Re: Impedance matching of quadband antenna

[kelvchan]

Re: Impedance matching of quadband antenna

Hi

I am looking to match a quadband antenna, basically, two resonances for a handset. I only have a pi configuration matching network to play. How should I match to both resonances since I have only one matching network to use? Please advise.

Many thanks

 

[vfone]

Re: Impedance matching of quadband antenna

Quad band antennas for handsets (as PIFA) usually don’t need any matching networks, because the antenna must be tuned in the way to provide reasonable VSWR for all four bands.

 

[kelvchan]

Re: Impedance matching of quadband antenna

But what if there is a need to do dual band matching then, not necessarily a handset, but an antenna with two band with a PI network? How would it be done then? Do you match to the upper frequency first or the lower one first or do you find an average of some sorts?

 

[vfone]

Re: Impedance matching of quadband antenna

A matching network usually is for one frequency band.
Is true, PI network can match broadband frequencies, but not the case of 900MHz and 1800MHz in the same time.
As I mentioned, the antenna shall be designed and tuned (not the matching network) in the way to provide resonances and reasonable VSWR on different bands.

 

[Joel-Tang]

Re: Impedance matching of quadband antenna

You could try a diplexing solution to isolate the two matching circuits. Just use the standard formulas for these filters and use the correct input and output impedances, and some tweaking to the values to achieve the desired match. This may not be practical at all and might not work, but this is something that can be verified rather quickly in a linear simulator.

 

[cybcad]

Re: Impedance matching of quadband antenna

You could try a diplexing solution to isolate the two matching circuits. Just use the standard formulas for these filters and use the correct input and output impedances, and some tweaking to the values to achieve the desired match. This may not be practical at all and might not work, but this is something that can be verified rather quickly in a linear simulator.

I don't think this is a good method, for either the LPF or HPF is need to be matched to 50 ohm, so if the match network is putting after the LPF and HPF ,the mismatch between the antenna and LPF/HPF will cause loss too.

 

[Joel-Tang]

Re: Impedance matching of quadband antenna

I don't think this is a good method, for either the LPF or HPF is need to be matched to 50 ohm.

The filters most certainly do not have to be 50 ohms. They can have different impedances for input and output, effectively "being" the matching network.
That being said, this particular approach is tricky and has other challenges not mentioned that would require some analysis.

---------- Post added at 03:34 ---------- Previous post was at 03:27 ----------

Quad band antennas for handsets (as PIFA) usually don’t need any matching networks, because the antenna must be tuned in the way to provide reasonable VSWR for all four bands.

this assumption is only (partially) valid if you are able to provide the recommended ground plane and follow all of the manufacturers recommendations. Even then, you still need a matching network. In my practical experience pifas don't do very well in less than ideal circumstances.