How to carry out the impedance of GSM mobile phone dual-frequency antenna

Label: GSM dual-band antenna impedance

This article refers to the address: http://

Usually the impedance matching at a certain frequency point can be performed by using the SMITH chart tool. The two devices can definitely get the impedance matching from one point to another through the series + shunt inductor or capacitor, but this It is single frequency. The mobile phone antenna is dual-frequency, and matching one of the frequency points will inevitably affect another frequency point, so the impedance matching can only be compromised in two frequency bands.

Matching at a certain frequency is easy, but the dual frequency is more complicated. Because the 900M is completely matched, then the 1800 will not reach the match, it is a suitable matching circuit. It is best to use the simulation software or a point to match, and adjust it under the S11 parameter on the network analyzer, because the matching point of the dual frequency is definitely not too far from here. Only two component matches are unique, but there are countless solutions for pi-type network matching. At this time, you need to simulate to pick, it is best to use experience.

Simulation tools are almost useless in the actual process. Because the simulation tool is not aware of the model of your component. You must enter the model of the actual component, that is, the various distribution parameters, and your results may be consistent with the actual. An actual inductor is not simply measured by the amount of inductance, it should be an equivalent network to simulate. I usually only use simulation tools to do some theoretical research.

In the actual design, you should fully understand the principle of the Smith chart, and then use the network analyzer's circle chart tool to debug more. Knowing the principle allows you to know exactly what to use. Multi-tone is to let you know how the components you use will move on the actual circle. (Because of the distribution parameters and the frequency response characteristics of the components, the movement of the actual part on the circle diagram will be different from the movement of your theoretical calculation).

Dual-frequency matching is indeed a compromise process. It must be purposeful to add one piece. In the case of GSM and DCS dual-band, if you want to adjust GSM and don't want to change DCS, you should choose the way of series capacitor and shunt inductor. Similarly, if you want to adjust the DCS, you should choose the string inductor and capacitor.

In theory, two parts need to be adjusted to one frequency point, so the actual mobile phone or mobile terminal usually arranges the matching circuit according to the following rules: for simpler, the antenna space is larger, the reflection is originally smaller, and the Pai type is used (2 and one String), such as conventional candy bar phone, conventional clamshell machine; slightly more complicated with double L type (2 string 2 and): For more complicated, use L+Pai type (2 string 3 and), such as mobile phone with rod antenna .

Remember that the matching circuit reduces reflections but introduces losses. In some cases, although the standing wave ratio is better, the efficiency of the antenna system will decrease. Therefore, the design of the matching circuit is somewhat taboo; for example, in the matching circuit of the GSM and DCS mobile phones, the series inductance is generally not more than 5.6 nH. Also, when the reflection of the antenna itself is relatively large and the bandwidth is insufficient, it is seen on the Smith chart that the boundary point of each frequency band is large from the center of the circle, and generally the matching does not improve the radiation.

The antenna's reflection index (VSWR, return loss) is generally used as a reference during the design process. The key parameters are transmission parameters (such as efficiency, gain, etc.). Some people emphasize the return loss, one port should be -10dB, and the standing wave ratio should be less than 1.5, which is meaningless. When I came across such a person, I joked that if you have a good reflection index, I will give you a 50 ohm matching resistor, so that the standing wave is less than 1.1. As long as your mobile phone can work, I will not care!

The SWR standing wave ratio only describes the degree of matching of the ports, that is, the degree of impedance matching. The matching is good, the SWR is small, and the power reflected back from the antenna input port is small. If the matching is not good, the power that is reflected back will be large. As for the part of the power that enters the antenna is not radiated, you are not sure. The efficiency of the antenna is the ratio of the total power radiated to the space to the total power at the input port. So SWR is good, can't judge the antenna efficiency must be high (take a 50ohm matching resistor to connect, SWR is good, but there is radiation?). However, the SWR is not good, and the power of reflection is large, and it can be confirmed that the efficiency of the antenna must not be high. SWR is a necessary condition for good antenna efficiency, not a sufficient condition. Good SWR and high radiation efficiency are necessary and sufficient conditions for high antenna efficiency. When the SWR is the ideal value (1), the port is ideally matched, and the antenna efficiency is equal to the radiation efficiency.

