Photonics and Optical Engineering Formula Sheet

This formula sheet collects first-pass equations used in photonics and optical engineering. Use it to screen optical power, photon flux, beam geometry, diffraction limits, fiber and waveguide coupling, photodiode response, noise, bandwidth, thermal drift, safety margins, and validation evidence.

The equations are not a replacement for optical design software, source and detector datasheets, safety standards, or calibrated measurement. They are a way to make assumptions visible before a design, calibration, commissioning, or troubleshooting review.

Constants and Symbols

Symbol	Meaning	Typical unit
c	speed of light in vacuum	\text{m/s}
h	Planck constant	\text{J s}
q	elementary charge	\text{C}
\lambda	wavelength	\text{m}
f	optical frequency	\text{Hz}
P	optical power	\text{W}
\Phi	photon flux	photons/s
R_\lambda	photodiode responsivity	\text{A/W}
\eta	quantum efficiency or coupling efficiency	dimensionless
w_0	Gaussian beam waist radius	\text{m}
z_R	Rayleigh range	\text{m}
NA	numerical aperture	dimensionless
B	electrical bandwidth	\text{Hz}
u_c	combined standard uncertainty	same as measurand

Use wavelength consistently. Optical engineering often mixes nm, um, mm, GHz, dBm, and linear watts in one calculation. Convert before combining terms.

Wavelength, Frequency, and Photon Energy

Optical frequency:

Photon energy:

Photon flux from optical power:

Worked Check

For an 850\ \text{nm} source with:

photon energy is:

Photon flux:

This is a photon arrival rate, not a detected carrier rate. Detector quantum efficiency, optical loss, aperture fill, reflection, and saturation still have to be included.

Optical Power Units

Optical power in dBm:

Power from dBm:

Power ratio in dB:

For losses in series:

Worked Check

If:

and total optical loss is:

then:

Linear output power:

A 3\ \text{dB} loss is about half the optical power. It is not a small correction in an optical power budget.

Optical Attenuation and Margin

Fiber, waveguide, window, filter, connector, splice, and coating losses can be treated as power losses for a first-pass budget.

Fiber attenuation:

where \alpha is in \text{dB/km} and L is in \text{km}.

Received optical power:

Margin above sensitivity:

Worked Check

For:

with:

received power is:

If receiver sensitivity is:

margin is:

The number should be guarded for aging, connector contamination, temperature, repair splices, measurement uncertainty, and future patching.

Laser Diode Output and Thermal Drift

Above threshold, a simplified laser diode output model is:

where \eta_s is slope efficiency and I_{th} is threshold current.

Wavelength thermal drift:

Electrical heat that must be removed:

Worked Check

If:

then:

If wavelength drift is:

and temperature rises:

then:

This drift can matter for filters, wavelength-division systems, spectroscopy, detector responsivity, and safety classification.

Irradiance, Exposure, and Absorbed Power

Irradiance:

Radiant exposure:

Absorbed optical power:

where A_{abs} is absorbed fraction, not area.

Temperature rise through a thermal resistance:

Worked Check

If:

is focused into:

then:

For a 2.0\ \text{s} exposure:

Whether this is acceptable depends on material, wavelength, absorption, cooling, exposure limit, and failure consequence.

Gaussian Beam Geometry

Rayleigh range:

Beam radius:

Far-field half-angle divergence:

Worked Check

For:

Rayleigh range is:

Divergence:

A small divergence number assumes a beam close to Gaussian and a well-defined waist. Real diode beams can be elliptical, astigmatic, multimode, clipped, or distorted by packaging.

Diffraction-Limited Spot Size

For a circular diffraction-limited aperture, the Airy disk first-zero radius is approximately:

For an optical system with f-number N_f:

Worked Check

For:

the first-zero radius is:

This is an optical limit, not a full imaging-system specification. Pixel size, aberrations, vibration, focus error, scattering, contrast, exposure, and reconstruction algorithms can dominate.

Numerical Aperture and Fiber V-Number

For a step-index fiber or waveguide:

Acceptance half-angle in air:

Fiber normalized frequency:

where a is core radius. A step-index fiber is single-mode when:

Worked Check

For:

the V-number is:

The fiber is below the step-index single-mode cutoff. The real cable still needs connector, bend, dispersion, wavelength, polarization, and launch-condition checks.

Coupling Loss from Lateral Misalignment

For a simple Gaussian-to-Gaussian coupling screen with equal mode radii:

Coupling loss:

Worked Check

If:

then:

Loss:

This model is a screening approximation. Angular error, defocus, mode mismatch, polarization, Fresnel reflection, contamination, and clipping may add more loss.

