Novel silicon architecture for increased dynamic range of infrared LED scene projectors

Abstract

Infrared (IR) imaging systems are important tools in many contexts, from academia to industry, but many of the conditions they are meant to operate in can be dangerous, expensive, or impossible to recreate for testing purposes. Further complications arise when such testing requires real-time response of the test scenario to the system under test. This type of testing is referred to as hardware-in-the-loop (HWIL). The test equipment employed in these situations consists of infrared scene projectors (IRSPs), which are comprised of an infrared microdisplay and optics to match the image to the detector. Since their introduction in the 1980s, most IRSP systems have been based around arrays of thin-film black-body emitters (resistors), a technology that is still in widespread use. This technology has a number of limitations preventing it from fully exercising modern imaging systems. These limitations are largely due to thermal emitters being required to operate at approximately the temperature at which they appear to the imaging system. This affects maximum apparent temperature and pixel response time, which, in turn, affects effective frame-rate. Other technologies have, therefore, recently been investigated to improve these metrics. ☐ One such technology is based around arrays of superlattice light-emitting diodes (SLEDs). The best of these arrays can exhibit apparent temperatures greater than 1400K to a narrow-spectral-band detector in the 3-5µm range, and their response times are limited primarily by their drive electronics. These SLED arrays are then indium bump-bonded to a read-in integrated-circuit (RIIC) to produce a microdisplay hybrid. This hybrid and appropriate support electronics constitute a complete system. The first of these systems was demonstrated by The University of Delaware in collaboration with the University of Iowa in 2015. This system was not, however, ready to be used to test real IR imaging systems for a number of reasons. Many of these issues have since been addressed, but many remain to be solved before this technology can reach its true potential. ☐ This work explores many proposed solutions to these remaining problems, the additional problems created by these solutions, and the solutions to these additional problems. The primary focus of this paper is a complete architecture redesign of the RIIC, but the requisite changes to the control electronics and packaging are also presented. The changes to the RIIC extend the concept of using two transistors of different sizes to drive each of the pixels, as implemented in the systems currently being developed, to instead using numerous transistors that can be controlled semi-independently. The most efficient ways to accomplish the specific maximum radiance and radiance resolution goals are discussed, as are the solutions to problems created by the proposed solutions.