Wednesday, April 25, 2012

MWIR Pixel Size Approaches the Wavelength

Photonics-online: Sofradir demonstrates a prototype of the first 10-micron pixel pitch HgCdTe-based MWIR imager at SPIE DSS in Baltimore, April 23-27, 2012.

Assuming the imager covers 3-5um band, the pixel size is just twice the wavelength, similar to 1.4um visible pixels for 0.4-0.7um band.

The focal plane array prototype was developed with the support of DGA (Direction Générale de l’Armement) by CEA-Leti at DEFIR, the joint laboratory of Sofradir and CEA-Leti. CEA-Leti is a leading European microtechnology, IT and health technology research center.

"It is critical for the military to see first and see the right target, whatever the weather conditions. This is what the higher resolution, higher range 10-micron pixel pitch infrared detector helps provide," said Philippe Bensussan, chairman and CEO at Sofradir. "Sofradir continues to build on its legacy of innovation. We were the first to introduce the 15-micron pixel pitch TV format IR detector, a compact high-resolution product that brought system integrators significant advantages in performance and footprint and has become an industry standard. We’re taking the lead once again by pushing the bar from 12-micron pixel pitch that exists today to 10-micron. Our customers can look forward to the ultimate performance in IR systems."

3 comments:

  1. Nice to see hybridization improvement in Europe.
    Raytheon already offer visible hybrids with 8 µm pitch: http://www.raytheon.com/businesses/ncs/rvs/Capabilities/ic_design/Capabilities/index.html#chip

    ReplyDelete
  2. FYI, One of the best papers presented at SPIE DSS 2012 in Baltimore:

    Christel-Loic Tisse et al., "An information-theoretic perspective on the challenges and advances in the race toward 12µm pixel pitch for megapixel uncooled infrared imaging".

    This paper explains first the effects of technology scaling on the SNR performance of a-Si microbolometer pixels and presents SNR simulations for different pixel sizes from 45μm to 12μm. SNR simulations confirm that, despite a reduction of the pixel’s structure and active area, a SNR performance greater than 30dB (for a blackbody at room temperature) can still be achieved while keeping a thermal time constant below 10ms. By using an information theory approach, the authors then model the imaging system (Lens + FPA + ISP) as a cascade of (analog and digital) communication channels. They show that, for a given die size and at constant SNR of FPA, the information capacity C of FPA increases quadratically with shrinking pixel geometry. They introduce the Wigner distribution function and defined the notion of space-bandwidth-products SBP of FPA and SBP of Lens in phase-space domain. They show that, to take full advantage of C of FPA, the lens must be designed so as to ensure that SBP of FPA remains contained within SBP of Lens over a large DOF. The authors argue that the z spatial dimension does not carry additional degrees of freedom. They analyze the relationship between lens complexity, DOF performance and SBP of Lens. They explain how pupil-phase engineering for DOF extension can be used to improve information efficiency of the imaging system, i.e. (C of System) / (C of FPA). They compare the performance of various extended-DOF techniques. They found that spectral sweep (or spectral focal coding) could be particularly beneficial in the race towards 12μm pixel pitch megapixel uncooled LWIR imaging.

    ReplyDelete
    Replies
    1. Christel-Loic Tisse et al., "An information-theoretic perspective on the challenges and advances in the race toward 12µm pixel pitch for megapixel uncooled infrared imaging"

      This paper is now available at:

      http://link.aip.org/link/doi/10.1117/12.918534

      Delete

All comments are moderated to avoid spam.