Hard drive being dropped or knocked over.

There have been numerous times where the drives have accidentally been pulled or knocked over by the user. Most external drives are engineered to stand upright on a flat surface, which could lead to it tipping over. Best way forward is to lay it flat or store it somewhere away from edges of a surface.

Twisted Electrical Cords and Cables

Loose cords and electrical cables build over time under desks and small corner tables where we tend to have our laptops and desktops plugged in. These are the perfect breeding grounds for a mishap for external drives being pulled off a surface onto the floor. It only takes a small fall for the drives to suffer a failure. It is important to keep these areas nice and tidy, maybe use cable ties or clips to keep things neat. Make sure power cables are not stretched taught to the wall socket as a jittery connection could cause a power surge. Always keep the cords away from traffic in the room to prevent the drive from being accidentally pulled down to the floor.

Inserting the wrong power adaptor

Many times over you will have three to four different power adaptors in the same areas as where you plug your external drive in, so it is important to make sure you use the correct one! There have been many instances where a user has plugged a different power cord into their external hard drive causing a power surge which will overheat the drive. A good idea is to label the cords with a small white label to ensure you are using the correct cord for the device you are using.

Power Surges from Lightning Strikes

If you live in an area that has a lot of lightning storms, it is important to unplug your external storage hard drive during these storms (or when not in use). A power surge from lightning strikes can cause external hard drive failure by overheating the drive or by causing the heads to crash to the disk surface. This can result in the dreaded clicking sound otherwise known as “click of death” and you will need hard drive data recovery.

Contact us if you have experienced external hard drive failure.

Great care is necessary with the replacement head stack to ensure that the air bearing surfaces of the heads (actually the “sliders”) do not touch each other. This is because they can easily damage each other or even stick together. When removing the damaged heads, it is important not to drag them across the disk and cause more damage. Ideally custom tooling, such as a special comb, will be used to lift the sliders off the disk surface before removing them.
Equal or greater care is necessary when loading the good heads back onto the damaged drive’s media.

A new manufacturing process that will put 640GB on a single HDD platter, has allowed manufacturers to dramatically increase storage capacities drastically. As the 3.5-inch drives used in desktop PCs typically hold four or five platters, that suggests a whopping 3.2TB hard disk space of storage on a single drive. Looking at the likely timing, the new manufacturing technology could put the giant drives in shops as early as February next year.

Manufacturers such as Western Digital, Samsung, TDK and Hitachi are all introducing 2TB+ drives and the other manufacturers will not be far behind. However if you are storing this amount of data, is a single HDD the answer? Given the well-documented manufacturing problems with many of the 1TB drives, should users be more cautious?

Well in our opinion the answer should be a resounding yes. New technology is always fraught with teething problems and it often takes several generations of media, before a reliable product is found. Although companies like Hitachi are spending large research budgets on developing this technology, the mean time before failure (mtbf) cannot be accurately calculated until the product has been on the market for a while and all failures carefully analysed by the manufacturer.

Given the minute tolerances within a HDD and the associated heat dissipation issues, my advice would be to stick with lower capacities for the time being and even RAID them together for added security. RAID data recovery is always a possibility if the RAID drive fails.  Hitachi still remain at the top of our list of the most reliable manufacturer, but I would sit tight until the new technology is proven.

The hard drive itself consists of two main parts.
The mechanical part including the heads and disk assembly.
The PCB electronic board.

As these parts are physical they can be subject to failure and result in the need for hard drive data recovery.

The PCB is made from different components, each chip, resistor, fuse etc all have their own tasks and duties. Later in the week we will discuss the PCB more in depth. At this stage you should be aware that as this is the electronic circuitry of the hard drive it is subject to electrical failures.
Inside the hard drive we can see many more physical components. See fig 1 for the view inside a hard drive.

Fig 1

The internal parts of the HDA (Hard Drive Assembly) are all mechanical and subject to mechanical failure.

How can these components fail? Wear and tear is common with any moving component; the platters themselves spin at a fast rate approx 7,200 RPM in a desktop hard drive. Motors can fail the heads can come off or become damaged as a result of dropping the hard drive. The heads can even fail due to mishandling and receiving electrostatic charge from a human via its PCB.

The hard drive is a piece of specialist hardware that should be handled with care, using all the precautions you can prevent many human caused problems.

Hard drives can also suffer serious damage from excess heat; this is something that can be prevented by the user, after all prevention is better than cure.

Common mechanical failures.
* Dust enters the hard drive assembly; this will affect the heads and / or platters.
* Dropping, knocking or shocking the hard drive whilst in use or turned off.
* Power surge or lightning.
* General wear and tear as each part wears over time.

If you ever suspect your drive is failing mechanically, please shut down the machine and contact us.

U Series 5
Typical problems with this drive family are diode failures which affect the voice coil and spindle, and failure of the read/write channel.

Barracuda ATA II
This family tends to suffer from problems with the stabilizers of the spindle motor circuit either by burning or shorting out. The latching transistors of the read/write channel circuitry, the symptoms of which are the drive produces knocking at a frequency of 5Hz.

Barracuda ATA III
It is not untypical of these drives to suffer from password protection due to an error or corruption accessing the RAM. The RAM and or ROM chips can fail. Pass through diodes in this family on the 12V rail are also common problems we can encounter, the diode does not have to be necessarily have to be replaced for functionality of the drive. Just removing it works in some cases.

Barracuda ATA IV
The 12V pass through diodes are again an issue in this family on the circuit where the spindle and VCM reside. These problems can result in head positioning errors if the read/write channel is also affected. This family suffers from spindle seizures and failure of the preamplifier chip. (In these cases there is usually no problem with the PCB). You will need to examine the preamp in the clean room and decide for yourself whether to replace or swap the head stack.

