Infrared equipment, camera setup and measurement techniques
Infrared Equipment
Detector families
Uncooled microbolometer arrays respond to absorbed heat in the longwave band, typically 8 to 14 um, and need no cryogenics. Cooled photon detectors (InSb or MCT in 3 to 5 um, QWIP structures, InGaAs for shortwave) offer faster response, better sensitivity and narrower spectral selection, at the cost of size, weight and cool-down time. Detector pitch, array format and thermal time constant set the practical resolution of the instrument.
Optics and geometry
Field of view follows from array size and focal length. Instantaneous field of view equals detector pitch divided by focal length, expressed in milliradians. Lens materials matter: germanium transmits the longwave and midwave but not visible light, silicon works in shortwave and midwave, sapphire and barium fluoride serve specialised bands. Close-up and telephoto lenses change both field of view and instantaneous field of view.
Performance specifications
Noise equivalent temperature difference describes sensitivity, not accuracy. Accuracy is usually quoted as a fixed offset plus a percentage of reading, over a stated ambient range. Spectral band, frame rate, dynamic range, digital resolution and the calibration source all affect what a reading means.
Traps
- Assuming more pixels means more accurate temperatures.
- Confusing noise equivalent temperature difference with measurement accuracy.
- Choosing the wrong spectral band for a low-temperature target or for measuring through a specific medium.
- Expecting germanium to pass visible light, or silicon to pass longwave radiation.
- Forgetting that a cooled camera needs stabilised cool-down before quantitative work.
Worked example
A camera with 17 um pitch and a 25 mm lens has an instantaneous field of view of 17/25 = 0.68 mrad. At 10 m the single-pixel spot is about 6.8 mm; applying the three-by-three-pixel rule gives a minimum measurable target near 20 mm. For a 5 mm fuse clip, that lens is unsuitable at that distance.
Camera Setup and Measurement Techniques
The setup sequence
Focus first, then set the temperature range, then level and span, then the input parameters: emissivity, reflected apparent temperature, distance, relative humidity, atmospheric temperature and, for long paths, atmospheric transmission. Only after these are correct do spots, boxes, isotherms, delta temperatures and profiles mean anything.
Geometry and focus
Sharp focus maximises high-frequency detail and is a precondition for detecting small thermal anomalies. Keep the line of sight within about 30 degrees of surface normal to limit angular emissivity effects and spurious reflection, and measure perpendicular to the surface, not along it.
Emissivity and reflection
For low-emissivity targets use high-emissivity tape, paint or a contact probe, allowing time to equilibrate. Estimate reflected apparent temperature with a crumpled aluminium foil or a diffuse reflector held at the same distance and angle, and enter it into the camera.
Traps in questions
- Leaving the camera on automatic range, which hides small delta temperatures.
- Setting emissivity too low on a warm target, which pushes the reported temperature up; too high does the opposite.
- Ignoring reflected apparent temperature on shiny metals or inside hot enclosures.
- Measuring through glass, acrylic guards or infrared-opaque sight glasses.
- Forgetting wind cooling, solar loading or a target that has not reached steady state.
- Recording a temperature without the conditions under which it was measured.
Worked example
A busbar with true emissivity 0.10 is read with emissivity set to 0.95. The camera assumes almost all received radiance is emitted, so on a target warmer than its surroundings the reported value is far too low, and a genuine 70 C joint may read near ambient. The rule of thumb: if emissivity is below about 0.9, correct it or change the surface before judging severity.