Verbatim USB 3.0 Desktop Hard Drive 2TB has Samsung HD204UI (reported by operating system).

1. That hard drive needs a firmware update, otherwise data might be lost according to Samsung:
http://www.samsung.com/global/business/ ... msg_id=386

However, it is not possible to apply Samsung firmware fix, because Samsung software (patch tool) does not recognize the drive, when it is in Verbatim USB enclosure.

I tried it even with USBASPI V2.20 MS-DOS Driver: http://hddguru.com/software/2006.02.09- ... OS-Driver/
to no avail.

It is not possible to remove the hard drive from the enclosure in order to apply the fix, without loosing warranty, as there is "Warranty void if broken" sticker.

Is there any software, which would allow to update the firmware of the Samsung HD204UI hard drive without removing it from the enclosure?

Because of USB enclosure it is also not possible to use any tool for testing the hard drive surface (and SMART status), including Samsung's diagnostic software: ESTOOL v. 3.01 and MHDD (with ASPI driver it can detect it, but says 'drive not ready')

Any idea how to get around this?

2. Verbatim in their Quick Start Guide says (in capital bold letters):
"Verbatim will not be liable for data loss or any incidental, consequential or special damages, however caused, for breach of warranties or otherwise."

However, I believe that Verbatim cannot deny its liability for data loss in this case, because Verbatim is selling a faulty product (with a known fault, which can be fixed only by firmware update or hard drive exchange to a newer revision), and Verbatim, so far, has not provided any solution.

3. Samsung HD204UI hard drive, as most of other 2 TB hard drives is built in Advanced Format:
http://www.samsung.com/global/business/ ... 72&type=94
&subtype=98&model_cd=552&tab=fea&ppmi=1219

This means that it needs to have partitions properly aligned, otherwise it might work considerably slower and cause other problems.
This is so because with the emulation of 512B sectors, there is the risk that a partition could be misaligned compared to the 4K physical sectors - where it would be unwittingly started in the middle of such a sector. As a result, the clusters of a file system on that partition would end up straddling 4K sectors, which would cause performance problems.

Verbatim is providing the drive pre-formatted to FAT 32. But have they aligned it properly?
I think they should warn their customers about that aligning problem (particularly if someone would like to change formatting for example to NTFS).
Partitioning the hard drive under Windows 7 should align it properly. But Verbatim could provide their customers with a special tool from Acronis and/or Paragon to align the partitions (like WD did).

You can read more about Advanced Format HDs and the alignment problem on those pages:
http://seagate.com/docs/pdf/whitepaper/ ... ectors.pdf
http://consumer.media.seagate.com/2010/ ... ve-primer/
http://seagate.com/docs/pdf/whitepaper/ ... on_faq.pdf
http://www.anandtech.com/show/2888
http://blogs.msdn.com/b/jimmymay/archiv ... plate.aspx

4. Last but not least: why is warranty for Verbatim product lasting only 2 years in the European Union (and in most of other places in the world), while in USA warranty last 7 years: http://www.verbatim.com/UserFiles/File/ ... rranty.pdf
http://www.verbatim.com/prod/hard-drive ... p/usb-3.0/ ?

Why do they discriminate Europeans regarding warranty so grossly?

The F4EG.EXE update tool does not appear to look for drives attached to USB ports. The following is the embedded documentation for the tool:


Depending on the USB-SATA bridge chip, you could probably retrieve the SMART report using one or more of the following tools:

http://www.hdsentinel.com/
http://hddscan.com/

smartmontools (Linux / Windows):
http://sourceforge.net/projects/smartmontools/files/
http://sourceforge.net/apps/trac/smartm ... i/Download

A surface scan should be possible using any scanning software, eg HDDScan.

To confirm whether your partitions are 4K aligned, you could use Microsoft's Sector Inspector to examine the partition table and boot sector:
http://www.users.on.net/~fzabkar/SecInspect.zip

Extract the above archive to the one folder and execute the SIrun.bat file. The procedure will generate a report file named SIout.txt. If the boot sector is located at LBA 2048 rather than 63, then this would suggest that the partition is aligned.

