Search

A backside-illuminated (BSI) near-infrared enhanced silicon single-photon avalanche diode (SPAD) for time-of-flight (ToF) light detection and ranging applications is presented. The detector contains a 2 $\mu$m wide multiplication region with a spherically-uniform electric field peak enforced by field-line crowding. A charge-focusing electric field extends into a 10 $\mu$m deep absorption volume, whereby electrons generated in all corners of the device can move efficiently towards the multiplication region. The SPAD is integrated with a customized 130 nm CMOS technology and a dedicated BSI process. The device has a pitch of 15 $\mu$m, which has the potential to be scaled down without significant performance loss. Furthermore, the detector achieves a photon detection efficiency of 27% at 905 nm, with an excess bias of 3.5 V that is controlled by integrated CMOS electronics, and a timing resolution of 240 ps. By virtue of these features, the device architecture is well-suited for large format ToF imaging arrays with integrated electronics.

A near-infrared (NIR) enhanced silicon single-photon avalanche diode (SPAD) fabricated in a customized 0.13 $\mu$m CMOS technology is presented. The SPAD has a depleted absorption volume of approximately 15 $\mu$m x 15 $\mu$m x 18 $\mu$m. Electrons generated in the absorption region are efficiently transported by drift to a central active avalanche region with a diameter of 2 $\mu$m. At the operating voltage, the active region contains a spherically uniform field peak, enabling the multiplication of electrons originating from all corners of the device. The advantages of the SPAD architecture include high NIR photon detection efficiency (PDE), drift-based transport, low afterpulsing, and compatibility with an integrated CMOS readout. A front-side illuminated device is fabricated and characterized. The SPAD has a PDE of 13% at wavelength 905 nm, an afterpulsing probability < 0.1% for a dead time of 13 ns, and a median dark count rate (DCR) of 840 Hz at room temperature. The device shows promising performance for time-of-flight applications that benefit from uniform NIR-sensitive SPAD arrays., Comment: 8 pages, 3 figures

Vignetting in camera lenses is generally seen as something to avoid. For spectral cameras with thin-film interference filters, however, we argue that vignetting can be an advantage. When illuminated by focused light, the bandwidth of interference filters increases with the chief-ray angle, causing position-dependent smoothing of the spectra. We show that vignetting can be used to reduce smoothing and preserve important spectral features. Furthermore, we demonstrate that by adding additional vignetting to a lens, measurements can be made more consistent across the scene. This makes vignetting a useful parameter during spectral camera design.

Spectral cameras with integrated thin-film Fabry-P\'erot filters have become increasingly important in many applications. These applications often require the detection of spectral features at specific wavelengths or to quantify small variations in the spectrum. This can be challenging since thin-film filters are sensitive to the angle of incidence of the light. In prior work we modeled and corrected for the distribution of incident angles for an ideal finite aperture. Many real lenses however experience vignetting. Therefore in this article we generalize our model to the more common case of a vignetted aperture, which changes the distribution of incident angles. We propose a practical method to estimate the model parameters and correct undesired shifts in measured spectra. This is experimentally validated for a lens mounted on a visible to near-infrared spectral camera.

Heart-rate estimation is a fundamental feature of modern wearable devices. In this paper we propose a machine intelligent approach for heart-rate estimation from electrocardiogram (ECG) data collected using wearable devices. The novelty of our approach lies in (1) encoding spatio-temporal properties of ECG signals directly into spike train and using this to excite recurrently connected spiking neurons in a Liquid State Machine computation model; (2) a novel learning algorithm; and (3) an intelligently designed unsupervised readout based on Fuzzy c-Means clustering of spike responses from a subset of neurons (Liquid states), selected using particle swarm optimization. Our approach differs from existing works by learning directly from ECG signals (allowing personalization), without requiring costly data annotations. Additionally, our approach can be easily implemented on state-of-the-art spiking-based neuromorphic systems, offering high accuracy, yet significantly low energy footprint, leading to an extended battery life of wearable devices. We validated our approach with CARLsim, a GPU accelerated spiking neural network simulator modeling Izhikevich spiking neurons with Spike Timing Dependent Plasticity (STDP) and homeostatic scaling. A range of subjects are considered from in-house clinical trials and public ECG databases. Results show high accuracy and low energy footprint in heart-rate estimation across subjects with and without cardiac irregularities, signifying the strong potential of this approach to be integrated in future wearable devices., Comment: 51 pages, 12 figures, 6 tables, 95 references. Under submission at Elsevier Neural Networks

We report first results of laboratory tests of Si:As blocked-impurity-band (BIB) mid-infrared (4 to 28 um) detectors developed by IMEC. These prototypes feature 88 pixels hybridized on an integrated cryogenic readout electronics (CRE). They were developed as part of a technology demonstration program for the future DARWIN mission. In order to be able to separate detector and readout effects, a custom build TIA circuitry was used to characterize additional single pixel detectors. We used a newly designed test setup at the MPIA to determine the relative spectral response, the quantum efficiency, and the dark current. All these properties were measured as a function of operating temperature and detector bias. In addition the effects of ionizing radiation on the detector were studied. For determining the relative spectral response we used a dual-grating monochromator and a bolometer with known response that was operated in parallel to the Si:As detectors. The quantum efficiency was measured by using a custom-build high-precision vacuum black body together with cold (T ~ 4 K) filters of known (measured) transmission., Comment: 11 pages, 8 figures, to appear in "High Energy, Optical, and Infrared Detectors for Astronomy" SPIE conference Proc. 7021, Marseille, 23-28 June 2008