Antenna Miniaturisation

An article just published on Nature Communications covers some interesting and quite exciting developments in antenna miniaturisation.



Electrical antennae rely on being resonant or partly resonant with the electromagnetic wave at the frequency of interest. For a dipole, a length of lambda / 4 is ideal but antennae can be designed with electrical lengths down to lambda / 10. The technique of antenna construction shown in this research means that effective antennae can be constructed that work at electrical lengths of nearly lambda / 600.

If you take a 950MHz (mobile phone low band frequencies) full dipole it will have a length = c / (f * 2) = 16cm. With this new technology, this antenna length could come down to sub 1cm distances. This decrease by over a factor of 10 is highly significant and with the application to sub-resonant antenna designs, further decreases in size could reasonably be expected.



The technology itself is very clever, sandwiching a layer of piezoelectric material (voltage to mechanical movement) with a layer of ferromagnetic material to form what the paper refers to as a “ME heterostructure” (I prefer sandwich… mmmm… sandwich). The magnetic part of an incident EM wave causes the ferromagnetic layer to change shape in response. In turn, this makes the bonded piezoelectric layer change shape, creating a voltage on the output terminals. The process of reception is now mechanical rather than electronic with the tuning of the antenna primarily performed by selecting the materials based on their mechanical properties, thereby tuning the mechanical resonant frequency.

Similarly, applying a voltage to the piezoelectric material will change its shape, causing the shape of the

It’s worth taking a step back to point out that you can perform a similar trick with a conventional antenna by increasing the dielectric constant of the surrounding material. This changes the speed of light within this medium which is dictated by the Velocity Factor (VF) = 1 / sqrt [Er] (where Er is the dielectric constant or relative permittivity). This is why placing your antenna closely coupled to a plastic enclosure can change the effective frequency. Two downsides that make this approach less useful for antenna miniaturisation:

  1. The range of dielectric constants of insulating materials. FR4 PCB material is typically around 4, silicon and alumina can be in the 11-12 region. Higher values for liquids or more exotic materials do exist.
  2. The square root term around the dielectric constant causing the reduction in VF to become less significant with increasing Er

The mechanical “stiffness” of the new antenna (analogous to dielectric constant) is suitably high that the resonant wavelength is much smaller in the materials used leading the the significantly smaller antennae producible using this technique.


CMOS Process

What is even more exciting is that these new antennae are produced using a CMOS process (detailed in the supplementary material).

This means that integration of the antenna and receiver could be integrated onto the same silicon die in future. Not only would this make things ridiculously compact but it would also allow the receiver circuitry to be placed electrically right next to the antenna which could reduce coupling of noise and improve sensitivity.

It also gives the scope, as mentioned in the paper, of being able to print arrays of antennae on the same die. This could be used to make compact phased arrays, wideband arrays or just the key antennae for different frequencies required for a mobile communications device.

IoT (the favourite industry buzzword at present) solutions will get smaller, cheaper and easier to integrate, perhaps leading to a further connected-ness revolution. The potential impact is fascinating.


Magnetic Coupling and the Near Field

Because these antenna operate on the magnetic component of the EM wave, it should make them more efficient when operating in the near field.

Original image can be found here.

The near field is the distance between antenna and source when the ratio of distance to wavelength is below the ratio of 1 / 2*pi (see above chart). In this region, the impedances of the magnetic and electric fields are not related by the impedance of free space (approx 377 ohms). The field from a magnetic dipole decays with the cube of the distance (not the square) so being able to place a small magnetically sensitive resonant antenna closer to the magnetic field source provides a new method for sensing small currents right next to the source.



The decrease in size could lead to problems in the capacity to handle higher transmit powers or currents. Having said that, the thermal coupling of the antenna to a solid material should be good meaning the thermal resistance of the antenna to ambient could be both quite low and controllable.

Magnetic materials have other characteristics such as saturation and hysterisis so it will be interesting to see how it handles large DC or low frequency AC magnetic fields caused by electrical power wiring. I wonder if this will mean it gets classified as a magnetically sensitive component and starts to fall under the remit of IEC 61000-4-8 testing for such things.

Having worked on designs involving high voltage differential piezoelectric transformers below, I know that they require careful mechanical handling as they will chip, crack and break if treated roughly (although, probably just as much as ferrite of the same dimensions). Conventional metallic structure antennae are generally quite robust or can be made so. I believe the small mass and the encapsulation of these antennae in a plastic or ceramic package could well solve many of these issues.

This also presents the possibility of mechanical interference to an electrical signal. I don’t have a handle on what sort of high frequency mechanical vibrations might exist in the real world and I can imagine their energy would be quite low at the really high frequencies with elasticity of materials starting to take over. I can imagine trying to debug an RF sensitivity issue with a stethoscope for a change!

The technology website Futurism speculates that these could form the basis of a brain-computer interface. Needless to say, the EMC immunity applications of such a technology, especially in our increasingly “EM-dense” environment would be of massive concern. Perhaps we will be seeing the imminent return of the tin-foil hat as a genuine reason for keeping the government out of your brain!

You can read the full article here.

Header image taken from the original article.