6.5 Detectors, Optics, and Imager Specifications

The ITC Level II outline opens its technical content with Camera Features and Specifications — detector types, IR resolution and optics, sensitivity, the advantages of different lenses, and the importance of IFOV and MFOV when measuring temperatures. Infraspection's electrical standard turns the same material into mandatory instrument requirements. Level II is where "which camera should we buy for this job?" becomes your question to answer.

Detector Technology

Why waveband matters at Level II:

• Long-wave (7.5–14 µm) suffers less from solar reflection and works better through smoke and some atmospheric conditions. It is the default for building and general condition-monitoring work.
• Mid-wave (3–5 µm) offers higher contrast at elevated temperatures and permits narrow-band filtering — the basis of flame-band, gas, and glass-surface measurement covered in Chapter 7.
• The choice is not cosmetic. Germanium optics, IR window materials, and emissivity values are all band-specific. An emissivity table value quoted for 8–14 µm may not hold at 3–5 µm.

Published Instrument Minimums

Infraspection's Standard for Infrared Inspection of Electrical Systems & Rotating Equipment (2016) Section 7 sets requirements that appear directly on exams:

That last row is the one candidates miss. An "Auto Image"-only camera is explicitly non-compliant for standard-conforming inspection, because the thermographer cannot set level and span to reveal a small ΔT against a wide scene range.

For comparison, ANSI/NETA requires imaging equipment "capable of detecting a minimum temperature difference of 1 °C at 30 °C" — a looser bar than Infraspection's 0.1 C°.

NETD in plain terms

Noise Equivalent Temperature Difference is the amount of infrared radiation, expressed in millikelvins, needed to produce a signal equal to the system's own electronic noise. Lower is better. A 50 mK camera cannot reliably resolve a 30 mK real difference — the signal is buried in noise. NETD is quoted at a stated scene temperature (usually 30 °C) and degrades at lower scene temperatures, which matters for building envelope work in cold weather.

Resolution, IFOV, and MFOV

Detector resolution is the pixel count (160 × 120 through 1280 × 1024 and beyond). It sets how many independent measurements exist across the scene.

IFOV (instantaneous field of view) is the angular size of one pixel:

IFOV = detector pitch ÷ focal length (radians; multiply by 1000 for milliradians)

At distance D, one pixel covers a spot of size IFOV × D. A camera with 1.3 mrad IFOV at 10 m resolves 13 mm per pixel.

MFOV (measurement field of view) is the crucial distinction. A single pixel that is just filled by a target does not produce an accurate temperature, because of optical spreading and detector crosstalk. Accurate radiometry commonly requires roughly a 3 × 3 pixel block on target — so MFOV ≈ 3 × IFOV.

Worked example. Camera IFOV = 1.3 mrad. Target: a 10 mm bolt head at 8 m.

• Spot per pixel = 1.3 × 10⁻³ × 8 = 10.4 mm — the bolt barely fills one pixel.
• MFOV spot = 3 × 10.4 = 31.2 mm — far larger than the bolt.
• The reading will be low, averaging the bolt with cooler surroundings.
• Fixes: move closer, fit a telephoto lens, or report the value as a minimum with the limitation stated.

This is the mechanism behind the classic under-reporting failure: a genuinely severe small connection graded as mild because it did not fill the measurement spot.

Lens Selection

The trade is strict: narrowing the field of view improves spatial resolution. When you cannot approach a target — energised switchyard, roof-level equipment, arc-flash boundary — the telephoto lens is the only way to satisfy MFOV.

Infraspection Section 9.2.5 requires the report to record the field of view of the imager lens, and 9.2.6 requires notation of any windows, filters, or external optics used. Optics are part of the measurement chain and must be documented like any other parameter.

Applying This on the Job

Scenario A — Spec review. A client proposes a 80 × 60 pixel imager with auto-only gain for NETA-referenced switchgear surveys. Both the resolution and the absence of manual level and gain fall below Infraspection's requirements. Recommend compliant equipment.

Scenario B — Distance limit. An overhead connection at 25 m with a 1.3 mrad standard lens gives a 32.5 mm pixel and a roughly 98 mm MFOV. A 20 mm connector cannot be measured. Fit a telephoto (for example 0.65 mrad, halving both figures) or report qualitatively.

Scenario C — Band mismatch. A germanium IR window rated for long-wave use is fitted, then imaged with a mid-wave camera. Transmittance is not what the window's data sheet claims. Match window material, camera band, and the transmittance correction of Chapter 6.

Quick Answer: Uncooled microbolometers work at 7.5–14 µm with 30–50 mK NETD; cooled photon detectors work at 3–5 µm with sub-20 mK NETD and microsecond response. Infraspection requires 2–14 µm, ≥ 120 × 120 pixels, NETD ≤ 0.1 C° at 30 °C, and manual level and gain — auto-only imagers are not sufficient. IFOV = pitch ÷ focal length sets what you can see; MFOV ≈ 3 × IFOV sets what you can measure. Telephoto lenses shrink IFOV and are the answer when you cannot get closer.