Encoding Magnetic Stripes

By Dr. Brad Paulson, Thor Engineering, ICMA Standards Representative

If we consider a magnetic stripe on a card as tiny bar magnets lying end-to-end, a portion of an encoded stripe could be illustrated by the representation in the first figure. In this example, the electric current in the encode head coil changed direction twice as the encode head passed over the magnetic stripe, as illustrated at the top of the figure. Notice the change in the magnetic polarity in the magnetic stripe where the change in encode current occurred. This change in polarity is called a flux transition and it is what a “read” head detects as it passes over the magnetic tape; the resulting signal from the read head is represented at the bottom of the figure. It is the change in direction of magnetization (flux change) that is used to encode information onto the card. Unlike an analogue recorder, which is used to record sound, a card reader ignores the strength of magnetization – the pulse only needs to be sufficiently strong and constant for reliable operation. It is the simple presence or absence of pulses that represent stored data.

As data is stored digitally, that is ZEROs and ONEs, an obvious convention would be to have one direction of magnetization for a zero, and another for a one. However, the consequence would be that adjacent ones would run together, as would adjacent zeros.

This would make it impossible to know how many of either were included in a long sequence of uniform magnetization, so a more reliable scheme was required. Although several schemes used to record binary data in the computer industry, the International Standards Organization (ISO), has defined two-frequency coherent phase recording, commonly called frequency – double frequency (F2F) recording, as the encoding scheme for cards in parts 2, 6, and 8 of ISO/IEC 7811. By placing a regular series of “clock” transitions on the magnetic stripe at regular distances, F2F encoding provides for self-clocking data; the spacing between clock transitions is known as the interval. No additional transitions are added to denote a ZERO, but another transition is placed between the clock transitions for a ONE, which is illustrated in the second and third figures; the spacing between a clock interval and the following data transition is known as a subinterval. The key feature of self-clocking data is that the data (bits) can be extracted from the data stream without the need to control the speed of the magnetic media past the encoding head. Thus, card swipe readers, where a human hand is passing the card through the read slot, can work regardless of how fast or slow the card is passing through the slot.

The bit spacing of encoded data is defined in the magnetic stripe parts of ISO/IEC 7811 Identification cards – Recording technique. Parts 2, 6, and 8 define the bit density of tracks 1 and 3 as 8,27 bits/mm (210 bits/inch), which is 0,12092 mm (0.0047619 inches) for zero bits and 0,060459 mm (0.0023809 inches) for one bits (half the zero bit distance). The same parts define the bit density of track 2 as 2,95 bits/mm (75 bits/inch), which is 0,33898 mm (0.013333 inches) for zero bits and 0,16949 mm (0.0066665 inches) for one bits (again, 1/2 the zero bit distance). As a point of interest, in 2004, ISO/IEC 7811-7 Identification cards - Recording technique - Part 7: Magnetic stripe - High coercivity, high density was published, specifying the modified frequency modulation (MFM) recording technique and defining six encoding tracks with bit density of 40 bits/mm (1016 bpi), which is 0,025 mm (0.00098425 inches) for zero bits and 0,0125 mm (0.00049213) inches for one bits.

During data recording, the encoding equipment is responsible for placing the flux transitions at the proper distances from each other. Variations in the proper distance between flux transitions is commonly referred to as “jitter”. Jitter is more precisely stated in ISOIEC 7811 as Flux Transition Spacing Variation (Bin), Subinterval Spacing Variation (Sin), Adjacent Bit Cell Variation (Bin+1), and Adjacent Subinterval Spacing Variation (Sin+1). Graphically, the definitions for Bin, Bin+1, Sin, and Sin+1 are illustrated in the fourth figure.

Signal amplitude and flux transition spacing variation provide the foundation of the encoding specifications included in parts 2, 6, and 8 of ISO/IEC 7811. To improve the reliability of decoding the encoded data, clock transitions need to be evenly spaced and data transitions need to be centrally located between clock transitions. Remember, cards are not just mini-billboards, but they are also critical parts of a worldwide identification and payment system, which depends on accuracy and reliability.