LWS
News Update for 5 May 1999

Contents:

LRES spectroscopic mode sensitivity determined

As part of the ongoing effort to characterize observing properties of LWS, we have measured the sensitivity of LWS in LRES (low-resolution) spectroscopic mode. We find that a signal level of 150 mJy yields S/N=1 per spectroscopic resolution element (3 pixels for the smaller 0.25 arcsec slit). Note that one could expect higher S/N with the wide (0.50 arcsec) slit. Also note that this measurement is based on a limited sample of data taken under optimal conditions. Your results may vary depending on transparency and other factors.

NODDELAY bug fixed

LWS has suffered from a few problems with regards to the operation of the instrument and quality of the resulting data. One such parameter is the noddelay system. Operating LWS in chop-nod mode demands excellent synchronization between the telescope, secondary chopper and data acquisition system. Noddelay (a pause measured in seconds) is a keyword that controls the amount of time LWS waits for the telescope to complete a nod operation (see timing parameters discussion for details).

Analysis of some early LWS engineering data revealed that the noddelay system was not functioning properly, causing some degree of image blur indicating that the exposure had started while the telescope was still moving. We confirmed the problem by analyzing the timing of the acquire program and comparing to the resulting images. The source of the problem was found in a section of low-level DSP (digital signal processor) instructions called the pass-thru code. This code helps coordinate the timing of the coadders, hardware that digitizes the analog pixel data and combines the data for each saved frame. The pass-thru code was not properly halting the coadders during the noddelay period.

Roger Anderson, a consultant to Keck Observatory and the SSL Berkeley on the LWIRC program, and Ed Wishnow, LWIRC project scientist, agreed to investigate the problem since LWIRC, a "sister" instrument to LWS, suffered from the same bug. The team used the existing LWIRC system in Berkeley to analyze the noddelay system. In April 1999, lab tests of modified pass-thru DSP code in Berkeley appeared to indicate that the system was fixed and ready for on-sky testing with LWS.

Noddelay=0 sec Noddelay=1 sec Noddelay=3 sec

On 5 May 1999, the modified code was installed on the LWS system and tested on-sky during an LWS engineering night. The code was evaluated under several different instrument configurations. One test was to simply observe a bright point source star and vary noddelay. Some of the resulting images from this test are included above. The noddelay=0 sec case clearly shows the effects of the telescope motion in the image. The noddelay=1 sec case has only a little telescope motion visible and the noddelay=3 sec seems to have none. Unfortunately the noddelay=2 sec case was not performed. These images demonstrate the the noddelay keyword functions as expected and that the telescope nod time overhead (at least for this individual case) was less than three seconds. Note that the time required to complete a telescope nod operation may depend on other parameters, including elevation and amplitude of the move, so it should not be concluded that 3 seconds will always be sufficient. However, the results of the this test confirm that the LWS noddelay system is now operating as expected.

Science with LWS

LWS comes off the bench for an injured MIRLIN. LWS has had several nights of science folded into its commissioning time this semester. These include a night of director's discretionary time and three nights of emergency substitution for MIRLIN observers that lost time on Keck II due to technical problems with the guest instrument. The time was available due to unused engineering time in late April that was originally set aside for the commissioning of ESI. The displaced MIRLIN observers were able to complete portions of their programs using LWS even though some of the capabilities and specifications differ between the two (see Keck mid-IR comparison). The experience was useful for the observers and for the LWS commissioning team, giving us a feel for the expectations and operating needs of LWS observers. The feedback we received has been very valuable and constructive towards improving the instrument user interface. The observing programs carried out included both imaging and LRES spectroscopy. The HRES spectroscopic mode of LWS remains the only significant mode where more observing time is needed in order to prepare the instrument for guest observing in semester 99B.

Go to: LWS Home Page - Instruments Home Page - Keck Home Page
Last modified: Fri May 14 14:09:30 HST 1999