In today's mobile phones, the space of the antenna is compressed smaller and smaller, at the expense of the performance of the antenna. For some multi-frequency antennas, even VSWR has reached 6. In the past, we used more external antennas, and the average efficiency was low at 50%. Now the efficiency of 50% or more is very good! Take a look at the mobile phones on the market, even if it is a famous company, such as Nokia, it is also inefficient. At 20%. Some mobile phones (sliding, rotating) are only about 10% efficient at some frequencies.

I have seen several test reports of the built-in antenna of the mobile phone. The antenna efficiency is basically around 30-40%. At that time, I felt that it was really bad enough (compared to the microstrip antenna I designed), and it seems to be okay now. However, in actual engineering, it seems that the loss due to S11 and the loss of the matching circuit are counted in the efficiency. According to the antenna principle, only the dielectric loss (including the substrate caused by the magnet in the mobile phone) and the metal loss (although very small) ) is in the loss of the antenna, and the loss of the return loss and matching circuit should not be recorded. But engineering is engineering, so it's easy to test.

Right, add another sentence, software simulation is helpful to the project to a certain extent: Of course, the accuracy of the simulation results can not be compared with the test, but the antenna performance obtained by parameter scanning simulation is useful with the trend of the parameters. This is much faster than getting data through testing, especially for some parameters that are not commonly used.

"The simulation tool has no use in actual engineering", which means that when designing the matching circuit, it is more specific when designing the dual-band GSM and DCS mobile phone antenna matching circuit. If you understand this sentence alone, it is undoubtedly wrong. In fact, I have been using HFSS for antenna simulation, and the results are based on simulation results.

By the way, welding components is really a laborious thing, and there are methods. The so-called practice makes perfect. Large companies may give you a special welder, so you may just have to say what to weld. However, what we are discussing here is how to effectively complete the design of the matching circuit. Pay attention to effectiveness! Validity includes the time spent and the accuracy of the selected components. If there is no hands-on experience, only a matching design is made by software simulation but the actual antenna input is used. Oh, I can say that in all likelihood your design will not be used, even different from your imagination!

In the actual design, there is another situation that you can't consider in the simulation (unless you measure it beforehand). That is, the effect of the distribution parameters on the PIFA. Since the height of the antenna is getting smaller and smaller, and the matching circuit is either below (inside) or below (outside) the antenna, anyway, the addition of an actual component introduces a change in the distribution parameters in practice. Especially if the board layout is not good, this effect will be obvious. In actual welding, even if one piece is not welded well, re-welding will bring about a change in impedance.

Therefore, in the design of PIFA, we usually do not use matching circuits (or 0ohm matching). This requires you to carefully adjust and optimize your antenna. In general, it is easier to implement today's flexible circuit board design (Flexfilm) because it is easier to modify the radiation film. It is relatively more difficult to stamp a metal for another design that is used more. First, the hardness is large, and the space cannot be fully justified by the limitation of the process. Second, it is difficult to modify the design of the radiation sheet once the mold is formed.

There is no great use of simulation tools in matching designs. Not many people can use simulation tools to calculate matches. Besides, is there any great effect on how to measure it? The engineering is fast and accurate. Simulation for simulation has no practical meaning. In order to get a 2, 3, up to 5 pieces of matching, you can build a model of inductance and capacitance, which is not worth it. Also, how do you consider the change in the distribution parameters of the PIFA match mentioned above? I also mentioned that some of the taboos of matching circuits are not derived from theory and are entirely derived from practice. Because the antenna is designed to improve its radiation efficiency (total efficiency)! I have not succeeded in finding an accurate matching circuit (say GSM, DCS) dual-frequency through simulation tools within 1 hour. The wrong method is ok).

There is also a question here: Many of the built-in antennas now have a speaker, how does the speaker affect the operation of the mobile phone (and the antenna) and how to model it in HFSS? (Only the first half of the problem is known to be built.)