Antenna - input impedance (Zo= 50 Ω)

1. Antenna - input impedance (Zo= 50 Ω)

Hello everyone, I'm having difficulty with the line input impedance of the circular planar antenna signal. Its impedance is 117 ohms and the correct is 50 ohms, I am not able to lower the value. I made modifications to the input increasing the line width and also made an impedance transformer. When I get it down, the whole process on the antenna I made is impaired. Example: reflection coefficient is lost, surface current loses efficiency and also its power and directivity. Please can someone help about this.

2. Re: Antenna - input impedance (Zo= 50 Ω)

Try to use impedance repeater: put additional half wavelength line between patch and quarter wave transformer. I would try thinnest possible line (0.16mm or 0.2mm). Repeater actual impedance may have arbitrary value, maybe need to optimize length to ensure reflection coefficient goes full circle on a Smith chart and gets back to 117 Ohm. And then there will be sqrt(117*50) part you tried before. Although repeater will reduce bandwidth, it may help with excitation. I solved similar problem before this way.

•

3. Re: Antenna - input impedance (Zo= 50 Ω)

Originally Posted by Georgy.Moshkin
Try to use impedance repeater: put additional half wavelength line between patch and quarter wave transformer. I would try thinnest possible line (0.16mm or 0.2mm). Repeater actual impedance may have arbitrary value, maybe need to optimize length to ensure reflection coefficient goes full circle on a Smith chart and gets back to 117 Ohm. And then there will be sqrt(117*50) part you tried before. Although repeater will reduce bandwidth, it may help with excitation. I solved similar problem before this way.
Are you asking me to do on the real model?
But I am using a virtual simulator to make this antenna work.

Note - I made two antennas: one that I did my calculations and the other.
I'm using the scientific article to see where my impedance error is. Yet the error continues.

Name: DESIGN AND PERFORMANCE ANALYSIS OF UWB CIRCULAR RING ANTENNA WITH DEFECTED GROUND STRUCTURE Neha Tanzeem Sagar , M. L. Meena and Pankaj Shukla

http://ictactjournals.in/paper/IJCT_..._1656_1663.pdf

Vídeo

4. Re: Antenna - input impedance (Zo= 50 Ω)

According to QUCS transcalc calculator and simple formula 7 GHz 2mm thick line on a 1.5mm thick FR4 have impedance around 60 Ohm. I am not sure how your software calculated line impedance of 115.7 Ohms "peak", which looks strange. I will try to simulate this structure and answer later.

•

5. Re: Antenna - input impedance (Zo= 50 Ω)

thank you, so far I'm breaking my head on this subject

- - - Updated - - -

Originally Posted by Georgy.Moshkin
According to QUCS transcalc calculator and simple formula 7 GHz 2mm thick line on a 1.5mm thick FR4 have impedance around 60 Ohm. I am not sure how your software calculated line impedance of 115.7 Ohms "peak", which looks strange. I will try to simulate this structure and answer later.
thank you, so far I'm breaking my head on this subject!

•

6. Re: Antenna - input impedance (Zo= 50 Ω)

Here is simple simulation result (PEC, no losses, no metal thickness)

2mm feeding line is 60 Ohm line, so i set lumped port to 60 Ohm, and also checked results if 115 Ohm lumped port is used. For some reason your software decided to use 115 Ohm impedance for the port. But i think that port impedance must be set to 60 ohm, and microstrip line calculators i've tried gave 60Ohm for 2mm wide line on a 1.5mm Er=4.3 substrate. Interestingly, S-parameters on your video look more like if it was 60 Ohm port. Try to plot S11 on a smith chart and see if it swirling around 60 Ohm.

Maybe your software decided to choose port impedance to minimize reflections at central frequency. We need someone with knowledge of software you are using to clarify how it determines port impedance, and why it also mentions "peak" word (notice dip at 7 GHz on my plot).

7. Re: Antenna - input impedance (Zo= 50 Ω)

I did some search on CST MWS, and it seems that "discrete face port" must be used, then impedance value must be set manually to 62 Ohm.

8. Re: Antenna - input impedance (Zo= 50 Ω)

Originally Posted by Georgy.Moshkin
I did some search on CST MWS, and it seems that "discrete face port" must be used, then impedance value must be set manually to 62 Ohm.
I'm asking for help, there are other forums, because it's scary, the values don't even come close to yours. Mine range from 80 to 107 ohms.

I am checking to see why these changes happen and if they do happen. Does not make sense. I will research deeply on waveguide and discrete face port.

•

9. Re: Antenna - input impedance (Zo= 50 Ω)

Do you discuss the line impedance or the input impedance into the device (line + antenna)?

And are the assumptions on the substrate (1.5mm FR4) correct for this example? The link to the PDF is broken.

10. Re: Antenna - input impedance (Zo= 50 Ω)

Originally Posted by volker@muehlhaus
And are the assumptions on the substrate (1.5mm FR4) correct for this example? The link to the PDF is broken.
paper: http://ictactjournals.in/paper/IJCT_..._1656_1663.pdf

lufer17 uses CST MWS, and his youtube video shows that software somehow calculated port impedance of 115.7 Ohms. And all confusion comes from the fact that port impedance is not set manually.

I took dimensions from paper and run some simulations, here is the result:

red curve: 60 Ohm lumped port - 60 Ohm transmission line - circular patch
blue curve: 115 Ohm lumped port - 60 Ohm transmission line - circular patch
both S11 looks well because antenna is very wideband, and with 115Ohm port impedance is "swirling" on the left part, crossing 115 Ohm point several times.
Maybe CST MWS took impedance from center frequency, and used real part as port impedance.

To obtain results similar to the paper port impedance must be manually set to 62 Ohm, or maybe:
1. Use 2.9mm wide line (50 Ohm)
2. Manually set port impedance to 50 Ohm
3. Add quarterwave transformer or linear taper to transform 50 Ohm (2.9mm) to 62 Ohm (2mm)

11. Re: Antenna - input impedance (Zo= 50 Ω)

Thanks Georgy!

FR4 1.5mm at these frequencies, really? I doubt the "gain" in the paper if loss is included in simulations.

--[[ ]]--