Inductive coupling and backscatter: near-field and far-field physics decide the read

01 / FIELD NOTE

Keep the decision tied to the operating context.

A passive tag carries no transmitter, so the question of how it answers is a question of coupling: the tag has to borrow energy from the interrogator and shape part of it into a reply. Which coupling exists is not a design preference but a fact of geometry. The field around an interrogator antenna is divided by a boundary at λ/2π, the carrier wavelength divided by two pi, roughly one sixth of a wavelength out. Inside that radius the field is still attached to the antenna and behaves reactively; outside it the field detaches and travels as a wave. Which side of the boundary a tag sits on decides the entire character of the read.

Everything downstream of the boundary depends on frequency, because wavelength sets the radius. At 13.56 MHz the carrier is about 22.1 m long, so the near field ends around 3.5 m from the antenna — far more than any practical tag-to-reader gap, yet close enough that the field there is already weak. At 135 kHz the wavelength is roughly 2400 m and the whole of any working volume lies deep inside the near field. The practical consequence is that frequency fixes the range budget before any antenna design or power choice is made: low frequencies buy a tiny, well-behaved region of operation, higher frequencies trade that away for a propagating wave that can reach further.

Inside the near field the field decays much faster than intuition from wave propagation suggests. The magnetic field strength falls as the inverse cube of distance, and because the energy a tag loop can harvest is drawn from that field, the power available to the tag falls as the inverse sixth power — the steepest of the fall-offs in this note. The field is intense immediately around the antenna and negligible a short distance out. Default read range follows: a typical 13.56 MHz system stays under a meter, a handheld loop reaches on the order of inches, and a larger antenna aperture extends the read to a couple of feet. The near field confines, and nothing in the physics argues against that confinement.

Because the near field is a magnetic field, it has vector character, and the orientation of the tag relative to the antenna matters in a way it does not in the far field. Coupling between two loops is strongest when their axes are aligned and the loops face each other; as the tag is tilted, the flux threading the tag loop falls away. The tag passes in and out of read depending on attitude, which is why inductively coupled systems care about how an item is presented — a badge walked past a reader, a card tilted, a tool pouch turned. Holding geometry steady is part of the read discipline, not an afterthought.

Material behaviour in the near field is decided by what a magnetic field couples to. It couples to loops of conduction: a piece of metal in the operating zone carries eddy currents whose own field opposes the interrogator’s field, absorbing energy that would otherwise reach the tag and detuning the tag loop itself. Water and tissue are largely transparent at these frequencies because the coupling is magnetic and they present no free-moving charges for an alternating magnetic field to drive; they matter mainly through electric coupling, which low-frequency inductive systems do not rely on. The near-field material rule is therefore the reverse of what wave thinking suggests: metal is the hazard, moisture is not.

Beyond the boundary the picture changes completely. The wave detaches from the antenna and propagates outward, and it never gives energy back to the radiating element: the interrogator cannot collect its own field once it has become a travelling wave. Power available at a tag now follows the square-law fall of free space, far gentler than the near-field collapse. The tag sits in the path of the wave, intercepting a small proportion of it, and rectifies the induced high-frequency voltage with diodes to power its chip. Range is no longer a fraction of a wavelength; it becomes a function of how much of the passing wave the tag can capture.

How much of that wave a tag captures is described by its reflection cross-section, and that cross-section is largest when the tag is in resonance with the wave front hitting it. The same physics that makes a half-wavelength object a strong reflector of a wave makes a resonant antenna a strong re-radiator: the antenna gathers the incoming field, and a share of it is reflected and reradiated outward. The tag antenna is therefore designed to resonate with the carrier the interrogator emits, at which point it behaves far less like a passive bystander and far more like a purposeful mirror that is larger than its own physical size.

Data travels from such a tag by changing the mirror, not by making a new signal. A load resistor connected across the antenna is switched on and off in time with the data stream, and each state changes the resonant properties of the antenna: in one state the tag is a good reflector, in the other a poor one. The alternation presents itself at the interrogator as a varying strength in the returning signal, which the receiver decodes as the tag’s reply. There is no transmitter anywhere in the chain — the tag modulates energy that the interrogator itself sent out, which is why the technique is called modulated backscatter.

The backscattered reply is expensive in signal budget. It crosses the path loss twice — outward to the tag and inward again — and on the outward leg some of the energy is absorbed by the tag just to power it, so what returns is a small residue of an already attenuated wave. Before it can be read it must be separated from the interrogator’s own transmitted signal, which is far stronger; the return enters the antenna in the reverse direction and a directional coupler steers it to the receiver input while largely suppressing the forward signal. The round trip explains why far-field range behaves as it does: doubling distance does not halve the reply, it pays the loss twice over.

The two mechanisms end in two placement disciplines. An inductively coupled tag wants the axis of its loop aligned with the loop of the antenna, a gap that is a small fraction of the operating wavelength, and an operating zone kept free of conductive material. A backscatter tag wants to sit where the interrogator’s wave actually reaches it, oriented so its antenna presents its resonant cross-section, with metal and water treated as obstructions or absorbers of the passing wave rather than as detuners of a local field. The λ/2π radius is a guide, not a wall: near-field effects fade gradually, and antennas large relative to the wavelength move the transition. What this note does not answer is how far a particular reader will reach in a particular room; that is antenna, power and environment. It answers the physics those answers obey.