cs is the field of view (FOV), i.e., what is the target size at a prescribed distance? A very common lens system, for example, would be a 1 in. dia. target size at a 15 in. working distance. Using the inverse square law, by doubling the distance (30 in.) the target area theoretically doubles (2 in. dia.). The actual definition of target size (area measured) will vary
depending upon the supplier, and it is price dependent. Other optical
configurations vary from small spot for close-up pinpoint measurement, to distant optics for distant aiming. It is important
to note that working distance should not affect the accuracy if the FOV is filled by the target. In one technique for measuring FOV, the variable is signal loss vs. diameter. A strict rule is a 1% energy reduction, although some data are presented at half power, or 63.2% Alignment (aiming) is another optical factor. Many sensors lack that capability; the lens is aligned to the surface and measures surface temperature. This works with sizable targets, e.g., paper web, where pinpoint accuracy is not required. For small targets that use small-spot optics,and for distant optics used in remot monitoring, there are options of visual aiming, aim lights, and laser alignment. Selective spectral filtering typically uses short-wavelength filters for hightemperature applications. This obviously fits the blackbody distribution curves, and there are some technological advantages. For example, high temperature/short wavelength uses a very thermally stable silicon detector, and the short-wavelength design minimizes temperature error due to emissivity variations. Other selective filtering is used for plastic films (3.43 μm and 7.9 μm), glass (5.1 μm), and flame insensitivity (3.8 μm). A variety of detectors are used to maximize the sensitivity of the sensor. Most detectors are either photovoltaic, putting out a voltage when energized, or photoconductive, changing resistance when excited. These fast-responding, high sensitive detectors have a tradeoff thermal drift that can be overcome in many ways, including temperature compensation (thermistors) circuitry, temperature regulation, auto null circuitry, chopping (AC vs. DC output),and isothermal protection. Drift-free operation is available in varying degrees and is price dependent. In the IR thermometer’s electronics package, the detector’s nonlinear output signal, on the order of 100-1000 μV, is processed. The signal is amplified 1000 x, regulated, and linearized, and the ultimate output is a linear mV or mA signal. The trend is toward 4-20 mA output to minimize environmental electrical noise interference. This signal can also be transposed to RS 232 or fed to a PID controller, remote display, or recorder. Additional signal conditioning options involve on/off alarms, adjustable peak hold for intermittent targets, adjustable response time, and/or sample-and-hold circuitry. On the average, IR thermometers have a response time on the order of 300 ms, although signal outputs on the order of 10 ms can be obtained with silicon detectors. In the real world, many instruments have an adjustable response capability that permits damping of noisy incoming signals and field adjustment on sensitivity. It is not always necessary to have the fastest response available. There are cases involving induction heating and other types of applications, however, where response times on the order of 10-50 ms are required, and they are attainable through IR thermometry.
