jaclaz

No. You are "looking at" sector 103355673 and 103363844 (as you can see at the top of the screenshots you posted ) Which means that you input 0x6291519 and NOT 6291519, i.e. you are using the Decimal number I gave you as if it were a Hex number. (Cannot say about you, but I have 5 fingers each hand and I am more familiar with decimal system ) 0x6291519 (Hex) = 103355673 (Dec) Try again. jaclaz

Sure , any app capable of a RAW image will do. I like when people ask what to use and then use ANOTHER app. If it was formatted by Vista (unpatched) the number of hidden sectors will be probably 2048 and thus the $MFT should be at (786432*8+2048)=6,293,504 The 1024 you mention would be "atypical", and I have never seen a first partiion starting at LBA 1024, only "good ol'" 63 or "new, stoopid" 2048. Once you have the image, get Tiny Hexer, install it, then: File-> Disk->Open disk image or large file as drive->(choose the image file)->in the "First Sector" box input 6291519 If the first few bytes are "FILE0" it should be the right place, to make sure at offset 240 (around half of the sector) there should be $.M.F.T. If not try: File-> Disk->Goto Sector/Position->input 6293504 and check if the "FILE0" and "$.M.F.T." is there Once you have a sector opened, you can use SHIFT+F7 to go one sector back or SHIFT+F8 to go one forward. (the sector just before the "right" one should be all FF's. jaclaz

Good. Very good. Yep. Suggested apps are:. DatarescueDD http://www.datarescue.com/photorescue/v3/drdd.htm or Clonedisk: http://erwan.l.free.fr/clonedisk/ http://reboot.pro/8480/ You want to image the \\.\Physicaldrive or "Drive" as RAW. The extension has really no importance, DatarescueDD uses .dd extension by default, clonedisk .img, but it's just a name. Let's wait until you have the image. Which OS are you running ? (2K/XP is "better" than Vista or 7, you know, "run as admin", "elevated privileges" and such) Under which OS was originally partitioned the disk? The general plan is: create the dd-like image verify where the $MFT is (if found) create a NTFS sparse file same size as the image mount the latter in a virtual drive and partition/format it (hopefully EXACTLY as it was partitioned/formatted before) copy the first million sectors from the newly created and partitioned/formatted image to the disk (or to a second copy of the original image) copy the last million sectors from the newly created and partitioned/formatted image to the disk (or to a second copy of the original image) The result should be a disk with a valid partition and filesystem with a number of $MFT entries pointing to "nowhere" (the ones that indexed now 00ed sectors) and hopefully a number of still valid entries that should allow recovering/copying the files that occupied non-00ed sectors. jaclaz

