High-Speed Imaging

Applications that require imaging with short effective shutter times are challenging from several points of view. First, a short effective shutter time may result in very few photons being collected, which means that often the device needs to have high fill-factor, high quantum efficiency, and low read noise. These requirements are at odds with the industry-standard front-illuminated interline-transfer-charge-coupled device (ILT-CCD) normally used for these applications because such a device depends on the photocharge being hidden behind opaque (metal) lines after the shutter is closed. These metal lines limit the pixel fill factor of the device and therefore its sensitivity.

Lincoln Laboratory has developed an electronic shutter that can be used with a high-quantum-efficiency and 100%-fill-factor CCD to allow higher sensitivity devices than are possible with standard ILT-CCD technology. We have successfully designed and built a number of small CCD devices, typically with multiple readout ports, both with and without the electronic shutter, for adaptive optics use.

An example of a 128 × 128, 16-port device is shown in Figure 1. Figure 2 shows this device mounted in a low-noise camera system developed at the Laboratory.

Figure 1. Sixteen-port high-speed, low-noise CCD.Figure 1. Sixteen-port high-speed, low-noise CCD.
Figure 2. Low-noise, high-speed camera.Figure 2. Low-noise, high-speed camera.

A second approach to high-speed imaging technology is to take a number of fast sample images of the scene during an event and then read out the individual samples after the event is over. One recent implementation of this concept is a four-sample high-burst-rate CCD imager developed at Lincoln Laboratory. This device is capable of taking a series of four images less than 1 µsec apart in time. It makes use of an electronic shutter technology we developed to store four samples in a single pixel and to ensure good isolation between samples taken at different times. A 50-sample device is a second example of this type of imager. This device is capable of capturing 50 images at an 80 ns exposure time and 2 MHz frame rate.

For the special case in which the image moves both laterally and vertically across the surface of the device during the exposure (for example, when there is measurable vibration, atmospheric turbulence, or time-delay-and-integrate (TDI) with generalized direction), we have developed the orthogonal-transfer CCD (OTCCD) concept. Although this concept dates back to shortly after the introduction of the CCD, practical difficulties in its fabrication limited its use until now. We have developed a robust OTCCD technology and are currently applying it to very large arrays (5 × 5 cm device size) for use in ground-based astronomy to partially compensate atmospheric distortions. We call these orthogonal-transfer arrays.

The complementary metal-oxide semiconductor (CMOS) active pixel-array device is also useful for high-speed imaging. We have developed a very fast shutter speed (~100 ps) CMOS imager.

Photon-counting technology is also a candidate for high-speed applications since it inherently has no read noise. We have recently developed a passive photon-counting array for a Hartmann adaptive-optics wavefront sensor. There is a 64 × 64 array of Geiger-mode avalanche photodiodes (GMAPDs) that constitutes this device. Each 2 × 2 neighborhood of GMAPDs represents a quad cell used to find the centroid of a spot of light that represents the tilt of a local portion of a wavefront being detected by the Hartmann sensor.


top of page