External Seagate 3TB issues.

TonyYamato

Honorable
Feb 1, 2014
4
0
10,510
One of my external hdds was starting to die, so I decided to buy a new one. Problem is the new one seems to have some very strange issues going on with it which may or may not be related to the fact it's a 3TB.

I noticed the problems when I wanted to test the hard disk speed in HD Tune.
http://imgur.com/3QYBEAN
As you can see it's mentioned as being a 375GB and does not allow any kind of test.

http://imgur.com/6eFLLud
Hard disk sentinel seems to recognize the hard drive fine, but gives some rather strange SMART values that are often higher than external hdd's I've had for years.

After all this I decided to simply try out the tools of the brand itself, Seagate in this case.
http://imgur.com/HCNs9TZ
But Seatools just doesn't recognize anything about the hard drive itself and fails every test I let it try on it. Because it was formatted as a 3TB MBR, I even tried converting it to a GPR and making it a 3TB volume but the problem still persists. The hard drive seems to work fine with copying stuff on it, am I just being paranoid or should I have it returned?

As for extra information, I'm running Windows 7 x64, Hard drive is connected with USB 3.0 ( Problem remains even with 2.0 slots ). Addressing on device manager and the fact my main disk is an MBR would make me assume it's in fact booting with BIOS even though the motherboard is UEFI capable.
 
You should have left the drive as it was. :-(

Seagate's 3TB external drives are configured with 4KB LBAs rather than the traditional 512-byte sectors. Your partitioning and formatting utility probably doesn't understand 4KB sectoring, so it appears to have made a mess of things.

4KB = 4096 bytes = 8 x 512 bytes

375GB x 8 = 3TB

I suggest you use Seagate's DiscWizard to restore your drive to its original out-of-box condition.

The HD Sentinel SMART results are perfectly fine. The raw values of the Seek Error Rate and Raw Read Error Rate attributes are sector counts, not error counts.

See http://www.users.on.net/~fzabkar/HDD/Seagate_SER_RRER_HEC.html
 

TonyYamato

Honorable
Feb 1, 2014
4
0
10,510
The conversion to GPT was done in Seagate's DiscWizard through the 'Add new hard drive' option. Which allowed me to choose either MBR or GPT. So I doubt it'd set me up with a faulty set up. Either way the issue on both SeaTools and HD Tune was there from before I made the change to GPT so I don't see how that would have helped anything. The fact that worries me the most is the fact SeaTools never recognized the hard drive at all.
 
I would examine sector 0 with a disc editor such as DMDE (freeware). This will tell us what is going on inside the box.

BTW, HD Sentinel will probably be able to see the drive behind the USB-SATA bridge PCB. HD Tune seems to have a lot of trouble doing this. CrystalDiskInfo might be able to do it, too.
 
DMDE is reporting a range of sectors from 0 to 732566644 and a GPT(4K) file system.

732 566 645 x 4096 = 3 000 592 977 920

This means that the enclosure is configured with an LBA size of 4KB.

There appear to be two partitions, a 134MB reserved Microsoft partition beginning at sector 6, plus a 3TB data partition beginning at sector 33024.

There is also the ghost of the original 3TB "Seagate Expansion" NTFS partition beginning at sector 2048.

In short, it all looks OK. I'm wondering if HD Tune is merely consulting the partition table and incorrectly assuming that the sector size is 512-bytes.

My next step would be to examine sector 33024. This should be an NTFS boot sector.

Using DMDE ...

select Editor -> Goto Offset
Sector = 33024
Sector Offset = 0
From Start/End
Dec

select Mode -> Hexadecimal/Text to see the raw data

or select Mode -> FAT/FAT32/NTFS Boot Record to see the NTFS file system parameters
 

TonyYamato

Honorable
Feb 1, 2014
4
0
10,510
http://imgur.com/shxSLnT
Here are the results of sector 33024. Isn't it odd that there's a ghost of the original partition if I removed it and formatted it before converting to GPT?
On a sidenote, I'm not so much worried about HD Tune not getting the correct results, as I know not all free tools can always give reliable information. But the fact their very own SeaTools doesn't recognize any information about it is rather worrying. Is it related to the same problem with assuming a wrong sector size?