Interferometry — Abridged Guide
Quick-reference guide to interferometric measurement. For full derivations and worked examples, see the Comprehensive Guide.
Comprehensive Interferometry Guide →
1.Introduction to Interferometry
Interferometry measures physical quantities by superposing two coherent beams and analyzing the resulting intensity pattern. The fringe pattern encodes optical path differences with sub-wavelength sensitivity — down to sub-nanometer precision with phase-shifting techniques.
For quick estimates, remember that one fringe = λ/2 of surface height change in reflection. At HeNe (632.8 nm), one fringe ≈ 316 nm.
2.Fundamentals of Interference
Two-beam interference
Constructive interference (bright fringes) occurs when OPD = mλ; destructive (dark fringes) when OPD = (m + ½)λ. Fringe visibility V = (I_max − I_min)/(I_max + I_min) must exceed ~0.2 for reliable detection.
The coherence length sets the maximum usable OPD. For a HeNe laser, l_c ≈ 20 cm. For white light, l_c ≈ 1 μm — which is exploited for absolute position sensing in WLI and OCT.
Coherence length
3.Interferometer Architectures
Six major architectures cover most applications. Michelson and Fizeau dominate surface testing; Mach–Zehnder for refractive index measurement; Fabry–Pérot for spectroscopy; Sagnac for rotation sensing.
| Architecture | Passes | Best For |
|---|---|---|
| Michelson | Double | Displacement, FTIR, surface testing |
| Mach–Zehnder | Single | Refractive index, fiber sensors |
| Fabry–Pérot | Multiple | Spectroscopy, wavelength filtering |
| Fizeau | Common path | Surface figure, optical shop testing |
| Sagnac | Common path | Rotation sensing (FOGs) |
| Twyman–Green | Double | Optical component testing |
Fizeau interferometers are preferred in optical shops because their common-path design inherently rejects vibration and source noise — both beams share the same optical path until the last surface.
Fabry–Pérot transmission
Free spectral range
4.Fringe Analysis and Interpretation
Fringe spacing from tilt
Fringes are contour lines of OPD. Straight fringes indicate tilt; rings indicate curvature mismatch; deviations from the ideal pattern reveal surface figure error. One fringe of deviation = λ/2 surface error in reflection.
In a Fizeau test, count fringes across the aperture for total tilt. Then assess fringe straightness — deviations of 0.1 fringe correspond to ~32 nm of surface figure error at HeNe.
5.Phase Measurement Techniques
Hariharan 5-step PSI
Phase-shifting interferometry captures multiple intensity frames at known phase offsets (typically 90° steps) to extract the phase at every pixel. This transforms fringe-counting (λ/20 resolution) into quantitative surface mapping (λ/1000 resolution).
Use the Hariharan 5-step algorithm as the default for PSI — it is insensitive to PZT step-size calibration errors to second order. For vibrating environments, switch to simultaneous PSI (pixelated-mask camera) or single-shot spatial-carrier methods.
| Algorithm | Frames | Step Size | Calibration Robustness |
|---|---|---|---|
| 3-step | 3 | 90° | Low — sensitive to step error |
| 4-step | 4 | 90° | Medium — cancels first-order error |
| Hariharan 5-step | 5 | 90° | High — insensitive to linear + quadratic error |
| Carré | 4 | Unknown (self-calibrating) | High — no step calibration needed |
6.Coherence and Source Requirements
Source coherence length must exceed the maximum OPD in the interferometer. HeNe lasers (l_c > 20 cm) suit most equal-path instruments. Frequency-stabilized HeNe (l_c > 300 m) extends to long-path applications. Broadband sources (l_c ~ 1–30 μm) enable absolute distance sensing via coherence gating.
For Fizeau surface testing, a standard HeNe is almost always sufficient — the OPD is set by the cavity gap (typically < 100 mm). For scanning Michelson or Mach–Zehnder with large OPD, upgrade to a frequency-stabilized source.
| Source | Coherence Length | Best For |
|---|---|---|
| HeNe (multimode) | ~20 cm | Fizeau, short-path Michelson |
| HeNe (stabilized) | > 300 m | Long-path, large-aperture testing |
| Diode laser (external cavity) | 1–10 m | Compact systems, wavelength tuning |
| SLD (broadband) | 10–30 μm | OCT, white-light interferometry |
| Halogen lamp | ~1 μm | VSI, thin-film metrology |
7.Error Sources and Limitations
OPD error from air temperature
dn/dT ≈ −0.93 × 10⁻⁶ /°C for air at 632.8 nm.
Environmental errors (vibration, air turbulence, thermal drift) dominate most interferometric measurements. A 1°C temperature fluctuation over a 100 mm air path introduces ~93 nm of OPD error — nearly λ/7.
Three practical rules: (1) isolate vibration (pneumatic table), (2) enclose the beam path (reduce turbulence), (3) let the system thermally equilibrate before measuring. For λ/100 work, all three are mandatory.
8.Components and Hardware
Critical components include the beam splitter (50:50 for maximum visibility), reference optic (surface figure sets accuracy floor), PZT actuator (sub-nm resolution, < 2% hysteresis), and camera (≥ 10-bit, global shutter, ≥ 2 pixels per fringe).
When specifying a reference flat, remember that its figure error is your accuracy floor — not your repeatability floor. A λ/20 reference limits absolute accuracy to λ/20 even if your PSI repeatability is λ/1000. For better absolute accuracy, invest in a calibrated λ/50+ reference or perform three-flat calibration.
9.Metrology Applications
Interferometry spans a vast application range: surface figure testing (Fizeau + PSI), displacement measurement (heterodyne Michelson), thin-film thickness (spectrophotometric), rotation sensing (fiber Sagnac), tissue imaging (OCT), and gravitational wave detection (LIGO).
| Application | Architecture | Typical Precision |
|---|---|---|
| Surface figure | Fizeau / Twyman–Green | λ/100 RMS |
| Displacement | Heterodyne Michelson | 0.3–1 nm |
| Thin-film thickness | Spectrophotometric / WLI | ±1–5 nm |
| Rotation rate | Fiber Sagnac | 0.001°/hr |
| Tissue imaging | SD-OCT / SS-OCT | 1–15 μm axial |
If the test surface is rough or has step features, skip PSI and go directly to vertical scanning interferometry (WLI) — PSI requires smooth, continuous surfaces with < λ/2 slope between adjacent pixels.
10.Interferometer Selection and Specification
Start with four questions: (1) What are you measuring? (2) What precision do you need? (3) What is the sample geometry? (4) What are the environmental constraints? The answers map directly to architecture and component choices.
For general-purpose optical shop testing, a Fizeau interferometer with HeNe source, λ/20 reference flat, megapixel camera, and Hariharan PSI is the default starting point. Deviate from this only when the application demands it (vibration → simultaneous PSI; rough surfaces → WLI; large OPD → stabilized laser).
Continue Learning
The Comprehensive Guide includes 6 worked examples, 6 SVG diagrams, and 10 references.