It's not very clear (to me at least) what are the actual problem/questions. I can give you for sure that the partition table on that drive looks like having been manually made by someone extracting Bingo numbers (or using several sets of dices) instead of rationally and logically putting together senceful numbers. Some more BAD news for you, most of what you think to know is either inaccurate, or incomplete or plainly wrong Namely: there is NO "need" (and I actually never seen) "a little partition about 8 Mb needed for booting". the first sector of a partition is the PBR or bootrecord there is NO need for NTLDR to be anywhere specifically in the filesystem (though some versions may need to be in the first 8 Gb ( maybe you are mistaking with DOS before version 7.x, where IO.SYS had to be first file in the partition ) the booting of a NT/2K/XP/2003 happens with THREE (or in some now rare cases four) files: NTLDR <. the actual bootloader BOOT.INI <- the configuration file for the bootloader, a simple text file NTDETECT:COM <- the detector of hardware that loads necessary drivers in protected mode NTBOOTDD.SYS <- optional Mass Storage device for SCSI drives (or however for drive cards with extension ROM) [*]The MBR contains the IPL CODE, the partition table DATA and the Disk Signature (and NOTHING else), the IPL code laods the bootsector of the Active (boot) partition) [*]then there are normally (not in your queer partitioning ) 62 empty sectors, so-called hidden sectors - these on "unpatched Vista/7/2008 become 2047 - for alignement to 4 Kb size - default now [*]then there is the bootsector or PBR of first partition (which starts at the first sector of the partition, and may be - depending on filesystem - one or more sectors long [*]The booting of Vista :ph34r:/7/Server 2008 happens with THREE files: BOOTMGR <- the actual bootloader \boot\BCD <- the configuration file for BOTOMGR, a binary file using the same format of a Registry Hive WINLOAD.EXE <- the actual OS loader You may want to spend some time on this: http://web.archive.org/web/20070608224106/http://www.ranish.com/part/primer.htm and this: http://thestarman.pcministry.com/asm/mbr/index.html and this: http://www.multibooters.co.uk/ in order to update and "fix" what you think you know. Then, you may come back here and try explaining better the problem/issue you are currently having. Or is the problem just that you have now a "standard" MBR instead of the original Fujitsu-Siemens one? Just as was discussed for the Dell of the OP, it is very possible that the Fujitsu Siemens uses a "peculiar" MBR CODE. Which EXACT model is that? (maybe with the model we can find the iutility that contains the "custom" MBR). The fact that it has a WINRE partition should however mean that the setup is not very different from this one: http://www.goodells.net/dellrestore/vista.htm (and the F8 has nothing to do with the MBR) This is the "normal" way to create a WinRE environment on HD: http://www.svrops.com/svrops/articles/winvistare.htm http://blogs.msdn.com/b/winre/archive/2007/01/12/how-to-install-winre-on-the-hard-disk.aspx jaclaz

I don't know about the "Quick" option. The DLGDIAG.TXT talks about the : and If whatever you used is the correspondent to "-FWRITE", the first and last million sectors in your hard disk should be 00's. 1,000,000x512=512,000,000 bytes at the beginning and at the end of the disk should be 00's. Basically if you had a single whole partition on the disk you have "lost" all the initial part and all the end part, this includes the MBR, the PBR and the PBR mirror (besides some initial data). Since I presume that the drive was partitioned in a single, biggish, NTFS partition, it is very likely that the $MFT has not been overwritten. Open the disk in a disk editor (suggested Tiny Hexer): http://reboot.pro/8734/ and try checking sector # 6,291,519 (786432*8+63) compare with: If the $MFT is found, it is possible that we can manually recreate the filesystem (at least enough to run a chkdsk on it or however get the files together with the filenames). You need to make an image of the disk "as is", before starting fiddling with it, you will need another hard disk bigger than 320 Gb (or however 320 Gb free on any hard disk). Mind you it won't be easy, it will take some time and patience, and there is NO guarantee it will work. jaclaz

It's strange that the drive doesn't work at all via SATA but it "works somewhat" via USB. With the drive USB connected you can see the \\PhysicalDisk in Disk Manager and the three partitions? It is possible that somehow the drive is still working but very, very slowly and when connected to SATA it simply "times out". It should be possible (I seem to remember reading something about it) to "force" a firmware upgrade. I'll check and see if I can find some reference). What does Seatools say about the drive? (try using the DOS version). The "latest" SD1A *should* be this one: http://seagate.custkb.com/seagate/crm/selfservice/news.jsp?DocId=207951 You should have ready a (say) 640 Gb hard disk so that if we find a way to access the disk, you can image it ASAP. jaclaz P.S.: Here is some reference: http://forums.seagate.com/t5/Barracuda-XT-Barracuda-and/ST31500541AS-force-upgrade/m-p/79104 http://niallbest.com/seagate-2tb-st32000542as-cc35-firmware-upgrade/ of course you need to adapt the procedure and command line to the actual .LOD you have.

