Photonics and Optical Engineering Exercises

These exercises practise photonics and optical engineering calculations as review evidence. They connect wavelength, optical power, photon flux, optical loss, beam geometry, diffraction, fiber coupling, laser drive current, detector response, noise-equivalent power, saturation, bandwidth, pulse fluence, alignment tolerance, thermal drift, uncertainty and validation.

The purpose is not only to calculate a number. The purpose is to decide whether an optical system has enough signal, margin, alignment tolerance, detector range and environmental stability to support an engineering decision.

Assume simplified screening models unless an exercise states otherwise. Real optical systems also require measured source spectra, calibrated power meters, connector inspection, beam profiling, stray-light control, safety review, detector linearity, thermal testing, mechanical tolerance analysis and uncertainty budgets.

Release Evidence Notes

Use the worked answers as optical release evidence only when the wavelength, source drive state, optical boundary, connector condition, alignment state, aperture geometry, detector bandwidth, calibration record, thermal condition and uncertainty budget match the system being released. Optical power at the source is not equivalent to useful signal at the detector, sample or reported software boundary.

The strongest release records separate source output, path transmission, beam geometry, detector range, noise floor, timing and environmental drift. Each gate should name the measured plane, calibrated instrument, uncertainty guard, contamination or alignment inspection, and the required action when signal margin, saturation margin, irradiance limit or thermal drift no longer supports operation.

How to Use These Exercises

For each exercise, define:

1. the optical boundary: source output, fiber launch, sample plane, detector surface, receiver input or reported software value;
2. whether power is linear watts, dBm, dB ratio, photon flux or irradiance;
3. wavelength, bandwidth, aperture, spot size, source drive state and detector state;
4. the acceptance rule: power margin, resolution, coupling loss, SNR, bandwidth or drift limit;
5. the validation evidence needed before release.

The common mistake is to treat the optical component as the whole system. A source, lens, fiber, detector or amplifier can be correct in isolation and still fail after alignment, contamination, temperature, bandwidth, background light or calibration are included.

Engineering Boundary Notes

These exercises use simplified optical and photonic screening models. They do not replace source characterization, beam profiling, fiber inspection, stray-light testing, detector linearity validation, optical safety review, thermal qualification, mechanical tolerance analysis or calibrated system-level testing. A calculated optical margin applies only to the stated wavelength, path, aperture, alignment, detector state, bandwidth and environmental condition.

Separate the optical planes before accepting a result. Source output, launched fiber power, free-space beam power, sample-plane irradiance, detector photocurrent, transimpedance output and reported software value can all differ. Release evidence should name the plane being accepted and explain how loss, contamination, polarization, saturation, noise and drift are transferred across the path.

Common Release Mistakes

• using source optical power as detector evidence without path loss and alignment checks;
• treating dB, dBm, watts, photon flux and irradiance as interchangeable quantities;
• accepting coupling efficiency without connector cleanliness, numerical aperture, focus and lateral offset records;
• checking detector signal while ignoring saturation, bandwidth, background light, dark current or transimpedance limits;
• releasing a laser or optical sensor without thermal drift, safety margin and calibration traceability;
• changing optics, fiber routing, polarization state, aperture or firmware scaling after validation.

Scenario Map

Scenario	Main calculation	Engineering decision
Optical power boundary	photon flux, dB loss, dBm power and uncertainty	Confirm that useful power exists at the correct plane.
Beam and image geometry	Gaussian propagation, diffraction, aperture clipping, irradiance and pulse fluence	Check whether the optical field fits the downstream system.
Fiber coupling	numerical aperture, lateral offset and alignment tolerance	Decide whether launch conditions preserve link margin.
Polarization-sensitive coupling	polarization loss factor, PDL and guarded detector power	Decide whether source, optics and detector polarization state preserve usable signal.
Laser source drive	threshold current, slope efficiency, optical path efficiency and thermal derating	Decide whether source output still meets the delivered-power requirement.
Detector chain	responsivity, shot noise, NEP, saturation and bandwidth	Verify that signal, noise and electronics range close together.
Environmental release	thermal wavelength drift, passband margin and calibration evidence	Decide whether the optical system remains valid in service.

