Thursday, November 29, 2012

Investigating my 1TB drive, 2 months later

Earlier this month I finally got my USB 3.0 ExpressCard... card. I say finally because it took over a month to finally get to me, from the time I ordered it to the time I realized something was wrong, to the eventual inquiry to my post office and what do you know, they had it and didn't notify me. Which is weird, they know who I am, they know my M.O., and it wasn't held at customs, they just... fell asleep I guess.

After plugging it in, setting it up, the big result is what you see below. Here's what the test intended:
  • To test whether the drive was locked to either USB2 or USB3
  • Whether or not having the drive on USB3 will properly regenerate it
  • To further conclude my (and other reviewers) theory that Western Digital's proprietary USB3 controller-fused-onto-the-HDD-motherboard is a faulty design
And my ExpressCard gave me a lot of answers. Before I had the ExpressCard, when I plug my 1TB into a USB2 port, the drive takes a while to post (when the drive shows up in My Computer). A REEEAAAL long time, I timed it to 28 minutes before I said screw it I'll wait till I get the ExpressCard. Guess what? On USB3, the drive posts almost immediately, with just the drive delays being what slows it down. This means WD's USB2/USB3 controller is faulty, if you ask me such  a drive was not intended for USB2, especially when this is WD's design that does not have a separate USB controller card from the HDD. Let me try to explain this if you are unfamiliar with external drives. A normal external HDD is comprised of the hard drive itself, and the USB controller card with a Mini-USB port to connect to a computer. This design ensures the controller is a separate unit, and in case of failure, only the USB controller needs to be replaced. WD's proprietary HDD+USB2/USB3 controller design means that if you experience a problem like I did (drive failed to remember it was on USB2 mode while committing write operations), the entire drive is at fault.

This is key because a hard drive has a maximum and minimum spin time depending on what task you are doing (write operations require more power, read operations require less). In my case, I was writing to the drive, and I believe that somehow, the drive's defective design did not utilize USB2's maximum 500ma output, but mistakenly wanted more power (USB3's 900ma) and because I was on a USB2 port, it did not have enough power to write to the drive and thus the drive froze and was unable to continue, resulting in bad sectors in the areas it was writing to. This is why I believe WD's USB2/USB3 controller design is defective because the drive (in USB3 mode) operates twice as fast as on USB2, and only then was I able to do the complete regenerator scan above. Remember that my drive is a 2010 model, and I know for sure USB3 was still in it's infancy, and for WD to have this bright idea to use a prototype'ish tech in a new design? Bad idea.

And an even worse problem now is that somehow I cannot regenerate the sectors. The red-white "X" you see in the beginning cannot be regenerated no matter what I do. I even tried the advice of someone on the HD sentinel forums to change start/end sectors to the bad area, and repeat the regen process until it works. And this isn't just the beginning, I targeted the bad areas in the lower left of the screenshot, and it results in the same "discovery of new sectors while regenerating" error. The bad news is I do not know where the bad sectors are, if you look at the above screenshot, you might think it's simply a matter of selecting the red B and then start regenerating, but it's not that simple. See, one B area is comprised of thousands if not millions of sectors, and it is these sectors that are bad. But that can be anywhere from 1-10,000, and when I left the drive regenerating overnight, somewhere in the middle the drive completely halted and HDD Regenerator threw an error. Then what I have to do is plug out then plug in the drive, start the regen program, let it "resume regenerating" and then I have to physically watch this, no going AFK while letting it run.

A very timely process I have to run, and I go AFK a lot, and I'm in no mood to babysit this thing, so I have to take the last resort move. I need to sacrifice the data that is in the bad areas and just copy what I can onto a new drive (that I don't have). I estimate this to lose me at least 200GB of the 800GB I am using. Sucks to be me, but that's as far as I can tolerate now. It's costing me valuable time that I don't have a lot of and it's also going to cost me crucial funds that I have been using for my health to buy a new drive, which is even worse because right now not only am I recovering from a cold, but I have a 7 day old pain above my left eye that I have no explanation for and no painkiller can alleviate. I'm not sure if I said this before, but I've had one hell of a year and I've got to turn that around.

(if you're coming over from XDA, you can read the previous posts about my HDD here, here and here.)

1 comment:

  1. You have my sympathy sir and I feel your pain, since my 1TB died recently too.
    Thanks for the info about the separate controller, hopefully mine can be saved at some point.