If the Zero fill command was carried, the drive is filled with zero's . Unfortunately there is NO way on Earth you can recover ANYTHING from that drive. If you managed to recover the 26Gb of Data or *any* data it means that either you stopped the execution of the zero fil, or it ddn't work as expected. How long did the program run? How was it terminated? jaclaz

Forgot to mention Swissknife. The (old) Freeware edition runs on 9x too: http://www.devsite.compuapps.com/index.php?option=com_content&view=article&id=112:swiss-knife-v322&catid=48:drive-managment&Itemid=193 The download is not anymore available from the homepage (as it is superseded by the new, commercial version that only runs on XP and later) Via Wayback Machine: PE edition: http://web.archive.org/web/*/http://files.compuapps.com/us/swissknife/CompuAppsSwissKnife-BartPE.exe "Normal": http://web.archive.org/web/*/http://www.compuapps.com/Download/swissknife/swissknife.exe jaclaz jaclaz

NO, it doesn't. It does that only if at (first) install time it finds an UNpartitioned system. See here, JFYI: jaclaz

If your OS was updated to SP3, the repair install from a SP2 disk would not have worked, so it's good you had a SP3. The conflict btween "repair" and "Internet Exporer >6" is known. See if anything here works: http://forums.techarena.in/windows-xp-support/1112240.htm http://help.iforumz.com/windows-xp-help/xp-home-windows-activation-problem-after-repair-install-1391.html jaclaz

I would anyway, first thing, create an install CD from the source you have. I mean, the general idea is: you must have an install CD you then make from it (or from a copy of it on hard disk) the USB install media Anyway, yes, the \I386 is all you need to "feed" the WinsetupFromUSB thingy with. For the record, the problem might be due to the source being OEM or modified (nlited, etc.). jaclaz

...firmware that was NOT released before the drives started falling down bricking like flies, and additionally the early versions of the firmware update were also no good AND the stupid seagate tool DID NOT (and doesn't still) list correctly drives affected. ... and I don't remember seeing anywhere on newspapers actual recalls.... http://en.wikipedia.org/wiki/Product_recall Though there are no safety issues involved, we have the Intel recall of 1994 as a precedent, they could well do somethng like that.... The real issue is not in the actual problem - it may (and it has and it will) happen to *any* hard disk manufacturer, it is in the way it was managed badly once found out. For the record, these topics have been already talked to death: http://www.msfn.org/board/index.php?showtopic=128807&view=findpost&p=935754 http://www.msfn.org/board/index.php?showtopic=128807&view=findpost&p=905666 jaclaz

Yes, it could be part of the problem, all the stupid drivers not signed, elevated privileges and what not may cause it. But more generally this kind of tools are developed for technicians and are very poorly documented, so it is possible also that you are simply doing something "wrong". And another thing is that the fact that the stick has a SMI chip is still an *assumption* on my part, I wasn't able to find any definite reference/data about it, only a generic SMI reference inside the tools that you can download from the PNY website. If I were you I would try on a machine running XP, if you really want to further experiment, I presume that you can find some friend crazy enbough to let you use his/her XP PC and installing to it an *experimental* driver. Please also keep in mind that even if you successfully manage to run the app, it is a very risky program to run, and if used incorrectly it may lead to an unrecoverable device. jaclaz

Just for the record, there is at least another tested and working FAT32 "large" formatting freeware. Here: http://tokiwa.qee.jp/EN/Fat32Formatter/ More generally Partition Magic and similar "advanced" tools are only *needed* when you need to do "advanced" things, (like converting a NTFS to FAT32) there is no need for them to simply create or format a partition. There is also an "outsider", which is represented by partition logic: http://partitionlogic.org.uk/ which, since it contains it's own OS can be a solution for those that have not a XP machine, but there are not AFAIK "extensive" test results for it's FAT32 creation and formatting. jaclaz

Well, you need to install the "factory driver" and set it to recognize the USB Vid & Pid, by default it is 090C/1000, READ the \smi_mptool_v2.03.34_v12_2\UFD_MP\FactoryDriver\Documents_for_SMI_Driver.doc jaclaz

The "base" windows files, BOOTMGR (and probably the BCD ) must be UNcompressed. Anything you did had nothing to do with the compression of the file. You need to UNcompress the BOOTMGR. The COMPACT command should be available once you have booted from the Windows 7 install disk (which is a kind of PE): http://technet.microsoft.com/en-us/library/cc976800.aspx jaclaz

