I have a Seagate Barracuda 7200.12 in a Windows Home Server 2011 machine that has gone bad. While the server was up and running, the OS locked up, presumably due to some failure with the drive (it was the system drive). On boot, the BIOS pauses for a moment while trying to detect the drive, then fails to identify it. Attached to another working Windows box via a simple USB to SATA interface, the drive is still not identifiable. I have a few other SATA drives that work fine being attached in the same way (in other words, pretty sure this isn't operator error here). Upon being connected to a working machine, the drive spins up without any abnormal sounds (no clicking, buzzing... just sounds healthy). The SATA LED lights up on my USB to SATA interface. After spinning up, it makes seemingly random head seek noises, typically in short 0.25 to 2 second bursts, with no head activity for about 5 seconds in between each burst. I say random because the sound seems to be a different pattern each time. The drive doesn't seem to ever stop seeking completely and stop spinning, but I've only let it stay on for about 30 minutes. I've removed and inspected the PCB, but I see no physical external signs of damage (though I know well that IC's can and do fail with no visible signs outside of the packaging).

I have a degree in electrical engineering, but I'm simply not in the hard drive business, so I'm looking for some general advice. I have done some general web searching and talked to a couple data recovery companies. I have a various backups of the drive's content, BUT there is some content on the drive that is simply not backed up. Thus, I'm in a situation where my world is not going to come to an end if I lose the unique data on that drive, but it would be worth a couple hundred dollars to me if I could get it back because it will save me some work (such as the re-ripping of my entire CD library). But it's definitely not worth $600 to $2000 to me.

The drive specs are:

Seagate Barracuda 7200.12 1000 GB
Model: ST31000528AS
Part: 9SL154-022
Firmware: HP35 (from an HP MediaSmart Server -- OEM firmware)
PCB: 100574451 REV A
PCB Sticker: 8552 C

I know enough to know that it could be the PCB or the "firmware" that's stored on the platters. I also know enough to know the price of quality and people who know what they're doing. I've found PCB replacements at donordrives.com (http://www.donordrives.com/seagate-1tb- ... 5-pcb.html) and hdd-parts.com (http://www.hdd-parts.com/12101102.html) but I haven't had much luck in finding any positive customer testimonials that aren't on the sites themselves. I'm willing to spend the $50 or so if it's a decent gamble that it might be successful, but I don't want to just throw $50 away, nor have sensitive information on the drive fall into nefarious hands (not that I have any reason to suspect such a thing). At the same time, shipping my drive to a pro clean room facility who would charge me a minimum of $500 or so doesn't make sense to me, as even if they could do it I wouldn't feel it was worth it, and thus I've wasted their time.

Any advice for me? Is there anyone with expertise in these drives who might give me a ballpark figure on the odds of it being *just* the PCB, or if that's a too-good-to-be-true dream?

Hi,

According to what you say, I don't think you will get lucky with just the "Swap PCB shoot". You will spend your money and in the end get frustrated, as that might be a Firmware issue.

First of all, what you should do is search here how to build a terminal for that drive in order to be able to check the drive log. With that information on hand, it will be easier to guide you on the correct direction afterward.

Buy one or make a RS232 to TTL convertor

http://hddguru.com/articles/2005.10.01- ... schematic/

connect the drive and post what you see in Hyperterminal

90% firmware / media/heads. IF the drive does this long 'seek' noise with only power on - no data cable - it's internal problem. As you should buy/get another drive to copy the content of the 'failing' one and the data is not worth $$, simply get another one, re-rip and carry on. Then search a pro in your area who can maybe refurb the drive (NOT recover data) for a small fee as in this state is unreliable even if with FW fixed. Everything else is a time waste.

Agreed, PCB is not at fault here. Most likely some f/w error or some internal mechanical damage.
As sathyan says, a terminal output would be very helpful in your case.

