Optical fibres
Optical fibres: total internal reflection, numerical aperture and acceptance angle, V-number and modes, dispersion and attenuation, fibre types and losses, sources/detectors, and typical applications.
Drafted with Aria, reviewed by the AiCanCode.org team · Updated . How we write and check our notes. Spotted an error? Use Give Feedback at the bottom of the page.
Why it matters
Optical fibres carry huge amounts of data over long distances with low loss and immunity to electromagnetic interference, so every branch of engineering touches them through communications, sensing, or control systems. Engineers use fibres in internet backbones, LANs, medical imaging (endoscopes), structural health monitoring (fibre sensors), and industrial automation where electrical noise would corrupt copper cables.
Key ideas
- Refraction, critical angle, and total internal reflection (TIR)
- Definition: Refractive index (n) is the ratio c/v, where c is the speed of light in vacuum and v is the speed in the medium.
- Law (Snell’s law): At a boundary between media of refractive indices n₁ and n₂, for incidence angle i and refraction angle r measured from the normal: n₁ sin i = n₂ sin r. Assumes homogeneous, isotropic, non-absorbing media and a smooth interface.
- Definition: The critical angle (i_c) is the angle of incidence in the denser medium (n₁ > n₂) for which the refraction angle is 90°. Exact statement: sin i_c = n₂ / n₁.
- Definition: Total internal reflection (TIR) occurs when light travels from denser to rarer medium (n₁ > n₂) and i > i_c; then all light is reflected back into the denser medium. TIR is the guiding mechanism in fibres.
- Numerical aperture (NA) and acceptance angle
- Definition: The acceptance angle (θ_a) is the maximum half-angle with the fibre axis at which light can enter the fibre from an external medium (index n₀) and still be guided.
- Definition: The numerical aperture (NA) is NA = n₀ sin θ_a. It measures light-gathering ability; larger NA means easier coupling.
- Derivation (step-index fibre): Consider a meridional ray entering from external medium (n₀) into core (n₁), then hitting the core–cladding boundary (n₂). Applying Snell’s law at input and TIR condition at the core–cladding boundary gives the exact result
- n₀ sin θ_a = √(n₁² − n₂²)
- Hence NA = √(n₁² − n₂²), and sin θ_a = √(n₁² − n₂²) / n₀
- Definition: The relative index difference (Δ) is Δ = (n₁ − n₂)/n₁ (Δ ≪ 1 in practice). For small Δ, NA ≈ n₁ √(2Δ).
- Modes and V-number
- Definition: An optical mode is a stable electromagnetic field distribution that satisfies Maxwell’s equations and boundary conditions in the fibre.
- Definition: The normalized frequency (V-number) is V = (2πa/λ) √(n₁² − n₂²), where a is the core radius and λ is the free-space wavelength.
- Result (step-index fibres):
- Single-mode condition: V < 2.405 (first zero of J₀ Bessel function).
- Approximate number of guided modes (meridional approximation) for multimode step-index fibre: M ≈ V²/2.
- Graded-index fibres have fewer modes than step-index for the same V, and intermodal delay is reduced.
- Fibre types and index profiles
- Step-index multimode (SI-MM): core has uniform n₁; abrupt step to n₂ at cladding. Supports many modes for large a or large NA.
- Graded-index multimode (GI-MM): core index decreases gradually from centre to edge, typically nearly parabolic: n(r) = n₁ [1 − 2Δ (r/a)^α]^(1/2) with α ≈ 2. This compensates path differences and reduces intermodal dispersion.
- Single-mode step-index (SMF): small core radius (≈ 4–5 μm) and low NA so V < 2.405 at operating λ; only fundamental mode propagates, minimizing modal dispersion.
- Dispersion and bandwidth
- Definition: Dispersion is pulse spreading due to different frequency components or modes traveling at different speeds, which limits data rate.
- Types:
- Intermodal dispersion (multimode fibres): different modes take different paths and arrive at different times; largest in SI-MM, much reduced in GI-MM, absent in single-mode.
- Material (chromatic) dispersion: refractive index depends on wavelength, so different spectral components travel at different group velocities; depends on source spectral width Δλ and fibre material.
- Waveguide dispersion: in single-mode fibres, confinement causes wavelength-dependent mode propagation constant; can cancel material dispersion at specific wavelengths.
