| MIKE
CCD system |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Each
MIKE
CCD detector system uses a 2K × 4K back-illuminated device,
with 15 µm pixels and two output amplifiers. At present,
the
red channel uses a SITe ST-002A
CCD, while the blue channel uses a MIT Lincoln Labs CCID-20. The
ST-002A has a standard broadband AR coating. The CCID-20 has
a
blue enhanced coating. Each detector
is mounted in a rectangular aluminum housing. The dewar
window on
each side is the field flattener of the camera.The CCD housing is coupled to an IR-Labs ND-5 cryostat, with the CCD control electronics box attached to the side of the dewar (see photo). Each CCD is liquid nitrogen cooled and has roughly 15 hour hold time in the horizontal position. Inside the housing, the CCD is connnected to a rigid-flex preamp board, which brings the signals out on a flex cable to a hermetic connector. Each CCD controller is a compact dual-channel digital signal processor based system, capable of both high-speed and slow-scan readout. We plan to operate in slow-scan mode, using only the better of the two CCD output amplifiers for the lowest noise performance. Lab tests show that read noise is about 4 e-/pixel for the SITe device. The LL device has a measured read noise of 1.8 e-. The two detector systems are mounted side-by-side on the instrument (see photo) and operate independently. That is, they have independent shutters, exposure times and readout times. [from R. Bernstein et al, "MIKE: A double-echelle spectrograph for the Magellan telescopes at Las Campanas Observatory," SPIE 4841 Astronomical Telescopes and Instrumentation, Kona (2002).] CCD Properties
CCD Quantum
efficiency measurements
Dewar Mechanical Internals The CCD housing for the MIKE detectors is machined from a block of aluminum, and measures 6.625 inches x 6.625 inches x 5 inches deep (see mechanical drawings). The MIKE_RED ccd is mounted on an Invar package by SITe, and has a 3 inch ribbon cable to connect the electrical signals. The ribbon cable terminates in a 40-pin RN plastic connector. The ribbon cable is not thermally isolating, so we have a 2-inch flex cable which connects the ccd ribbon cable to the 37-pin preamp connector. The MIKE_BLUE ccd is mounted on an AlN substrate, and has a short flex cable for the ccd signals. The flex cable terminates in a 37-pin micro-D connector. A short cross-connect cable of constantan wire connects the ccd flex-cable to the preamp connector. We mount both the SITe package and the LL package onto a small Invar substrate (or boat) which is then screwed down onto a baseplate. The baseplate has a hole in it for a cold probe to contact the boat from beneath. The Invar baseplate is supported on three G-10 standoffs which provide thermal isolation and permit the ccd tilt to be adjusted. The G-10 standoffs are screwed into the base of the dewar head.
One side of the dewar head has a recessed sideplate (with o-ring) where we mount a preamplifier rigid-flex circuit board. The component area of the pcb is rigid, and has a flex cable which solders to the 79-pin connector. The internal board has two preamp circuits, filters for the clock and bias lines, and two temperature sensor circuits.
The cold probe from the dewar cold surface contacts the ccd+boat from underneath. A small temperature sensor and heater resistor are fitted onto the bottom of the boat to stabilize the temperature.
Test results -- Known Issues
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Greg Burley (burley@ociw.edu) Ian Thompson (ian@ociw.edu) |