_________________
http://www.northwind.gr
SandForce SSD Recovery
Ransomware Reverse Engineering - NoMoreRansom! partners

+2

PCB is 99.99% not going to be your problem, so don't waste your $$

My bet is for firmware issue (offline activity)

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

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

Hey guys,

I've got my USB to TTL going, terminal configured and I'm talking to the HDD. I haven't been able to pull a log though. I get "DiagError 00008000" as a response. Again, I'm an educated fellow, but I'm not in the hard drive business and I have no idea what this means, nor really if I'm giving the right commands (which yes, does feel a bit scary, but hey, the Internet is great for finding multiple sources of information that agree with each other, giving me some comfort).

Here's what I'm getting on the terminal:

Upon power up:



Upon hitting Ctrl-z, then "/L" and "D":



I can spin down and spin up the motor with success:



But still no log...



And Ctrl-c successfully resets:



And the rest of my repetitions have given me the same results.

I think "/L" takes it to level L, and at that level, "D" is the command to display the log. But I admit I may be wrong. Do you guys have any advice from here? Am I using the right terminal commands to produce the log?

By the way, I have tried powering up the drive with the pins connecting the heads masked off, and the motor pins remaining connected. This results in the motor spinning up upon receiving power, then the motor spinning down a few seconds later. This results in a "No HOST FIS-ReadyStatusFlags 2002B001" after the initial power-up, then "Failed to load overlay 00000004", "LED:000000CC FAddr:002887E5", "LED:000000CC FAddr:002887E5", ... (repeats) when trying to access the log (with the same "/L" and "D" commands).

And for those Googling with similar issues and happening upon this thread, I'm looking at this list of Seagate terminal commands, using this USB to TTL UART bridge, and this procedure for masking the head connector pins.

Happy tinkering then.

I have a feeling it is just another thread where data is no longer important / want just to do it for fun or learning / another backup was found.

P.S. some commands of just 1 character can f**k your drive in a flash.

No, like I said in the OP, the data is worth something to me. While I do find the tinkering a bit fun -- and I have since I was a kid -- the main thing I'm looking for here is if there's a reasonably priced solution for recovering the data on the drive. And don't me wrong, I'm certain that DR services are reasonably priced in general -- professional technicians with care and know-how, clean room, the right tools, the donor drive parts, shipping, customer service / sales... all that costs money and there needs to be profit on top of that. I'm just trying to figure out what's wrong with the drive so that I don't waste someone's time analyzing it themselves only to tell me it will be $1000, which I'm not going to think is worthwhile -- I'll sit and re-rip my music library and re-write some software projects instead. But if we're talking about a $50 PCB, or even a couple hundred for something... that's interesting.

Are /L and D the right commands for displaying the log? And if so, what might that "DiagError 00008000" mean? Would that be indicative of a board failure? Or a firmware-on-the-media / head issue? Any advice now that I have terminal output?

Hi,

If you want a DIY solution, you can find a lot of commands here on the forum, as an experienced user will not risk in giving you commands that might take your data off for good.

I appreciate that. But displaying the log isn't going to risk wiping the data, is it? Certainly telling me what "DiagError 00008000" means wouldn't be risky, would it? I do understand very well that the wrong letter or wrong level or wrong capitalization can destroy the data on the drive, so I plan to be careful. That's why I'm here asking questions instead of randomly pushing buttons.

You had said, "... build a terminal for that drive in order to be able to check the drive log. With that information on hand, it will be easier to guide you on the correct direction afterward." Then, sathyan and northwind both echoed that terminal output would be helpful. I have done some searching on the commands to check the drive log, but obviously I'm not getting it right.

You expect to see the failure displayed by 'log' in clear words, but it is not the way it works. And after that, let's say there is a damage in some fw area : how to work on it without dedicated tools ?

Hi,

The problem is that the log itself can happen due to several reasons, so it's not a sure shoot to tell anything based on that.

Bro i got same problem and before trying to do anything risky, i need some knowledge so i also wanted to know what happened after you're searching for that error please let me know..

... difficult to tell after 2 years of silence ....

mayur-c,
Hi Where are you from in India ?
Give me your email id.

Regards,
Peter

PCB is not at fault, so don't waste your moeny.