Author Topic: pi-network Impedance Matching for an RF antenna  (Read 2533 times)

0 Members and 1 Guest are viewing this topic.

Offline roogadget

  • Contributor
  • Posts: 27
  • Country: au
pi-network Impedance Matching for an RF antenna
« on: November 17, 2018, 11:39:12 am »
I need to select the capacitor and resistor (or inductor) in order to match the RF antenna impedance (50 ohm).
I am confused about the "source impedance".
Should I also consider the parasitic capacitance and inductance of the trace?
If the source impedance is simply the module to the antenna, shouldn't the resistor just be 50 ohms to match the output impedance?
I am using an online calculator to find out what my capacitor values and resistors need to be (pi matching calc google..)
Designing for 2.6GHz.

I've updated the circuit image.

I've watched a great video from Michael Ossmann about RF circuit design .

Do I actually need a pi-matching network at all if the impedance of the trace is approx. 50 ohm.
(See the image attached "Microstrip Impedance Calculator input.png")

Also, if I do use a pi network, how do i find the Q-factor? (see pi_network_impedance_matching.png).
I've been all over the web and it isn't yet clear what they are asking for. Is it the inductor's internal resistance? Or the formula from qfactor.png.

Side note:
The 0 ohm resistor from my schematic was just taken from the datasheet. I believe the 0R ohm resistor was placed there to tell me that it should be left unpopulated when putting the board together for "tweaking".

Thanks for your help!
« Last Edit: November 17, 2018, 11:35:42 pm by roogadget »

Offline Benta

  • Super Contributor
  • ***
  • Posts: 3082
  • Country: de
Re: pi-network Impedance Matching for an RF antenna
« Reply #1 on: November 17, 2018, 12:03:04 pm »
No resistor, you need an inductor. A resistor will just waste power.
If you don't know the output impedance of U2 (source impedance), there's not a lot you can do.
If it's different from 50 ohms, you need the pi matching.
If it's 50 ohms, you don't really need a pi matching circuit.
The following users thanked this post: roogadget

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 5964
  • Country: au
Re: pi-network Impedance Matching for an RF antenna
« Reply #2 on: November 17, 2018, 12:47:52 pm »

As Benta says, you should not have a resistor there, a pi network as used for impedance matching is an LC circuit.
I'm wondering if whoever you got the circuit off, purely by serendipity, found that a zero ohm resistor of some kind looked like an inductor of the correct value for his/her particular application.
That doesn't mean that any such resistor will do in anyone's application.
The following users thanked this post: roogadget

Online Bud

  • Super Contributor
  • ***
  • Posts: 4758
  • Country: ca
Re: pi-network Impedance Matching for an RF antenna
« Reply #3 on: November 17, 2018, 01:46:41 pm »
As well as he found the two electrolytic caps looked like inductors  :-DD
Facebook-free life and Rigol-free shack.

Offline cdev

  • Super Contributor
  • ***
  • Posts: 6449
  • Country: 00
Re: pi-network Impedance Matching for an RF antenna
« Reply #4 on: November 17, 2018, 02:30:01 pm »
You are always likely to be using non-polarized capacitors, In an HF antenna matching network, for a fixed frequency device they may be fixed, for a ham radio antenna tuner they will typically be variable capacitors or combinations of the two of some kind.

Similarly with the inductor, (not resistor) you would likely want to be able to change whatever inductance you use there to optimize the match. You might do that with a switch on a "transmatch" or antenna tuner box, on a PCB you would usually use fixed values.

In microwave devices its common to incorporate matching and other RF-active transformations into the PCBs design in the form of various modifications to the traces. (Way more complicated than you'd need to do here).

One kind of Pi-matching network - assuming an inductor there in the signal path instead of resistor, also helps filter out harmonics if its on a transmitter. It also acts as a low pass filter. Another kind of pi-network, one where a capacitor works with two inductors to ground acts as a high pass filter.

The impedance of the device there may be nominally 50 ohms, but in a real world situation it in fact may vary a lot.  MMICs may specify 50 ohms but in fact many of them in fact have a real input impedance for the purpose of receiving that is much higher than that and when that is the case, indeed, they will perform markedly better on receive when you match the input from the antenna to their real input impedance, not 50 ohms. The real load impedance on a transmitter?  I suspect that in a school question they'd want you to match the output to a resistive load first. (this is common practice in the real world too) In a real world device, it depends on whether you know in advance what the antenna is going to be and what its immediate environs are in its use cases. (proximity to other objects will impact its performance and impedance a lot)

if you are using coax feedline to the antenna, they mean typically 50 ohms (more common) or occasionally 75 ohms in the case of some equipment that you would be expected to provide a match to under all circumstances.

But then an antenna matching circuit, for HF use by amateur radio operators, an ideal antenna tuner is there to provide a match to a wide range of antennas. Even electrically short or long resonant and non-resonant wires, loops and so on. So there, the wider the range your components can adjust to, often the better.

At HF no the traces matter but unless they are made in such a way that ignores basic rules of what an RF trace should do you can likely just do that and figure in the nominally small values only if it would seem to matter right there in your circuit (if they were high enough to relative to the filter values). At VHF/UHF and above yes, they matter, or even make a huge difference - more and more the higher you go in frequency.

« Last Edit: November 17, 2018, 03:07:36 pm by cdev »
"What the large print giveth, the small print taketh away."
The following users thanked this post: roogadget

Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo