Electromagnetic compatibility (EMC) usually is not the first thing in mind while working with mm-wave antennas. An engineer has a lot of challenges to solve within the antenna itself, let alone thinking about noise or signal distribution. Moreover, over-the-air losses are huge in the mm-wave range, meaning that as soon as the signal leaves the antenna, it brings no significant effects to surrounding devices. But what is happening in the radiating device itself? How does beam steering affect noise coupling to the device? These questions I am answering below.
5G mm-wave USB dongle
To perform an evaluation of EMC on an mm-wave device, the concept of a 5G USB dongle was created. Figure 1 shows the overview of the final device. Dimensions are 5x3x2cm.
Figure 1. 5G mm-wave USB dongle
To enable an end-fire radiation cavity antenna array with edge-soldering was placed on the edge of the device. The antenna concept was described thoroughly HERE.
A couple of random signal traces were created to evaluate beam steering effects (fig. 2). The upper trace has vertical placement, and the bottom one is vertical+horizontal. Both lines are terminated with 50Ohm load on one end and 10kOhm on another. 10kOhm represents digital input, and 50Ohm is the low power driver of the signal. The upper line is cut to an L shape to find out the coupling when a trace is parallel to the array. The lower line represents a trace with a longer diagonal section. The 50Ohm termination was selected to represent low impedance, where noise travels more easily, and the 10kOhm termination was chosen due to of high impedance input. The length of these traces was selected randomly.
Figure 2. Setup for EMC simulations. Port 1 is antenna port.
Radiation patterns and steering
The specification for the device has +/- 30 degrees beam steering to provide good horizontal and vertical coverage and acceptable CDF EIRP (cumulative distribution function) values for the mm-wave devices. 3D radiation patterns for 0-, 15- and 30-degree positions can be seen in figures 3 and 4. More about CDF simulations can be read HERE.
Figure 3. End fire radiation.
Figure 4. 15-degree steering (left) and 30-degree steering (right).
The 30-degree steering is the most concerning from an EMC point of view as side lobe levels decrease, meaning that more power leaks in unpredictable directions.
Simulation was done using Time Domain solver. Results are collected in table 1.
A small decrease in realized gain can be observed, when comparing antenna array with different steering angles. The bigger angle the more decrease in gain there is. The maximum loss that has been observed in simulation was at 30-degree position – around 0.2dB.
The most interesting situation is with coupling, as it can be clearly noticed, that coupling between antenna and signal traces drops for approximately 11 dB while scanning.
Table 1. Simulation results
It can be noted that the line closer to the device’s edge has bigger current amplitudes. Moreover, both horizontal edges of the device have significantly increased surface currents compared to the 0-degree position. Higher current amplitudes travel on the main board, which creates a possibility for more noise coupling on the traces, resulting in a drop in the coupling, that was presented above.
Figure 5. Surface current comparison
Mm-wave brings in a lot of challenges and electromagnetic compatibility is one of the things an engineer should consider. For some time now it was assumed that an mm-wave device needs to be tested for EMC only at a 0-degree beam position, neglecting scanning. As proved by this work, there is an effect from beam steering to noise coupling, so an antenna array should be tested at least at the most extreme angles in addition.
This blog article was originally posted on the author’s LinkedIn profile.