Antenna gain is directionality, not amplification. dBi = dBd + 2.15, and every 3 dB of extra gain halves the coverage angle — which is why a 3 dBi omni usually beats a 9 dBi omni on a device whose orientation you cannot control.
Gain compares an antenna's radiation intensity in its best direction against a reference antenna fed with the same power:
| Unit | Reference | Typical use |
|---|---|---|
| dBi | Isotropic radiator (perfect point source, radiates equally in all directions) | Datasheets, link budgets, EIRP calculations |
| dBd | Half-wave dipole | Legacy VHF/UHF and broadcast specs |
The conversion is exact and worth memorising:
dBi = dBd + 2.15
A half-wave dipole — the classic "rubber duck" reference element — has 2.15 dBi of gain simply because it is not isotropic. So a "3 dBd" antenna is a "5.15 dBi" antenna. When a supplier quotes dBd and you budget in dBi, you silently lose 2.15 dB of margin. Always convert before comparing.
Gain and coverage angle are reciprocal. For a single main lobe, a working approximation is:
G ≈ 41 253 / (θ_E × θ_H) — G in linear units, beamwidths in degrees.
| Nominal gain | What happened to the pattern | Consequence |
|---|---|---|
| 2–3 dBi | Near-omnidirectional, fat doughnut | Works at any device tilt; short-to-medium range |
| 5–6 dBi | Doughnut flattened | More range on the horizon, weaker above and below |
| 8–9 dBi | Flat, narrow vertical slice | Long horizontal reach, real nulls overhead and underneath |
| 12 dBi+ | Tightly collimated | Point-to-point only; alignment is now part of the install |
+3 dB of gain = half the beam area. That is the sentence to take into a design review.
A common field failure looks like this: a water meter with a 3 dBi antenna is upgraded to 9 dBi to "fix" a weak link, and some units get worse. The reason is geometry, not gain. A meter in a pit may sit at any tilt; the 9 dBi pattern has nulls off the horizon, and a tilted meter now points its null at the gateway. The 3 dBi part radiated less power toward the horizon but covered every tilt. If device orientation is uncontrolled, the lower-gain antenna usually wins on link availability — not on peak range.
Two more traps hide behind the gain number:
1. Gain is often peak gain at one frequency. Ask for gain versus frequency across the band. 2. Small embedded antennas can show decent peak gain at poor efficiency. A lossy but well-matched 2.4 GHz chip antenna may claim 2 dBi peak while radiating only 40–50 % of accepted power. Efficiency — not gain — sets real range. See the companion page on efficiency and radiation patterns.
The honest comparison is realised gain: gain(dBi) + 10·log10(total efficiency). If a datasheet only gives peak gain, you are comparing half a number.
| Situation | Sensible gain | Why |
|---|---|---|
| Handheld / wearable / moving asset | 0–3 dBi | Orientation never controlled |
| Meter in a pit, cabinet, or basement | 2–5 dBi, or relocate the antenna | Coverage beats peak; metal and soil detune anyway |
| Fixed gateway, endpoints all around at similar height | 5–8 dBi omni | Flat pattern matches a horizontal endpoint plane |
| Fixed gateway on a mast, wide area | 8–10 dBi + low-loss feeder | See cable-loss page before buying gain |
| Point-to-point link, both ends fixed | 12 dBi+ directional | Alignment is acceptable, so collimation pays |
Contact: Mr.Li
Phone: 0755-33169935
Tel: +86-755-33169935
Email: cs#antennahome.com (change # to @ )
Add: Build-2 201#, ShangXue Industry Park, Bantian, Longgang District, Shenzhen, Guangdong Province, China