What Is A Hard Disk?
Click on image to view bigger version
Once known as the Winchester drive, the hard disk or hard drive is an essential hardware that every computer desktop and server contain. And its functionality, is to store vast amount of digital data in a non-volatile form, so that data can be retained when the computer is powered off. Data is again accessed when the power is on.
How Does A Hard Disk Work?
If you are to dismantle the hard disk drive by opening the top casing (after removing all the necessary screws), the first thing you'll see is a spindle holding one or a number of mirror-like hard rotating platters (commonly called data platter). The platters could be made to spin at an extremely high speed, technically between 5,400 to 10,000 revolutions per minute (RPM). An extremely thin magnetic coating is layered onto the surface of the platter that is polished to mirror-type smoothness. The platter is usually made of glass or ceramic (modern platter may use titanium). Unlike a floppy disk, the platter cannot be bent or flipped and hence we coin it as "hard disk" or "hard drive".
Commonly a hard disk contains 1 to 10 identical platters that are stacked in parallel to form a cylinder. There is usually one Read Write (RW) head designated per platter face, and each head is attached to a single actuator shaft which moves all heads in unison and performs a uniform synchronous motion during reading or writing of data. Data on a hard disk is stored in microscopic areas called magnetic domains on the magnetic material. Each domain stores either a 1 or 0 value. Similar to a floppy disk, a hard disk records its data in concentric circles or tracks, which are numbered from the outermost edge to the innermost edge of the platter. These tracks are further subdivided into smaller units called sectors which typically store 512 bytes of data each. Zoning may be needed to further optimise the data storage as the outer circumference would normally pack more sector units than the inner circumference.
Read Write Head
The RW head is the key component that performs the reading and writing functions. It is placed on a slider which is in term connected to an actuator arm which allow the RW head to access various parts of the platter during data IO functions by sliding across the spinning platter. The sliding motion is derived by passing a current through the coil which is part of the actuator-assembly. As the coil is placed between two magnets, the forward or backward sliding motion is hence derived by simple current reversal. This location of the platter (just like the landmark along the road) is identified and made possible by the embedded servo code written on the platter.
To write a piece of information to the disk, an electromagnetic flux is transmitted through the head which hovers very closely to the platter. The RW head suspends on a thin cushion of air which the spinning platter induces. This designed distance between the head and platter is called the flying height. It can literally measure to a few millionths of an inch. A good analogy is try to imagine flying a Boeing 747 with about 6 inches above ground level.
It is therefore easy to understand that if the RW head happens to "knock" on the spinning platter out of design specification, a Read Write head crash occurs.
Read Write Function of Disk
As the head writes data onto the disk, it changes its magnetic polarization to induce either a one or zero value. During a read request, data is interpreted when the magnetic fields on the platter brings about an electrical change (as a result of change in electrical resistance of some special material property) in the read-head that passes over it. These electrical fields are then encoded and transmitted to the CPU to be processed and read by the system.
Parking of RW Head
Click on image to view bigger version
When the computer is switched off, the head is usually pulled to a safe parking zone to prevent the head from scratching against the data zone on platter when the air bearing subsides. This process is called a parking and different techniques have been implemented in various hard disks to handle the take offs and landings. In a Ramp load/unload design, a lifting mechanism parks the head outside of the platter onto a "parking bay" prior to a shutdown. It then automatically unparks and relocates itself above the disk platter when the platter spins up to appropriate rotational speed.
Hard Disk Controller PCB Board
A hard disk also contains a pcb controller circuit board that regulates data traffic. It ensures massive data to be streamed in and out of the disk smoothly. A logic board that sits under the drive controls and connects the spindle, head actuator, and various disk functions of the disk. Embedded with a micro-controller, it executes self-diagnostics test and cleans up data working area in the memory and all internal chip bus in the hard drive when it powers up.
Today's Disk & Data
Storage capacity is measured in terms of GB (Gigabyte). 1 Gb roughly equals to 1000 MB (Megabyte). A typical desktop machine purchased today carries a hard disk of at least 80 to 400 gigabyte. Computer data is stored onto the disk logically in the form of files or simply, a string of bytes. When a program requests for a file, the hard disk retrieves its bytes and sends them to the CPU to be processed.
There're two popular ways to measure the performance of a hard disk:
Data rate is also known as data transfer rate or throughput. It refers to the speed with which the data can be transmitted from one device to another. Data rates are commonly measured in megabytes (MB) per second. The data rates for a typically hard disk ranges between 5 to 40 megabytes per second.
Seek time refers to the total span of time between a file request by the CPU to the point when the first byte of the file is sent to the CPU. This is one of the several delays associated with reading or writing data on a computer's disk drive. In order to read or write data in a particular place on the disk, the read/write head needs to manoeuvre to the precise location of the disk platter. This process is known as "seeking", and the time it takes for the head to move to the right place is referred to the "seek time". A typical seek time for a hard disk is about 9 to 15 ms.
Majority of all hard disk today support a technology known as S.M.A.R.T. (Self-Monitoring, Analysis, and Reporting Technology) which helps to predict imminent disk failures so that users can be alerted to take preventive actions before the disk fails completely.
Generally, there are two classes of failures, namely Non-Predictable failures and Predictable failure. Non-predictable failure associates with abrupt and unpredictable damages such as a sudden power surge that burns an electronic component. On the other hand, Predictable failure monitors mechanical wear and aging which takes place over time. In a nutshell, S.M.A.R.T functions like a monitoring device that reads and analyses the performance of a disk. If a gradual defection is observed (which can lead to a likely disk failure), a message will notify you to backup your data and and replace your disk.
You could read more details on hard disk SMART here.
|DATA STORAGE MEDIA|
|Information on the different types of storage media|
HARD DISK DRIVE
|DATA RECOVERY RESOURCES|
|Do-it-yourself troubleshooting tips and knowledge on storage media types.|
|ADRC provides data recovery services for a wide range of storage media.|