IR Pyrometers & Thermal Imagers

Every object above absolute zero radiates energy, and a measurable share of it falls in the infrared band of the electromagnetic spectrum. An IR sensor — a/k/a a pyrometer, infrared thermometer, or thermal imager (across a field of view) — collects that radiation through a lens, optically filters it to a specific wavelength range (its spectral response), and focuses it on a sensing element. The detector converts the radiated energy into an electrical signal proportional to the target's temperature, which is then scaled and sent to a local display or a control system. Because it reads from a distance with nothing touching the target, an IR sensor measures surfaces a thermocouple or RTD cannot — moving webs and rolls, energized equipment, and hot or hard-to-reach objects. Two setup decisions govern accuracy: choosing the right spectral response for the material, and entering the correct emissivity. Emissivity is the fraction of energy a surface radiates compared with a perfect blackbody (the standard of which emissivity is based) at the same temperature. On an opaque surface the remainder is ambient heat sources adding to the measured energy via reflections onto the surface, or in some cases transmissions from behind the material. So emissivity plus reflectivity plus transmission (if any) equals one, although it is further affected by surface finish, oxidation, and wavelength choice — which is why it has to be specified and set deliberately, not left at a default.

Each spectral range reads surface temperature based on the material. Match the wavelength to the target:

• 8 to 14 µm — plastics (thick/opaque), paper, painted and organic surfaces; general-purpose for low to medium temperatures.
• 7.9 µm — thin polyester films (PET, PETG), polyurethane, and other plastic films that are transparent at 8–14 µm but absorb strongly at this wavelength.
• 5 µm — glass surface temperature. Glass is opaque at 5 µm, so the reading represents the surface rather than energy transmitted from behind it. Used on flat glass, container glass, and similar.
• 3.43 µm — thin polyethylene and polypropylene films; polyolefins that transmit at longer wavelengths but absorb at 3.43 µm.
• 1 to 2 µm (typically 1.0, 1.6, and 2.2 µm) — high-temperature metals, molten metals, induction-heated parts, and semiconductors. Short wavelengths reduce the temperature error caused by emissivity uncertainty and work well at high target temperatures, where signal strength at these wavelengths is strong.

Picking the wrong wavelength is the most common cause of a reading that looks wrong no matter how the emissivity is set — for example, sighting a thin film with an 8–14 µm sensor and unknowingly reading the roll behind it.

There are two distinct exercises, and people use "calibrate" for both. The first is dialing in emissivity to a known material temperature. If a surface thermocouple/RTD reading is possible, adjust the pyrometer's emissivity setting up or down until the IR reading matches. For a reference temperature, three field methods work: (A) — a thermocouple or RTD touching the material at process temperature — (B) — coat a patch of the target with a material of known high emissivity, such as flat black high-temperature paint, read the patch, then adjust emissivity on the bare surface to match — (C) apply Kapton® tape, which has a known emissivity of about 0.95 in the 8–14 µm range, onto a material appropriate for this wavelength range to a section of web material at process temperature — preset the pyrometer to 0.95, read the tape, then adjust the emissivity setting while looking at the bare material after the tape section has passed. The second use of "calibrate" is verification against a blackbody calibration source — a lab device with a known, traceable emissivity of 1.0. The blackbody is, in fact, what the emissivity scale is referenced to. NIST- and DAkkS probe based calibration certificates are available as a line item on any Fluke Process Instruments and Raytek sensor and should be renewed on the interval your quality system requires.

Typical mounts include single-axis "L" brackets, two-axis "double-L" brackets, and swivel or flange-mount accessories. Beyond the hardware, three things matter: for single-color pyrometers aim the sensor so that the target fully fills the measured spot at the working distance (see distance-to-spot, below), keep the line of sight clear of steam, dust, and obstructions (2-color can tolerate these). Where the optics face a dirty or hot process, add an air purge to keep the lens clean and a cooling accessory to keep the housing in range. Avoid sighting through a window unless the window material is transparent at the sensor's wavelength and set the window transmissivity value into the pyrometer interface.

The sensor's electronics must stay within their rated ambient temperature. That limit varies by series and detector type, but as a guide: most fixed sensors are rated roughly 0 to 65 °C uncooled, with the longwave Thermalert 4.0 spot sensors among the highest in the category at −20 to +85 °C uncooled. Past that, cooling accessories extend the envelope:

• With an air-cooling jacket and clean instrument air — often to about 120 °C ambient.
• With a water-cooled housing — often to about 175 °C ambient.
• With a water-cooled housing plus an air-purged over-jacket (for example the ThermoJacket) — often to about 315 °C ambient.

The target temperature the sensor reads is a separate spec — pyrometers measure up to 3000 °C+. The ambient ratings above are about protecting the instrument itself.

Cost generally scales with the optics (the distance-to-spot ratio), the feature set, and the accuracy rating.

If kept within their rated ambient temperature and not physically damaged, IR sensors often last many years — in some cases decades. They have no moving parts in the measurement path. The practical life-limiters are heat soak on the electronics, optics damage or contamination, and cable wear, all of which good mounting and cooling prevent.

Fixed industrial pyrometers are built for the plant floor. Sensing heads are housed in stainless-steel bodies sealed to IP65 / NEMA-4 against dust and washdown; the ThermoView fixed thermal imagers carry an IP67 rating and are qualified to IEC shock (50 g) and vibration (3 g) standards for harsh installs. Portable instruments such as the Raynger 3i Plus are drop-rated (about 1 m / 3 ft). The vulnerable part is the optics, not the housing — an air purge that keeps the lens clean and a cooling accessory that holds the body in its ambient range are what let these instruments survive years in rolling mills, glass plants, and kilns, short of physical damage.

