The impedance matching calculator above does two related jobs. It tells you how badly mismatched a resistive load is today, expressed as standing wave ratio, return loss and mismatch loss, and it designs the two-element L-network that removes that mismatch at a chosen frequency. Both halves matter, because the amount of power a mismatch actually costs is usually much smaller than the standing wave ratio makes people assume.
Arb Digital publishes free engineering calculators that name their assumptions instead of burying them. This page assumes both impedances are purely resistive and both components are lossless, which is the textbook case that the closed-form L-network solution applies to. Real components have finite Q and real loads have reactance, so treat the values here as the starting point for a bench adjustment rather than as final part numbers.
Transmitting Needs a Licence
Matching networks are usually built to feed an antenna, and deliberately radiating radio-frequency energy requires a licence in most jurisdictions. Amateur, commercial and broadcast transmission are all licensed activities with allocated bands, power limits and technical conditions, and the rules differ by country. The ARRL guide to getting licensed is the usual starting point for the amateur service in the United States; elsewhere your national regulator publishes the equivalent. Power limits are usually written as an effective radiated figure rather than as transmitter output, so use the EIRP calculator to check where you stand once feeder loss and antenna gain are accounted for.
What This Impedance Matching Calculator Does
Maximum power transfer between a resistive source and a resistive load happens when the two resistances are equal. When they are not, some of the incident wave reflects back down the line. The reflection coefficient Γ measures that reflection, the standing wave ratio is a restatement of the same information in a form that a slotted line or an SWR meter reads directly, and mismatch loss converts it into the fraction of power that never reaches the load.
An L-network fixes it with two reactances. One sits in series with the lower resistance, raising the impedance seen looking into it; the other sits in shunt across the higher resistance, lowering it. Choose the two values correctly and the transformation is exact at the design frequency.
The hero figure is the standing wave ratio you have before matching. The grid gives the network's loaded Q, the series and shunt component values for the topology you selected, and the mismatch loss in decibels. The notes underneath give the alternative topology and the approximate bandwidth.
How to Use It
- Enter the two resistances. Order does not matter to the arithmetic; the tool works out which side is high and which is low and places the elements accordingly.
- Set the design frequency. The reactances are frequency-independent, but the inductance and capacitance that produce them are not, so this is what turns ohms into henries and farads.
- Pick a topology. Low-pass is the default because it suppresses harmonics as well as matching. High-pass is useful when you need DC isolation or want to reject low-frequency interference.
- Check the loaded Q. A large impedance ratio forces a high Q, which means a narrow match and high circulating currents. If the Q comes out above about ten, a single L-network is probably the wrong answer.
- Adjust on the bench. Fit the nearest standard values, then trim for a minimum reflection reading. Stray inductance in leads and self-capacitance in components both shift the answer.
The Formula: How the L-Network Is Derived
Call the larger resistance Rhigh and the smaller Rlow. The network's loaded quality factor is fixed entirely by their ratio: Q = √(Rhigh ÷ Rlow − 1). You do not get to choose it in a two-element network, which is the main limitation of the topology.
The series reactance is Xseries = Q × Rlow and the shunt reactance is Xshunt = Rhigh ÷ Q. Converting to components at frequency f, an inductive reactance gives L = X ÷ (2πf) and a capacitive reactance gives C = 1 ÷ (2πf X).
The mismatch figures come from the reflection coefficient Γ = (RL − Z0) ÷ (RL + Z0). Standing wave ratio is (1 + |Γ|) ÷ (1 − |Γ|), return loss is −20 log10|Γ|, and mismatch loss is −10 log10(1 − |Γ|²). The ARRL transmission line reference covers where these definitions come from.
Work the defaults through by hand. Matching 50 Ω to 10 Ω gives a ratio of 5, so Q = √4 = 2. The series reactance is 2 × 10 = 20 Ω and the shunt reactance is 50 ÷ 2 = 25 Ω. At 100 MHz that is an inductor of 20 ÷ (2π × 108) = 31.83 nH in series with the 10 Ω load, and a capacitor of 1 ÷ (2π × 108 × 25) = 63.66 pF across the 50 Ω source. Unmatched, Γ = (10 − 50) ÷ 60 = −0.6667, so the standing wave ratio is 1.6667 ÷ 0.3333 = 5.00, the return loss is 3.52 dB and the mismatch loss is 2.55 dB. A transformer would need a turns ratio of √5 = 2.236 to do the same job.
Why a High SWR Costs Less Power Than People Expect
A standing wave ratio of 2:1 sounds alarming and is often treated as a fault condition. In power terms it is not. The magnitude of the reflection coefficient at 2:1 is one third, so one ninth of the incident power reflects and eight ninths reach the load. That is a mismatch loss of 0.51 dB, which is about a tenth of an S-unit and entirely inaudible.
Even 3:1, which many operators would refuse to transmit into, only reflects a quarter of the power and costs 1.25 dB. The real reasons to care about standing wave ratio are different: solid-state amplifiers fold their output back or shut down when the reflection rises, the extra voltage and current on the line stress components and connectors, and a high standing wave ratio on a lossy feeder multiplies the feeder loss because the reflected wave travels the length of the cable a second time.