Barracuda ATA V
The +12V and +5V diodes are prone to burning out. This family tends to suffer from malfunctions which cause the commutator and preamplifier chip to have the power supply disrupted. The motor control chip and voice coil end up burning out along with the other components required to function.

U Series X
This drive also has problems with the +5V and +12V diodes, the contact pads nearby also may suffer and the tracks can become damaged. If the power supply is +5V greater than the +5V required then sometimes this family will not operate or may produce scratching sounds from its heads. To repair a fault of that type we just need to lower the voltage -5V.

The drive came from a external hard drive casing. This was tested and we found it to be faulty as well. We can say that the drive suffered from a intense power spike. This spike has caused damage to the external hard drive casing, the PCB of the hard drive and also the spindle motor of the hard disk drive.

We will have to replace the spindle bearing motor of the hard drive and various power chips on the PCB. Once this is complete, we can mount the drive back into the secondary machine and image the disk over to a new working hard drive.

Very important when suffering from a power spike, if you see or smell any burning, please switch the hard drive off and unplug the power. Do not attempt to replace any of the parts on the drive, and even more important, DO NOT open the hard drive itself. This will lead to a complete data loss or corruption!

For more info, please visit us at MVI Data Recovery.

Background: Failures on WD hard disks can sometimes be attributed to a corruption of service data located in the firmware zone of the hard disk. This corruption often involves a table commonly known as the P-List and in many cases, it is impossible to read the original contents of the P-List once the drive has failed.

What is the P-List?: The P-List is a growing table of defects that the drive uses to avoid bad sectors that have developed since its factory formatting, the drive marks a sector as bad and then allocates a reserve sector from a special area for the drive to use so that the LBA of the drive does not change.
Operation: At start up, the WD hard disk loads the service data from its firmware zone into RAM and calculates the checksums of certain critical modules before allowing the drive to report on readiness. One of these modules is the P-List. When one of these drives fail with a P-List fault it will:

• not report readiness
• will be recognised by the PC BIOS by its factory alias
• refuse to allow access to user data

Possible repair and data recovery: Despite P-List inaccessibility, successful recovery of the drive and the data on it is often possible as a complete copy of the defects stored within the P-List is also located in the Pivot Defects table, the same drive table that is actually used by the drive when loading the defects and translating the Logical Block Addressing (LBA) of the drive at start-up.

The only reason that the drive will not start with a corrupt P-List is that it does not have the ability to repair the P-List itself and thus has nowhere to store any new defects that may develop during use. Because of this unique feature of the WD hard disk it is almost always possible to perform a near 90% data recovery of the data from the drive by using a multi-stage repair process.

MVI Data’s customers have sent in a much higher numbers of failed Seagate 2.5in SATA drives made in China and loaded with firmware version 7.01 than of any other current hard drive model. We’re getting 20-30 times more failed drives of this kind than others.

The drives of this kind sent to MVI Data for data recovery exhibited the same damage: the read/write heads have failed mechanically causing them to gouge deep scratches in the delicate data-storing surface of the disk platters. In almost all cases, this renders the drive useless.

And the drives in question all came from Apple machines. MVI Data only gets the drives, not the host computers, so we can’t specify exactly which Macs they came from, but since they’re all 2.5in models, that suggests not only laptops like the MacBook and MacBook Pro, but also potentially desktop Macs that use laptop-oriented components, like the Mac Mini.

We blame the problem what the described as “poor quality control in Chinese hard drive factories” – an issue we maintain affects other hard drive makers in addition to Seagate. We also warn all hard drive buyers to avoid HDDs manufactured in China.

“We believe that any sizeable manufacturer would by this stage be aware of such a problem and issue a product recall notice, or an offer to have the drive exchanged for a suitable alternative at their own expense.”

Users who find they have such a drive, should back up their data and consider replacing it with “an alternative drive, or a retail-version Seagate drive”.

Mac OS X’s System Profiler utility provides a way to identify the make and type of HDD installed in a Mac, but not directly its place of origin.

Update System Profiler also shows the HDD’s firmware: under the Serial ATA section, look up the Revision entry – that’s the firmware version.

Apple and Seagate did not respond to our requests for comment.

A discussion thread on Apple’s support site highlights a number of MacBook users who have experienced what appear to be hard drive failures. However, only five cases explicitly identify the failed drive as a Seagate model, and of these none can be stated categorically to be the product of Chinese factories.

Some of the failures are attributable – indirectly – to Mac OS X’s Safe Sleep feature. This copies the contents of a laptop’s memory to the hard drive whenever the computer is put to sleep.

Related stories

• Apple ‘looking into’ duff Chinese hard drive claims
• Buffalo unveils ‘world’s most capacious’ pocket hard drive
• WD ‘first to ship’ 320GB laptop hard drive
• Seagate to close Northern Ireland plant
• Dealer debuts campest MacBook Pro in the universe
• Apple nears MacBook revamp release?
• Seagate admits Flash-fitted HDDs fall short of promised benefits
• Apple sells one in six US laptops
• Alleged Apple ultra-thin MacBook rumour restoked

To be fair to Apple, it does warn users not to move the machine until this process is complete – at which point the light on the lid catch begins to pulsate – but it’s easy for users to grab their laptops and go as soon as the lid’s down.

Unfortunately, at that point, the drive’s heads may by busily moving back and forth across the platters, and any movement risks the two physically touching, with the potential of data loss and even disc damage of this kind . This is true of all hard disks, as proponents of solid-state storage like to point out.

Mac users can disable Safe Sleep by following Mac OS X Hints’ procedure here. However, doing so risks losing data from memory should the laptop’s power supply be cut.