Thermal Light Valve Brings IR Imaging to the Masses
- Created on Sunday, 01 October 2006
Technology uses MEMS structure with refractive properties sensitive to thermal radiation and reads out signal with inexpensive laser diode and CMOS sensor.
RedShift Systems, Waltham, Massachusetts
The desire to “see” in complete darkness or through obscurants such as smoke or fog has driven the development and adoption of thermal imaging technology. Thermal imaging is the translation of a scene’s heat signature — the 8-μm to 14-μm or long-wavelength infrared (LWIR) energy an object emits — into a visible image or data that can be interpreted by a computer.
Because the thermal energy of a scene is largely independent of reflected light and because it can travel through many obscurants, thermal imaging is the technology of choice for imaging in the dark or other difficult environmental conditions. The demand of infrared vision is there but has been hampered by the high cost of conventional thermal imaging cameras. A new passive optical component — the thermal light valve (TLV) — has entered the thermal imaging market with the intent of reducing the cost of thermal imagers.
Optical thermal imaging based on thermal light valve technology does not rely on the change of resistance, or electrical effects, to measure temperature changes. Instead, the thermal light valve imaging technique relies on changes in optical properties when exposed to temperature changes. These optical-property changes are detected using standard digital camera electronics rather than electrically reading the signal from the sensor itself.
The TLV is composed of narrow-band optical filter pixels standing on thermally isolating posts on a standard MEMS (microelectromechanical system) substrate (see Figure 2). Each pixel acts as a passive wavelength converter. Long-wavelength infrared (LWIR) radiation from the scene is imaged onto and absorbed by the TLV. This heats up thermal pixels on the array in direct relation to the thermal signature of the scene. The reflective wavelengths of the pixels shift based upon the thermal energy incident on each. A narrow-band near-infrared (NIR) light source is used to “probe” the temperature of the pixels across the TLV. This NIR probe signal is reflected off the TLV in varying amounts, depending on the pixel temperature, onto the CMOS imager. The intensity of the light received by the CMOS imager is therefore “modulated” by the heat signature of the scene.
The TLV tunable optical filter is a Fabry-Perot (FP) structure. It is constructed from amorphous silicon (a-Si) and silicon nitride (SiNx) thin films, which have been used extensively for many years in solar cells and flat-panel displays. These materials are deposited using plasma-enhanced chemical vapor deposition (PECVD), which is capable of producing uniform, dense materials in high-volume manufacturing environments. The optical filter’s minimum reflective wavelength depends on the optical thickness of the cavity — a product of physical thickness and index of refraction.
The TLV method achieves tunability by changing the index of refraction. The materials are characterized by a high thermo-optic coefficient, which is defined as the change of index of refraction per degree of temperature change.
Some important key differences compared to other technologies are:
- The sensing array is not an electronic device. It is purely a passive layer of optical thin films on glass. This greatly simplifies manufacturing and packaging.
- The sensing array is manufactured in a standard MEMS foundry, thereby taking advantage of foundry economies of scale to dramatically reduce manufacturing cost over that obtainable with custom fabrication lines.
- The readout circuit in the optical thermal imaging system is not physically coupled to the sensing array, nor is it a custom design. It consists of off-the-shelf parts, such as laser diodes and CMOS sensors that can be sourced from high-volume optical mouse and consumer camera applications, and managed independently from the sensor array. This reduces cost, increases yield, and reduces development cycle time.