What is your drive's current firmware version? AFAICT, the update changes the revision level to 1AQ3C0PM. I have managed to extract the 1AQ3C0PM.163 firmware image from the F4EG.EXE file, which means that, if your USB-SATA bridge chip can handle the standard ATA firmware download command, then it may be possible to update the drive with the aformentioned image.

_________________
A backup a day keeps DR away.

Suggestions and a warning below.

IMHO you should be addressing your concerns to Verbatim - you bought a Verbatim product; and despite you mentioning them several times, Samsung are under no obligation to assist (you didn't buy anything direct from them and hence have no contract with them) and for some of your questions, Samsung are clearly not responsible.


Yes, it's a nasty one which I first came across due to using smartmontools.

One area where it seems that Samsung have screwed-up IMHO, is that the "fixed" f/w version reported the same number as the "affected" one; so without specific testing to try to provoke the bug, I don't know of a way to tell whether your drive has the updated f/w or not. (I can't say whether that is still the situation regarding f/w version numbering, as I'm no longer involved in that issue, but it certainly used to be.)

But again, this is a Verbatim product, they got your money, so they need to be helping you to find out whether your drive has the "affected" or the "fixed" f/w version.

The good news is that the command which triggers the drive f/w bug (the ATA Identify Device command) can be thought of as an administrative command, which is not issued during normal I/O. However it is issued by various tools, including (but not limited to) SMART monitoring tools - and smartmontools (link already kindly given by fzabkar) is one example of them.

Therefore until you have updated the f/w in that drive (or got confirmation from Verbatim that your drive already has the updated f/w version), I would not use any utilities to get the SMART status of that drive (or any other unusual drive monitoring utilities), when there is any possibility of I/O in progress. Effectively that would mean that you need to have that disk's filesystems unmounted (easy in Linux, I don't know/care about Windows) when you run a SMART check, until the drive f/w is updated.


I'm not surprised; since Samsung can't test & qualify their download utility with all possible 3rd-party USB enclosures, they may be (sensibly) trying to limit their support exposure to those (direct-attach) config which can be tested. If I was in your place, I would be dealing with Verbatim - they are responsible for the consequences of the enclosure into which they installed the Samsung drive, they chose the Samsung drive to use, and it is their enclosure which you bought.

My girlfriend has the same type of issue with an external enclosure containing a Seagate drive; that USB-to-SATA bridge doesn't support ATA f/w download. Of course I took the drive out of the enclosure to update the f/w using direct SATA attach and then put it back in again but my point is that drive f/w updates in 3rd-party USB enclosures are not under the control of the drive manufacturer.


Ask Verbatim. You won't be the only customer in this situation, so if I was them, I'd be investigating this as a matter of priority.


Incorrect. My girlfriend's system runs something that I put onto there ages ago (might have been from WD - I'd have to check), that will do a surface scan of any USB-attached drive - that functionality is not limited by the USB-to-SATA bridge since a basic surface scan can be done using just Read commands.


Yes, some (many?) programs might not work for SMART monitoring the drive in that USB enclosure (depending on the enclosure), but you are making a very bold claim that it is not possible to use any tool to do that! Have you checked every possible program?

If it was safe to do so, then as fzabkar suggested, I would start by trying the various USB-related options on smartmontools - but until you have the drive f/w updated, I suggest that you heed my warning above and do not try any such program with a live filesystem mounted on that drive.


I agree - you need to be dealing with them.


Ask them! The folks here can't tell you why a company has made a commercial decision like that In my experience in the UK, a 2 or 3 year warranty is not unusual for typical external USB enclosures from "brand names" (although in the UK we are not specifically limited by such a warranty). In fact it is the 7 year US warranty which looks to be unusually generous IMHO.

Good luck...

Thank you fzabkar and Vulcan for your replies.
Of course I e-mailed Verbatim, before I even posted on the forum. Still waiting for replies.
Thank's for the warning. I won't be using smart tools then. Since people on internet report that even currently purchased internal Samsung HDs are Rev. A , then I doubt Verbatim would already contain Rev. B of the Samsung drive.

So the only time these problem with f/w can occur, when the disk is in such enclosure, is when I try SMART reading tools, right? Microsoft's Sector Inspector and surface scanning software I can try to use safely?



What do you mean?



There is often lifetime warranty over there for some products.

