Lick Infrared Camera User's Guide

Instrument Characteristics

Detector Specifications

Array size:256x256
Pixel size:40x40-microns with 3-microns between pixels
Dark Current:less then 2e- sec-1
Readnoise:approximately 50e-
Full well:300,000e-
K-band QE:62% @ 77K
Cut-on wavelength:0.9-micron
Cut-off wavelength:2.5-micron
Gain:12 electrons DN-1

Flux Information

Approximate Count Rates

BandpassMagnitudeDN sec-1
H N/A 2200
Table 2. Approximate exposure rates These numbers are based on 3-meter observations of the white dwarf standard GD 140. Narrow-band filters will be about 10% of the broad-band flux rates.

Approximate Sky Brightness

The Mt. Hamilton sky is approximately 13th to 14th magnitude per square arcsecond at K', and about one to two magnitudes fainter at J and H-bands. K and K' sky brightness, and consequently sky-limited exposures times, vary dramatically with local conditions, especially temperature. Table 3 gives approximate times to complete saturation of the array (about 23,000 DN) for two air temperatures at K'.

Outside Air Temperature (F)Camera Field of ViewApproximate Time to Saturation
39narrow 165 secs.
39intermediate 70 secs.
39wide 20 secs.
55narrow 85 secs.
55intermediate 35 secs.
55wide 10 secs.
Table 3. Times to saturation at K' These tests were made at the 1-meter under clear, dark skies, with the TUB diagonal mirror out. The same numbers should apply at the 3-meter.

Array Linearity

The array's linearity is plotted in Figure 9. The measurements were made by integrating a small photon flux for varying times from near zero to near saturation. Two different flux levels were used to correlate the relative slopes: about 50,000 e- sec-1 for the upper measurement, about 4,000 for the lower. The latter consists of four traces, each representing the response of one quadrant of the array.

Figure 9. Array linearity

NICMOS 3 Focal Plane Array Technical Description

The focal plane array (FPA) consists of a HgCdTe detector array hybridized to a silicon multiplexer using indium columns (bumps). The detector array converts the incoming infrared radiation into electrical charge and the multiplexer processes the charge. The two components are fabricated separately, resulting in the independent optimization of the performance of each device. Indium columns are evaporated on the detector and multiplexer array after each is fabricated. The two devices are then aligned and cold-welded together to form the hybrid focal plane array. The hybrid is illuminated through the substrate--in this case, sapphire.

The key component of the hybrid FPA is the readout multiplexer, which converts the integrated detector current into an output voltage and multiplexes the detectors into a serial data stream. The multiplexer was designed using 3-micron, single polysilicon, single metal, p-well CMOS design rules. The multipexer circuit consists of an array of MOSFET switches arranged so that each detector element is read out through a single unit cell repeated as a 128x128 array, connecting to each detector through an indium column. Four such arrays are mosaiced to form the NICMOS 3 256x256 detector.

The detector is biased to a voltage VR - DetSub. In our case, this is a reverse bias level of 0.5 volts. After the detector is read, the reset switch is turned on, forcing the potential at the cathode of the detector to voltage VR. When this switch is then opened, the cathode of the detector is free to float in potential. Each detector has a depletion capacitance into which the photogenerated and leakage currents are integrated. As charge integrates, this potential discharges toward DetSub. LIRC-II uses the reset gate for each pixel so that correlated double sampling can be used to reduce kTC noise. For correlated triple sampling and multiple readout of the same exposure, the reset gate is not used. This software-controllable feature is not yet an option in the LIRC-II camera system. Unfortunately, the process of resetting the detector bias is not uniform. This shows up as a non-uniform DC offset in the output signal. This is due to threshold variations caused by non-uniformity in the silicon MUX and processing variations. Because these non-uniformities are the result of materials variations, and not random electrical noise, they are repeatable from one readout of the array to the next and can therefore be removed by calibration.

The array uniformity is determined by the uniformity of the MUX and the detector material. Due to the way the MUX is laid out, there is a systematic odd/even offset built into the MUX. This offset is removed by most sampling schemes or by simple flat field calibration.

Array Quantum Efficiency

The following QE curve is for the IRIM camera at KPNO, but should be reasonably close to the curve for LIRC II.