Ham Radio Antenna Modeling & Software

What NEC2 Actually Does

NEC (Numerical Electromagnetics Code) is a method-of-moments computational solver originally developed at Lawrence Livermore National Laboratory. It solves for the current distribution along thin-wire antenna structures by dividing each wire into small segments and solving a matrix equation that satisfies Maxwell's equations at every segment junction.

From this current distribution, NEC calculates:

• Radiation patterns in three dimensions — azimuth and elevation
• Gain in any specified direction — in dBd or dBi
• Feedpoint impedance — R + jX at each frequency
• SWR relative to any specified reference impedance
• Near-field and far-field electromagnetic field strength
• Efficiency including ground loss and loading element loss

NEC4 extends NEC2 with improved ground modeling (buried wires, insulated wires in soil) and better accuracy for antennas close to real ground. For most wire antenna modeling above ground, NEC2 and NEC4 produce essentially identical results.

Wire Segmentation — The Most Important Modeling Parameter

Every wire in a NEC model is divided into segments — short sections over which the current is assumed to vary linearly. The number of segments per wire determines both modeling accuracy and calculation time. Getting segmentation right is the single most important factor in producing accurate NEC results.

• Minimum segments per half-wavelength: 10 — this is an absolute floor
• Recommended segments per half-wavelength: 15–25 for most antenna work
• Segment length should be uniform within each wire where possible
• At wire junctions, connected segments should be similar in length — avoid ratios greater than 3:1
• Short wires connected to long wires (loading coils, stubs) require careful handling
• Adding more segments increases accuracy but increases calculation time quadratically

A common modeling error is using too few segments on elements that are short relative to wavelength — loading coils, top-hat capacitors, and matching networks. These should be modeled with at least 5–10 segments regardless of physical length.

Ground Models — Choosing the Right One

How you model the ground beneath the antenna significantly affects the predicted radiation pattern and gain. NEC2 offers several ground models:

• Free space — no ground. Use for initial design and when comparing antenna types on equal footing
• Perfect ground — a perfectly conducting infinite plane. Maximum theoretical performance. Not realistic but useful for understanding pattern shapes
• Real ground (Sommerfeld-Norton) — models actual soil conductivity and dielectric constant. Most accurate for antennas close to real ground. Requires soil parameters (conductivity σ and permittivity ε)
• High-accuracy ground (NEC4) — required for buried radials, insulated conductors in soil, and antennas very close to lossy ground

Typical soil parameters: average ground σ=0.005 S/m, ε=13. Good agricultural soil σ=0.02, ε=20. Poor rocky soil σ=0.001, ε=5. Seawater σ=5.0, ε=81. Getting these right matters most for verticals and low dipoles — antennas at λ/2 height and above are relatively insensitive to ground quality.

What Models Cannot Tell You

NEC models are powerful but have important limitations that every modeler must understand to avoid misinterpreting results:

• Infinite ground plane — NEC models assume the ground extends to infinity. Real terrain — hills, valleys, buildings — is not modeled (use HFTA for terrain effects)
• No feedline radiation — NEC does not model common-mode current on coax shields. A model may show good gain while the real antenna radiates differently due to feedline effects
• Thin wire assumption — NEC2 is not accurate when wire diameter approaches wavelength. For thick conductors, NEC4 or FDTD (OpenEMS) is required
• No connector or matching network loss — model results assume ideal connections. Real-world loss in connectors, baluns, and matching networks reduces efficiency
• No mechanical effects — models assume perfect geometry. Sag in long wire antennas and wind-induced element movement are not modeled
• Frequency extrapolation — model accuracy degrades when the antenna structure has features smaller than λ/100 at the modeled frequency

Common Modeling Mistakes and How to Avoid Them

• Too few segments — always check that each half-wavelength has at least 10 segments, preferably 15–25
• Mismatched segment lengths at wire junctions — connected wires should have segment length ratios no greater than 3:1
• Modeling in free space and forgetting that real ground changes everything — always add a realistic ground model before drawing conclusions
• Trusting the model without a convergence check — run with double the segment count and verify results do not change significantly
• Modeling the antenna without the feedline — common-mode current on the coax affects real antenna patterns but cannot be shown in a simple NEC model
• Using perfect ground for anything other than theoretical comparisons — perfect ground dramatically overstates gain and distorts patterns
• Ignoring wire radius — element diameter affects resonant frequency and bandwidth; always enter the actual conductor radius

Interpreting Radiation Pattern Plots

A radiation pattern plot shows signal intensity in all directions relative to the antenna's maximum — it is a normalized relative plot, not an absolute power measurement. Key things to read from a pattern plot:

• Main lobe direction — the direction of maximum radiation. For a dipole this is broadside; for a Yagi it is forward
• Takeoff angle — the elevation angle of maximum radiation. Lower is better for long-distance HF propagation
• Front-to-back ratio — difference in dB between forward gain and rearward radiation. Important for directional antennas
• Sidelobe levels — secondary lobes that may cause or receive interference from unwanted directions
• Gain value — peak gain in dBi or dBd. Remember to verify whether the software is reporting dBi or dBd
• Null depth — how completely the antenna rejects signals from null directions. Deep nulls can help reject interference

When to Use NEC4 Instead of NEC2

For most amateur radio wire antenna work, NEC2 produces accurate and reliable results. NEC4 is specifically required when:

• Modeling buried radials for ground-mounted verticals — NEC2 cannot accurately model conductors in soil
• Modeling insulated wires — loading coils with insulated wire, insulated guy wires, or any conductor with significant dielectric coating
• Antennas very close to real ground — within a few percent of a wavelength above ground
• Junction modeling accuracy — NEC4 handles T-junctions and Y-junctions more accurately than NEC2
• High-accuracy modeling of thick conductors near ground

In practice: use NEC2 (MMANA-GAL or 4NEC2 free) for horizontal wire antennas, loops, Yagis, and verticals with above-ground radials. Use NEC4 (4NEC2 with NEC4 engine, or EZNEC Pro/4) when modeling buried radials or accurate ground-mounted vertical performance.

Modeling Workflow Best Practices

• Start with free space to verify basic antenna geometry and resonance before adding ground effects
• Use a convergence test — double segment count and verify results are stable before trusting them
• Model at the actual operating frequency, not a scaled frequency
• Enter real wire dimensions — conductor diameter, actual heights, realistic element spacing
• Save a baseline model before each optimization iteration
• Always validate with a physical measurement — build the antenna and compare NanoVNA results to the model prediction
• Use frequency sweeps rather than single-frequency calculations — bandwidth and SWR behavior across the band reveal design trade-offs
• When optimizing, change one parameter at a time and record the effect before changing another
• Document your models with comments — note the design goals, the date, and what each version changed