- Definition: The dispersion parameter (D) quantifies chromatic dispersion: Δt = D L Δλ, where Δt is pulse broadening, L is length, and Δλ is source spectral width. Typical units: ps/(nm·km).
- Bandwidth limit (rule of thumb): For non-return-to-zero (NRZ) on–off keying, bit rate B_max ≈ 0.5/Δt, where Δt is total pulse broadening.
- Attenuation (loss)
- Definition: Attenuation is the reduction in optical power with distance. In linear units: P(z) = P(0) e^(−αz), where α is the attenuation coefficient (Np/m). In decibels: Loss_dB = 10 log₁₀(P_in / P_out).
- Unit conversion: α_dB/m = 4.343 α (Np/m). Often quoted as dB/km.
- Mechanisms:
- Material absorption: intrinsic (electronic/vibrational) and extrinsic (impurities like OH−).
- Rayleigh scattering: from microscopic density fluctuations; ∝ 1/λ⁴; dominant in low-loss windows.
- Mie scattering and bending losses (macrobending, microbending): curvature or small bends cause radiation of guided power.
- Telecom windows: around 850 nm, 1310 nm, and 1550 nm; lowest attenuation near 1550 nm for silica, with dispersion management at 1310/1550 nm.
- Coupling, splicing, and connectors
- Splicing joins fibres permanently: fusion splicing melts ends together (low loss, typically <0.1 dB) and mechanical splicing aligns ends mechanically (slightly higher loss).
- Connectors provide removable joints (e.g., FC, SC, ST, LC). Loss depends on alignment, end-face quality, and cleanliness.
- Sources and detectors
- Optical sources: LEDs (broad spectrum, lower power, used with MM at 850 or 1300 nm), laser diodes (narrow spectrum, higher power, used with SMF at 1310 or 1550 nm).
- Detectors: PIN photodiodes (fast, low noise), avalanche photodiodes APD (higher sensitivity with gain but more noise and bias voltage). Responsivity in A/W converts received optical power to current.
- Typical exam diagrams (describe to draw and label)
- Step-index fibre ray diagram: Draw a long rectangle core (label n₁) surrounded by cladding (label n₂ < n₁). Show fibre axis as a dashed line. Draw two meridional rays zig-zagging via TIR at core–cladding boundary and a central axial ray. Label incidence angle at core–cladding boundary and indicate TIR.
- Acceptance cone at input face: Draw fibre end-face circle; axis as a line. Draw a cone of rays with half-angle θ_a entering from external medium (label n₀). Mark and label θ_a and NA = n₀ sin θ_a.
- Refractive index profiles: Plot n (vertical) vs radius r (horizontal). For SI: a flat line at n₁ until r = a, then a step down to n₂. For GI: a curve decreasing from n₁ at r = 0 to n₂ at r = a (nearly parabolic). Label a, n₁, n₂, Δ.
Formulas
Relative index difference: Δ = (n₁ − n₂) / n₁— dimensionless; small for telecom fibres. For small Δ:NA ≈ n₁·√(2Δ).V-number: V = (2·π·a/λ)·√(n₁² − n₂²)— a: core radius (m), λ: wavelength in vacuum (m); V dimensionless.Single-mode condition: V < 2.405— step-index fibres.Number of modes (SI-MM): M ≈ V² / 2— approximate, dimensionless.Attenuation (linear): P(z) = P(0)·e^(−α·z)— P in watts, α in Np/m, z in m.Attenuation (decibels): Loss_dB = 10·log₁₀(P_in / P_out)— dimensionless dB; per length:α_dB/km = (10/L_km)·log₁₀(P_in/P_out).Np to dB: α_dB/m = 4.343·α— converts α (Np/m) to dB/m.Chromatic dispersion: Δt = D·L·Δλ— Δt in s, D in s/(m·m) but commonly ps/(nm·km); L in km, Δλ in nm when using those units.Bit rate limit (NRZ): B_max ≈ 0.5 / Δt— B in Hz; Δt in s; rough rule of thumb.
Worked examples
- NA, acceptance angle, V-number, and number of modes Given: n₁ = 1.48 (core), n₂ = 1.46 (cladding), external medium air n₀ = 1.00, wavelength λ = 850 nm, core radius a = 25 μm. Steps:
- Formula for NA: NA = √(n₁² − n₂²). Compute n₁² − n₂² = (1.48)² − (1.46)² = 2.1904 − 2.1316 = 0.0588. NA = √(0.0588) = 0.2426 (dimensionless).
