L-Network Matching
An L-network is the workhorse impedance-matching circuit of ham radio — two reactive components, one in series and one in shunt, arranged in an "L" shape to transform almost any real feedpoint resistance up or down to your coax's 50Ω. It shows up at the base of loaded verticals, inside antenna tuners, and anywhere a feedpoint lands somewhere other than 50Ω. This guide covers how the two-component transform actually works, low-pass vs high-pass topology, a fully worked component-value example, a calculator, and troubleshooting.
Why almost nothing feeds directly at 50Ω
A resonant half-wave dipole in free space presents about 73Ω at its center. A quarter-wave vertical over a good ground system presents roughly 36Ω. A shortened, loaded vertical can present anywhere from a few ohms to several hundred. None of these are 50Ω, and feeding a mismatched load directly with 50Ω coax reflects power back toward the transmitter, raising SWR and — at high mismatch — causing real power loss in the line and stressing the transmitter's output stage. Something has to sit between the antenna and the coax to transform the actual feedpoint resistance into the 50Ω the line and radio expect. The L-network is the simplest circuit that can do this for a purely resistive load.
How the series and shunt elements do the transform
An L-network places one reactive element in series with the line and a second reactive element in shunt (parallel) at one end of the network. The shunt element always goes on the higher-impedance side of the transform; the series element sits between the shunt node and the lower-impedance side. Together they form a single-frequency impedance transformer: the "Q" of the network (a measure of how hard it has to work) is set entirely by the ratio between the two impedances being matched, and that Q in turn sets both component values.
Low-pass vs high-pass topology
Swapping which element is the inductor and which is the capacitor changes the network from low-pass (series inductor, shunt capacitor) to high-pass (series capacitor, shunt inductor). Both transform the same impedances at the design frequency — the difference is what happens off-frequency.
- Low-pass (series L, shunt C): the standard choice for antenna matching — it attenuates harmonics on transmit instead of passing them, which is exactly the behavior you want between a transmitter and an antenna.
- High-pass (series C, shunt L): passes harmonics more readily, so it is rarely used for antenna feedpoint matching. It shows up more in receive-only or specialized applications where a physical shunt inductor is impractical to mount.
Where to put the network — feedpoint vs shack
An L-network built and tuned at the antenna feedpoint transforms the mismatch right where it occurs, so the coax running back to the shack always sees a flat 50Ω and carries no SWR-related loss regardless of run length. An antenna tuner in the shack does the same electrical job, but the coax between the antenna and the tuner still carries the original mismatch — on a long run of lossy coax at HF, especially on bands where the mismatch is severe, that adds up to real, measurable power loss that a feedpoint-mounted L-network avoids entirely. The tradeoff is that a feedpoint network has to survive outdoors and can't be retuned from the operating position without a remote-control mechanism.
| Method | Elements Needed | Bandwidth | Best For |
|---|---|---|---|
| L-Network | 1 series + 1 shunt (L and C) | Narrow, single frequency, field-tunable | General-purpose matching of any real feedpoint resistance, at the antenna or in the shack |
| Quarter-Wave Transformer (Q-Section) | 1 transmission-line section, ~λ/4 long | Narrow (roughly 10–15% of center frequency) | Matching two known real impedances with no lumped components; phased and stacked arrays |
| Stub Match (open/shorted λ/4 stub) | 1 shorted or open transmission-line stub | Narrow | Canceling reactance or providing a DC ground path at a specific point on the line, e.g. J-pole feeds |
| Series Match | 1 series L or C | Narrow | Canceling small residual reactance when the resistance is already close to Z0 |
L-Network Component Value Calculator
Materials for an L-network matching box
An L-network matching box at a vertical antenna's base, with the air-wound series inductor and shunt capacitor visible inside the weatherproof enclosure.
Building and Tuning an L-Network
Measure before you build — an L-network designed from an assumed impedance instead of a measured one almost never matches on the first try.
Measure the actual feedpoint impedance
With the antenna tuned to resonance (or as close as you'll get it), connect a NanoVNA at the feedpoint and measure R and X at your operating frequency. The L-network math in this guide assumes a purely resistive load — if X is more than about ±j15Ω, resonate the antenna further (trim length or adjust a loading coil) before designing the network, or the network will need to absorb that reactance too and the calculator's values will be off.
Calculate L and C values
Enter the measured resistance and your line impedance (almost always 50Ω) into the calculator above, or work through the Q-based equations by hand. Note which side gets the series element and which gets the shunt element — the calculator states this explicitly, and getting it backward is a common first-build mistake.
