SWIRC Engineering and Commissioning Reports
SWIRC Commissioning
Comments or Questions? Contact wbrown@cfa.harvard.edu

PI: Warren Brown (SAO)
Co-I's: Steve Amato, Maureen Conroy, Giovanni Fazio, John Geary, Joe Hora, Brian McLeod, Michael Pahre, Steve Willner, Bill Wyatt (SAO), and Emily Bowsher (Wellesley)

Nov 2012
Marc Lacasse and I opened the dewar to replace the charcoal getter. Our goal was to improve vacuum and detector operating temperature. Tests showed that the detector is working like normal after handling.

June 2010
One channel on an A/D board failed. We repaired the channel, and upgraded components on all of A/D boards to prevent this from recurring.

January 2007
The new Y-band filter is installed and works great! See the exposure time calculator. The filter wheel problem is fixed.

October 2006
The new Y-band filter has arrived. Transmissions scans confirm that the filter has excellent out-of-passband rejection at high incidence angles. The Y-band filter will be installed before the next SWIRC run.

September 2006
SWIRC's filter wheel became stuck in H-band during the second night of the run, but the instrument continued to obtain excellent science data (in one filter). The filter wheel problem will be fixed before the next SWIRC run.

January 2006
The overhead time between exposures has been halved to nearly 4 seconds, a result of telescope motion flags being updated more rapidly by the mount computer. However, we had a few streaked images which is unusual for SWIRC. The streaked image problem was later tracked to primary mirror hard-point slippage.

November 2005
The new J-band filter is installed and gives a factor of ~3 improvement in SWIRC's J-band sensitivity. The instrument performance numbers have been updated accordingly. SWIRC operated in 0.3" seeing at times during the run.

April 2005
We bonded a blocking filter to the old J-band filter, and measured a 3.8x drop in background; the total J-band throughput is down by a factor of 0.88. The exposure time calculator has been updated accordingly.

We verified the world coordinate system header information in the 32-channel read-out mode, and exercised a number of new MICE control features. Our biggest problem, the telescope server repeatedly crashing, was traced to a Megacam guider server that was unrelated to SWIRC. A few minor fixes to the MICE interface will be implemented for the May 2005 run.

A new J-band filter, originally scheduled for delivery March 25, 2005, was postponed by the manufacturer to sometime in June.

October 2004
The 32 channel electronics were completed in September 2004, and SWIRC's read-out time reduced from 5.7 seconds to 0.7 seconds. The new electronics provide a reduced readnoise of 31 electrons, and a gain of 4.0. The linearity of the detector is excellent (at 0.1\% level) to 160,000 electrons, and the dark current is very stable. However, one channel has RMS 75 electron bias fluctuations. Cross-talk is not apparent between channels.

We confirmed the J-band filter red leak with transmission scans taken at large incidence angles. We purchased a blocking filter to reject this light, and bonded it onto the J-band filter. The composite filter successfully survived cryogenic temperature cycling. We joined a consortium filter purchase that has allowed us to purchase a new J-band and a new Y-band filter (with more stringent specifications) at a significant cost savings; we expect delivery sometime mid-2005.

We were allocated 5 nights, Oct 22 - 26, for SWIRC engineering and shared-risk science observations. All 5 nights on the telescope were lost due to weather. SWIRC worked flawlessly taking darks and flats. A 60 Hz noise problem was found, traced to a faulty power supply fan (probably caused by shipment), and eliminated.

Comparison of dome flats suggest that the J-band background has been reduced by a factor of >3. By comparison, we expected the background to decrease by a factor of ~15, and the throughput of the composite J-band filter to decrease by a factor of 0.8. However, until we go on-sky we cannot verify the new J-band performance or verify the correct orientation of SWIRC images (we use new software to de-commutate the 32 channel read-out). These issues will be resolved during engineering in Spring 2005.

July 2004
The new SAO Wide-field InfraRed Camera (SWIRC) was commissioned at the MMT 29 June - 5 July 2004. It was proposed, built, and shipped in 13 months at a modest cost ($139,000) paid for by Smithsonian IR&D funds.

Throughput is 2.5-3 times that of FLAMINGOS on the MMT. The H-band (1.6 um) sensitivity is comparable to FLAMINGOS, as expected. However, the J-band background during commissioning was substantially higher than expected.

Sustained seeing of 0.4 arcsec FWHM (2.7 pixel) was observed during commissioning, so the instrument's optics have good image quality. Maintaining f/5 secondary collimation and preventing telescope drive oscillation appear to be the major concerns in maintaining image quality. SWIRC was returned to the CfA following to implement a 32 channel read-out.

Globular Cluster M3
Commissioning image of the globular cluster M3, observed in 0.4 arcsec seeing.