Microwaves
EAS systems in the microwave range exploit the generation of harmonics at components with nonlinear characteristic lines (e.g. diodes). The harmonic of a sinusoidal voltage A with a defined frequency fA is a sinusoidal voltage B, whose frequency fB is an integer multiple of the frequency fA. The subharmonics of the frequency fA are thus the frequencies 2fA, 3fA,4fA etc. The Nth multiple of the output frequency is termed the Nth harmonic (Nth harmonic wave) in radio engineering; the output frequency itself is termed the carrier wave or first harmonic.

In principle, every two-terminal network with a nonlinear characteristic generates harmonics at the first harmonic. In the case of nonlinear resistances, however, energy is consumed, so that only a small part of the first harmonic power is converted into the harmonic oscillation. Under favourable conditions, the multiplication of f to n × f occurs with an efficiency of η = 1/n2. However, if nonlinear energy storage is used for multiplication, then in the ideal case there are no losses (Fleckner, 1987).

Capacitance diodes are particularly suitable nonlinear energy stores for frequency multiplication. The number and intensity of the harmonics that are generated depend upon the capacitance diode’s dopant profile and characteristic line gradient. The exponent n (also γ) is a measure for the gradient (= capacitance–voltage characteristic). For simple diffused diodes, this is 0.33 (e.g. BA110), for alloyed diodes it is 0.5 and for tuner diodes with a hyper-abrupt P–N junction it is around 0.75 (e.g. BB 141; ITT, 1975).

The capacitance–voltage characteristic of alloyed capacitance diodes has a quadratic path and is therefore best suited for the doubling of frequencies. Simple diffused diodes can be used to produce higher harmonics (Fleckner, 1987).

The layout of a 1-bit transponder for the generation of harmonics is extremely simple: a capacitance diode is connected to the base of a dipole adjusted to the carrier wave (Figure 3.5). Given a carrier wave frequency of 2.45 GHz the dipole has a total length of 6 cm. The carrier wave frequencies used are 915 MHz (outside Europe), 2.45 GHz or 5.6 GHz. If the transponder is located within the transmitter’s range, then the flow of current within the diode generates and re-emits harmonics of the carrier wave. Particularly distinctive signals are obtained at two or three times the carrier wave, depending upon the type of diode used.

Transponders of this type cast in plastic (hard tags) are used mainly to protect textiles. The tags are removed at the till when the goods are paid for and they are subsequently reused.

Figure 3.6 shows a transponder being placed within the range of a microwave transmitter operating at 2.45 GHz. The second harmonic of 4.90 GHz generated in the diode characteristic of the transponder is retransmitted and detected by a receiver, which is adjusted to this precise frequency. The reception of a signal at the frequency of the second harmonic can then trigger an alarm system.

If the amplitude or frequency of the carrier wave is modulated (ASK, FSK), then all harmonics incorporate the same modulation. This can be used to distinguish between ‘interference’ and ‘useful’ signals, preventing false alarms caused by external signals. In the example above, the amplitude of the carrier wave is modulated with a signal of 1 kHz (100% ASK). The second harmonic generated at the transponder is also modulated at 1 kHz ASK. The signal received at the receiver is demodulated and forwarded to a 1 kHz detector. Interference signals that happen to be at the reception frequency of 4.90 GHz cannot trigger false alarms because these are not normally modulated and, if they are, they will have a different modulation.

Frequency Divider
This procedure operates in the longwave range at 100–135.5kHz. The security tags contain a semiconductor circuit (microchip) and a resonant circuit coil made of wound enamelled copper. The resonant circuit is made to resonate at the operating frequency of the EAS system using a soldered capacitor. These transponders can be obtained in the form of hard tags (plastic) and are removed when goods are purchased.

The microchip in the transponder receives its power supply from the magnetic field of the security device (see Section 3.2.1.1). The frequency at the self-inductive coil is divided by two by the microchip and sent back to the security device. The signal at half the original frequency is fed by a tap into the resonant circuit coil.

The magnetic field of the security device is pulsed at a lower frequency (ASK modulated) to improve the detection rate. Similarly to the procedure for the generation of harmonics, the modulation of the carrier wave (ASK or FSK) is maintained at half the frequency (subharmonic). This is used to differentiate between ‘interference’ and ‘useful’ signals. This system almost entirely rules out false alarms.

Frame antennas, similar to those known from RF systems, are used as sensor antennas.

Electromagnetic Types
Electromagnetic types operate using strong magnetic fields in the NF range from 10 Hz to around 20 kHz. The security elements contain a soft magnetic amorphous metal strip with a steep-flanked hysteresis curve (see also Section 4.1.12). The magnetisation of these strips is periodically reversed and the strips taken to magnetic saturation by a strong magnetic alternating field. The markedly nonlinear relationship between the applied field strength H and the magnetic flux density B near saturation (see also Figure 4.52), plus the sudden change of flux density B in the vicinity of the zero crossover of the applied field strength H , generates harmonics at the basic frequency of the security device, and these harmonics can be received and evaluated by the security device.

The electromagnetic type is optimised by superimposing additional signal sections with higher frequencies over the main signal. The marked nonlinearity of the strip’s hysteresis curve generates not only harmonics, but also signal sections with summation and differential frequencies of the supplied signals. Given a main signal of frequency fS = 20 Hz and the additional signals f1 = 3.5 and f2 = 5.3 kHz, the following signals are generated (first order):

f1 + f2 = f1+2 = 8.80 kHz

f1 − f2 = f1−2 = 1.80 kHz

fS + f1 = fS+1 = 3.52 kHz and so on

The security device does not react to the harmonic of the basic frequency in this case, but rather to the summation or differential frequency of the extra signals.

The tags are available in the form of self-adhesive strips with lengths ranging from a few centimetres to 20 cm. Due to the extremely low operating frequency, electromagnetic systems are the only systems suitable for products containing metal. However, these systems have the disadvantage that the function of the tags is dependent upon position: for reliable detection the magnetic field lines of the security device must run vertically through the amorphous metal strip.

For deactivation, the tags are coated with a layer of hard magnetic metal or partially covered by hard magnetic plates. At the till the cashier runs a strong permanent magnet along the metal strip to deactivate the security elements (Plotzke et al ., 1994). This magnetises the hard magnetic metal plates. The metal strips are designed such that the remanence field strength of the plate (see Section 4.1.12) is sufficient to keep the amorphous metal strips at saturation point so that the magnetic alternating field of the security system can no longer be activated.

The tags can be reactivated at any time by demagnetisation. The process of deactivation and reactivation can be performed any number of times. For this reason, electromagnetic goods protection systems were originally used mainly in lending libraries. Because the tags are small (minimum 32 mm strips) and cheap, these systems are now being used increasingly in the grocery industry (Figure 3.9).

In order to achieve the field strength necessary for demagnetisation of the permalloy strips, the field is generated by two coil systems in the columns at either side of a narrow passage. Several individual coils, typically 9 to 12, are located in the two pillars, and these generate weak magnetic fields in the centre and stronger magnetic fields on the outside (Plotzke et al., 1994). Gate widths of up to 1.50 m can now be realised using this method, while still achieving detection rates of 70% (Gillert, 1997).