Validation Package Checklist

• wavelength, optical plane, source drive state and modulation condition are documented;
• source output, path transmission, beam geometry, aperture and detector state are separated;
• connector inspection, alignment, focus, polarization and contamination evidence are recorded;
• detector responsivity, noise, saturation, bandwidth, dark condition and electronics gain are bounded;
• thermal drift, mechanical tolerance, safety limit and environmental condition are included where relevant;
• calibration records and uncertainty guards map to the same plane as the acceptance limit;
• final release decision states accept, clean, realign, derate, recalibrate, redesign or hold.

Exercise 1: Photon Energy and Photon Flux

An optical source emits:

at wavelength:

Estimate photon energy in joules and electronvolts, and estimate photon flux.

Solution

Convert wavelength:

Photon energy is:

Convert to electronvolts:

Photon flux is:

Engineering Comment

The photon flux is large even for a milliwatt-scale source. That does not automatically mean the detector signal is adequate. Coupling loss, detector quantum efficiency, bandwidth, background light and saturation must still be checked at the measurement boundary.

Plausibility Check

An 850\ \text{nm} photon has energy near the near-infrared scale of about 1.5\ \text{eV}. Dividing milliwatts by a 10^{-19}\ \text{J} photon energy should give a photon flux on the order of 10^{15} to 10^{16}\ \text{photons/s}.

Exercise 2: Optical Loss in dB and Output Power

A source launches:

through an optical path with the following losses:

Loss item	Loss
connector pair	2.0\ \text{dB}
filter	0.7\ \text{dB}
route and bend allowance	1.2\ \text{dB}

Estimate output power in mW and dBm.

Solution

Total loss is:

Linear output power is:

Input power in dBm:

Output power in dBm:

Engineering Comment

The dB calculation is clean only because all losses are at the same optical boundary and wavelength. In a real link, connector loss, bend loss, filter loss, polarization effects and source spectrum should be measured or bounded separately. A dirty connector can easily consume more margin than a formula budget assumes.

Plausibility Check

A 3.9\ \text{dB} loss is a little less than a 4\ \text{dB} loss, so output power should be a little above 40\% of input power. The computed 0.611\ \text{mW} from 1.5\ \text{mW} matches that scale.

Exercise 3: Gaussian Beam Rayleigh Range and Spot Growth

A laser beam has wavelength:

and waist radius:

Estimate Rayleigh range and beam radius at:

using:

Solution

Convert units:

Rayleigh range:

Beam radius at 2.0\ \text{m}:

So the beam radius is about:

Engineering Comment

The radius more than doubles by 2\ \text{m} from the waist. That matters for aperture clipping, detector spot size, irradiance, safety enclosure design and alignment tolerance. A beam diameter stated at the laser package is not enough for a downstream optical design.

Plausibility Check

The propagation distance is about twice the Rayleigh range, so the radius should be w_0\sqrt{1+2^2}, a little more than twice the waist. The 1.0\ \text{mm} result is consistent.

Exercise 4: Diffraction-Limited Spot Size

An imaging lens operates at:

with numerical aperture:

Estimate Airy-disk radius using:

and estimate Airy-disk diameter.

Solution

Substitute:

The Airy-disk radius is:

Diameter is:

Engineering Comment

This is an optical diffraction screen, not a complete image-quality statement. Pixel size, aberrations, focus error, motion, vibration, illumination coherence, sample contrast and detector noise can all dominate before the diffraction limit is reached.

Plausibility Check

Visible light divided by an NA of 0.25 should produce a diffraction radius on the micrometer scale. A 2.60\ \mu\text{m} Airy diameter is therefore plausible.

Exercise 5: Fiber Numerical Aperture and Acceptance Angle

A step-index fiber has core index:

and cladding index:

Estimate numerical aperture and air-side acceptance half-angle.

Solution

For a step-index fiber in air:

Acceptance half-angle in air is:

Engineering Comment

The fiber accepts only a limited angular cone. A source can have enough optical power and still couple poorly if its divergence, focus, lateral offset or connector geometry does not match the fiber. Launch-condition control is part of optical validation.

Plausibility Check

The core and cladding indices differ by only 0.006, so the numerical aperture should be modest. An air-side half-angle below 10^\circ is consistent with that small index contrast.

Exercise 6: Gaussian Lateral Coupling Loss

A single-mode coupling estimate uses the lateral-offset approximation:

where w is the mode radius. Assume:

and lateral offset:

Estimate coupling efficiency and loss in dB.

