Optical Components¶

5-mark answer and 10-mark extension

For 5 marks — core system chain¶

A fiber-optic communication system carries an information-bearing electrical signal by converting it to light, guiding that light through a dielectric fiber and converting it back to an electrical signal. Its physical path is

1. Input/encoder: formats voice, video or data and may add line coding.
2. Source driver and optical source: the driver modulates an LED for economical short multimode links or a laser diode for high-speed single-mode links.
3. Coupler, connectors and splices: launch light into the core and join cable sections while adding small insertion losses.
4. Optical fiber channel: core and lower-index cladding guide light by total internal reflection; attenuation reduces power and dispersion broadens pulses.
5. Photodetector: a reverse-biased PIN or APD converts received optical power to photocurrent, \(I_p=RP_r\), where \(R\) is in A/W and \(P_r\) in W.
6. Receiver: a transimpedance amplifier, filter/equalizer, clock recovery and decision circuit reconstruct the output bits.

The basic power test is

with every subtracted term in dB; \(P_r\) must exceed receiver sensitivity. The system provides huge bandwidth, low loss, EMI immunity and electrical isolation, but needs careful source/fiber matching, precise joints and electro-optic equipment.

Add for a 10-mark system explanation¶

At the transmitter, the line coder produces a timing-suitable waveform, the current driver sets source bias and modulation, and the LED/laser converts carrier recombination energy to photons. Coupling optics align the emitted field with the fiber numerical aperture and core. Direct intensity modulation is common; externally modulated lasers are used when chirp and very high bit rate matter.

The channel includes fiber, connectors, splices and possibly optical amplifiers. At \(850\), \(1310\) or \(1550\,\text{nm}\) the designer selects fiber loss, source spectrum and dispersion appropriate to reach. Repeatable connectors are used at equipment; low-loss fusion splices are used along permanent cable routes.

At the receiver, photons generate primary electron-hole pairs. A PIN has no internal multiplication and low noise; an APD supplies avalanche gain at high reverse bias but adds excess noise. The TIA converts small current to voltage, filtering limits noise, equalization compensates channel response, and a threshold sampler decides each bit using recovered clock timing.

Two independent quantitative checks are required. If \(P_t=-2\,\text{dBm}\), total path loss is \(8\,\text{dB}\) and receiver sensitivity is \(-14\,\text{dBm}\), then \(P_r=-10\,\text{dBm}\) and margin is \(4\,\text{dB}\). Timing must also satisfy

for NRZ bit rate \(B\) in bit/s. Thus enough optical power does not guarantee freedom from dispersion-induced ISI. OTDR and optical-power-meter tests verify installed length, event loss and end-to-end power. Applications include FTTH, LANs, metro/backbone and submarine systems; limitations include bending, dirty end faces, source aging, receiver noise and the repair skill required for glass fiber.

Practice target: 9 minutes for the basic chain or 18 minutes for the detailed transmitter/channel/receiver operation and both budget checks.

Exam-ready answer

An LED is a forward-biased semiconductor junction in which injected electrons and holes recombine by spontaneous emission. Its active layer may emit through the surface or edge; there is no resonant feedback requirement, so photon phase and direction are random. A laser diode is a forward-biased heterojunction active region placed between reflecting cleaved facets. Above threshold current, population inversion and optical feedback make stimulated emission dominate, producing photons of nearly identical frequency, phase and direction. In either source photon energy is

so the semiconductor bandgap is chosen for fiber windows such as \(850\), \(1310\) or \(1550\,\text{nm}\).

Source spectral width creates chromatic broadening \(\Delta t=|D|\Delta\lambda L\). For the same fiber with \(D=17\,\text{ps/(nm·km)}\) over \(20\,\text{km}\), a \(30\,\text{nm}\) LED spectrum gives about \(10.2\,\text{ns}\) broadening, whereas a \(1\,\text{nm}\) laser gives \(0.34\,\text{ns}\). Therefore LED reliability and low cost suit short multimode systems; laser power, narrow linewidth and speed justify its cost, threshold behavior, eye-safety requirement and thermal sensitivity on long high-capacity links.

Practice target: 8 minutes; explain both emission mechanisms, sketch the source position in the link and reproduce at least six comparison rows.

5-mark answer and 10-mark extension

For 5 marks — detector comparison¶

A fiber receiver uses a reverse-biased photodiode to convert photons into electron-hole pairs. Detector responsivity is

where \(\eta\) is quantum efficiency and \(I_p\) is primary photocurrent.

A PIN photodiode has thin p and n regions separated by a wide intrinsic absorption layer. The reverse field sweeps photoelectrons and holes across this layer rapidly, producing \(I_p=RP_{opt}\) with multiplication \(M=1\).

An APD has an absorption region followed by a very-high-field multiplication region. Primary carriers gain enough kinetic energy for impact ionization; the carrier cascade gives \(I_{out}=MRP_{opt}\).

Add for a 10-mark detector explanation¶

In a PIN, the intrinsic region gives high photon absorption, low junction capacitance and short carrier transit time. The receiver TIA supplies electrical gain after detection. Performance is limited by quantum efficiency, dark current, thermal/shot noise and bandwidth, but its linearity and low excess noise make PIN the default detector when received power is adequate.