AFAICT, the documentation for the bug suggests that data corruption can occur if an ATA Identify Device command is issued during a concurrent write access. Therefore, ISTM, that you need some way to ensure that no application writes to the drive while it is being analysed by your SMART application. To this end you could temporarily hide the drive's partitions with a disc editor. For example, the partition type byte for an NTFS partition is 0x07, so you could temporarily change it to 0xD7, and then reboot.

List of partition identifiers for PCs:
http://www.win.tue.nl/~aeb/partitions/p ... pes-1.html

HxD - Freeware Hex Editor and Disk Editor:
http://mh-nexus.de/en/hxd/

Examination of Windows 2000 ( NT5.0 ) and Windows XP ( NT5.1 ) MBR:
http://mirror.href.com/thestarman/asm/mbr/Win2kmbr.htm

As for Vulcan's statement that "in the UK we are not specifically limited by such a warranty", I can say that the same applies in Australia. Here we have two warranties, the manufacturer's voluntary warranty, and a statutory "implied warranty of merchantability". For example, there was a case where a manufacturer warranted a TV for one year, but a consumer whose TV failed within two years successfully challenged this on the latter grounds.

_________________
A backup a day keeps DR away.

You're welcome


OK - you didn't mention you'd done that, so I didn't know... My previous comments still apply about Verbatim being the right people for you to talk to IMHO, especially since you seem to be concerned about not opening the USB enclosure, and so that appears to be your limiting factor.


No - that's just the best example. Anything which causes the USB-to-SATA bridge chip to send ATA Identify Device commands to the drive, while SATA NCQ Writes are also in progress, is a potential risk.

Has that bridge chip enabled SATA NCQ (i.e. is it sending SATA Write FPDMA Queued commands, and not PIO Write commands, to the drive)? No idea - those chips are not my specialist area. Ask Verbatim


Not necessarily (though IMHO it is unlikely that Sector Inspector and basic surface scanning programs would cause ATA Identify Device commands to be sent).

However I can think of a couple of ways in which software (even those which you mention above) could potentially cause the bridge chip to send ATA Identify Device commands - but the details are too long for now, and would just be a hypothesis anyway, without having a USB trace (or preferably SATA trace) when running whatever software you would specifically want to test.

Also, if the USB-to-SATA bridge chip is not sending SATA NCQ commands, then you're safe, even if ATA Identify Device commands are being sent to the drive at the same time.


Very similar to what fzabkar kindly explained applies in Australia. In the UK, the consumer's contract is not with the manufacturer; it's with the retailer. So any warranty from the manufacturer is in addition to (and not instead of) the consumer's rights for merchantable quality & "reasonable" durability, against the retailer. IANAL so I won't go into more detail here - you can see the various revisions of the UK "Sale of Goods [& Services]" Acts if you are interested


That just shows that the typical expectations are different in different countries. To me, here in the UK, a 7 year warranty for an external USB drive would be very, very generous, as I said before. This difference in length of warranties isn't a "Verbatim thing", it's a "different countries thing". Here is a link to the (now discontinued) external 1TB USB disk which a friend of mine bought a little while ago, here in the UK - again a 2-year warranty, this time from Hitachi:

http://www.overclock.co.uk/product/Hitachi-1TB-XL1000-USB-2.0-External-Hard-Drive,-Retail-Kit_28474.html

I hope that explains why I'm not getting excited about Verbatim "only" offering a 2 year warranty also

So in summary, your Verbatim-branded USB enclosure might not create the conditions to trigger that disk f/w bug, if it doesn't use SATA NCQ commands. IMHO the best people to answer that question are Verbatim, unless you can find enough details on the bridge chip to completely eliminate that possibility of it using NCQ yourself, from its datasheet.

Have you considered collecting the USB bridge chip VID & PID, and attempting to identify the bridge chip? You might succeed in identifying that chip, but you might not (depending on whether Verbatim use a custom VID/PID or leave it at the manufacturer's default). If you physically looked at the chip, then you'd have a much better chance to see what it "really" is (without relying on the VID/PID), but I remember that you don't want to open the USB enclosure.