Wind the inductor
For an air-wound coil on PVC pipe, a workable starting approximation is Wheeler's formula: L (µH) ≈ (d² × n²) / (18d + 40ℓ), where d is the coil diameter in inches, ℓ is the coil length in inches, and n is the number of turns. Solve for n at your chosen form diameter, wind a couple of extra turns beyond the calculated count, and plan to trim turns off during tuning rather than trying to hit the value exactly on the first wind.
Mount the capacitor and wire the network
Mount the shunt capacitor and series inductor in the enclosure with the shortest practical leads — even a couple of inches of extra lead adds stray inductance that shifts your carefully calculated values. Wire per the low-pass topology (series inductor between the antenna feedpoint and the shunt node, shunt capacitor from that node to the coax shield/ground) unless you have a specific reason to use high-pass.
Tune and verify
Sweep SWR with the NanoVNA. If the SWR dip is at the right frequency but not deep enough, the capacitor value is likely slightly off — adjust it first, since it's usually the easier element to fine-tune. If the dip is at the wrong frequency, the inductor value needs adjusting — remove or add a partial turn and re-sweep. Once SWR is acceptable across your working bandwidth, weatherproof the enclosure and lock all connections against vibration.
| Symptom | Most likely cause | Diagnosis | Fix |
|---|---|---|---|
| SWR won't come down no matter how the capacitor is adjusted | Feedpoint resistance used for the design wasn't accurate, or the antenna carries significant reactance the network wasn't designed to absorb | Re-measure R and X directly at the feedpoint with a NanoVNA rather than relying on an assumed or textbook value | Re-resonate the antenna first if X is large, then recalculate L and C from the freshly measured R |
| SWR dip is sharp but sits at the wrong frequency | Inductor value is off from the design target | Compare the dip frequency to the design frequency — dip low usually means too much inductance, dip high means too little | Remove or add partial turns on the coil and re-sweep; small turn changes near the target usually converge quickly |
| Network runs noticeably warm or hot under moderate power | Inductor Q is too low (often from a compact, tightly-wound coil or a lossy core), or the network is being operated well off its tuned frequency | Check coil construction against the spaced, air-core recommendation, and confirm SWR is actually low at the operating frequency, not just close | Rewind with wider turn spacing and larger-diameter wire, or move to a larger-diameter coil form to raise Q |
| Capacitor arcs or shows scorch marks | Voltage rating too low for the actual RF voltage present, especially during tuning when SWR is still high | Estimate peak RF voltage from your power level and the SWR seen during tuning, and compare against the capacitor's rated working voltage | Replace with a higher-voltage-rated capacitor and reduce power during the initial tuning sweep |
| SWR was good at first but has drifted over weeks or months | Moisture ingress changing capacitor value, or a loosened connection/turn shifting inductance | Open the enclosure and inspect for corrosion, condensation, or a shifted turn on the coil | Improve weatherproofing (conformal coating or a better-sealed enclosure) and re-tighten or re-secure the coil winding |
Should I mount the L-network at the feedpoint or use an antenna tuner in the shack?
A feedpoint-mounted L-network transforms the mismatch right where it occurs, so the entire coax run back to the shack sees a flat 50Ω and carries no SWR-related loss. A shack tuner does the same job electrically but leaves the coax between the antenna and the tuner carrying the original mismatch, which costs real power on long or lossy runs. If the antenna is reachable and can host a weatherproof enclosure, a feedpoint network is usually the better long-term choice.
Low-pass or high-pass topology — which do I need?
Use low-pass (series inductor, shunt capacitor) for essentially every transmit antenna-matching application — it attenuates harmonics instead of passing them, which is the behavior you want between a transmitter and an antenna. High-pass topology exists but is rarely the right choice for this use case.
Can I use a toroid-wound inductor instead of air-wound?
Yes, for lower-power installations a toroid-wound inductor is more compact and just as effective — see the Core Selection Guide for choosing an appropriate core and mix. Air-wound coils remain the more common choice at higher power because they avoid the risk of core saturation and core heating that a toroid can suffer under high circulating current.
Why is the bandwidth so narrow?
An L-network is designed to solve one specific impedance ratio at one specific frequency — move far enough off that frequency and both the antenna's own impedance and the network's reactances shift, degrading the match. This is normal and expected; if you need multiband coverage from a single feedpoint, either accept the L-network's usable bandwidth on each band with retuning, or use a remote-controlled network that can retune itself per band.
My measured feedpoint has significant reactance — can I still use a simple L-network?
The equations in this guide assume a purely resistive load. If the antenna carries substantial reactance at your operating frequency, the cleanest fix is to resonate the antenna itself first (trim length or adjust a loading element) so the L-network only has to transform resistance. An L-network can be designed to absorb reactance too, but the math becomes more involved and is easy to get wrong by hand — resonating first is simpler and more reliable for a homebrew build.