the same with a WD drive and I really got deep into breaking it.

I went ahead and downloaded both WDMarvel23 and WDR-Demo. I couldn't do much with WDR-Demo other than doing the translators rewrite and a logic test (which revealed a lot of slow sectors), however, WDMarvel23 let me dig pretty deep on the drive's firmware. At some point, I found out how to test each of the modules and realized that MOD24 and MOD6F had invalid CRCs, so I just looked up at HDDGuru's files and picked up a firmware from the same model (having this exact filename, to be specific: 00030010 WD5000AAKX-001CA0-15-01H15-WD-WCAYUR041974.7z). I flashed these modules without any luck, the interface was still super slow and I couldn't get the drive to read any of its files properly (not that it's important though, I'm trying to repair it to the point that it works again just for fun and to learn more about it). Tired of trying common stuff, I just went ahead and decided to flash ROM.bin from this firmware backup and FW Restart the drive.

Here goes the fun part: now the drive has nothing defined (no tracks, no passport, no ATA driver). It's simply there, spinning and doing nothing, the only good thing about it is that I can still access it through the ATA port, it's just that it does none of the functions that would be relevant to a user (like, actually working as a hard drive), but the interface still does its proper functions like allowing to write, read and erase the flash ROM. I thought that WDMarvel23 could be broken so I tried flashing the same ROM file through WDR-Demo, but it's exactly the same result.

My question is: is there anything I could do to at least get it back to its old behavior?

[facuarmo]

Alright people, so a little bit of an update about this issue I've been experiencing...

I spent a couple hours playing around with WD Marvel 2.3 and realized how to do the whole thing properly, so if anyone else is having trouble with the same sort of stuff, you can take the following steps:

• Download this firmware backup: https://files.hddguru.com/download/Non- ... D776233.7z
  
  Why? Because it contains the directory structure expected by WD Marvel and it's compatible with it, so it won't cause corruption if used.
• Unpack it anywhere accessible.
• Switch the view from Menu to Buttons.
• Click Select folder and select the path where you unpacked the firmware backup.
• Click Download overlays in memory and select Download both overlay. Click Open for each of both files (it'll automatically open and close, and the path will be chosen from the root of the selected firmware backup).
• Click Write rom and select PZU\Flash.bin.
• WARNING: From now on, NONE of the data in the device will be recoverable, unless super expensive tools are used or you're smart enough to create your own. Whatever the case is, just take my word, it'll make recovery super hard, this is just an attempt to revive the drive for learning purposes.
• Click Entry and initialization DIR and load the automatically marked file (just hit Open).
• Click Working with modules and hit From (this is a horrible translation that means "load from the working directory"). Once the list updates, close the window.
• Hit Restart the firmware and, once it's done, just hit Soft reset. Hit Define and you should see the normal passport again (model, S/N, FW Ver., etc.).
• Now, we gotta record the tracks again. To do that, hit Working with tracks and then click Write from a folder, Tracks will be selected automatically, just accept it.
• Once again, hit Restart the firmware and, once it's done, just hit Soft reset.
• Click Translator and format and, from there, pick Translator regeneration without lists.
• Once the device is ready, finally, click Restart the firmware and, then, hit Soft reset.

That's it, the device should be working as normal again. However, I'm still unable of using it, since all sectors that I try to write to are presented with a AMNF message, these were readable before I started messing around with the drive, so it's not bad media.

Now, if someone can give me a hand with this specific last issue, please let me know.

[pepe]

you screwed it up quite a bit when you wrote foreing firmware to the platters. Original defect lists are gone, so now it needs some selftest to be run to format the user area again and rebuild the defect lists and the translator.

pepe

[fzabkar]

"MOD24 and MOD6F had invalid CRCs"

These modules commonly have bad CRCs, even on perfectly functioning drives. There was no need to fix this.

"the interface was still super slow"

WDMarvel has a workaround for data recovery purposes - "deal with slow responding". It modifies modules 02h and 32h, but is not a fix. That's probably all you needed to do to stabilize the drive.

Using the "firmware backup" from a donor drive means that your patient has inherited the defect lists, SMART data, etc from the donor. Now you will need to perform a "selfscan" to recertify your drive. There are some examples at hddoracle.com.

In short, you have made a mess of things.

[facuarmo]

"MOD24 and MOD6F had invalid CRCs"

These modules commonly have bad CRCs, even on perfectly functioning drives. There was no need to fix this.

"the interface was still super slow"

WDMarvel has a workaround for data recovery purposes - "deal with slow responding". It modifies modules 02h and 32h, but is not a fix. That's probably all you needed to do to stabilize the drive.

Using the "firmware backup" from a donor drive means that your patient has inherited the defect lists, SMART data, etc from the donor. Now you will need to perform a "selfscan" to recertify your drive. There are some examples at hddoracle.com.

In short, you have made a mess of things.

Sorry, but I can't seem to find that option, are there any chances that you could show me some pictures of the process? Also, how would I go about getting the drive to let me write again? I tried using the track format option but it says DISK_SPBA_ERROR for the first track and then DISK_DAM_ERROR for all the rest.

[fzabkar]

Sorry, I have never done a selfscan. However, there are numerous threads at hddoracle.com on this subject.

[facuarmo]

Sorry, I have never done a selfscan. However, there are numerous threads at hddoracle.com on this subject.

It's fine, the hint about trying a selfscan was pretty cool though. I left it running for about 12 hours already and it seems to be doing well, now I'm back at the Xmit Blink test but it's not going through cylinders, it's got stuck on cylinder 256 and it's running through its LBA. I suppose it's fine, just wondering if anyone could let me know about what's it exactly doing and for how long should I wait.

At some point, the HDD seemed to crash (was stuck on cylinder 256 of this last step, the Xmit Blink), doing a soft reset got it back up and it kept running of from where it was, however, it's still going and it's just at LBA 4029, I'm kinda worried that it'd take weeks to complete if it wants to go up to the max LBA.

[facuarmo]

Sorry, I have never done a selfscan. However, there are numerous threads at hddoracle.com on this subject.

It's fine, the hint about trying a selfscan was pretty cool though. I left it running for about 12 hours already and it seems to be doing well, now I'm back at the Xmit Blink test but it's not going through cylinders, it's got stuck on cylinder 256 and it's running through its LBA. I suppose it's fine, just wondering if anyone could let me know about what's it exactly doing and for how long should I wait.

At some point, the HDD seemed to crash (was stuck on cylinder 256 of this last step, the Xmit Blink), doing a soft reset got it back up and it kept running of from where it was, however, it's still going and it's just at LBA 4029, I'm kinda worried that it'd take weeks to complete if it wants to go up to the max LBA.

Just a quick update here, since my post has just been approved and a couple hours have passed already...

The HDD is done, I omitted the last step of the selfscan since it was Lifeguard Diagnostics and I didn't really care about it, these are the steps that did succeed:

• Scan and add defect to log
• Gain Cal
• Ts Read
• Cluster
• Tlist
• Test ID 3300 (unnamed)
• Test ID 3301 (unnamed)
• Test ID 3402 (unnamed)
• Plist

So far, hardware-side, everything looks pretty good, still, trying to read sectors just gives a timeout, even raising the timeouts to 1500 ms doesn't do it (I thought that maybe the sectors were really worn out, but it's not even trying), both heads tests return GOOD, is there anything else that could be done here?