You can also review Charles Claver's SPIE paper on WTTM performance
which provides ample detail to the instruments performance.
User's Manual (PDF) (McDougal, Howell, Corson, et.al...)
- User Manual
- Tutorial on WTTM Telemetry
The user's manual is complete and has been released to the user community. The manual contains a quick start guide for using the instrument and information required for observer run preparation and configuration.
Photometry has been successfully demonstrated using WTTM and the 85/15 beam splitter. Links to Abi Saha's report are provided above and is provided as an appendix to the WTTM User's manual. Due to the similarity of the 85/15 and 95/5 beamslitters, and the general uniform dependence on color beyond approximately 430 nm, photometry with the 95/5 beamsplitter is judged to be unnecessary.
Claver provides further examples of photometry in the Spring 2004 NOAO
Newsletter using the 85/15 spare beamsplitter and a color magnitude diagram
from data is shown here. There are no unusual issues noted
in reducing photometry standards by Saha or Claver. Claver does note
a need to use a 2nd order term in I-band, likely a result of the wide range
of magnitude for this pass band.
Figure 1: Color Magnitude Diagram NGC-1193, BVI photometry using
At present, WTTM images can be flatfielded adequately as judged by the ability to perform photometric calibration and the ability to produce a 'flat' or uniform sky across the CCD. Analyzing how well the flatfields perform for all scientific goals has not been rigorously undertaken, as no specific criteria was given. Some details regarding flatfielding with WTTM are noted below:
- Limited testing of the flatfield repeatability due to the insertion/removal of the WTTM pickoff mirror indicated no effect. Flatfields taken before and after the pickoff mirror's insertion/removal were identical as determined by taking the ratio of the individual flatfield images and determining that there is no gradient across the illuminated area nor were any shifts observed for any distinct artifacts due to dust particles.
- One can obtain useful flatfields at any position of the tip-tilt mirror. There is a subtle effect due to the position of the tip-tilt mirror, but only at the very edge of the useful imaging area where the illumination gradient is strongest. It is recommended that this region, roughly 50 pixels in width, be avoided for your reductions.
REPORT: Flatfield as a function of tip-tilt position
- We recommend that one obtain the exposures with the WTTM control system on in order to render the tip-tilt mirror stable during the flatfield exposures, but it is not necessary to position the mirror to some nominal orientation.
- There is flexure in the instrument, at about a 0.25 arcsec or 2 pixel level and is therefore not predicted to induce significant flatfielding errors.
NOTE: The key issue in this discussion is that there was no criteria given for determining if WTTM images can be adequately flatfield.
Shutter Correction Report, Howell: [html]
STATUS:Bad Pixel Mask
Exposures that are less than 5 sec in duration should also include a shutter correction, the error in exposures time of ~1% are seen at the 3 sec level. The rather conservative limit of 5 sec is due to an uneven shutter correction image; there are 'bright' areas of the shutter correction with the shutter correction increasing in magnitude as one move radially from the center. WTTM users should not use exposures times of less than 1sec.
Users who will be using short exposure times should obtain shutter correction data and produce the correction image.
Steve Howell has kindly identified obvious bad pixel columns and pixel areas and provided the data below. This ascii data can be used as input to the IRAF routine 'proto: text2mask' to produce a bad pixel mask image.# WTTM CCD bad pixel map# 2K x 4K EEV CCD#Column Line#Start End Start End204 204 1469 2500255 260 1207 25001097 1097 263 25001445 1449 2005 20101447 1447 2011 25001793 1793 1079 25001837 1837 1192 2500
CCD Header Content
All header content is valid and updated for the instrument. IRAF's CCDPROC correctly uses the relevant keywords. As of this report, the World Coordinate System parameters have not been verified, on-sky.
DIQ Improvement with Tip-Tilt
There has been no evident dependency on the DIQ improvement and color (B&I), when using tip-tilt corrections. Though one expects some dependency, variations in seeing (what the site and facility delivers) have dominated when attempts to measure these have been made. Therefore, efforts have concentrated on obtain a more general result which indicates that a 10-15% improvement in seeing (FWHM) is obtained for the expected seeing at WIYN, see Figure 3 below. One should interpret this as a general result for the B, V, R, and I pass bands. An interpretation of this result might be that the source of tip-tilt errors at WIYN are generally very localized to the telescope (ex. mount jitter and mirror seeing). It is also observed that using WTTM will not always provide the stated 10-15% improvement in DIQ. It may be less, possibly no improvement, and it maybe more depending on site conditions.