Therefore the 3 options which I can think of, for you to choose between are:

- Change your mind and open the USB enclosure to look at the bridge chip; but if you're going to do that, then you might as well take out the Samsung drive, update its f/w via SATA direct attach, and sleep happily
or
- Try to identify the bridge chip via USB VID/PID and/or Google searches from people who have opened exactly the same enclosure. Then search for a datasheet on that bridge chip, and hope that SATA NCQ is not a feature or even an option (because if it is an option, you won't know if that option has been enabled by Verbatim in your enclosure or not). If you can rule out SATA NCQ being used, then you should be safe, even without updating the drive f/w.
or
- Ask Verbatim

Right, I'm off back to the day job - have fun

You could use Microsoft's UVCView to examine your USB devices:
http://www.users.on.net/~fzabkar/USB_ID ... ew.x86.exe

Hopefully the external drive identifies itself with the Vendor and Product ID of the bridge chip rather than as Verbatim's product.

_________________
A backup a day keeps DR away.

I wonder how do they interpret "reasonable" durability? A computer chair broke after one year warranty given by the UK retailer, but it should have lasted many, many years more to come, and it didn't. So how can one convince retailer that they should exchange it?

I think you're going a bit "off topic" for HDD Guru forums, don't you? However, assuming that you're resident in the UK (your profile doesn't seem to say so), I'll give you an answer based on my experiences with consumer law - but as I said, IANAL.


(a) Any warranty from a retailer is in addition to, and not instead of, the rights under the Sale of Goods Acts. Some retailers offer more than a 1 year period (e.g. John Lewis). Therefore this period is not a hard limit to your rights (see below).
(b) During that period, it is easier to get the retailer to repair/replace an item, since there can be no argument about how long their responsibilities last. However last time I checked, it is not actually required for them to offer a warranty (it's just that they all do, due to commercial pressure of competition from the retailers who do offer one).
(c) This leads onto the final part - your statutory rights are the Sale of Goods Act, irrespective of any warranty from a retailer (or manufacturer), and that Act (and related ones including Services etc.) only specifies a maximum period in which to take legal action (6 years in England). The problem if you come to relying on this statutory protection (i.e. after the end of a retailer's warranty period), is that you have to prove what is reasonable, if necessary, by going to court (e.g. Small Claims Court).

In deciding what is a reasonable time for a product to work, a court will take into account things like the price paid, how the item was used, the report of an independent repairer confirming the fault and its likely cause (which you will normally have to pay to obtain, if the item is over 6 months old, but this cost can also be claimed against the retailer as part of your court action).

The more something costs, and the less it has been used, and if it has not been mistreated, then the greater the chance that a court might agree it should have lasted a longer time


Often just the threat of going to court will encourage a retailer to enter into negotiations with you. (You also potentially have a claim against the credit card company, if the item was bought on a credit card and cost over £100 - see the Consumer Credit Act, section 79). However you are not entitled to an exchange after a period of time like a year (after a shorter time, then you'd have had a much better case) - see the webpages below.

You are now entering into the realm of "contract law" and IANAL (did I say that already? ), but in essence the product has to conform to the terms of that Act - and it did for a year; so you have had a year's use from the chair, and cannot now demand an exchange for a new one. That would be unreasonable against the retailer. Instead, it is more likely that you would claim damages for the cost of a repair, and they might instead offer you an exchange if that is likely to cost them less than paying for the repair.

Although they've changed the website and now archived this one, this page is one of the best official summaries of UK consumer's rights that I have seen:

http://webarchive.nationalarchives.gov.uk/+/http://www.berr.gov.uk/whatwedo/consumers/fact-sheets/page38311.html

Notice the link to Consumer Direct on the right side of that page - those are the people who I recommend you should contact for more (free) advice; that government organisation are specialists in advising UK consumers on their rights. Or, before many are shut down due to lack of cash then a CAB (Citizens Advice Bureaux) office would give you (free) advice too - see an example of their advice on this subject here:

http://www.adviceguide.org.uk/index/your_world/consumer_affairs/buying_goods_your_rights.htm

You'll see it's the same basic information on your rights, as on that (archived) BERR page I linked above.

Hope that helps. Now can we get back to hard disk drives?

Thank you for your reply. I wish everyones reply was always so thorough.



