Skip to content
RFrftools.io
Antenna

Microstrip Patch Antenna Calculator

Calculate rectangular microstrip patch antenna dimensions (width, length) using the Transmission Line Model. Outputs effective dielectric constant, edge-feed impedance, and nominal gain for common substrates like FR4 and Rogers.

Loading calculator...

Formula

W=c2f2εr+1,L=c2fεr,eff2ΔLW = \frac{c}{2f}\sqrt{\frac{2}{\varepsilon_r+1}}, \quad L = \frac{c}{2f\sqrt{\varepsilon_{r,\text{eff}}}} - 2\Delta L

Reference: Balanis, "Antenna Theory: Analysis and Design", 4th ed., Chapter 14

WPatch width (m)
LPatch length (m)
εrSubstrate relative permittivity
εr_effEffective relative permittivity
ΔLEnd-effect fringing extension (m)
cSpeed of light (299 792 458 m/s) (m/s)
fOperating frequency (Hz)

How It Works

Patch antenna calculator computes resonant length, width, feed position, and bandwidth for microstrip patch antennas on any PCB substrate — wireless device engineers, GPS receiver designers, and phased array architects use this to design low-profile integrated radiators and scalable arrays. The rectangular patch resonates when its length L approximately equals lambda_eff/2, where lambda_eff = lambda_0/sqrt(epsilon_eff) accounts for the substrate's effective dielectric constant, per Balanis's 'Antenna Theory' (4th ed.) and Pozar's 'Microwave Engineering'.

Patch dimensions for 50-ohm edge feed: width W = c/(2*f*sqrt((epsilon_r+1)/2)) provides good radiation efficiency (typically 90%+); length L = c/(2*f*sqrt(epsilon_eff)) - 2*delta_L corrects for fringing fields at radiating edges, where delta_L approximately equals 0.412*h*(epsilon_eff+0.3)(W/h+0.264)/((epsilon_eff-0.258)(W/h+0.8)). For FR-4 (epsilon_r = 4.4) at 2.4 GHz: W approximately equals 38 mm, L approximately equals 29 mm.

Bandwidth is inherently narrow: BW = (VSWR-1)/(Q*sqrt(VSWR)) where Q approximately equals c*sqrt(epsilon_eff)/(4*f*h). Typical 1.6 mm FR-4 patch at 2.4 GHz has Q approximately equals 30 and 2% bandwidth (48 MHz). Thicker substrates and lower epsilon_r increase bandwidth: 3.2 mm Rogers RO4003 (epsilon_r = 3.55) achieves 5% bandwidth. Gain is typically 6-9 dBi for single elements, increasing 3 dB per doubling of array elements.

Worked Example

Problem: Design a 2.4 GHz WiFi patch antenna on standard 1.6 mm FR-4 substrate (epsilon_r = 4.4, tan_delta = 0.02).

Dimension calculation per transmission line model:

  1. Substrate parameters: h = 1.6 mm, epsilon_r = 4.4
  2. Calculate patch width for good efficiency:
W = c/(2*f*sqrt((epsilon_r+1)/2)) = 3e8/(2*2.4e9*sqrt(2.7)) = 38.1 mm

  1. Effective dielectric constant:
epsilon_eff = (epsilon_r+1)/2 + (epsilon_r-1)/2 * (1+12*h/W)^(-0.5) epsilon_eff = 2.7 + 1.7*(1+0.504)^(-0.5) = 2.7 + 1.39 = 4.09
  1. Length extension for fringing:
delta_L = 0.412*h*(epsilon_eff+0.3)(W/h+0.264)/((epsilon_eff-0.258)(W/h+0.8)) delta_L = 0.412*1.6*(4.39)(24.1)/((3.83)(24.6)) = 0.74 mm
  1. Resonant length:
L = c/(2*f*sqrt(epsilon_eff)) - 2*delta_L L = 3e8/(2*2.4e9*sqrt(4.09)) - 1.48 = 30.9 - 1.48 = 29.4 mm

Performance analysis:

  1. Q factor: Q = c*sqrt(epsilon_eff)/(4*f*h) = 3e8*2.02/(4*2.4e9*0.0016) = 39.5
  2. Bandwidth (VSWR < 2): BW = 1/(Q*sqrt(2)) = 1.8% = 43 MHz (covers single WiFi channel)
  3. Gain estimate: G = 4*pi*W*L*radiation_eff/lambda^2 = 6.5 dBi
  4. Efficiency: radiation efficiency approximately 85% (limited by FR-4 tan_delta = 0.02)
Feed design (inset feed for 50 ohms):
  1. Edge impedance: Z_edge approximately equals 200-400 ohms for this geometry
  2. Inset distance: y_0 = L/pi * acos(sqrt(50/Z_edge)) approximately equals 8-10 mm from edge
  3. Verify with VNA: adjust inset by +/-1 mm to minimize S11 at 2.4 GHz

