A different principle from the microbolometer
In a photon detector, each incoming infrared photon frees a charge carrier in a semiconductor, and the resulting current is the signal. Unlike thermal detectors, which measure the heating of the material, these respond directly to the radiation.
Why it has to be cooled
At room temperature, the semiconductor itself generates carriers through thermal agitation, and this dark current is greater than the desired signal. If the detector is cooled to cryogenic temperatures, this noise falls drastically and the real signal becomes visible. The usual materials are indium antimonide, mercury cadmium telluride and quantum well structures.
How it is cooled
The detector sits in a vacuum enclosure with a window. A mechanical Stirling-cycle cooler built into the camera extracts the heat. Some older units use liquefied gases.
Practical consequences
- High sensitivity: it allows very small temperature differences to be seen.
- Short integration time: fast events can be frozen and images captured at a high frame rate.
- Adjustable band: specific filters isolate narrow intervals, useful for seeing gases, flames or through certain materials.
- Cost and maintenance: these are more expensive units, with a cooldown time before use, and the cooler has a limited service life.
- Higher power consumption and size than uncooled units.
When it does not apply
For routine electrical and mechanical maintenance, where temperature differences are several degrees and the scene is static, an uncooled detector is usually sufficient. A cooled core is justified when speed, a specific band, gas detection or the highest sensitivity is required.