Infrared radiation itself was discovered by astronomer William Herschel in 1800, when he measured a temperature rise just beyond the red end of a prism's spectrum. Practical non-contact thermometry followed a century later: the disappearing-filament optical pyrometer, in which an operator matched the brightness of a hot target against a calibrated lamp filament, was developed around 1901 by Holborn and Kurlbaum. Fully electronic photoelectric pyrometers — the direct ancestors of today's instruments — came into industrial use through the mid-20th century, and the uncooled focal-plane detector made compact fixed thermal imagers practical from the 1990s onward.

Specify a two-color (ratio) pyrometer when the target may be partially obscured (dust, steam, smoke), when the optics will get dirty in service, when the target underfills the field of view, or when emissivity drifts during the process. A ratio instrument measures at two wavelengths at once and cancels common-mode attenuation, so it reads accurately through those conditions. Specify single-color for clean, well-defined targets that fully fill the field of view, where emissivity is stable and known — single-color delivers higher absolute accuracy and faster response on those targets. The Endurance line, E1R/E2R (integrated) and EF1R/EF2R (fiber-optic) cover the ratio cases; E1M/E2M/E3M, the Thermalert 4.0, Raytek Mi3, GP, CI/CM & 3iPlus portable cover single-color.

All three measure radiated temperature; they differ in how much of the scene they cover. A spot pyrometer reads one point — fastest response, highest absolute accuracy, simplest to install; right for a single control point such as an induction-heated billet or an extruder melt. A line scanner sweeps a single line across the process at 150–300 Hz and reports up to 1,024 points along it — right for moving webs and surfaces where you need a temperature profile across the width, such as rolling mills, glass lines, paper drying, and thermoforming. A thermal imager captures a full two-dimensional image (320×240 or 640×480 pixels on the ThermoView line) — right for monitoring for process & R&D applications, and for safety of areas or assets where hot spots can appear such as electrical gear or furnace exteriors. As a rule: one control point → spot; a profile across a moving product → line scanner; a thermal profile for process, R&D or area/assets safety monitoring → imager.

Through a window — yes, but only if the window material is transparent at the sensor's wavelength, and the window transmission value is dialed into the pyrometer settings. Ordinary glass blocks longwave IR, so an 8–14 µm sensor cannot read through it; shortwave (1–2 µm) instruments can use glass or quartz windows, and purpose-made windows exist for other bands.

A sight tube is the standard way to give a sensor a clean, defined line of sight into a furnace or a dusty enclosure. Measuring a flame or measuring a target through a flame is a specialized case: dedicated flame-measurement sensors (CO₂, CO, NOX bands) read combustion, while reading a target behind a flame calls for a wavelength where the flame is transparent. For flare stack monitoring, the pyrometer watches and alarms on cessation of the flame. Tell us the obstruction and we'll specify the wavelength around it.

This is the most common IR support call, and it's often not a broken sensor. Work through, in order: (1) Emissivity (see "How do IR sensors work?", above) — wrong value is the usual culprit; a bare metal surface set to a default 0.95 will read far low. (2) Reflected or transmitted ambient energy — a low-emissivity target near a hotter surface reflects that heat into the sensor and reads high; for visually transparent materials heat sources from behind can transmit through and skew the reading — Proper Emissivity adjustments & in some cases background-temperature compensation corrects it. (3) Cone of view from the pyrometer face to the material — if the target doesn't fully fill the measured spot (for single-color pyrometers), the sensor averages in whatever is behind or around it. (4) Wavelength mismatch — sighting a thin film or glass with the wrong band reads the wrong layer entirely. (5) Dirty optics or an obstructed path — steam, dust, or a filmed lens attenuate the signal (single-color or in extreme conditions two-color pyrometers). And remember the reference contact sensor can be wrong too — poor surface contact reads low. A ratio pyrometer sidesteps several of these failure modes at once.

Yes — discontinued Ircon Modline, Maxline, and ScanIR units, and legacy Raytek sensors, all have defined migration paths to the current Endurance and Thermalert 4.0 lines. Prater recommends the correct successor based on the legacy nameplate and/or p/n based process data, and supports existing installations still in service.

Distance-to-spot ratio (D:S) tells you how large a spot the sensor measures at a given working distance. A 100:1 sensor reads a 10 mm spot at 1 m; a 10:1 sensor reads a 100 mm spot at the same distance. The rule is simple: the measured spot must be fully filled by the target (for single-color use) — if the spot is larger than the target, the reading is skewed by whatever surrounds it since cooler or hotter surrounding items factor into the reading. Choose the D:S from your two fixed facts: how far the sensor must mount from the target, and how small the target is. Small targets or long standoffs need a high ratio; close, large targets can use a modest one. The IR line spans roughly 4:1 close-focus up to 250:1, with line scanners detecting hot spots at 600:1 — Tighter optics generally correlate to cost, which is why D:S is the first thing to nail down. 2-color (ratio) pyrometers can lose as much as 95% of combined energy from an underfilled and/or obstructed view spot, retaining as little as 5% in some cases, and still get a good measurement.

Most Fluke Process Instruments and Raytek lines — Endurance, Thermalert 4.0, GP, Thermoview imager, & the MP line scanners — are rep-firm, quote-only, and built to the configuration. The exception is the Raytek compact sensor families (Mi3, CI/CM, 3iPlus), sold through Prater Technical Partners, authorized distribution. Select SKUs ship from Brooklyn stock, others are special order contact for lead time. Tell us the p/n or application and we'll confirm whether a stocked SKU fits before you commit.