QUESTION

Question

Hi,

I designed a PCB where I planned to use a SR4G013 antenna.

Initially I tried the antenna on a hand milled FR4 piece (2 layers) and had great results, but then I produced a series of PCB samples of the product (4 layers) and the results were completly off...

During the tests I tried to hand trim the width of the ground short and the return loss of the antenna improved.

Are there any guidelines on the size of the short?

The prototype PCB was a 4 layer [TOP(signal + GND pour) / IN1(GND) / IN2(Signal + Power + GND pur) / BOTTOM (Signal + GND pour)]

The distance between TOP and IN1 layers is 0.18mm
The width of the FEED trace is 0.2mm (I will increase to 0.3mm in the next revision)
The distance between signal (feed) and ground pour on the TOP layer is 0.2mm
The width of the SHORT between the FEED pad of the antenna and GND was 0.5mm (no good result)

Thank you very much

Answer 1

Hi Stefano,

When you say results do you mean the return loss? Have you tried matching the antenna on the new PCB?

Best regards,

Yu Kai Yeung

Comments

Hi, 

Yes we measured the return loss to start tuning the antenna, without any component in the pi filter.

And the first measurement gave us in the center band only -2.17dB @1641.86MHz

After that I reduced the width of the short circuit track to about 0.1mm, after doing this the measurement improved slightly, to -10.29 @1620.93MHz

Since varying the width of this short circuit track substantially affects the return loss, I wanted to know if there were any guidelines for choosing this width? 

Because I would like to know if I have to redesign the PCB to be able to do a better tuning and get better results.

Right now the PCB is this:

---

Hi Stefano,

A -10dB return loss is a really good result so for now keep that width and use the Pi matching network to tune the resonance down. Try using an inductor in the matching network it should bring the resonance down.

Best regards,

Yu Kai Yeung