Solution

Efficiency:

Loss:

Engineering Comment

A 2\ \mu\text{m} lateral offset produces more than 1\ \text{dB} loss in this simplified single-mode estimate. That is large enough to consume meaningful optical margin. The real package also needs angular alignment, longitudinal focus, mode-field mismatch, connector cleanliness and thermal drift checks.

Plausibility Check

The offset is about 0.38 of the mode radius. For a Gaussian overlap, that is large enough to create a visible dB loss but not so large that coupling collapses, matching the 0.744 efficiency.

Exercise 7: Photodiode Current and Shot-Noise-Limited SNR

A photodiode receives:

at a wavelength where responsivity is:

Dark current is:

Electrical noise bandwidth is:

Estimate photocurrent and shot-noise-limited current SNR.

Solution

Photocurrent:

Shot-noise current:

Current SNR:

In decibels for an amplitude ratio:

Engineering Comment

The shot-noise-limited SNR is very high, which means another limit may dominate in practice: transimpedance amplifier noise, resistor thermal noise, ambient light, ADC range, source intensity noise, EMI, dark-current drift or saturation. A high theoretical SNR is not a substitute for installed noise measurement.

Plausibility Check

A 50\ \mu\text{W} optical signal at 0.55\ \text{A/W} gives tens of microamps. Shot noise over only 100\ \text{kHz} should be below a nanoamp, so a very high shot-noise-limited SNR is expected.

Exercise 8: Receiver Bandwidth, Rise Time, Sampling and Thermal Drift

An optical receiver has electrical bandwidth:

Use the first-pass rise-time relation:

The digitizer samples at:

A laser wavelength drifts with temperature at:

Temperature changes by:

Estimate receiver rise time, sample spacing and wavelength drift.

Solution

Rise time:

Sample spacing:

Wavelength drift:

Engineering Comment

The sampling interval is shorter than the receiver rise time, so the digitizer can represent the receiver-limited edge in this simple screen. The 1.44\ \text{nm} wavelength drift can be harmless for a broad detector and serious for a narrow filter, wavelength-division channel, interferometric sensor or calibrated spectroscopy system. Thermal drift must be compared with the actual optical passband and calibration model.

Plausibility Check

A 5\ \text{MHz} bandwidth corresponds to a rise time of tens of nanoseconds, and a 25\ \text{MS/s} digitizer samples every 40\ \text{ns}. The timing values are therefore in the same range and need to be checked together.

Exercise 9: Photodiode Saturation and Required Attenuation

A photodiode monitor receives:

Responsivity is:

The transimpedance gain is:

The amplifier output limit is:

Estimate photocurrent, amplifier output voltage, maximum optical power before saturation and attenuation required to avoid saturation.

Solution

Photocurrent:

Amplifier output:

The amplifier saturates because:

Maximum photocurrent:

Maximum optical power:

Required attenuation:

Engineering Comment

More optical power is not always better. A saturated photodiode channel can report a stable but wrong value, distort modulation, hide source drift or invalidate calibration. Receiver range should be checked before interpreting high SNR as good evidence.

Plausibility Check

The received optical power is about 47\% above the maximum unsaturated value. An attenuation requirement below 2\ \text{dB} is therefore plausible.

Exercise 10: Guarded Optical Power Uncertainty

A link margin measurement reports:

Independent one-sigma uncertainty components are estimated as:

Source	Standard uncertainty
power-meter calibration	0.20\ \text{dB}
connector repeatability	0.10\ \text{dB}
polarization dependence	0.15\ \text{dB}
source stability	0.08\ \text{dB}

Compute combined standard uncertainty, expanded uncertainty with k=2, and guarded margin.

Solution

Combined standard uncertainty:

Expanded uncertainty:

Guarded margin:

The link has positive guarded margin, but it is thin.

Engineering Comment

An optical margin below 1\ \text{dB} should be treated carefully. Connector cleaning, reference setting, polarization state, meter calibration and source stability can decide whether the result is real. A small positive measured margin is not strong release evidence unless uncertainty is included.

Plausibility Check

The largest uncertainty term is 0.20\ \text{dB}, and several smaller terms contribute. A combined one-sigma uncertainty near 0.28\ \text{dB} and a two-sigma value near 0.56\ \text{dB} are reasonable.

Exercise 11: Alignment Tolerance for Coupling Loss

Use the Gaussian lateral coupling relation:

with mode radius:

The package budget allows no more than:

of lateral-offset loss. The assembly process has lateral placement tolerance:

Find maximum allowed lateral offset and decide whether the process is adequate.