Of course.
The firmware reported by HDDScan is 1AQ10001.
I did not try to check whether it would be able to read SMART.
However, I started surface scan - it is very slow - only 1% scanned so far, but already 5 blocks are < 500 ms.
Have you got similar result in your scans of 2 TB drives?

I will PM you the result of Sector Inspector.

Here is the result of UVCView.x86:

---===>Device Information<===---
English product name: "Desktop USB3.0 Drive"

ConnectionStatus:
Current Config Value: 0x01 -> Device Bus Speed: High
Device Address: 0x01
Open Pipes: 2

===>Endpoint Descriptor<===
bLength: 0x07
bDescriptorType: 0x05
bEndpointAddress: 0x81 -> Direction: IN - EndpointID: 1
bmAttributes: 0x02 -> Bulk Transfer Type
wMaxPacketSize: 0x0200 = 0x200 max bytes
bInterval: 0x00

===>Endpoint Descriptor<===
bLength: 0x07
bDescriptorType: 0x05
bEndpointAddress: 0x02 -> Direction: OUT - EndpointID: 2
bmAttributes: 0x02 -> Bulk Transfer Type
wMaxPacketSize: 0x0200 = 0x200 max bytes
bInterval: 0x00

===>Device Descriptor<===
bLength: 0x12
bDescriptorType: 0x01
bcdUSB: 0x0210
bDeviceClass: 0x00 -> This is an Interface Class Defined Device
bDeviceSubClass: 0x00
bDeviceProtocol: 0x00
bMaxPacketSize0: 0x40 = (64) Bytes
idVendor: 0x18A5idProduct: 0x022A
bcdDevice: 0x0100
iManufacturer: 0x0A
English (United States) "Verbatim"
iProduct: 0x0B
English (United States) "Desktop USB3.0 Drive"
iSerialNumber: 0x03
English (United States) "*************"
bNumConfigurations: 0x01

===>Configuration Descriptor<===
bLength: 0x09
bDescriptorType: 0x02
wTotalLength: 0x0020 -> Validated
bNumInterfaces: 0x01
bConfigurationValue: 0x01
iConfiguration: 0x04
English (United States) "USB Mass Storage"
bmAttributes: 0xC0 -> Bus Powered
MaxPower: 0x01 = 2 mA

===>Interface Descriptor<===
bLength: 0x09
bDescriptorType: 0x04
bInterfaceNumber: 0x00
bAlternateSetting: 0x00
bNumEndpoints: 0x02
bInterfaceClass: 0x08 -> This is a Mass Storage USB Device Interface Class
bInterfaceSubClass: 0x06
bInterfaceProtocol: 0x50
iInterface: 0x06
English (United States) "MSC Bulk-Only Transfer"

===>Endpoint Descriptor<===
bLength: 0x07
bDescriptorType: 0x05
bEndpointAddress: 0x81 -> Direction: IN - EndpointID: 1
bmAttributes: 0x02 -> Bulk Transfer Type
wMaxPacketSize: 0x0200 = 0x200 max bytes
bInterval: 0x00

===>Endpoint Descriptor<===
bLength: 0x07
bDescriptorType: 0x05
bEndpointAddress: 0x02 -> Direction: OUT - EndpointID: 2
bmAttributes: 0x02 -> Bulk Transfer Type
wMaxPacketSize: 0x0200 = 0x200 max bytes
bInterval: 0x00

You're welcome, although I can't promise to always be able to spend that amount of time writing a reply...

In fact this has to be a 5 minute reply from me now, so let me cut to the bad news You'll remember when I said:


The bad news is that Verbatim have taken the option to make the USB-to-SATA bridge report a Verbatim VID/PID of 0x18A5 / x022A. From your output of UVCView.x86, you can see:


Since I don't believe that Verbatim make any such chips, this just hides whatever the real chip identification is, and hence prevents further investigation of what functionality it might be using, specifically NCQ. Without you doing lots of Google work (to see if anyone else has found which bridge chip is really used in that enclosure), or without you opening it yourself to look at the chip's markings (and even then you aren't guaranteed to find enough public info about how the bridge chip is being programmed in that enclosure, even if you can read its part number), I can't think of how you will be able to eliminate the possibility of that enclosure allowing the circumstances which would trigger the Samsung f/w bug. Of course someone else on the forum might think of a way - but I can't at the moment...


