SMA Connector Impedance

Question

I am new to antennas, and am trying to create a monopole that connects to an SMA female to transmit a 2.4ghz wifi signal. I know that the antenna needs to be a multiple of 3.1cm in order to have the maximum signal strength, and have also looked into having ground plane radials (which I don't completely understand the point of). However, one thing I don't understand at all is matching the impedance. I read online that I need to match the antenna impedance to the rest of the system, but I have no idea where to even start. Online it states that SMA connectors have 50 ohm impedance, however, it also states impedance is measured in an imaginary reactance component and a real resistance component. What do those mean and how can imaginary and real components add to 50. Lastly, how do I go about changing the impedance of the monopole to match the system.

The system: Router sma male-> SMA female to male cable -> sma female to female bulkhead -> antenna

Answer 1

Lots of questions here, let me try to cover them point by point.

• A monopole is half of an antenna. A real antenna is a dipole. The monopole needs a "ground plane" to replace the missing half of the dipole. Playing with the shape and angle of the ground plane radials changes the radiation pattern.
• Impedance is complex resistance. With an antenna, you can separate the resistance into loss resistance (which is turned into heat) and radiation resistance (which is turned into radio waves that escape). There is also reflection, which is what the antenna rejects and reflects back into the radio. The measure of what the radio sends vs. what the antenna bounces back is SWR. If your impedance is perfect (50+0i ohms), then there is no reflection and the SWR is 1:1.
• Matching impedance is a bit of black magic, but there are some obvious things. If the antenna is the wrong length, then the imaginary part of the impedance will not be zero. Generally, we would say that it is or is not resonant. Obviously, this can be fixed by changing the length. If the real part isn't 50 ohms, you can do things like change the shape of the ground plane. It is also possible to add a network of capacitors and inductors to electrically change the length. This is called a matching network. Generally, the matching network has a conjugate impedance from the antenna to try to cancel the imaginary part, but also possibly get it closer to 50 ohms if that isn't right either.
• Matching networks add loss resistance to the system. Tweaking the antenna directly is more efficient if it is convenient. If the bandwidth of the antenna isn't wide enough, you might have to just get close, and then use a matching network to get it closer when you change frequencies. However, at higher frequencies, it may just be easier to bend the ground plane a little to tune it.

Answer 2

More examples:

Characteristic Impedance may be matched by lossy resistance source or loads but it is the lossless impedance defined a square root ratio of positive and negative lossless reactance simply $$Z_o= \sqrt{\dfrac{L}{C}} ~~~~[\Omega]$$ [nH/mm / pF/mm or H/F] or both.

Lossless reflections requires the conjugate match of complex reactance. E.g. inductance feeding capacitance or x+jy(f) matching with x-jy(f). That is useful for RF channels.

For baseband channels, in logic for example a 25 ohm 74ALCxx CMOS driver step pulse may have DC ~ 1GHz BW which causes edge ringing on long xx ohm wires from the echoes. So adding a small series R, Rs of xx ohms dampens the ringing by getting closer to matching the path impedance = Zo whether it is PCB trace and gnd plane or coax or twisted pair each with known ratios. But this also attenuates level, so the choice of matching impedance is a tradeoff with ringing and attenuation and risetime, if overdamped. Yet smaller lithography has faster rise times with smaller C and smaller FET resistance and thus smaller Vol/Iol= Ron of Nch switch in CMOS.

Impedance matching on antenna is done by measuring reflections in a splitter or the maximum Return Loss by trimming dipoles for example.

A matched impedance network is always 50% lossy yet maximum power is transferred.

Consider a DC photovoltaic Solar cell. It is a current source (hi-Z) driving a voltage sink (low-Z) . Although it does not dissipate heat if open circuit, maximum power transfer (MPT) occurs only when the load impedance is matched. This is also the calculated impedance of open cct. Voltage over short-circuit current or Zmpt= Voc/Isc = Vmpt/Impt where V*I=Power.

Answer 3

If you're just starting out a recommended DIY build for such a "monopole" antenna would be something like in this Ground-Plane antenna for 2,4 Ghz photograph. (Don't miss the links to http://www.jensenjensen.com/radio/GP24antenna.htm and http://www.csgnetwork.com/antennagpcalc.html which may be helpful as well.)

I know that the antenna needs to be a multiple of 3.1cm in order to have the maximum signal strength, and have also looked into having ground plane radials (which I don't completely understand the point of)

The size will be somewhere in that ballpark but you may need to trim it a bit to get the best match. This will be a difficult to do unless you have tools to measure how well the antenna is absorbing power after you make it.

The ground "plane" (radials in this case) sort of has two purposes:

For one it helps "complete the circuit" of the antenna. Of course an antenna isn't exactly a "circuit" since current has to just displace in and out rather than going through/around in the "electric circuit" sense, so more accurately the ground radials help balance the antenna. Meaning, when the center conductor of the coax/connector is pushing electrons into the monopole the outer conductor wants something to pull from to keep things even overall.

The other purpose is to "hide" the antenna from being interfered with by some of its surroundings, in your case the cable itself. See https://practicalantennas.com/designs/verticals/gp2/ for some more discussion but it's often easier to use a monopole plus four more radials than to build a dipole (monopole plus another monopole) in the first place simply because it means you don't have to worry about keeping the feeder coax out of the way.

However, one thing I don't understand at all is matching the impedance. I read online that I need to match the antenna impedance to the rest of the system, but I have no idea where to even start. Online it states that SMA connectors have 50 ohm impedance, however, it also states impedance is measured in an imaginary reactance component and a real resistance component.

So 50 Ω impedance of the SMA connector or any coax cable connected is only its "characteristic" impedance. It helps to have it correct but ultimately your transmitter will end up only caring that the impedance of the antenna alone is correct.

Impedance sort of means "how does it absorb energy". A resistive ("real") impedance means the energy simply "goes away". The energy leaves the circuit, either as heat or as radio waves. But there's also reactive impedance, which is what the "imaginary" part of the math is used for. This reactive part represents energy going into a capacitor or inductor. That energy is absorbed too, but it's not really gone — it will bounce back into the circuit sooner or later.

Lastly, how do I go about changing the impedance of the monopole to match the system.

You're trying to get the feedpoint to behave approximately as a 50 Ω resistive impedance, across as much of the frequency range needed. For a quarter-wave ground plane the biggest adjustments will be:

1. the length of the main "monopole" element (so it's not acting like much of an inductor or capacitor at the center frequency)
2. the angle of the radials (bending them down tends to lower the overall impedance)
3. the length of the radials (honestly I'm not sure how much of an effect this would/wouldn't have)

But you'll need some sort of way to measure if the changes you make are working or not. This would be an antenna analyzer of some sort or another: an SWR meter or a VNA or a dip meter or…. The received signal strength at a certain distance might sort of provide insight but won't directly or definitively tell you if you have a good match.