How would UNIATA for Win9x fit in the equation? http://reboot.pro/2384/ jaclaz

Well NO, this is not "more" info, it is COMPLETELY DIFFERENT, (most probably RIGHT this time ) info. It is more probable that the chip is SMI. It is likely that one of these two may be the "right" tool: SMI MPTool V2.03.34 v12 [sM3255AB] SMI MPTool V2.03.34 v12 [sM3257AA] Besides the already given link, here is another page: http://tagievara.narod.ru/pages/programs/flash.html via google translate: http://translate.google.it/translate?u=http%3A%2F%2Ftagievara.narod.ru%2Fpages%2Fprograms%2Fflash.html&sl=ru&tl=en&hl=&ie=UTF-8 jaclaz

Very strange. 04FC should be "Sunplus Technology Ltd.", which does not produce AFAIK memory sticks, and HP should be 03F0. Please describe that stick, brand, is it this one? http://www.pny.com.tw/pny_en/products_info2.asp?idno=97&pidno=41&pidno3=81 that is from PNY (HP v175w) PNY should have Vid 154B. though there are reports of using different Vid's. The PNY's I have seen use Phison chips: http://www.flashboot.ru/index.php?name=Files&op=cat&id=6 but I have also seen references to some of them using SMI chips: http://www.flashboot.ru/index.php?name=Files&op=cat&id=10 You may want to try some of the apps listed above, but first check if the PNY software: http://www.pny.com.tw/pny_en/download.asp has any reference to the actual chip manufacturer. jaclaz

Yes , I mean you are posting here, so you do have another computer where you can prepare from scratch the USB device using this method: Try using the "stable" 0.2.3 AND NOT the 1.0 beta7 version on first attempt. jaclaz

There are 2 (two) ways to have multiple partitions accessible on a USB stick on NT based systems: flip the removable bit install a filter driver To flip the removable bit you may find - besides the "Lexar" utility, the manufacturer's "Mass Production Tool" for the USB controller used in your USB stick. You can install the filter driver "on the fly", but this is considered "advanced" and requires a good understanding of the procedure and of English. I am sorry for your English (or lack of it), but this is an English forum, and it is unlikely that you will find here help in your native language (whatever it is), You may want to: find a suitable technical forum in your language or find a friend that better understands English that can help you In other words, you are trying to do an "advanced" something without the needed tools (basic English knowledge), it will take you some time and a lot of efforts, maybe it would be easier if you buy a "Lexar compatible" new USB stick. If you want to try finding the appropriate Mass production tool, get chipgenius: http://reboot.pro/4661/ fint the Vid/Pid of your stick see if you can find a suitable tool here: http://flashboot.ru/index.php?name=iflash or post an EXACT description of the stick AND it's Vid/Pid, and I may try helping you in the search. jaclaz

If we are allowed to go Commercial , here: http://www.dosusb.net/ (the "original" home page is sort of botched): http://www.georgpotthast.de/ I missed if these also were tried : http://bretjohnson.us/ (same info on the page given by dencorso direct links for the record) jaclaz

Read the read-me-first AND the FGA's: leave the disk alone, try a loopback test. (this won't be affected by the "wrong" TTL levels -if any) Take into account to try powering the adapter at 3.3v (NOT with a 3V battery), but it "looks" like a 5V level TTL only ("wrong"). About the adapter, this one: http://cgi.ebay.de/FTDI-USB-UART-seriell-Adapter-AVR-PIC-TTL-RS232-/280604136107?pt=Wissenschaftliche_Ger%C3%A4te&hash=item41554f52ab jusdging from it's manual: http://www.sigma-shop.com/manuals/ftdi_manual.pdf is of the "right" kind. jaclaz

You can use a filter driver. See here (please read the WHOLE thread): jaclaz

Again, no. Well, actually it is a PE 3.x (i.e. it is a Pre-installation Environment), you can think of it as a minimal, portable OS that is used to actually run the installers, and yes, of course it can be modified/customized. The general idea is given in this topic (sticky): (which is however seemingly abandoned by original Author and might use new ideas/contributions) ...hint, hint... jaclaz