What issues regarding antenna isolation should we pay attention to?
[Copy link]
This article will be divided into three parts to explain the definition of antenna isolation, several key factors affecting antenna isolation, and how to improve the isolation of terminal antennas. I hope it will be helpful to everyone.
Part 1: Definition of Antenna Isolation
Not long ago, our Dianchao RF team received a RF-related consulting project. The customer needed to solve the isolation problem of a terminal antenna, and their requirements were relatively high, requiring the isolation to reach 30dB. The simulation data obtained by the customer through CST simulation did not match their measured data, so they found Dianchao and hoped to solve the problem of the mismatch between simulation and measurement. We did some research on antenna isolation technology issues for this project.
Today's mobile communication services have entered the 5G era. I believe that everyone is familiar with the terms that come with the 5G era, such as multi-antenna technology, massive MIMO technology (that is, multiple-transmit multiple-receive technology), multi-band carrier aggregation, and so on.
The introduction and application of these technologies cannot avoid a key problem, that is, the number of RF bands and standards that work simultaneously has increased. For example, the RF terminal wide area network bands include LTE bands 12345678, as well as 38, 39, 40, 41, etc. In addition to the same frequency band divisions as LTE, 5G NR also has 3.5G bands and millimeter wave bands.
In addition to WAN, terminals usually have WIFI and Bluetooth functions. As the frequency bands increase, the number of antennas on the terminal also increases. A terminal may have multiple antennas that transmit and receive different signals at the same time. These signals may work in adjacent frequency bands or even the same frequency band, such as WIFI and Bluetooth.
If the operating frequency of one of these simultaneously operating RF signals happens to fall within the receiving frequency band of another signal, the transmitted signal will cause serious interference to the received signal; even if the out-of-band spurious signal of the transmitted signal falls within the receiving frequency band of other signals, it may bring noise effects that cannot be ignored.
Let me tell you a real case. I once encountered a very serious signal interference problem while working on a project. At that time, the spurious emission of LTEB41 seriously interfered with the receiving signal of WIFI2.4G, causing the coexistence test to fail. Although the problem was finally attributed to an RF filter component and was eventually solved by software time division. At that time, the isolation between the LTE antenna and the WIFI antenna was about 10dB, which had met the basic requirements for terminal antenna isolation. However, I think that if the isolation between the LTE antenna and the WIFI antenna can be improved, it will be another effective way to solve the problem.
What is the isolation between antennas? As the last load on the RF wireless path, the antenna carries the mission of sending and receiving signals. It is essentially a bidirectional passive device.
It does not only transmit useful signals, but also all signals provided by the source, whether useful or not. It just transmits signals of different frequencies with different efficiency. The signal transmission efficiency is high at the resonant frequency, and lower at other frequencies.
While transmitting, it also receives signals, no matter what the signal is. Of course, the reception efficiency is high at the resonant frequency, and low at other frequencies. There are two antennas here. The signal transmitted by antenna A will be received by antenna B, and similarly, the signal transmitted by antenna B will also be received by antenna A.
Professionally, this physical phenomenon is called antenna mutual coupling. Isolation is a physical quantity used to measure the degree of antenna mutual coupling. To put it more directly, or in a more down-to-earth way, assuming that two antennas form a two-port network, the isolation between the two antennas is the S21 between the antennas. Therefore, the method to test isolation is to connect the two antennas to the two ports of the network and directly measure S21.
Part 2: Key factors of antenna isolation
Next, I will demonstrate several test examples. These experiments were all completed in our electric nest sharing laboratory.
As shown in the figure, there are 8 antennas on this board, named No. 1 to No. 8. How is the isolation between these 8 antennas? Let's first test the two antennas No. 1 and No. 2 that are farther apart, as shown in the red box.
It can be seen that the isolation between antennas 1 and 2 is very good, reaching about 30dB.
We next tested two antennas that were closer together, number 2 and number 7.
From the data, the isolation of these two antennas is quite different, which means that the distance should affect the isolation of the antenna. Next, we test No. 2 and No. 3. The distance between 2 and 3 is the same as the distance between 2 and 7. Let's see if the isolation is the same.
It can be seen that the isolation is much worse. In absolute terms, the isolation is more than 4 times worse than that of the previous set of antennas. This shows that distance is not the only factor affecting isolation, and there must be other reasons affecting the isolation of the antenna.
Yes, the radiation direction of the antenna is also an important factor affecting the antenna isolation. The radiation pattern of a standard antenna can be derived from theoretical analysis. When the strongest radiation directions of two antennas are opposite, even if the distance between the two antennas is relatively far, there will still be a strong mutual coupling effect between them, resulting in poor isolation.
So how to improve the isolation between antennas?
For individual antennas, there are four main ways to improve isolation. Let’s demonstrate them one by one.
The first way is to increase the distance between the antennas. Let's demonstrate it in a more intuitive way. These are two 2.4G dipole antennas. Let's see how distance affects isolation. First, the isolation at a distance of 5cm.
