First Quarter 2000

In January, an excellent Aladdin 3 array became available from the PAIDAI project. Aladdin 3 devices employ an improved (206) multiplexer and have generally improved readnoise, speed and reduced odd/even column structure compared to Aladdin 2 devices (152 mux). After comparing performance of the Aladdin 2 array SCA47672 with the Aladdin 3 array SCA412202, we decided to take delivery of SCA412202 and return SCA47672 to the PAIDAI project. Testing of SCA412202 in lab test dewar during January confirmed the expected performance at 30K and 0.4V bias: dark current 0.2e/sec, 6Hz readnoise 40e CDS scaling to 10e at slower speeds and with Fowler multiple sampling, well depth 100,000e which is linear to 0.5% up to 60,000e. Cosmetically, the array has two long one-pixel-wide cracks (which will have undetectable impact on data quality), and very good uniformity.

Meanwhile the fifth cold test of the instrument cryostat again using the muxes, was started on January 10. The objective of this test was to cold test the modifications to reduce flexure and to test a modification to the Fused Silica/ZnSe prism mount. Flexure testing confirmed the warm results (see last quarter) and that no further work was required. During the previous test 10% fringing was observed at 0.8-0.9 microns in the 0.8-2.5 micron cross-dispersed mode. The fringing was completely removed by increasing the spacing between the two prisms from 0.1mm to 1.0mm. (Apparently, AR coats don't work when separated by 0.1mm or less.) Unfortunately the grating turret mechanism failed during the test and the turret had to be positioned manually (an advantage of using warm motors). The turrets' flex stage mechanism proved to be too weak. Work to replace it with an anti-backlash worm mechanism was begun.

Following characterization of SCA412202 in the lab test dewar the science arrays were installed in SpeX, the Aladdin 3 1024x1024 array in the spectrograph and the Aladdin 2 512x512 array in the imager. Cooldown number six - the first all-up test of the instrument, was started on February 15. Overall the test went extremely well. The spectrograph background stabilised at 0.2e/sec after about 48 hours of cooling. This is the same as the blanked off detector dark current measured in the test dewar and indicates that the background from the instrument due to light leaks and thermal emission from the 80K cold structure is insignificant. Placement of spectra on the array in each of the five spectral modes was close to that predicted. The spectrograph array will be moved about 50 pixels (1.35mm) in the long wavelength direction. Two optical problems requiring fixes were discovered. In the 0.8-2.5 micron cross-dispersed and 0.8-2.5 micron low-resolution modes a deep absorption was observed at about 2.15-2.3 micron. This was tracked down to the prism vendor supplying us with optical grade Fused Silica instead of IR grade Fused Silica. Optical grade Fused Silica contains a deep water band at about 2.1-2.3 microns. The vendor is remaking the two Fused Silica prisms from IR grade material. We hope to install them by early April. In the 2-5.5 micron cross-dispersed mode fringing was observed at the 5-10% level. This should be removable by wedging the Ge/LiF prisms. Flat fielding and lamp line calibration using the calibration unit worked very well. The new grating turret mechanism worked without incident. After about a week of cold tests the two-day warm-up sequence was started on February 25.

Following modifications to the Ge/LiF prism mount to wedge the prisms by 0.3 degrees and the slight shift in the spectrograph detector position, cooldown number seven was started on March 3. Wedging the prisms was completely successful in removing the fringes and the spectrograph detector is now in the correct location.

As of this writing (March 9) SpeX is undergoing more cold tests and calibrations. The current testing includes experimenting with different read out procedures to reduce the effects of persistence from which all Aladdin arrays suffer, testing the high speed occultation mode and slit-viewer guide modes. We are still using the Aladdin 2 readout code which leaves two rows unclocked in the centre our Aladdin 3 array (which has a different mux architecture). The Aladdin 3 readout code is currently being tested and will be implemented by the end of March.

IRTF graduate student Mike Cushing and IRTF astronomer Bill Vacca are writing an IDL spectral extraction package which will allow observers to view extracted and calibrated spectra while observing. A preliminary version of this package should be available in June.

0.8-2.5 micron XD mode Quartz-Tungsten-Halogen lamp 0.5sec flatfield exposure with 0.3x15arsec slit. The spectrum runs from 0.80um at top-left to 2.47um at bottom-right ( see Science Modes). The faintness of the 0.8um order is due the decrease in intensity of the QTH lamp towards the visible. The absorption seen at about 2.2um will be remedied when the IR grade Fused Silica prism is installed (see above). The narrow absorption lines at the edges of some orders are due to the air in the calibration box. The two diagonal lines are due to narrow cracks in the InSb. The horizontal line will be removed when the Aladdin 3 clocking code is implemented.
1.9-4.2 micron XD mode QTH lamp 0.5sec flatfield exposure with 0.3x15arsec slit. The spectrum runs from 1.9um at top-left to 4.2um at bottom-right (see Science Modes). The same range with a 1100K IR lamp. Note the atmospheric features at 1.8-2.0um and 2.5-2.9um.
2.4-5.5 micron XD mode 1100K IR lamp 0.5sec flatfield exposure with 0.3x15arsec slit. The spectrum runs from 2.4um at top-left to 5.5um at bottom-right (see Science Modes). Note the atmospheric features at 2.5-2.9um, and the fringes at 5.3-5.5um. The fringes occur in the InSb as the substrate becomes more transparent with increasing incident wavelength.
0.8-2.5 micron low-res mode Argon lamp spectrum with 0.3x60arcsec slit. The re-imaged slit lines are slightly curved due to the double pass through the single prism. This mode is matched to one quadrant of the array. This permits faster read out when required (occultation mode - see Science Modes).
3.1-5.5 micron single-order mode 1100K IR lamp 0.5sec flatfield exposure with 0.3x60arsec slit and with a 3.13-3.53um blocking filter in the order sorter wheel (see Science Modes). Different order sorting filters can be selected for other spectral ranges.
0.9-2.5 micron single-order mode Argon lamp spectrum with 0.3x60arcsec slit and with a 1.92-2.52um blocking filter in the order sorter wheel (see Science Modes). Different order sorting filters can be selected for other spectral ranges.
IR guider/slit-viewer A K-band image of the lab ceiling (a flat field) taken with the IR guider/slit-viewer with the 0.3x15arcsec reflective slit mirror in the slit wheel. The 512x512 Aladdin 2 array provides a 60x60arsec FOV (0.12arsec/pixel). A similar view but with the 0.8x60arsec slit in the slit wheel.