Antenna Downtilt Calculator
Find the downtilt that puts a sector antenna's upper half-power edge on the cell edge, the ground footprint of the half-power beam, and the gain toward the cell edge from the 3GPP vertical pattern, with Earth curvature included.
Effective-Earth geometry (ITU-R P.526-15 Eq. 44) · tilt = θ_e + θ₃dB/2 · 3GPP TR 38.901 vertical patternAPI slug antenna-downtilt
Results
LiveRecommended Downtilt 5.1°
- Depression Angle to Cell Edgeθe
- 1.6°
- Inner Half-Power Edge (recommended tilt)din
- 187.7m
- Inner Half-Power Edge (applied tilt)din
- 190.7m
- Outer Half-Power Edge (applied tilt)dout
- 1.091km
- Relative Gain Toward Cell Edge (applied tilt)AV
- −2.8dB
- Line-of-Sight Distancedlos
- 27.62km
Recommended Downtilt against Cell-Edge Distance
Drag the operating point or type a value.
Formula
3GPP TR 38.901 V17.0.0, Table 7.3-1; ITU-R P.526-15, Eqs. (21) and (44); ITU-R P.834-9, §2
| Symbol | Quantity | Your value |
|---|---|---|
| Δh | Antenna height minus user-equipment height (m) | — |
| d | Ground distance to the cell edge (m) | 1 km |
| kR | Effective Earth radius, R = 6371 km (m) | — |
| θe | Depression angle to the cell edge (°) | 1.6° |
| θ3dB | Vertical half-power beamwidth (°) | 7.0° |
| AV | Relative gain of the vertical pattern toward the cell edge (dB) | −2.8 dB |
Assumptions
6 items
- 01The ground is a smooth sphere of effective radius k × 6371 km, and the depression of a point at user height a distance d away is atan(Δh/d + d/(2kR)) (ITU-R P.526-15 Eq. 44); terrain, buildings and clutter are not modelled.
- 02The recommended tilt puts the beam’s upper half-power edge on the cell edge, θ_e + θ₃dB/2; interference into neighbouring cells and the SINR are not optimised.
- 03The footprint is where the beam edges at the tilt ± θ₃dB/2 meet the user-height surface. An edge at or above the horizontal, or shallower than the grazing angle √(2Δh/kR), never comes down and is reported as unbounded, not as a number.
- 04The gain toward the cell edge is the vertical cut of 3GPP TR 38.901 Table 7.3-1 with your beamwidth, −min(12((θ_e − tilt)/θ₃dB)², 30) dB: a parabolic main lobe on a 30 dB floor, with no side lobes, null fill or horizontal pattern.
- 05The applied tilt is the total of electrical and mechanical tilt, taken as a rotation of the whole vertical pattern; the azimuth effects of mechanical tilt are not modelled.
- 06The line-of-sight distance is √(2kR)·(√h_BS + √h_UT) (ITU-R P.526-15 Eq. 21), and a cell edge beyond it is flagged; both heights are above the same smooth ground.
Worked example
A 30 m macro sector for a 1 km cell
A sector antenna with a 7° vertical beamwidth sits 30 m up and should cover users 1.5 m above the ground out to 1 km, at k = 1.33333. It is tilted 2° today. The calculation gives the tilt that puts the upper half-power edge on the cell edge, and what the 2° tilt does instead.
| Step | Expression | Result |
|---|---|---|
| Height of the antenna above the users | 30 m − 1.5 m | 28.5 m |
| Effective Earth radius | 1.33333 × 6 371 km | 8 494.7 km |
| Depression angle to the cell edge | atan(28.5 m / 1 km + 1 km / (2 × 8 494.7 km)) | 1.6° |
| Recommended downtilt | 1.6° + 7° / 2 | 5.1° |
| Inner half-power edge at that tilt | 2 × 28.5 m / (tan φ + √(tan² φ − 2 × 28.5 m / 8 494.7 km)), φ = 5.1° + 7° / 2 | 187.7 m |
| Upper half-power edge at the applied 2° | 2° − 7° / 2 < 0 | ∞ m |
| Inner half-power edge at 2° | 2 × 28.5 m / (tan φ + √(tan² φ − 2 × 28.5 m / 8 494.7 km)), φ = 2° + 7° / 2 | 296 m |
| Gain toward the cell edge at 2° | −min(12 × ((1.6 − 2) / 7)², 30) | −0.032 dB |
| Line-of-sight distance | √(2 × 8 494.7 km) × (√30 m + √1.5 m) | 27.62 km |
How it works
A sector antenna on a mast concentrates its power in a narrow vertical beam. Pointed at the horizon, the strongest part of that beam flies over the cell and lands in the neighbours' cells as interference. Tilting it down puts the energy where the users are. The usual rule is to tilt until the upper half-power edge of the beam meets the ground at the cell edge, so the cell edge sees the beam 3 dB down and everything beyond it sees less.
