How to use
- Measure the antenna at its feed point with a vector network analyzer (VNA) and read R + jX at the operating frequency, or enter S11 in dB with its phase angle.
- Enter the frequency and system impedance (normally 50 Ω) and press Calculate.
- Pick a solution: prefer a low-pass network (series inductor, shunt capacitor) because it also attenuates harmonics; choose high-pass when you need DC blocking.
- Fit parts closest to the calculated values in the table, then re-measure and fine-tune with the network analyzer.
How it works
When an antenna is not matched to the 50 Ω line, part of the power is reflected. The reflection coefficient is Γ = (ZL − Z0) / (ZL + Z0), VSWR = (1 + |Γ|) / (1 − |Γ|) and return loss RL = −20·log₁₀|Γ|. A matching network uses lossless inductors and capacitors to transform the antenna impedance to Z0 so that Γ approaches zero.
An L-network uses only two components and always has at least two solutions for any R > 0: shunt-then-series (from the antenna) works when the antenna conductance G ≤ 1/Z0, and series-then-shunt works when R ≤ Z0. A Pi-network adds a third component and lets you set Q: higher Q gives a narrower bandwidth, roughly f / Q. Inductance is L = X / (2πf) and capacitance C = 1 / (2πf·|X|); shunt parts are converted from susceptance B.
Frequently asked questions
Why match an antenna to 50 Ω?
Most RF modules, coaxial cables and test instruments are 50 Ω. If the antenna deviates from 50 Ω, transmit power is reflected back into the module, reducing range, increasing current draw and, in severe cases, destabilising the power amplifier. VSWR ≤ 2 (return loss ≥ 9.5 dB) is usually acceptable and VSWR ≤ 1.5 is good.
Should I use an L-network or a Pi-network?
An L-network has the fewest parts and the lowest loss, which suits most antenna tuning. A Pi-network adds one part but lets you control Q and bandwidth. Many production boards reserve a Pi footprint (a 0 Ω link plus two unpopulated pads) and populate only the parts the measurement calls for.
Which matching network has the lowest loss?
Ideal parts are lossless; in practice most of the loss comes from inductors (0402 multilayer inductors have Q of roughly 30–50, while capacitors are usually above 200). An L-network has the fewest parts and the lowest Q, so it normally has the lowest loss. A Pi-network adds a part and a higher Q, so its loss is larger and can even exceed the loss of leaving the antenna unmatched. Putting the inductor in the shunt-to-ground position and a series capacitor in the signal path is usually the lowest-loss option. The calculator estimates the percentage of power reaching the antenna for each solution from the inductor and capacitor Q you enter and marks the lowest-loss one. If harmonics must be suppressed, choose a low-pass network anyway.
Why does the measured result differ from the calculation?
The calculation assumes ideal components. Real inductors and capacitors have parasitics and a self-resonant frequency, and PCB traces and pads add reactance. At UHF (for example 915 MHz) and above, use 0402 or smaller RF-grade parts and re-measure at the feed point once or twice to fine-tune.
How do I convert S11 to impedance?
Γ = 10^(S11dB/20)·∠angle and the antenna impedance is ZL = Z0·(1 + Γ) / (1 − Γ). When you select the S11 input mode, the calculator converts it automatically and fills in R and X.
Does matching make the antenna efficient?
Matching only removes reflection; it does not improve the antenna's radiation efficiency or pattern. An antenna that is too small, lacks ground plane or sits next to metal must be addressed in the antenna design and placement. WENSHING can help with antenna selection and custom design.