Practical Tips

  • For prototyping, design patch 5% larger than calculated and trim with razor blade while monitoring S11 on VNA — much faster than iterating PCB fabrication
  • Use coaxial probe feed for narrow bandwidth applications (simpler) or aperture coupling for wider bandwidth (more complex but better performance)
  • For arrays, space elements 0.5-0.7 lambda_0 center-to-center to balance gain, sidelobe level, and mutual coupling — closer spacing increases coupling, wider spacing creates grating lobes

Common Mistakes

  • Ignoring effective dielectric constant — using epsilon_r directly gives wrong resonant length; epsilon_eff is always lower than epsilon_r due to fringing fields in air above substrate
  • Neglecting substrate loss in efficiency calculation — FR-4 (tan_delta = 0.02) limits radiation efficiency to 80-90%; PTFE substrates (tan_delta < 0.001) achieve > 95% efficiency
  • Using thin substrates for wideband applications — 0.8 mm substrate has Q approximately equals 80 (1% BW); need 3.2+ mm substrate for 5%+ bandwidth suitable for WiFi bands
  • Expecting accurate resonant frequency from formulas alone — manufacturing tolerances in epsilon_r (+/-5%) and h (+/-10%) cause 2-5% frequency shift; always include tuning margin in design

Frequently Asked Questions

Bandwidth is inversely proportional to Q factor, which scales as Q approximately equals epsilon_r^(3/2)/(h/lambda_0). Three factors increase bandwidth: (1) Thicker substrate — doubling h approximately doubles bandwidth. (2) Lower dielectric constant — foam (epsilon_r = 1.1) provides 3x bandwidth of FR-4 (epsilon_r = 4.4). (3) Lower epsilon_r — also increases patch size. Typical bandwidth: 1-2% for 1.6 mm FR-4, 3-5% for 3 mm Rogers, 10-15% for stacked patch or U-slot designs. For WiFi (100 MHz BW at 2.4 GHz = 4%), use 3+ mm low-loss substrate.
Yes — the transmission line model is frequency-independent. Key scaling: patch dimensions scale inversely with frequency. At 5.8 GHz versus 2.4 GHz: dimensions shrink by 2.4x. At 915 MHz versus 2.4 GHz: dimensions grow by 2.6x. Practical limits: at 5.8 GHz, patch is approximately 12 mm on FR-4 (easy to fabricate); at 915 MHz, patch is approximately 85 mm (may need air-dielectric for manageable size). Above 10 GHz, tight fabrication tolerances require +/-0.1 mm etching accuracy.
Transmission line model provides +/-5% resonant frequency accuracy per Balanis. Error sources: (1) Dielectric constant variation: FR-4 is 4.0-4.8 depending on glass content and resin — get actual value from laminate datasheet. (2) Fringing field approximation: accurate for W/h > 1, less accurate for narrow patches. (3) Substrate thickness tolerance: +/-10% typical for 1.6 mm FR-4. For production, simulate with 3D EM solver (HFSS, CST) before fabrication. For prototypes, design with 5% tuning margin and iterate.
Common feeds ranked by complexity and performance: (1) Microstrip line edge feed: simplest, narrow bandwidth, poor for thick substrates due to surface wave. (2) Inset feed: direct 50-ohm match by insetting feed line into patch; most common for single-layer designs. (3) Coaxial probe: drills through substrate to patch interior; best for thick substrates but narrow bandwidth. (4) Aperture coupling: two-layer design with slot in ground plane; widest bandwidth (10%+), best isolation, most complex. (5) Proximity coupling: two-layer with no galvanic connection; good bandwidth, moderate complexity. For WiFi/Bluetooth, inset feed on single-layer is standard practice.
Single patch gain is limited to 6-9 dBi by aperture size. Improvement options: (1) Array: each doubling of elements adds 3 dB. 2x2 array = +6 dB. 4x4 array = +12 dB. (2) Larger patch on lower-epsilon substrate: increases aperture but also resonant frequency. (3) Stacked patches: parasitic element above driven patch increases gain 1-2 dB and bandwidth. (4) Reflector plane at lambda/4 spacing: adds 3 dB but increases profile. (5) Higher-efficiency substrate: Rogers RO4003 (tan_delta = 0.0027) versus FR-4 (tan_delta = 0.02) adds 0.5-1 dB. For maximum gain, use array on low-loss substrate with proper corporate feed network design.

Shop Components

As an Amazon Associate we earn from qualifying purchases.

SMA Right-Angle Connectors

Edge-mount and right-angle SMA connectors for antenna feeds

RTL-SDR Dongle

Wideband SDR receiver for antenna and signal experiments

Magnet Wire (22 AWG)

Enameled copper wire for winding custom antennas and coils

Related Calculators