The angle below the antenna's horizontal at which a user at distance appears depends on the height difference and, slightly, on the curvature of the Earth. Over an effective Earth of radius , with km and for the standard atmosphere, it is the elevation formula of ITU-R P.526-15 (equation 44) with the arctangent kept:
The recommended downtilt adds half the vertical beamwidth :
The half-power footprint
The two half-power edges of the beam point at below the horizontal. Each meets the ground at the nearer distance that has that depression angle. The upper edge never comes down if it points at or above the horizontal, which happens whenever the tilt is no greater than half the beamwidth. The beam then reaches the horizon, and the calculator reports no outer edge rather than a distance.
The vertical pattern
For the gain toward the cell edge at the tilt you actually apply, the calculator uses the vertical cut of the antenna element model in 3GPP TR 38.901, Table 7.3-1, with your beamwidth in place of the table's 65°:
It is exactly −3 dB half a beamwidth from boresight, −12 dB one beamwidth away, and never below the 30 dB side-lobe floor.
Validity
The ground is a smooth sphere: terrain, buildings and clutter are not modelled. The cell edge should lie within the line-of-sight distance between the antenna and user heights, (ITU-R P.526-15, equation 21). Beyond it the user is below the radio horizon, and the calculator warns. The pattern is a parabolic main lobe with a floor. A real antenna has nulls and upper side lobes, and its datasheet pattern is the one to check before the final tilt is set.
Practical tips
- Start from the recommended tilt, then add a degree or two where neighbouring cells overlap heavily; the edge gain output shows what that costs at the cell edge.
- Prefer electrical tilt for most of the angle. It tilts the whole pattern uniformly, while mechanical tilt narrows the horizontal coverage at large angles.
- Check the inner edge too. With a narrow beam and a tall mast the near ground may fall outside the half-power beam and rely on the side lobes.
- For long rural cells keep k = 4/3, and check with k = 2/3 to see how sub-refraction moves the edge angle.
Common mistakes
- Setting the tilt equal to the depression angle of the cell edge. That points boresight at the edge, so the upper half of the beam, with nearly full gain, carries on into the next cell.
- Using the horizontal beamwidth. The tilt depends on the vertical half-power beamwidth from the datasheet, typically 5° to 15° for a macro panel.
- Measuring the antenna height from sea level or the building's base for some sites and from the ground for others. Use the height above the ground the users stand on, and the user height above the same ground.
- Treating mechanical and electrical tilt as the same everywhere. They agree in the main direction, which is what this calculator models, but mechanical tilt reduces the tilt toward the sector edges and tilts the back lobe upward.
Frequently asked questions
5 questions
How much should a cell antenna be tilted down?
Enough that the upper half-power edge of the vertical beam meets the ground at the cell edge: the depression angle to the cell edge plus half the vertical beamwidth. For a 30 m mast, a 1 km cell and a 7 degree beam that is about 5.1 degrees.
What does it mean when the outer edge shows no value?
The upper half-power edge of the beam points at or above the horizontal, so it never meets the ground. The beam then covers everything out to the horizon, including other cells. This happens whenever the applied tilt is no greater than half the vertical beamwidth.
Where does the edge gain come from?
From the vertical pattern of the antenna element model in 3GPP TR 38.901, Table 7.3-1: a parabolic main lobe that is 3 dB down half a beamwidth from boresight and 12 dB down one beamwidth away, with a floor at 30 dB. The calculator uses your beamwidth in place of the table's 65 degrees.
Does Earth curvature matter for downtilt?
Only a little at macro-cell distances. At 1 km it changes the angle to the cell edge by about 0.2 percent; at 10 km and beyond it becomes noticeable. The calculator includes it through the effective Earth radius, k times 6371 km.
Is electrical tilt different from mechanical tilt?
In the main direction of the sector they are the same, and that is what this calculator models. Away from boresight, mechanical tilt gives less tilt and raises the back lobe, while electrical tilt lowers the whole pattern evenly.
References
Study on channel model for frequencies from 0.5 to 100 GHz · Propagation by diffraction · Effects of tropospheric refraction on radiowave propagation
- Study on channel model for frequencies from 0.5 to 100 GHz — 3GPP TR 38.901 V17.0.0 (2022-03), §7.3, Table 7.3-1 — radiation power pattern of a single antenna element
- Propagation by diffraction — Recommendation ITU-R P.526-15 (10/2019), §3.2 Eq. (21) and §4.4 Eq. (44) — line-of-sight distance and elevation over an effective Earth
- Effects of tropospheric refraction on radiowave propagation — Recommendation ITU-R P.834-9 (12/2017), §2 — effective Earth radius, k = 4/3
The pattern points (−3 dB at half the beamwidth from boresight, −30 dB floor) and the horizon limit are exact; as k grows the geometry reduces to the flat-Earth tangent.
Shop Components
As an Amazon Associate we earn from qualifying purchases.