 The magazine of the Melbourne PC User Group The good old days - magnetic tapes Ewart Matthews ewartm@melbpc.org.au |
Magnetism has fascinated scientists for a long time, from days long ago when they observed the behaviour of a certain kind of stone, known as a "lodestone" (probably from an old Saxon word for "leading stone").
Known now as "magnetite," a piece of this, suspended on a floating disc of wood would always align itself in a certain direction. In the northern hemisphere this pointed to the North Star; the ancients found that they could use this to navigate accurately, hence the name "leading stone."
The lodestone also attracted iron particles to itself. As the schemes of nature became better understood they realised that the north-seeking lodestone was attracted by the earth's magnetic field.
A magnetic field can be readily seen if you put a child's toy magnet under a sheet of white paper and then file a soft iron object over it, causing a rain of iron filings to fall freely onto the paper. The child's toy magnet has a horseshoe shape to bring the two ends of the bar magnet close together, concentrating what is a very weak field across a small gap.
Very early in the investigation of electricity, scientists discovered that a soft iron bar could be magnetised by winding a coil tightly round the bar and passing a current of electricity through the coil. However, the field of the magnet collapsed if the current ceased.
If they used a harder iron, or an alloy of steel, more current was needed to induce magnetism but when the current was switched off, vestiges of magnetism remained.
As understanding of the nature of matter increased and molecular theories emerged, they realised that the magnetic flux (the field round the magnet or energised coil) aligned the molecules within the iron in the one direction, much as the iron particles deposited by the file aligned along the lines of force surrounding the magnet.
Magnetism can also be induced into a hard iron bar by stroking a strong magnet along its length, aligning the molecules in the same direction.
During the 1960s, experiments were made in the production of what are now called "ferrite" materials.
An alloy of steel ground into a fine powder could be included in a flexible compound to make a paste that could be readily shaped and allowed to harden. This process was used to form ferrite rods in the antenna coils of small radios - a tiny rod could replace long wire aerials, often strung across the rooftops.
The development of plastics created all sorts of products, such as wrapping sheets for food. It was a short step for someone to coat a sheet of tough plastic with a flexible paste containing the steel filings that could be magnetised in specific areas. This formed much more intense fields than metallic iron and the fields lasted much longer, almost indefinitely.
One bit of information consists of a tiny group of magnetised particles with all its molecules aligned magnetically in the same direction, as in a bar magnet.
An early commercial application was in the magnetic "stripe" on accounting machine cards where monthly balances were recorded.
The principle is still alive and well, each credit card in your wallet carries a magnetised stripe embedded in its surface, containing your account information.
Magnetic tapes soon replaced paper tapes and cards, and the appearance of magnetic tape handlers in computer rooms quickly followed.
Tapes have one weakness. If the information on the tape is sorted and is in serial form ready for processing all is well, and they perform very effectively. If, however, the tape contains "raw data," neither sorted nor processed in any way, it is time consuming to search through a 2400 foot (730 m) tape for a record.
This would take up to three minutes and computer rooms were cluttered with rows of tape handlers, all frantically winding and rewinding as the processor searched for the necessary data.
Many solutions were sought, raw data was searched and arranged (or collated) offline, requiring yet another extension of the computer room for this purpose and great storehouses of racks of reels of tape began to appear on the sites.
The "read" and "write" heads on each unit contained small horseshoe magnets, similar to the child's toy, except that each tiny head - about 1/64 inch (0.4 mm) thick - had coils around its centre and the magnetism could be reversed at will.
Imagine 12 of these packed together into a block about 1.5 inches (37 mm) and visualise a tape half-an-inch (12 mm) wide (a standard tape), moving past the gaps in the heads at about 150 inches-per-second (3.8 metres-per-second).
Two sets of these, one behind the other, gave "read" and "write" ability and this could be carried out at quite respectable speeds passing 12-bit "bytes" of information to and from the processor at about 10 microsecond intervals.
When operating on tapes the internal processor clock would be switched to emit 10 microsecond timing pulses and the CPU would slow down to tape speed.
The "write" head was placed first so that the data it wrote was read and checked just after it was written, as it passed the "read" head. All data was checked across the tape by computing and recording parity checks for each character. As the record was written along the tape a Longitudinal Redundancy Check was calculated including every bit. When the record was complete this LRC character was written after the last data character. As each record was read from tape these checks were recomputed and compared - if the tests failed, an error was indicated preventing faulty data getting to the processor.
Usually the 12 bits across the tape formed one character but there were many variations among manufacturer.
Some tape units wrote data using one head for each record, so it was recorded in one uninterrupted line of bits down the length of tape. This meant that 12 records could be processed at one time, each quite independent of the others and so the recording rate could be increased and the data packed more tightly.
Punched paper tape and punched card machines disappeared when data collection switched to tapes at point of entry. As accounts or wages, for example, were processed, the data was recorded on a small magnetic tape, in the office as part of an accounting machine.
Try to imagine all the data transmitted today over the phone lines being handled as described above!
Reprinted from the May 1996 issue of PC Update, the magazine of Melbourne PC User Group, Australia
