The Messenger
Total Page:16
File Type:pdf, Size:1020Kb
Progress on the VLT deformable secondary mirror The Messenger Commissioning of KMOS Ultra-faint dwarf galaxies The VIPERS redshift survey No. 151 – March 2013 March – 151 No. Telescopes and Instrumentation Upgrading VIMOS — Part II Peter Hammersley1 Figure 1. VIMOS on the Nasmyth Ronald Brast1 platform of VLT UT3. Paul Bristow1 Pierre Bourget1 Roberto Castillo1 Hans Dekker1 Michael Hilker1 Jean-Louis Lizon1 Christian Lucuix1 Vincenzo Mainieri1 Steffen Mieske1 Dan Popovic1 Claudio Reinero1 Marina Rejkuba1 Chester Rojas1 Ruben Sanchez-Jannsen1 Fernando Selman1 Alain Smette1 Josefina Urrutia Del Rio1 Javier Valenzuela1 Burkhard Wolff1 1 ESO VIMOS is the powerful visible (360– – integral field unit (IFU) spectroscopy and operational procedures; and opera- 1000 nm) imager and multi-object/inte- with three fields of view: 13 by 13, tional efficiency. gral field spectrometer mounted on 27 by 27, and 54 by 54 arcseconds the VLT Unit Telescope 3, Melipal. Its depending on the specific mode This article discusses the results of this high multiplex advantage makes it ideal requested. second phase. More technical details for undertaking large-scale spectro- on the overall upgrade programme can scopic surveys of faint sources. In order VIMOS and its commissioning on the VLT be found in Hammersley et al. (2012). to extend the life of the instrument, im - in 2002 were described in Le Fevre et al. prove its performance and prepare for (2002) and in operation by D’Odorico et possible large-scale surveys, in 2009 it al. (2003). After eight years of operation it Improving image quality was decided to upgrade VIMOS. The became necessary to upgrade the instru- first phase of the upgrade, which in - ment in order to address various issues Following the change of the detector, it cluded replacing the detectors and the and to extend its useful life. The first became apparent that the optimum focus fitting of an active flexure compensation phase of the upgrade was implemented was changing across the detector due system, has been previously reported; in 2010, (Hammersley et al., 2010) and to a tilt between the detector and the re- this article describes the second stage included: imaged focal plane. This meant that if the of the upgrade, which has improved – Replacement of the shutters, which detector was well focused at its centre, the delivered image quality and stability. were worn out. then the edges would not be as well – Replacement of the CCD detectors, focused. Due to the optical configuration which has improved the sensitivity of VIMOS this could lead to the images The instrument and upgrades in the red and reduced the fringing. at the corners of each detector becoming – Reduction of the instrument flexure. significantly elongated. VIMOS has four identical arms, each – Provision of new mask cabinets that with a 7 by 8 arcminute field of view on keep masks in position more reliably. In May 2011 an intervention was made the sky with a gap between the fields – Improvement of the data reduction to correct this effect by moving one of 2 arcminutes. The instrument offers pipeline. of the lenses and the detector laterally three main modes: with respect to the rest of the optics. – UVBRIz-band imaging covering four The second upgrade phase took place in This reduced the tilt between the tele- fields each 7 × 8 arc-minutes; 2011–12 and addressed a number of scope focal plane and detector by typi- – slitlet-based multi-object spectroscopy issues which became apparent following cally a factor five. At the correct focus with spectral resolutions from a few the first phase: image quality; the focus only the extreme corners of the array hundred to 2500 in each of the four mechanisms and their control; the effi- showed a small amount of astigmatism. imaging fields; ciency of the HR-blue grism; calibration Figure 2 shows the evolution of average 2 The Messenger 151 – March 2013 Figure 3. Variation of the FWHM in pixels with rotator angle for VIMOS channel 3. The dashed lines show the case with no focus compensation and the solid lines with compensation. "G@MFDC OHW CDSDBSNQ ', %6 "NQQDBSDC 6', SHKS % l l 8D@Q 1NS@SNQ@MFKDCDFQDDR Figure 2. Evolution of the image FWHM in pixels very complicated and it was very easy to so a further intervention in October 2012 since 1 January 2009 for VIMOS channel 4. The red inadvertently make mistakes which were was required to correct this. points correspond to the average FWHM of the pin- holes within 800 pixels of the centre; blue between not immediately obvious. 