Solution

Minimum allowed efficiency:

Solve for offset:

Assembly tolerance excess:

The assembly tolerance is not adequate for a 0.50\ \text{dB} lateral-loss budget.

Engineering Comment

Micrometer-scale tolerances can dominate optical coupling. A drawing tolerance should be checked against the optical mode field, not only against mechanical fit. If the process cannot hold the tolerance, the design needs active alignment, a larger mode field, relaxed loss budget or stronger production test.

Plausibility Check

Exercise 6 showed that a 2.0\ \mu\text{m} offset caused 1.28\ \text{dB} loss. A 0.50\ \text{dB} limit should therefore require an offset well below 2.0\ \mu\text{m}, matching the 1.25\ \mu\text{m} result.

Exercise 12: Laser and Filter Thermal Drift Margin

A narrow optical receiver uses a laser and filter centered at:

Laser wavelength drift is:

Filter center drift is:

The temperature change is:

The filter half-width is:

The release rule reserves:

for calibration and aging. Compute relative drift and passband margin.

Solution

Relative drift rate:

Relative wavelength shift:

Usable half-width after guard:

Margin:

The laser/filter combination fails the guarded drift screen.

Engineering Comment

Absolute wavelength drift is less important than relative drift between source and passband. A laser and filter can both be temperature dependent but drift at different rates. Narrowband optical systems need thermal control, matched components, wider passbands or calibration compensation.

Plausibility Check

A relative drift of 0.075\ \text{nm/K} over 25\ \text{K} is nearly 1.9\ \text{nm}, larger than the guarded 1.2\ \text{nm} half-width. The negative margin is therefore expected.

Exercise 13: Aperture Clipping and Irradiance Margin

A measurement head sends a centered Gaussian beam toward a mechanical aperture. At the aperture plane, the optical power is:

and the 1/e^2 beam radius is:

The clear aperture radius is:

Use the centered Gaussian transmitted-power approximation:

After the aperture, the beam reaches a sample spot with radius:

The sample-facing release limit for average irradiance is:

Calculate aperture transmission, clipping loss, transmitted power, average irradiance, irradiance margin, and the margin if misfocus reduces the spot radius to 1.0\ \text{mm}.

Solution

Aperture transmission:

Clipping loss:

Transmitted power:

Convert the nominal sample radius:

Spot area:

Average irradiance:

Nominal irradiance margin:

For a misfocused spot radius of 1.0\ \text{mm}=0.10\ \text{cm}:

Misfocused irradiance margin:

The nominal condition passes the simplified irradiance screen, but the misfocused condition fails.

Engineering Comment

Aperture clipping and irradiance must be checked at the same physical boundary as the acceptance rule. The aperture reduces transmitted power, but focusing that transmitted power into a smaller spot can still exceed a sample, detector, coating or process limit. This exercise uses a simplified device/sample limit, not an eye-safety classification; real safety release needs the applicable standard, exposure geometry, time basis and administrative controls.

Plausibility Check

The aperture radius is slightly smaller than the beam radius, so a transmission near 75\% and a loss a little above 1\ \text{dB} are plausible. Reducing spot radius from 1.5 to 1.0\ \text{mm} cuts area by more than half, so irradiance should more than double and can flip the margin negative.

Exercise 14: Laser Diode Threshold and Slope-Efficiency Margin

A laser diode is driven at:

At the reference thermal state, its threshold current is:

and its slope efficiency is:

The optical path from diode facet to the declared output plane has efficiency:

The delivered-power requirement is:

For a hot operating condition, use temperature rise:

threshold-current coefficient:

and slope-efficiency derating:

Calculate nominal delivered power, hot delivered power and release decision.

Solution

Nominal facet power:

Nominal delivered power:

Nominal margin:

Hot threshold current:

Hot slope efficiency:

Hot facet power:

Hot delivered power:

Hot margin:

The nominal condition passes narrowly, but the hot condition fails the delivered-power requirement.

Engineering Comment

Laser-diode source power is not just a drive-current setting. Threshold current, slope efficiency, thermal path, aging, optical feedback, driver compliance and monitor-photodiode calibration can all move delivered power. A release based only on a room-temperature L-I curve can be wrong if the output plane, temperature state and optical path losses are not included.