- Acceptance half-angle: sin θ_a = NA / n₀ = 0.2426 / 1.00 = 0.2426. θ_a = arcsin(0.2426) ≈ 14.0°.
- V-number: V = (2πa/λ)·NA. Convert: a = 25 μm = 25 × 10⁻⁶ m; λ = 850 nm = 850 × 10⁻⁹ m. 2πa/λ = 2π × (25 × 10⁻⁶)/(850 × 10⁻⁹) = 2π × 29.4118 ≈ 184.80. V = 184.80 × 0.2426 ≈ 44.83 (dimensionless).
- Number of modes (step-index MM): M ≈ V² / 2 = (44.83)² / 2 ≈ 2009 / 2 ≈ 1005. Answer: NA = 0.2426, θ_a ≈ 14.0°, V ≈ 44.83, modes ≈ 1.01 × 10³ (multimode).
- Fibre attenuation from input/output powers Given: Input power P_in = 2.0 mW, Output power P_out = 0.50 mW, fibre length L = 5.0 km. Steps:
- Formula: α_dB/km = (10/L_km)·log₁₀(P_in/P_out).
- Compute power ratio: P_in/P_out = 2.0 / 0.50 = 4.0.
- log₁₀(4.0) = 0.60206.
- α_dB/km = (10 / 5.0) × 0.60206 = 2.0 × 0.60206 = 1.204 dB/km. Answer: Attenuation = 1.204 dB/km.
- Pulse broadening and bit-rate estimate from chromatic dispersion Given: D = 17 ps/(nm·km), fibre length L = 50 km, source spectral width Δλ = 0.20 nm. Estimate max NRZ bit rate. Steps:
- Formula: Δt = D·L·Δλ.
- Compute Δt: Δt = 17 × 50 × 0.20 ps = 170 ps = 170 × 10⁻¹² s.
- Rule-of-thumb bit rate: B_max ≈ 0.5 / Δt = 0.5 / (170 × 10⁻¹²) Hz.
- Calculate: 0.5 / 170e−12 ≈ 2.941 × 10⁹ Hz = 2.94 Gbps. Answer: Pulse broadening ≈ 170 ps; estimated B_max ≈ 2.94 Gbps (chromatic-dispersion-limited).
Common mistakes
- Units in dispersion: Using D in ps/(nm·km) but L in metres and Δλ in nm without converting; keep consistent units.
- Counting modes: Applying M ≈ V²/2 to graded-index fibres without noting it is for step-index and is approximate.
- dB arithmetic: Subtracting powers in mW directly to get dB; must use logarithm of the ratio.
- Ignoring bending losses: Tight bends in lab setups cause unexpected high loss; macrobending cutoff depends on fibre and bend radius.
- Overstating bandwidth: Quoting B ≈ 1/Δt (too optimistic) for NRZ; use a realistic factor (≈0.5) unless a specific pulse shape is stated.
For your exam
- Diagrams: step-index ray diagram; acceptance cone; refractive index profiles.
- Numericals: compute NA and θ_a; compute V and number of modes; attenuation in dB/km; pulse broadening from D, L, Δλ; bit rate estimate from Δt.
- Full-marks answers include: clear statement of laws/definitions with assumptions, neat labeled diagrams, correct formulas with units and stepwise calculations, and a concluding sentence interpreting the result (e.g., multimode vs single-mode, practical implication).
Quick check
- State the condition for total internal reflection at the core–cladding boundary.
- A fibre has n₁ = 1.50, n₂ = 1.48, n₀ = 1.00. Find NA and θ_a.
- Why do graded-index fibres have lower intermodal dispersion than step-index fibres? Answers: 1) Light must travel from higher to lower refractive index (n₁ > n₂) and the incidence angle at the boundary must exceed the critical angle i_c, where sin i_c = n₂/n₁. 2) NA = √(1.50² − 1.48²) = √(2.25 − 2.1904) = √(0.0596) = 0.2442; θ_a = arcsin(0.2442) ≈ 14.1°. 3) The graded (approximately parabolic) index makes outer rays travel faster than inner rays, compensating longer paths so different modes arrive closer in time, reducing intermodal delay.
Finished this topic? Mark it so your progress, study plan and readiness keep up.
Stuck on something here?