That last effect is the one worth calculating. On a short run of good cable it is negligible. On a long run of thin coax at VHF it can dominate everything else, which is why an antenna tuner at the transmitter end hides the mismatch from the amplifier without recovering the power already lost in the feeder.
What the Loaded Q Tells You About Bandwidth
Because a two-element network has only two degrees of freedom and both are spent achieving the match, the Q is whatever the impedance ratio dictates. Matching 50 Ω to 25 Ω gives a Q of 1 and a very broad match. Matching 50 Ω to 1 Ω gives a Q of 7, and matching to 0.5 Ω gives 9.95.
A rough guide to the usable bandwidth is the design frequency divided by the loaded Q. At a Q of 2 and 100 MHz that is roughly 50 MHz of useful width, which is generous. At a Q of 10 it is 10 MHz, which is narrow enough that component tolerances start to matter more than the design.
When the Q is uncomfortably high there are two standard answers. Split the transformation into two cascaded L-networks through an intermediate resistance, typically the geometric mean of the two ends, which halves the Q of each stage. Or move to a three-element pi or T network, where the extra component lets you set the Q independently of the impedance ratio — usually to make the match deliberately narrower for filtering, but the freedom works both ways.
Where This Sits Next to the Other RF Tools
This page designs a matching network between two resistances. It does not tell you what impedance a transmission line presents — the cable impedance calculator handles the line geometry, and the RLC impedance calculator and reactance calculator handle a load that has reactance you need to tune out before the L-network applies. For antennas, the dipole antenna calculator gives the physical length of a centre-fed half-wave element and the J-pole antenna calculator covers the end-fed case, where a quarter-wave stub does the matching instead of a lumped network. Convert any of the ratios here into decibels with the decibel calculator. The guided-wave theory behind all of it is covered in MIT OpenCourseWare's 6.013 Electromagnetics and Applications.
Arb Digital builds free tools like this one because useful pages earn attention. If you want tools, calculators or content built for your own audience, we can help.
Browse All Free Tools Talk to Arb DigitalCommon Mistakes to Avoid
- Putting the shunt element on the low-impedance side — the shunt always goes across the higher resistance, and reversing the two elements makes the mismatch worse rather than better.
- Feeding a reactive load into a resistive design — an antenna off resonance is not a resistor, and its reactance must be measured and tuned out or absorbed before these values mean anything.
- Reading SWR as a power loss — 2:1 costs about half a decibel, which is inaudible; the real cost is amplifier foldback and extra loss on a long lossy feeder.
- Ignoring component Q — a lossy inductor dissipates real power inside the network, and at high loaded Q the circulating current makes that loss much larger than the mismatch it was fixing.
- Assuming an antenna tuner recovers feeder loss — it presents a good match to the transmitter, but power already lost in the cable before the tuner is gone.
Related Free Tools From Arb Digital
Pair this with the EIRP calculator to check radiated power against a regulatory limit, and the free space path loss calculator for what happens to the signal after it leaves the antenna. The reactance calculator converts your component values back into ohms at any frequency, the RLC impedance calculator handles a full series or parallel network, and the wavelength calculator sets physical dimensions. For antenna geometry see the dipole antenna calculator and the J-pole antenna calculator. Everything Arb Digital publishes is on the free online tools hub.
Frequently Asked Questions
Across the higher of the two resistances, always. The series element goes on the lower side. The shunt reactance lowers the apparent resistance of the high side and the series reactance raises the apparent resistance of the low side, so putting them the other way round drives the two impedances further apart instead of together.
About 0.51 decibels, or roughly 11 per cent of the incident power. The reflection coefficient at 2:1 has a magnitude of one third, so one ninth of the power reflects. The practical problems with a high standing wave ratio are amplifier protection circuits folding back the output, component voltage stress, and extra loss on a long lossy feeder, not the reflected power itself.
Low-pass in most transmitting applications, because the series inductor and shunt capacitor also attenuate harmonics that the amplifier produces. High-pass is the better choice when you need to block DC, when you want to reject strong low-frequency interference, or when the low-pass component values come out impractically large or small.
It is set entirely by the ratio of the two resistances in a two-element network, and it controls the bandwidth. Usable bandwidth is roughly the design frequency divided by the Q. A high Q also means large circulating currents inside the network, so component losses and voltage ratings become significant well before the match itself fails.
Not directly with this design, which assumes both impedances are purely resistive. Measure the load's complex impedance first, then either resonate the reactance out with an opposite reactance and match the remaining resistance, or absorb it into the network by adjusting the component value that sits next to it. A vector network analyser makes both approaches straightforward.
Split the transformation into two cascaded L-networks through an intermediate resistance, typically the geometric mean of the two ends, which roughly halves the Q of each stage. Alternatively move to a three-element pi or T network, where the extra degree of freedom lets you choose the Q independently of the impedance ratio.
Deliberately radiating radio-frequency energy requires a licence in most jurisdictions, and the bands, power limits and technical conditions vary by country. The network itself is just passive components, but connecting it between a transmitter and an antenna puts you inside the licensing rules. Check with your national regulator before transmitting.
This tool is provided for educational and preliminary design use. It assumes purely resistive source and load impedances and lossless components, and does not account for component Q, self-resonance, stray reactance or voltage and current ratings. Transmitting requires a licence in most jurisdictions; confirm your obligations with your national regulator, and verify any design with a vector network analyser on the bench.