Plausibility Check

At the reference state, the current above threshold is 58\ \text{mA}, so a slope near 0.4\ \text{mW/mA} gives facet power near 22\ \text{mW}. A 22\ \text{K} temperature rise both raises threshold and lowers slope efficiency, reducing delivered power from 15.4 to 13.1\ \text{mW}. That is a plausible failure for a source with only 0.4\ \text{mW} nominal margin.

Exercise 15: Photodiode NEP and Minimum Detectable Power

A photodiode receiver has responsivity:

The measured input-referred current-noise density, including amplifier and detector contributions, is:

The selected measurement bandwidth is:

The detection rule requires:

for the optical signal at the detector plane. The expected low-light signal is:

Calculate RMS current noise, noise-equivalent power, minimum detectable optical signal and margin. Then repeat the result if bandwidth is reduced to 5\ \text{kHz}.

Solution

Convert noise density:

RMS current noise at 20\ \text{kHz}:

Noise-equivalent power:

Minimum optical signal for SNR=10:

Margin at the detector plane:

The 20\ \text{kHz} case fails the minimum-detectable-signal rule.

For B=5\ \text{kHz}:

Updated margin:

The lower-bandwidth case passes the signal-detection rule, provided the reduced bandwidth is compatible with the measurement timing.

Engineering Comment

NEP is only meaningful at a stated detector boundary and bandwidth. Reducing bandwidth can recover low-light SNR, but it may also slow the measurement, smear pulses or hide modulation. A defensible detector release must state bandwidth, responsivity calibration, dark current, amplifier noise, ambient light, filtering, sampling, timing requirement and whether the expected signal is measured at the actual detector plane.

Plausibility Check

Noise integrated over bandwidth scales with \sqrt{B}. Reducing bandwidth from 20\ \text{kHz} to 5\ \text{kHz} cuts RMS noise by a factor of two, so the required optical power for the same SNR also halves from 7.20 to 3.60\ \text{nW}. That makes the 6.0\ \text{nW} signal fail at high bandwidth but pass at lower bandwidth.

Exercise 16: Pulsed Laser Fluence and Peak Irradiance Margin

A pulsed laser inspection head delivers average optical power at the sample plane:

The pulse repetition rate is:

and the optical pulse width is:

The focused spot radius on a coating is:

The coating release rules require:

for single-pulse fluence and:

for peak irradiance. Calculate pulse energy, spot area, fluence, peak power, peak irradiance and release decision. Then check whether increasing the spot radius to:

passes both limits with the same average power and pulse width.

Solution

Convert average power:

Pulse energy is:

So:

Convert spot radius:

Spot area is:

Single-pulse fluence is:

Fluence margin is:

Peak power is:

Peak irradiance is:

or:

Peak-irradiance margin is:

The fluence rule passes, but the peak-irradiance rule fails. The optical head should not be released in this focused condition.

For the larger spot:

Updated fluence:

Updated peak irradiance:

or:

Updated peak-irradiance margin:

The larger spot passes both simplified coating limits, provided the relaxed focus still meets the inspection-resolution and signal requirements.

Engineering Comment

Average optical power does not define pulsed optical stress. A modest average power can create high pulse energy, high peak power or high peak irradiance when repetition rate, pulse width and spot radius are considered. Release evidence should state the pulse waveform, measurement bandwidth, spot-size method, coating or sample limit, focusing tolerance, scan dwell time and whether the limit is for single-pulse damage, cumulative exposure, thermal load or a safety classification.

Plausibility Check

At 20\ \text{kHz}, 180\ \text{mW} average power becomes 9.0\ \mu\text{J} per pulse. Dividing that energy by an 8 ns pulse gives more than a kilowatt of peak power, which explains why peak irradiance can fail even though average power sounds moderate. Increasing radius from 60 to 70\ \mu\text{m} increases area by about 36\%, enough to bring peak irradiance below the 8.0\ \text{MW/cm}^2 screen.

Exercise 17: Quantum Efficiency, Responsivity and Detector Output

A silicon photodiode is used at:

The supplier reports quantum efficiency:

at that wavelength. The expected optical power at the detector plane is:

The transimpedance amplifier gain is:

The ADC input is valid up to:

and the release rule requires detector output of at least:

for the low-signal test condition. Use:

with:

Estimate responsivity, photocurrent and amplifier output. Then check a degraded detector lot with \eta_q=0.45.

Solution

Convert wavelength:

Responsivity is:

Photocurrent at 80\ \mu\text{W}:

Amplifier output:

ADC headroom:

Low-signal release margin:

The nominal detector passes both the ADC headroom and minimum-output screens.