The intrinsic image quality delivered by the WTTM optics is as designed, producing diffraction limited images in the lab but also achieving 0.3 arcsec images in R and I. Data obtained to compare MiniMo and WTTM, in static or no tip/tilt mode, show the same delivered image quality to within the variations of seeing. Also, an experiment comparing WTTM to Lowell Observatory DIMM instrument was carried out and demonstrated that WTTM was able to deliver seeing at the 0.3 arcsec level and take advantage the excellent seeing provided by the site DIMM Comparison Results .
There has been no formal study during this commissioning of DIQ gain
versus guide star magnitude or correction speed. However,
it has been demonstrated that the best performance is obtained when the
user samples and corrects at the greatest possible speed that the star
will allow, so long as a minimum level of signal maintained.
In general, it is recommended that one sample and correct at rates greater
than 100 Hz (99% of the tip-tilt error occurs at 10 Hz or below but WTTM
requires sample and correction speeds at 100 Hz or more to adequately remove
these errors). Adequate signal to noise is found to be the strongest
influence on the tip-tilt correction performance. On-sky data
has shown that one will require a minimal signal level of 30-40 cnts above
background at any sample rate to achieve optimum performance. Signal
below about 30 cnts produces a 'noisier' determination of the guide star's
centroid, and rapidly degrades with less signal.
Focus Error Sensing
As designed, the WTTM error sensor provides for the ability to measure changes in focus by sensing the change in astigmatism of the tip-tilt guide star. As simple as this method is, it is susceptible to various influences that can over-whelm the astigmatic signal. Therefore, the WTTM focus sensing is not recommended for observer use. We find that the WIYN IAS autofocus sensor is adequate for sensing focus error, that stars are generally available for this sensor when WTTM is in use, and there is no degradation in DIQ for WTTM when used.
The factors that may influence the focus sensing are typically changes in the telescopes static aberrations, seeing, and varying sky background. Also a limitation is the rather narrow 'capture range' of the sensor, typically +/- 15 um. This limited capture range, which is not symmetric, results in the occasional condition where a sudden focus change is out of the capture range and results in the corrective signal sent to the telescope driving the secondary further out of focus. The WIYN port's autofocus sensor is far more reliable and well understood and we therefore choose to use this sensor for focus error sensing.
An astronomer's interface to WTTM will not be provided, but the engineering interface as been simplified and improved to suffice as the default user's interface. As of this report, no telemetry tutorial has been provided.
Demonstration of WTTM throughput
Using both on-sky photometry, and an independent comparison using a near V pass band, it is determined that the throughput of WTTM is approximately 70-65% relative to MiniMo using the 85/15 beamsplitter. This easily beats the SAC defined criteria of 50% using the 95/5 beamsplitter.
Distortion Map and WCS Header Parameters
STATUS:BeamSplitter and Dichroic Wavelength Response
A report on the distortion as measured at the WTTM CCD has not been completed at this time. Valdes has provided World Coordinate System (WCS) constants for the image header using a small subset of data obtained for the distortion map, but these constants have not been verified on-sky.
STATUS:CCD and Filter Characteristics
Two types of beamsplitters are available for WTTM, one with approximately 90/10% reflectance/transmission and one with approximately 95/5% reflectance/transmission. There is no dichroic (no vendor submitted a bid for the dichroic). We refer to these as the 85/15 and 95/5 beamsplitters respectively based on their original specifications of reflectance and transmission properties.
The wavelength response of the two beamsplitters is uniform from about 4300 angstroms to 1 um. Below 4300 angstroms there is a strong 'leak', or peak in transmission, but WTTM's error sensor is not sensitive to these wavelengths, it's peak response is centered at wavelengths of 6700-7000 angstroms, see the 85/15 transmission curve and the APD QE curve below. NOTE: there is a discontinuity in the transmission data at 8700 angstroms and is a measurement artifact of the spectrophotometer and not a property of the beamsplitter.
All beamsplitters have AR coatings and effectively eliminate ghosting and reflections for most observations. Bright objects, about 10th magnitude and brighter, will produce a small, reflected image but at very low signal, and at a predictable location from the source.
One of the 85/15 beamsplitters has already been lost due to an apparent coating loss. We are working with the vender to resolve this issue and will explore the acquisition of another '85/15' beamsplitter.
Figure 4: WTTM 95/5 BeamSplitter Transmission Curve
Figure 5: WTTM APD QE Curve
There will be no further CCD characteristic testing due to the anticipation of new CCD system to be delivered at some future date.
- linearity (75,000 e- @0.1%)
- gain (1.3e-/adu )
- readnoise (4.5 e- )
- dark current
Due to column traps and other defects in the CCD, engineering grade chip, there are two led's on the chip. Though dark current is not expected to be an issue, indeed the dark-current is measured to be less than 0.002 e- / second, the led's will be. The two led's are located at pixel locations 198,1423 and 1447,1984. The later position is within the imaging area of the CCD and illuminates an area that is centered within a 50 pixel radius.