Photodiode Responsivity and Quantum Efficiency

Photodiode current:

Responsivity from quantum efficiency:

Quantum efficiency from responsivity:

Worked Check

For:

responsivity is:

If:

photocurrent is:

With a transimpedance gain:

output voltage is:

The calculation must be checked for detector area, linear range, dark current, capacitance, amplifier swing, wavelength mismatch, and calibration boundary.

Shot Noise and Signal-to-Noise Ratio

Shot-noise current:

where I_d is dark current and B is electrical noise bandwidth.

If amplifier current noise i_{amp} and other independent current-noise terms are available:

Current-domain SNR:

Worked Check

Use:

Shot noise:

Shot-noise-limited SNR:

In practice, amplifier noise, ambient light, source relative intensity noise, digitization, electromagnetic pickup, and drift may dominate before shot noise does.

Transimpedance Bandwidth

A first-pass feedback-pole screen for a photodiode transimpedance amplifier is:

where:

Rise time for a single-pole response:

Worked Check

For:

bandwidth is:

Rise time:

This screen does not prove stability. Transimpedance amplifiers require op-amp gain-bandwidth, input capacitance, feedback capacitance, phase margin, layout leakage, overload recovery, and noise analysis.

Sampling, Quantization, and Timing Jitter

Sampling theorem screen:

Quantization step:

RMS quantization noise for an ideal converter:

Jitter-limited SNR for a sinusoidal signal:

Worked Check

For:

jitter term:

Therefore:

The jitter is unlikely to be the first limit at 100\ \text{kHz}, but the same timing uncertainty can matter at much higher modulation frequencies.

Optical Modulation Bandwidth

Small-signal bandwidth is often screened with a single-pole relationship:

Magnitude loss in dB:

For a first-order system, the -3\ \text{dB} point occurs at:

This model is useful for early review. Real laser drivers, LEDs, photodiodes, transimpedance amplifiers, cables, packages, and digitizers may show resonances, peaking, nonlinear distortion, bandwidth compression, or pattern-dependent response.

Thermal Optical Drift

Optical power temperature coefficient:

Relative power drift:

Focus or position drift from thermal expansion:

Worked Check

If an optical spacer is:

with coefficient:

and temperature changes:

then:

That shift may be negligible in a large beam and unacceptable in a fiber, microscope focus, slit, or imaging sensor.

Guarded Optical Margin

For a nominal optical margin:

and combined uncertainty:

guarded margin is:

Release condition:

Worked Check

Suppose an optical receiver measures:

with sensitivity limit:

Nominal margin:

If:

then:

If the required guarded margin is 5.0\ \text{dB}, the link misses release by:

This is a small numerical miss, but it is a real release finding if the uncertainty model and acceptance requirement are valid.

Validation Evidence to Preserve

Photonics calculations should leave evidence that another engineer can reproduce:

1. wavelength, spectral width, power boundary, aperture, polarization, and optical path;
2. source drive current, thermal state, modulation condition, and safety limits;
3. lens, mirror, filter, fiber, waveguide, connector, coating, and alignment state;
4. detector responsivity, dark current, bias, bandwidth, gain, saturation and calibration;
5. noise bandwidth, sampling rate, timing, averaging, and signal-processing basis;
6. optical loss budget, coupling tolerance, contamination state, and service margin;
7. temperature, vibration, humidity, background light, cleaning and aging tests;
8. uncertainty budget, guard band, release decision and recalibration triggers.

Common Mistakes

Common mistakes include:

1. using source output power instead of power at the target or detector;
2. mixing dBm, dB, watts, photon flux, irradiance and exposure without a boundary;
3. selecting a photodiode from responsivity alone while ignoring capacitance and bandwidth;
4. assuming a Gaussian beam model for a clipped or multimode laser diode;
5. treating diffraction-limited resolution as the full imaging-system resolution;
6. ignoring connector contamination, back-reflection, and bend loss in optical fibers;
7. validating at one wavelength while using responsivity at another;
8. omitting thermal drift from wavelength, focus, source power, and detector dark current;
9. reporting optical SNR without bandwidth, averaging and background-light condition;
10. accepting a link or measurement with no guarded uncertainty margin.

The strongest optical calculation is tied to the physical boundary. It states where the power is measured, what wavelength applies, which optical path is included, how noise bandwidth is defined, and which validation evidence proves that the margin survives alignment, temperature, contamination, aging, and use.