As you may know, first step of diagnostics on Seagate drives is to check terminal output after power on and check is drive able to read system files.
But there a lot of firmware errors when you can’t access terminal – got T> string after pressing Ctrl+Z. In terminal log you have last message something like:
LED:000000CC Faddr:0026A86F
or
LED:000000EE FAddr:00358846
At this situation you should try to short read channels.
Compact Flash cards – are the most fastest memory cards in the market. The main secret of CF – is a very powerful controller, which has direct access to several physical memory chips, which are placed on the CF PCB. The architecture of Compact Flash controllers is much closer to SSD CPU than to common USB Flash drive controllers.
One of the most popular Compact Flash card controller is Silicon Motion SM2234H. Usually this controller works with 4 physical memory chip, and provide a high speed of reading and writing operations.
When customer get the recovery task based on this controller, he make a wrong steps of recovery and usually never get a good result. In this article we will try to explain, how to deal with this controller, and what exactly you should do for getting user data from such case!
Sometimes PC-3000 database can get a problem with the folders location.
One or more folders from the root are moved to the other place.
For example on this screenshot you can see that HDD profile folders are moved in to the SSD tool profile folder:
Such behavior caused by problems with database service in OS or critical shutdown of this DB.
How to solve such problem?
Using Data Extractor RAID Edition, we are able to identify special RAID configurations and extract data from virtual machines on RAID arrays.
Data Extractor RAID Edition supports SPECIAL-VMFS level for manipulations with VMFS spanned across multiple disks
These virtual machines created with VMware or other products typically use virtual disks.
The virtual disks, stored as files on the host computer or on a remote storage device, appear to the guest operating systems as standard disk drives.
We will show in this topic how to assemble RAID with virtual machines on 4 disks with VMFS volume.
We do this using a little knowledge of the RAID array architecture which we then combine with the metadata autodetection of Data Extractor RAID Edition.
Nowadays a lot of modern NAND flash devices use a new type of architecture, where the interface, the controller and memory chips are integrated into one common ceramic layer. We call it Monolithic structure.
Until recently, all memory cards like SD, Sony MemoryStick, MMC and others, contained a very simple “classic” structure with separated parts – a controller, a PCB and a NAND memory chip in the TSOP-48 or LGA-52 package. In such cases, the full process of recovery was very simple – we just unsoldered the memory chip, read it in PC-3000 Flash, and made the same preparation as with common USB Flash drives.
PC-3000 Circuit Board adapter is a really good and useful thing. First of all with the help of this adapter you can read the data from monolithic devices like mSD, SD, xD, MS and UFD. But at the same time, Circuit board works great in case if for some reason, we need to read data from common chips like TSOP-48, TSSOP-48 (short package) or BGA-52 with physical damages of pin legs.
Lets imagine that during chip unsoldering you damaged several legs of TSOP-48 memory, and you are not able to put it inside PC-3000 Flash reader. What should you do in this case? Of course you can try to repair this legs, make a serious job under microscope and restore the functionality of damaged legs. But this process will be very complicated, and during next chip installation into Flash reader socket, you can damaged them again.
The method can be used to recover original ROM image and then write it into donor PCB when original PCB was damaged, lost, or original ROM was lost.
The “Building ROM image from SA data” tool was fully revised in the PC-3000 6.0 software version so the process is different from 5.X versions.
We have another article about the process for PC-3000 v6.X that can be found here.
The procedure allows to recover full image of original ROM and all ROM modules.
The procedure is applicable to almost all WD Marvell architecture drives with internal (integrated into processor) ROM, that has SA modules 102, 103, 105, 107 and 109. Some firmwares has SA module 109 that doesn’t contain necessary data – ROM image. If we open such 109 module in HEX-editor we’ll see that it is filled with zeroes. In such cases the utility can’t built the image and gives error.
This procedure is not applicable to drives with external ROM IC (U12) as well as to classical WD Caviar architecture drives(non-Marvell).
To perform the procedure the SA and at least one system head should be OK.
Block rotation – it’s a special operation which is typical for some of Phison, noname, Phison-clone and Sandisk controllers. Basically – it’s some kind of operations where inside block, one group of pages are always rotated. Controller of Flash drive just choose the position inside each block, split the block in this position, and move the end part of it in the beginning – as the result, all headers which must situated in beginning of block like for example – Master Boot Record, moving somewhere in the middle of block. All files become split too – and their integrity usually become not complete even after all preparations!
Sometimes after heads stack replacement the new heads stack works not stable. That occurs because every heads stack has its unique adaptive parameters. Those parameters are stored in module 47 of ROM.
The adaptive parameters are more different from one batch of drives to another and is connected with date of manufacture. That is why it is recommended to select donor drive for heads stack with the closest date of manufacture possible. Some heads stacks from exactly the same drive as your patients can be totally incompatible due to these parameters.
In some cases a drive with one or more malfunctioning heads can finish the initialization process without any visible errors. But when we try to read user area it results in error and no user data can be read. For example it can occur when one or more heads have problems with writing.
The procedure allows to get access to data on all normally working heads of the drive when SA and at least one of system heads are OK.
The main idea of this method is to to trick the drive calibration procedure through substitution of a normal head instead of a malfunctioning one. Consequently it works with drives using two or more heads.
Also it can be applicable when one of the system heads is damaged. In such cases the patient drive may produce clicks, knocks and turn down the spindle. This case will be reviewed in details in the next article.