The 'fat-zero' characteristics have not been measured.
- noise isolation
No noise coupling to WTTM or the telescope has been observed, but we would advise to follow normal observing practices.
As this is a one amplifier chip, no cross-talk is observed. Bright stars in the presence of low sky background, will produce a residual in the overscan. This is a known issue of all KPNO CCD systems.
- QE (~88% at 5000 angstroms, peak): See WTTM CCD QE Filter Set links below
- WIYN WTTM Filter List:
WTTM CCD QE & BroadBand Filter Set Transmission
WTTM CCD QE & Sloan Filter Set Transmission
Note: Nominal filter thickness is 5 mm +/- 0.1 mm. The optical performance of WTTM cannot be guaranteed, nor can the use of the WTTM focus sensor, for filters that deviate from the nominal filter thickness. Users should be aware that the PSF will degrade due to the necessity to shift the telescope focus at the WTTM re-imaging plane, the instrument entrance, to accommodate the non-standard filter, focused for the CCD focal plane.
WTTM # Cwl, A FWHM, A Filter Thickness, mm Focus Offset
%T max @ wavelength,A Name & Comments Date Measured Age Data Tables Graph 1 4340 1051 5.105 +10 +/-5 73.5 @4,360 Kron B 02-2003 2003 Ascii Graph 2 5353 829 5.055 -10 +/-5 87.6 @5,200 Kron V 02-2003 2003 Ascii Graph 3 6298 1208 5.029 - 82.1 @5,920 Johnson R 02-2003 2003 Ascii Graph 4 8158 1770 4.902 0 +/-2 91.4 @8,375 Kron I 02-2003 2003 Ascii Graph 5 4744 1518 4.902 0 +/-10 93.0 @5,010 Sloan g' 07-2003 2003 Ascii Graph 6 6172 1300 4.953 0 +/-10 96.2 @6,436 Sloan r' 07-2003 2003 Ascii Graph 7 7685 1259 5.055 0 +/-10 98.9 @8,068 Sloan i' 07-2003 2003 Ascii Graph 8 5.131 0 +/-5 97.5 @10,250 Sloan z' 07-2003 2003 Ascii Graph Table 1: WTTM 2x2in Filter Set
(All filters are 2-inch square and measured with the NOAO Lambda 9 Spectrophotometer with a ~f/13 beam)
WTTM Integration into WIYN TCS/Router Environment
STATUS:WTTM Focus Offset with respect to MiniMo
Complete and functional. All interaction with the WIYN environment and WIYN TCS are operational (telescope guiding, headers, secondary focus control..)
NOTE: No provision has been provided to add WTTM telemetry to the WIYN telescope environment due to work loads on other personnel (MPG).
STATUS:Optical Alignment with Telescope and MiniMo
WTTM CCD is co-focal to MiniMo to within 50 um of secondary focus, R-band. Please see the WTTM user's manual for detail.
STATUS:Scattered Light Measurements
The WTTM CCD center is well matched to the WIYN port optical axis and to MiniMo. WTTM's CCD center and the WIYN port optical axis are co-aligned to within 10 arcsec. WTTM's CCD and MiniMO's centers are 30 arcsec mis-aligned, but this works in one's favor as the center on the WTTM CCD will fall left of the gap on MiniMo.
Measurements to determine the repeatability of the WTTM pickoff mirror and optical alignment have produced repeatability to better than 3pixels, or ~0.5 arcsec.
There is no evidence of any scattered light issues. However, it is noted that bright stars, brighter than about 10th magnitude or so, will produce weak reflections of the in-focused star, see user's manual. Fortunately, the reflection occurs in a repeatable and predictable fashion.
Many images have been obtained with WTTM with bright stars in the field of view and just outside of the field of view. There is no evidence to indicate any serious or detrimental issues with light scattered off of reflective surfaces.
The PSF is very uniform over the small, 4arcmin field of view. Variations observered are largely a function of radial distance from the guide star, a very weak dependence: see Figure 5 below. Variations are also seen at the extreme edge of the field where the focal plane distortion is greatest, or where the illumination gradient is largest and represents only the outer edge of the WTTM CCD field of view. As for the uniformity of the PSF itself, we do note a ellipticity in some of the images and a likely cause has been uncovered. This cause is due to a slight inbalance in the tip-tilt X/Y amplifier response, ideally they would be balanced. Resolution of the issue is not part of the commissioning effort and will address it as time and resources permit. Note: it worth stating that uniformity of the PSF is also dependent on the selection of an isolated, single star that has no near neighbors closer than about 4arcsec.
Figure 5: FWHM variation as a function of distance from the guide star
The flexure of the WTTM instrument has been measured on-sky and is determined to be less than 0.25arcsec over the entire range of the WNIR rotator.