NRF52DK antenna pattern orientation has nulls to side?

[butch]

Since the antenna on the NRF52DK is oriented horizontally, and the theoretical 3D antenna pattern is a torus (donut) whose hole is pointed in the horizontal axis of the strip antenna, does the NRF52DK have nulls (weak signal strength areas) towards the sides having the pin headers, when the DK is just laid on a table?

So that in most applications, a PCB board having a similar antenna should be mounted oriented vertically in the product, so that the strong signal radiates along the floor, uniform as you walk around it?

Is there an omni-directional antenna, with a spherical (non-toroidal) pattern?

[Jørn]

Hello Butch

Yes, the antenna would typically have weaker signal emanating from the tip of the antenna. In a monopole antenna you will have a voltage maxima, and current zero, at the tip. The first voltage zero, and current maxima, is located 1/4th of a wavelength in. This implies that the electric field lines stretch from the tip of the antenna to the point of the quarter wavelength. The magnetic field is created by the current in the antenna (and the changing electric field). Due to this the electromagnetic far field will mainly originate from the "body" of the antenna, while very little is radiated by the tip. Unfortunately we do not have a measurement of the nRF52 DK antenna field, but you can safely assume it radiates a weaker field "looking out" of the antenna tip.

You are correct in your assumption that the second PCB with a similar antenna should ideally be oriented the same way as the first. Monopole antennas, among others, have a specific polarity, usually defined by the direction of the electric field lines, in this case parallel to the antenna length. In theory, with an ideal system, if a second monopole is oriented 90 degrees compared to the first antenna it will not receive any of the signal transmitted by the first antenna, as the electromagnetic field will not induce any currents along the antenna length. In reality you will in most cases still get a signal, but it can be attenuated. This attenuation is often called polarization loss in your link budget.

Floors and walls can cause reflections, especially if there is metal within them, which can cause both destructive and constructive interference at the receiver due to reflection. You can read up on the Fresnel zone and multipath fading for more information on this. So whether to mount your antenna horizontally or vertically depends on your application and environment.

An antenna with a perfect spherical pattern is called an isotropic antenna, however they only exist in theory. The ideal isotropic antenna is used as reference when characterizing the antennas gain. The gain is given in dBi, and is the ratio between the main lobe strength compared to that of an isotropic antenna. However, you can still have an antenna with 0dB gain without it having a perfect spherical pattern, due to losses.

Best regards

Jørn Frøysa

[butch]

Thanks. I noticed this when I turned the TX strength to -40dBm. (Doing that helps you to test various configurations of units in a small area e.g. on a desktop.) I noticed that the relative placement of units made a difference. And now I need to redesign my product: if the PCB lays flat, I can't use an antenna printed on the PCB and expect a uniform signal in the horizontal plane of the earth's surface.