800 and 1200 pixels and green to those beyond Figure 3 shows the variation in FWHM 1200 pixels (see the schematic of the regions used In March 2012 the original focus stepper with rotator angle when observing the on the detector to the upper left ). Arrows mark when motors were replaced with DC motors, pinhole mask for the three colour-coded the detector was changed and the tilt corrected. each with an encoder. The software was regions designated in Figure 2. The updated to allow the focus offset to now dashed lines show the FWHM with no image quality in three regions of detector take into account rotator angle and compensation with rotator angle and 4 since 1 January 2009. The data were grism as well as temperature and filter for the solid lines indicate the cases with obtained using daytime calibration images each channel. After implementing the compensation. This plot demonstrates of a pinhole mask placed in the focal new motors and encoders, it was found that the instrument is now far closer plane. The dates when the detector was that the mechanism did not have suffi- to the optimum focus in normal operation changed and the tilt corrected are marked cient range to focus all of the modes cor- where the rotator angle is constantly with arrows on Figure 2. The degradation rectly, affecting in particular the IFU, and changing. of image quality following the detector ex - change is clear, but following the correc- tion of the tilt, the image quality to wards Figure 4. R-band image of the globu- the edge of the detector has significantly lar cluster M55 taken in October 2012, after the focus upgrade. This shows improved. Currently, approximately 90% the whole field of view in channel 3 of the imaging area has an image full (7 by 8 arcminutes). The dark area to width at half maximum (FWHM) < 2.4 pix- the top right is caused by the guide els and an ellipticity below 0.1, whereas probe vignetting the image. prior to the upgrade it was closer to 70%. Focus control The original focus mechanism and con- trol software also limited the perfor- mance of the instrument. Among the problems were: the stepper motor could lose steps and so lead to a drift in the focus position of the camera; the camera focus changed significantly with the rota- tor angle and this could not be cor- rected; it was not possible to control the focus for each grism; updating the focus parameters in the control software was The Messenger 151 – March 2013 3 Telescopes and Instrumentation Hammersley P. et al., Upgrading VIMOS — Part II -DVENBTR Figure 6. The spectral ,DBG@MHRL resolution in the red, blue and central regions "NQQDBSDC of lamp spectra, colour- 3HKS coded red, blue and green, taken with the "G@MFDC HR-orange grism in IFU #DSDBSNQ mode for channel 1 is shown. The dates when the detector was BNMCR RQ changed, the tilt cor- B rected and the new @Q focus mechanism were implemented are all 6', % marked. It should be noted that this is one of the more extreme examples. 7OHWDK 8D@Q Figure 5. The variation of the stellar FWHM, in arc- tre of the globular cluster (top right). focus was not optimum then the resolu- seconds, as a function of the X position on the There is some elongation in the bottom tion was degraded, particularly at each image shown in Figure 4. corners of the image and the FWHM end of the spectral range. The new focus in those regions increases from about control allows the focus to be set for each Figure 4 shows a VIMOS R-band image 0.55 to 0.60 arcseconds. grism and this has led to the resolution of the globular cluster M55, taken with being more uniform between the four 0.5-arcsecond seeing during the last quadrants and over time. The red end of intervention. This uses the full imaging Spectral resolution the spectrum now always has a higher field of view on channel 3. Figure 5 shows resolution than the blue (as it should be), how the image quality varies with posi- The grisms used by VIMOS are placed which was not always the case before- tion on the horizontal axis (X direction). in the pupil and it was originally assumed hand. The sources with FWHM ~ 0.55 arcsec- that these would not significantly affect onds are stars, while extended back- the focus. Tests in May 2011, however, As an illustration of the implication of ground objects and blended stars have showed that this was not the case, and this change for observations, we estimate larger FWHM, in particular near the cen- the high-resolution grisms in particular re - that programmes to detect Lyman-α quired a significant focus offset if the emitters at high redshift would have up Figure 7. The total VIMOS efficiency when using the spectral resolution was to be maintained to a 40% gain in sensitivity just because old (purple) and new (green) HR-blue grisms is shown.