How to match impedance for GSM/DCS dual band antenna?

how to match impedance

The GSM/DCS or GSM/PCS antenna is dual band.

I want to know how to design the impedance matchment for it.

I can only do it for single frequency, just transform from one point to another on the Smith chart by 2 lump component (L or C).

But for 2 bands, when I match one frequency , the other will also change.

What should I do? Especially by 3 lump components (Pi or T).

Is there any software to design it?

What is the rule ?

Thank U!

The design of the dual-band mobile antenna (for example PIFA) shall give a proper VSWR at both bands without external matching components.
Implementing a wide-band antenna match for this situation is not worthed, due to high insertion loss.

You can do broadband matching by given terminal impedance. There are a few IEEE papers talking about this issue, such as "a new theory by broad-band matching" by Youla. However, I have to say, it is not simple, since they didn't talked much about the implementaion, except the theory.

Good luck,

Matching antenna is basically matching not only through space and also through usability modes. Note that in mobile applications, natural resonance should be typ. slightly offseted to the higher frequencies. In actual scenario when using use the antenna, we present a load to it and resonance shifts to the lower freuqencie side. So typically this is the TX band should be centred sincethe PA want a better load that present to it and hence better power consumption. Hope that helps.

vfone said:

The design of the dual-band mobile antenna (for example PIFA) shall give a proper VSWR at both bands without external matching components.
Implementing a wide-band antenna match for this situation is not worthed, due to high insertion loss.

Click to expand...

Thank U very much!


I have to use the antenna produced before, which is a little shift on DCS frequency.

So I must do impedance matchent, or all the antenna will be waste.


Now I simulate is with MWO, and get two L to get good VSWR.

But just like U said, the insert loss is bigger, so the efficiency of the antenna is bad.

Added after 7 minutes:

dunwill said:

Matching antenna is basically matching not only through space and also through usability modes. Note that in mobile applications, natural resonance should be typ. slightly offseted to the higher frequencies. In actual scenario when using use the antenna, we present a load to it and resonance shifts to the lower freuqencie side. So typically this is the TX band should be centred sincethe PA want a better load that present to it and hence better power consumption. Hope that helps.

Click to expand...

In our handset design, the antenn is conected to the switch, the switch to the LNA and PA.

So, to get better power, the impedance match can be do from PA to switch.


And I wonder how do U design your antenna not at 50 ohm, for the network analyzer just has 50 ohm port.


How do U test the output impedance of PA ??

It's really hard to measure the PA output impedance in active state, because you can never know what it would be like in the real network. And if tested in passive state, it makes nonsense.

For GSM/DCS band matching design, if your DCS center frequency not changed much, maybe you can choose the right type of element. like series an inductor or shunt a capacitor, to tune the high band while not influence the low band much.