Then, when the distance between the two antennas is 10cm, the isolation between the two antennas is
It can be seen that when the distance is 10cm, the isolation between the antennas is better than that of 5cm. Therefore, if conditions permit, we try to increase the distance between the two antennas, which is an effective way to improve isolation. However, the small structural space of the terminal often limits the distance between the antennas. So how to improve the isolation of the antenna in a limited structural space? In the early days, base station antennas would use isolation walls to improve isolation.
Let’s try to see if this method works!
From the experimental results, adding a wall does have an effect, but the height of the wall will affect the final isolation effect. It is difficult to draw a conclusion based on simple experience to determine how high the wall should be to achieve the isolation index. In addition, whether there is space on the terminal to add the wall is also a problem. So we need to try other methods, such as making the polarization directions of the two antennas perpendicular.
The two antennas with perpendicular polarization directions still achieved extremely high isolation even when the distance was only 5cm, which shows that this method is very effective. It seems that we only need to determine the polarization directions of the two antennas and then make them perpendicular to each other, so easy!
How to determine the polarization of an independent antenna? The polarization direction of an antenna is the direction of the electric field radiated by the antenna. Therefore, the polarization direction of the antenna can be easily determined by the direction of the circuit on the antenna. For an independent wire antenna, its current direction can also be simply determined by the shape of the antenna.
Since vertical polarization of antennas can improve isolation, can this method be used to improve isolation on the terminal? Let's test it directly. Antennas 2 and 6 on this board
From the test results, we can see that the isolation between the two seemingly vertical antennas is very poor. What is the reason?
This is because we think that the antenna is just a small piece of copper structure, but in fact, the radiator is the entire board. Indeed, the current is vertical on the two metal surfaces that we think are the antenna body, but the current in other places on the board is not vertical. The following two pictures show the direction of the current on the board when the antenna is working.
It can be seen that when the two antennas are working, the currents on the board are parallel, so the condition of vertical polarization is not met, and the isolation will naturally not be good.
The environment of our board is actually very simple, clean, with only a PCB board. The current direction is also very regular, but even so, it is impossible to make the polarization directions of the two antennas perpendicular. The environment in a real terminal is much more complicated, and the polarization direction is more certain. Therefore, it seems a bit unreliable to improve the isolation by making the two antennas perpendicular to each other.
Part 3: How to improve the isolation of terminal antennas
In the article “A Brief Discussion on Antenna Isolation, Part 1: Definition of Antenna Isolation”, we mentioned that the antenna radiation pattern will also affect the antenna isolation. By simply aligning the directions where the two antennas radiate the weakest, we can obtain a better isolation index.
However, the antenna radiation pattern sometimes cannot be obtained through simple empirical judgment, especially the PCB antenna, PIFA antenna, and IFA antenna in our terminal .
The radiation patterns of these antennas are affected by the environment around the antenna and the ground plane, and you can't tell 123 by just looking at it. If you want to predict the radiation pattern of the antenna in advance, you can only get the result through accurate 3D electromagnetic field simulation .
For example, we used CST to simulate the design of this board and predicted the antenna pattern and antenna isolation. You can see how different our simulation and measured results are.
Comparison between simulation and actual measurement of antenna isolation of No. 1 and No. 2
1. Comparison between simulation and actual measurement of antenna isolation of No. 4
Through simulation, we can predict the antenna's radiation pattern and modify the shape and position of the antenna in advance to achieve the purpose of improving the antenna isolation.
However, if the antenna position is fixed and the isolation cannot be improved by changing the antenna form, is there any other way to solve this problem?
Yes. The mutual coupling between antennas will affect the isolation, so if it is adjusted by matching decoupling, it is theoretically possible to optimize the isolation again. The decoupling network topology is as follows.
As a four-port network, the D network plays a decoupling role, and its goal is to change S21 to 0 through network transformation. In the process of network transformation, S11 and S22 will inevitably degrade, so a matching network M is required to match the antenna to a suitable value.
We treat these two antennas as a two-port network, then use network analysis to test the S parameters of this two-port network, save it as an SNP file and import it into ADS simulation.
ADS simulation is used here to quickly find the appropriate lumped parameter device. The actual matching situation may be slightly different from the simulation result. The following is the simulation result.
Before improvement
Improved
After obtaining the device values through simulation, we soldered these devices on the real motherboard to see if the isolation can be truly optimized. The matching devices actually used are slightly different from the simulated devices. The simulation results are only used as qualitative data, and the test results must be based on the actual devices.
By comparing the measured results with the simulation results, we can see that the isolation curves are basically consistent and have been greatly improved compared to before, from -10dB directly optimized to -20dB, while the VSWR of the antenna itself has not deteriorated too much. This shows that improving the antenna isolation performance by adding a decoupling network is real and effective.
The isolation problem of terminal antenna is indeed a difficult point in antenna design, but we have many ways to circumvent it. However, no matter which method is used, it is necessary to do preliminary design in the early stage of development and fully consider the possible situations that may arise in the later debugging.
No engineering experience is universal. Solving problems requires engineers to think more and accumulate more experience.
| 
提升卡
变色卡
千斤顶