In an APD, the absorption field creates primary carriers and a separate high field creates secondary pairs. Multiplication improves signal before following amplifier noise, but it is statistical. A simplified multiplied shot-noise relation is

where \(B\) is noise bandwidth in Hz, \(I_d\) is the relevant primary dark current and \(F(M)>1\) is the excess-noise factor. Raising \(M\) therefore does not improve SNR indefinitely; there is an optimum gain, and bias must track temperature to avoid breakdown.

For \(R=0.8\,\text{A/W}\) and \(P_{opt}=2\,\mu\text{W}\), a PIN gives

An APD at \(M=10\) gives \(I_{out}=16\,\mu\text{A}\) before considering excess noise. This quantitative gain can extend a power-limited span, while a PIN may still give the better cost, linearity and noise result on a strong short link. Both require wavelength matching, low-capacitance packaging, reverse-bias safety and a bandwidth compatible with the bit rate; neither corrects fiber dispersion.

Practice target: 9 minutes for structures and comparison or 18 minutes with multiplication noise, the current calculation and receiver-selection reasoning.

Exam-ready answer

An optical transmitter accepts information in electrical form and launches a corresponding modulated optical signal into fiber. Its physical chain is line coder/encoder → current driver and bias → LED or laser diode → coupling optics/connector. The source converts carrier-recombination energy to photons; an LED is economical for short multimode links, whereas a narrow-linewidth laser gives greater power and speed for long single-mode links.

An optical receiver performs the inverse conversion. Its chain is fiber/coupler → reverse-biased PIN or APD → transimpedance amplifier → filter/equalizer → clock recovery and threshold decision → electrical output. The detector produces primary current \(I_p=RP_r\) in amperes, where \(R\) is responsivity in A/W and \(P_r\) is incident power in W.

They are linked quantitatively by \(P_r=P_t-\sum L\) in dB units: the transmitter launch power minus fiber, splice and connector losses must leave \(P_r\) above receiver sensitivity. For example, \(-3\,\text{dBm}\) launch and \(9\,\text{dB}\) path loss give \(-12\,\text{dBm}\) at the receiver; a \(-15\,\text{dBm}\) sensitivity leaves \(3\,\text{dB}\) margin. Together they enable telecom, FTTH and data links; neither alone overcomes excessive channel attenuation or dispersion.

Practice target: 7–8 minutes; draw both conversion chains, give five differences and close with one power-margin check.

5-mark answer and 10-mark extension

For 5 marks — joints and OTDR principle¶

An optical connector is a detachable precision alignment between fiber end faces and equipment or another fiber. SC push-pull and compact LC connectors dominate telecom/patch panels; ST uses a bayonet and FC a threaded coupling. Lateral/angular offset, end gap, core or NA mismatch, dirt and poor polish cause loss and reflection. Insertion and return loss are

low \(IL\) and high \(RL\) are desirable.

A splice permanently joins two prepared fibers. Fusion splicing strips, cleans, cleaves, aligns and arc-melts the ends, then protects and tests the joint. Mechanical splicing holds cleaved ends in an alignment fixture with index-matching gel. Fusion gives lower loss/reflection and greater long-term strength; mechanical splicing is quicker and cheaper for restoration but has higher loss and reflection.

An OTDR launches a short pulse through a directional coupler and records Rayleigh backscatter plus Fresnel reflections versus round-trip time. Event distance is

where \(c\) is in m/s, \(t\) in s and fiber index is \(n\).

Add for a 10-mark testing explanation¶

The physical connector uses polished ferrules and an adapter sleeve to center the cores. Clean inspection is essential because a particle comparable with the roughly \(9\,\mu\text{m}\) SMF core can cause major loss. APC polish reduces reflected power where lasers, analog video or high-split PONs demand high return loss. A connector is reusable and convenient but normally lossier and less stable than a sound fusion splice.

On an OTDR trace, normal fiber is a downward backscatter slope whose gradient gives dB/km. A reflective connector causes a spike followed by a step; a good fusion splice gives a non-reflective step; a bend gives wavelength-sensitive loss; an open break/end gives a large reflection and then the noise floor. Event loss is the level difference across the event, while end-to-end attenuation is confirmed with a light source and power meter because OTDR loss can be direction-dependent.

For \(n=1.5\) and a return at \(t=100\,\mu\text{s}\),

Longer OTDR pulses increase dynamic range but worsen spatial resolution and dead zones; shorter pulses resolve close events but have less reach. A launch fiber reveals the first connector and a receive fiber reveals the last. Testing at \(1310\) and \(1550\,\text{nm}\) helps distinguish bends. These methods install and maintain FTTH, LAN, metro and long-haul plant; limitations include bidirectional testing needs, strong-reflection dead zones, incorrect index settings and inability of OTDR alone to certify live service BER.

Practice target: 9 minutes for definitions and principles or 18 minutes with the comparison, labeled trace, distance calculation and test limitations.