For the degraded detector lot, responsivity scales with quantum efficiency:

Photocurrent becomes:

Amplifier output becomes:

Low-signal margin:

The degraded lot should not be released for this low-signal condition unless optical power, amplifier gain, detector selection, bandwidth/noise requirement or acceptance criteria are changed and validated.

Engineering Comment

Responsivity is not a universal photodiode constant. It depends on wavelength, quantum efficiency, temperature, bias condition and detector construction. A detector chain that passes with one lot can fail when quantum efficiency changes, even if optical power and electronics are unchanged. Release evidence should preserve wavelength, source spectrum, detector lot, responsivity calibration, bias voltage, temperature, dark current, TIA gain, ADC range and the exact optical plane where power was measured.

Plausibility Check

At 940\ \text{nm}, a perfect photodiode would have responsivity around 0.76\ \text{A/W}, so 0.515\ \text{A/W} at 68\% quantum efficiency is plausible. An 80\ \mu\text{W} signal then produces tens of microamps, enough for a volt-scale TIA output with 40\ \text{kV/A}. Reducing quantum efficiency to 45\% cuts the output below the 1.20 V minimum, so the failed release decision follows directly from detector physics.

Exercise 18: Polarization Mismatch and Detector-Power Release Gate

A polarization-sensitive optical receiver expects power at the detector package before the analyzer of:

The source polarization and detector analyzer are misaligned by:

Use the linear-polarization mismatch screen:

The optical package also has polarization-dependent loss:

The detector release rule requires guarded optical power at the photodiode to satisfy:

with:

and:

Calculate the delivered detector power and release decision. Then check a corrected alignment with:

Solution

Polarization loss factor at the original angle:

Equivalent polarization loss:

Convert the package polarization-dependent loss to a linear factor:

Detector power after polarization mismatch and PDL:

Guarded detector power:

The original alignment fails because:

Required polarization loss factor for release:

Maximum allowed misalignment:

The original 32^\circ alignment is outside the allowable range.

For the corrected alignment:

New detector power:

New guarded detector power:

Release margin:

The corrected alignment passes the detector-power release gate.

Engineering Comment

Polarization mismatch is an optical boundary condition, not only a radio-link issue. A source can have adequate total power and a detector can have adequate responsivity, while the polarization-sensitive part of the receiver sees too little usable power. Polarization-dependent loss, rotated fiber pigtails, stress birefringence, polarizer mounting error, connector keying, package orientation and temperature can all move the actual state away from the bench setup.

Release evidence should state the optical power plane, analyzer orientation, source polarization state, PDL measurement, alignment tolerance, connector/keying convention, environmental condition and uncertainty guard. If the receiver is not intentionally polarization-sensitive, the validation should still prove that polarization drift does not create hidden margin loss.

Plausibility Check

A 32^\circ linear polarization mismatch keeps about 72\% of the power before other losses, which is a little over 1.4\ \text{dB}. Adding 0.55\ \text{dB} of PDL leaves roughly 82\ \mu\text{W} from a 130\ \mu\text{W} input, so failing an 85\ \mu\text{W} requirement after a 6\ \mu\text{W} guard is plausible. Tightening the angle to 12^\circ leaves about 96\% of the polarization component and restores comfortable margin.