Hmm, I'm not quite sure why you did that - I didn't ask for it, did I? If someone else asked, perhaps you should either send it to them, or attach it to another post as a file attachment (zip'd first) to make it publically available.

I've got to rush now... So in summary - unfortunately for you, the only route I can see for you to get the confidence of not being able to trigger that Samsung drive f/w bug (which depends on finding intimate knowledge of whichever USB-to-SATA bridge that they have used), or to be able to update the drive f/w while leaving the drive in that Verbatim enclosure, you either have to do lots of searching to try to find that bridge chip (which might be a waste of time if you don't find the original manufacturer & chip), is for you to rely on Verbatim to answer your questions. If only they hadn't chosen to hide the original bridge manufacturer, then further investigation would have been possible, but they aren't the only "enclosure brand" to do so

I think you are back to a position of: Ask Verbatim - unless you are willing to open the enclosure and upgrade the drive via direct SATA attach, or unless someone else can think of a way to get that Samsung f/w file across the [unknown] bridge chip and into that drive, along with the risks which performing that untested process would carry.

Good luck...

I thought you'd be curious to know. I sent it to both of you i.e. to you and fzabkar. Actually I have some troubles interpreting those results, because they are not that obvious - where is what; how to check if the boot sector is located at LBA 2048 rather than 63? There are obiously two HDs in the results, and I can see LBN 0, LBN 63, LBN 20964825, LBN 20964888, LBN 2048, LBN 2147485696. I believe that LBN 2048 is for the extrenal HD, so it suggests that it is aligned. Nevertheless I am simply guessing. The internal disk on the other hand is not aligned, but it shouldn't be a problem, should it?
This article, however, seems to suggest that also not aligned disks in non Advance Format also would cause worse performance:
http://blogs.msdn.com/b/jimmymay/archiv ... plate.aspx



I'll wait for Verbatim's reply, but in the meantime I have a question: has anyone tried to remove that sticker from Verbatim enclosure and attach it again, so that they could not tell that the sticker was removed?

History of the firmware bug discovery:
http://sourceforge.net/apps/trac/smartm ... GBadBlocks

From another forum:

"Here is the reason from the Samsung HDD department for the firmware number staying the same:

I just wanted to let you know the firmware version will not change after flashing. The version did not change because it takes a long time to get customers approval for change notices. It would have taken over a month before we could have released the fix.

Please make sure that for each drive the flashing process is successful…

We have already implemented the fix to our production. So drives available on the market in 2 months time will have the fix applied. To be sure check the manufacturing date on the drives, January 2011 would be safer…"

http://lime-technology.com/forum/index. ... 9#msg90249

According to Sector Inspector, here is the partition table for PhysicalDrive1:


There are two NTFS partitions (type 0x07). They begin at LBA 2048 and LBA 2147485696, and their sizes are 2147483648 and 1759539200, respectively. Both boot sectors (LBA 2048 and LBA 2147485696) report 8 sectors per cluster.

Since all these numbers are divisible by 8 (4KB = 8 sectors), then both partitions are 4KB aligned.

_________________
A backup a day keeps DR away.

Quote:
"Also, the SMART output from all of the drives has now changed. I think Samsung added some new fields or otherwise changed the data, because now the output is quite messy."
http://hardforum.com/showpost.php?p=103 ... stcount=24

Good. Thanks.

Is aligning non Advanced Format drives beneficial in some ways? Or shouldn't I bother (particularly if they already contain data, so it is better not risk loosing it)?

When HDDScan is identifying the drive and its firmware does it sent the IDENTIFY DEVICE commands? Because it is identifying that without using SMART option. It does it by default.

I don't think so.

The reason you need to align an Advanced Format drive is that, if you wish to write to one particular sector, you first need to read its entire 8-sector block, modify the block's contents, recompute the ECC bits, and then write the whole block to the platters on the next rotation. The above happens when an 8-sector cluster is not aligned to an 8-sector block. OTOH, if the clusters are aligned, then the data can be written on the first pass.

_________________
A backup a day keeps DR away.

Then why this guy:
http://blogs.msdn.com/b/jimmymay/archiv ... plate.aspx
was writing about aligning HDs even before Advanced Format disks appeared?