Review Table

Check	Result	Interpretation
Photon energy at 850\ \text{nm}	1.46\ \text{eV}	Confirms detector material and wavelength compatibility.
Photon flux at 2.0\ \text{mW}	8.56\times10^{15}\ \text{photons/s}	Flux is large, but coupling and detector range still matter.
Optical path output	0.611\ \text{mW}	A few dB of loss can halve received power.
Gaussian beam radius at 2\ \text{m}	about 1.0\ \text{mm}	Downstream aperture and safety calculations need propagated beam size.
Diffraction diameter	2.60\ \mu\text{m}	Real image quality may be worse due to aberration and detector limits.
Fiber acceptance half-angle	7.62^\circ	Launch alignment controls coupling.
Lateral coupling loss	1.28\ \text{dB}	Micrometer offsets can consume optical margin.
Polarization-mismatch detector power	82.4\ \mu\text{W}	A rotated polarization state plus PDL can fail a detector-power release gate.
Photodiode current	27.5\ \mu\text{A}	Check amplifier range and noise, not only detector responsivity.
Receiver rise time	70\ \text{ns}	Must be compared with pulse width and sampling.
Thermal wavelength drift	1.44\ \text{nm}	May exceed narrow optical filter tolerance.
Saturation threshold	0.818\ \text{mW}	High optical power can overload the receiver chain.
Guarded power margin	0.136\ \text{dB}	Positive but too thin for casual release.
Coupling offset limit	1.25\ \mu\text{m}	Mechanical tolerance must be tighter than the optical budget.
Laser/filter drift margin	-0.675\ \text{nm}	Relative thermal drift fails the guarded passband screen.
Aperture transmission	0.751	Mechanical clipping removes about one quarter of the power.
Nominal irradiance margin	18.1\ \text{mW/cm}^2	The nominal spot passes the simplified sample limit.
Misfocused irradiance margin	-21.7\ \text{mW/cm}^2	Focus error can turn a passing optical power case into a failed irradiance case.
Hot laser delivered power	13.1\ \text{mW}	Thermal threshold and slope changes can erase nominal source margin.
Detector minimum signal at 20\ \text{kHz}	7.20\ \text{nW}	The expected 6.0\ \text{nW} signal fails at the wider bandwidth.
Detector minimum signal at 5\ \text{kHz}	3.60\ \text{nW}	Lower bandwidth restores SNR if timing still works.
Focused pulse peak irradiance	9.95\ \text{MW/cm}^2	Average power looks moderate, but pulse width and spot size fail the peak coating limit.
Expanded-spot peak irradiance	7.31\ \text{MW/cm}^2	A slightly larger spot passes the peak limit if inspection resolution remains acceptable.
Photodiode responsivity at 940\ \text{nm}	0.515\ \text{A/W}	Quantum efficiency converts photon absorption into detector current.
Degraded detector output	1.09\ \text{V}	Lower quantum efficiency fails the minimum-output release rule.

Common Mistakes

• mixing dB loss, dBm power and linear watts without a clear reference;
• using source output power instead of power at the detector or sample plane;
• ignoring connector contamination, bend loss and alignment after a bench test;
• claiming diffraction-limited performance without aberration, pixel and focus evidence;
• checking beam power without checking aperture clipping and spot irradiance at the real boundary;
• accepting pulsed optical power from average watts without pulse energy, pulse width, fluence and peak irradiance;
• using laser drive current as proof of delivered optical power without threshold, slope and temperature evidence;
• using fiber numerical aperture without checking mode-field diameter and lateral offset;
• assuming total optical power is usable when the detector, analyzer or waveguide is polarization-sensitive;
• treating photodiode responsivity as fixed while wavelength, quantum efficiency, lot and temperature change it;
• calculating photodiode shot noise while ignoring amplifier noise and saturation;
• quoting NEP without stating detector boundary, bandwidth and required SNR;
• accepting measured optical margin without uncertainty guard band;
• choosing sampling rate without checking receiver bandwidth and trigger jitter;
• treating laser wavelength as fixed when temperature, current and aging can shift it.

Review Checklist

Before accepting an optical calculation, confirm:

1. wavelength, optical boundary and power units are explicit;
2. every dB term is a ratio and every dBm value has a power reference;
3. beam size is checked at the actual aperture, detector or safety boundary;
4. coupling assumptions include lateral, angular, focus, mode-field and cleanliness effects;
5. polarization state, analyzer orientation and PDL are checked when the optical path or detector is polarization-sensitive;
6. detector response includes wavelength-dependent responsivity, quantum efficiency, dark current, bandwidth, saturation and noise;
7. laser-source release includes threshold current, slope efficiency, optical path efficiency, thermal state and driver limits;
8. digitizer settings match receiver bandwidth, pulse shape and timing tolerance;
9. detector-floor calculations state NEP, bandwidth, required SNR and actual signal boundary;
10. measured optical margin includes calibration, connector, polarization and source-stability uncertainty;
11. alignment tolerance is checked against mode-field and coupling-loss requirements;
12. thermal drift is compared with the actual passband, calibration equation and alignment budget;
13. aperture clipping and irradiance are checked at the sample, detector, coating or safety boundary that owns the limit;
14. pulsed-source release separates average power, pulse energy, pulse width, fluence, peak irradiance and scan dwell time;
15. validation evidence includes calibrated power, beam profile, connector inspection, noise measurement and environmental repeatability where relevant.